ML20247K938
| ML20247K938 | |
| Person / Time | |
|---|---|
| Issue date: | 05/13/1998 |
| From: | Advisory Committee on Reactor Safeguards |
| To: | |
| References | |
| ACRS-T-3037, NUDOCS 9805220270 | |
| Download: ML20247K938 (326) | |
Text
': : c.
.' ' ' _ ^ * '. .'.-_...:.i-:
d'^ '
9:.f. 4,k k iM.7 ';j' ' # '.
- f. '. ' '" E' 4$ ' .' ..
Q,cg 3;
- s:- .' 9 ' M 3 7 1 gg%:4f., '
. .g. . g - [
.- . L g
( .
af 9 a .
, e . ., ' $ . : "
,e
- s. D
- h. f, jf(
4
.$ .3k D v ; ' , ' ? ?
- f. '
a ! I' - SM ' i 6 . jg .t .
L,
f,'.. ap
', 4,:.3_;' 9 p , q.
, . ! !' M il U I ' T -
.l cl - ' ~ '
, 8
}.$ 7 ' i
~~.87:r ..
' [':"% ' ' .
NN -
~
, Q '.'.'. . ..
.y b
~~
, . lg . 3. .y - . q, . . . ' g. : , . .
.7 4 .
4.f . ..4
,h b ;w p 7
=
'. ~!. i ' '-
!.%:. m gwdog. i?
- . s . . -- -
7 ' ..
,f
'.y.Af.f., ..; , 2
. r
.~ l _.
.. l
^
_. ; ' l .
s' - ' ,,
~ , ;c. . .
- g. e
.u- .
W:a, ..,. i., h, . ; i 1
' ' * ^
,, i 2. 6 :. . . .:
>..+
. .~. _.ug.L: . o .
- %,.o ,g .
- ::v's; 2 s. h.. .
hhk "p:. :a
- Nt 3
.. :e
. . 3n.
~
f'
- I
$W.n.hlih*f u; :
~
l_
f
.k;w'ay5 h..p a'
n .. - 4 .
,..;..t . . + -
- .; w.
- 4.) ", *
? g 4
=
'< .- .'I ,
, ..p_
.p .:1; .gk.,q. %. ,
- 4 ~'
- $ i' '. .. %
e.;
E hyd;i
. f.* f'.., -
' l. -l < z. b v 1 .
I n ; .&g. j' ' :l r '. '" w a
~. _ -
a p:;. Aq .' .
- v. ;.- . .. .
. , ; ._ ,.. , .. ,g, . .
y ', + = ' ' # ,
- .j e , . . . .
.g
.. J -
%; # W. -.
- Q h.
% :.y ,
f . s.
f.N ~,, >$ -
e
- s. . > s. ... ;: :. . .. . . 7 -
h e 5
O$
4
- m 7 .
- g .
--' 4>-.
[} , y
'e : .
. .[
9_
..*b .
- .. - . 4 . . , ' .. -
. . , . . j.- ' c - -
3,; G
~
^'
~ -
- c. . .- .
.$ . 6 I
[.'- p' T
~-
?.h ..j ' '
- .:.-, . ~ .-
+ - -
ru.',e. : .- . e
-:s s ,
p .
' *' ' - ' ' ,- , :' _. f'.*
- } ~l }'? .
a.- s v
[8
.. p
- .x,. , '
-4 -
,i'.w;...,$
- .:1, .
.x.., . ..;. . .
- pn.; . -
.= . . ,
..,.l.,., -
'8.. ..
- o
....p.-
^* ,
y..
i' 9' .
.. .a , . ^ .
. . . . ..,..i :~ , ',
- yg
'- .6 {
~ *
.i_ .' .C. ,- ;
{ ; .,. ,]_ '-5,
[$ -
l m .,
3 .f 77 .
m v.],, ,f.. .., .
, +
. . . .ap...,.
- ' . Y y; _ l ,l '
-,w ...wa . :.. .
. >J ..
- s. .
. w.. ,
4
- l .+ [ , " ,[
-- +*,- _- .'..-,. *,.
m ,
ORIGIN _fCR$/-So37
,q. OFFICIAL TRANSCRIPT OF PROCEEDINGS I LNUCLEAR REGULATORY COMMISSION 4;
- ADVISORY COMMITTEE ON REACTOR' SAFEGUARDS l I
- ,a ;
1 j
Title:
- SUBCOMMITTEE ON.
. ADVANCED REACTOR DESIGNS l
Docket No.: Ta 4 'Acas agras cua:NAL l Tc BTdHI". ' l K3 . celt J 4;5 7 '. 3 )
.. Work Order No.: ASB-300-278 THANKS: !
.3
\
(f
- LOCATION: Rockville,hfar3 1and DATE: Wednesday,hiay 13,1998 PAGES: 1 - 258 g
98CS220270 980513 Ulm ACRS T-3031 PDR V [}
.; V ,v4
. 3 bs o-ANN RILEY & ASSOCIATES, LTD.
. ).
1250 I Street, NW, Suhe 300 e' N Washington,D.C. 20005
.i
's )- aq f -
p ' m-y ) bpp , J292) 842-0034Ce , sopy-p ,0F.2e j8 0f Me Congpp j
l DISCLAIMER UNITED STATES NUCLEAR REGU.'.ATORY COMMISSION' S ADVISORY COMMITTEE ON REACTOR SAFEGUARDS MAY 13, 1998 The contents of this transcript of the proceeding of the United States Nuclear Regulatory Commission Advisory Committee on Reactor Safeguards, taken on May 13, 1998, as reported herein, is a record of the discussions recorded at the meeting held on the above date.
This transcript had not been reviewed, corrected and edited and it may contain inaccuracies.
1 O
l 1
1 UNITED STATES NUCLEAR REGULATORY COMMISSION 2 ADVISORY COMMITTEE ON REACTOR SAFEGUARDS 3 ***
4 SUBCOMMITTEE ON' 5 ADVANCED REACTOR DESIGNS 6
7 8 U.S. Nuclear Regulatory Commission 9 Two White Flint North 10 11545 Rockville Pike 11 Rockville, Maryland 20852-2738 12 13 Wednesday, May 13, 1998 14 15 The Committee met pursuant to notice at 8:31 a.m.
16 17 MEMBERS PRESENT:
18 JOHN BARTON, ACRS Chairman l 19 MARIO FONTANA, ACRS Member
, 20 ROBERT SEALE, ACRS Member 21 THOMAS KRESS, ACRS Member 22 DON MILLER, ACRS Member i
23 DANA POWERS, ACRS Member l 24 WILLIAM SHACK, ACRS Member 25 ANN RILEY & ASSOCIATES, LTD.
Court Reporters 1250 I Street, N.W., Suite 300 Washington, D.C. 20005 (202) 842-0034
2 1 PROCEEDINGS
() 2 [8:30.a.m.]
3 CHAIRMAN BARTON: The meeting will now come to 4 order. This is a meeting of the ACRS Subcommittee on 5 Advanced Reactor Designs. I am John Barton, Chairman of the
[
l 6 Subcommittee.
ACRS members in attendance are Mario Fontana, i
l 8 Thomas Kress, Don Miller, Dana Powers, Robert Seale, and j 9 William Shack. We also have in attendance ACRS consultants, 10 James Carroll and Ivan Catton.
11 The Subcommittee will continue its review of the 12 Westinghouse AP-600 design. Specifically, the Subcommittee l 13 will review chapters 3, 6, 14, 16 and 17 of the AP-600 l 14 standard safety analysis report, the probabilistic risk 15 assessment, regulatory treatment of non-safety systems, and 16 the associated NRC staff's evaluations.
- 17 The Subcommittee will gather information, analyze i
l 18 relevant issues and facts and formulate proposed positions 19 and actions as appropriate for deliberation by the full 20 committee.
21 Noel Dudley is the cognizant ACRS scaff engineer 22 for this meeting.
23 The rules for participation in today's meeting 24 have been announced as part of the notice of this meeting 25 previously published in the Federal Register on April 22, ANN RILEY & ASSOCIATES, LTD.
Court Reporters l 1250 I Street, N.W., Suite 300 Washington, D.C. 20005 (202) 842-0034
3 1
1 1998. A transcript of the meeting is being kept and will be (q) 2 made available as stated in the Federal Register Notice.
3 It is requested that the speakers first identify l
4 themselves and speak with sufficient clarity and volume so {
5 that they can be readily heard.
6 We have received no written comments or requests 7 for time to make oral statements from members of the public.
8 This is the first day of a three day meeting. l 9 Today, the Subcommittee will hear presentations on and l 10 discuss SSAR Chapter Six, engineered safety features, 11 Chapter 14, initial test program, the level two and three 12 probabilistic risk assessment, and severe accidents, I 13 including external reactor vessel cooling.
14 Copies of the AP-600 SSAR and PRA are available at l
' (, 15 the front of the room. l 16 DR. CARROLL: John, what became of the level one 17 PRA? When are we going to do that?
18 MR. DUDLEY: That was covered during the 19 Subcommittee back in 1997, I believe, June 1997. We will 20 hear additional information on the level one PRA during the 21 June meeting.
22 CHAIRMAN BARTON: We will now proceed with the 23 meeting and I call upon Tom Kenyon of the staff for an NRC i 24 status of the AP-600 review.
25 MR. KENYON: Thank you, Mr. Barton. My name is (O'j ANN RILEY & ASSOCIATES, LTD.
Court Reporters
- 1250 I Street, N.W., Suite 300 Washington, D.C. 20005 (202) 842-0034 i
l 1
4 j 1 Tom Kenyon and I'm the senior project manager for the
() 2 AP-600. I just wanted to spend a few minutes talking about 3 the status of the staff's review to date, and reminding you 4 of the schedule which the staff is working under.
5 By way of history, in November of 1994, the staff 6 issued a draft SER that contained 1,100 open items, and then 7 in 1996, we issued a supplement to that draft SER that 8 talked about the test program and Westinghouse's 9 co-development program.
10 Of course, on May 1 of this year, we issued an 11 advance copy of the final SER that contains eight open items 12 and 15 confirmatory items. We have come a long way.
13 The schedule that we have in front of us now is we l 14 will be meeting with ACRS Sub and Full Committees over the 15 next couple of months and our schedule right now is to 16 finish up the final SER and publish it in August 1998 and 17 make our final decision on the final design approval in 18 September. i 19 The next slide I have in the package is --
20 DR. CARROLL: That neans the Commission makes the 21 decision or staff?
22 MR. KENYON: No, the FDA is a decision made by the 23 Director of NRR, Mr. Collins, at this time, although the l
24 Commission will be informed of our decision.
25 The next slide --
ANN RILEY & ASSOCIATES, LTD.
Court Reporters 1250 I Street, N.W., Suite 300 Washington, D.C. 20005 (202) 842-0034
5 1 DR. CARROLL: They have no action?
() 2 MR. KENYON: They have no action per se, unless 3 they choose --
4 DR. SEALE: Negative consent.
5 MR. KENYON: Right. The next slide is a listing 6 of the meetings that we have in front of us with ACRS Sub 7 and Full Committees. As you can see, we have a large number 8 of meetings ahead of us. We are going to be meeting wit the 9 committees at one time or another on an average of every two 10 to three weeks until July.
11 As Mr, Barton said, today's topics cover a wide
- 12 diversity of subjects including structures, engineered 13 safety features and DRA and a number of other issues.
14 Although there are a few open items on these 15 subjects identified in our SER, we have been working with 16 Westinghouse to get closure. In fact, in some cases we feel 17 that we have closure and we are just finishing up on the l
18 documentation.
19 I think it's best to characterize it that we don't j 20 have any significant open items in these areas. It's just a 21' matter of finishing up our review.
22 Staff will be present to answer questions during 23 all of Westinghouse's presentations. We have prepared some 24 presentations to address some specific questions by the 25' Committee. In addition to that, we will be prepared to talk ANN RILEY & ASSOCIATES, LTD.
Court Reporters i 1250 I Street, N.W., Suite 300 l Washington, D.C. 20005 l' (202) 842-0034
1 6
I 1 about the status.of the open items on the particular '
() 2 technical areas. In some cases, we have a pret antation and 3 in other' cases, we will probably speak from our desk and 4 just inform you of the status.
5 Unless there are any questions, that completes my l 6 presentation.
7 DR. CARRCLL: You' don't have a summary of the 8 eight open items?
9 MR. KENYON: If it's not in your package, I have a 10 listing of them. I l
11 DR. CARROLL: I'm just curious what they are. !
l 12 Every time you make a presentation, you say we have no open 13 items on this chapter.
14 MR. KENYON: I say no significant --
15 DR. CARROLL: Okay. So, some of these are?
16 MR. KENYON: .Let me go down the items. Number one 17 is the review of the proprietary information, and 18 Westinghouse has submitted some proprietary information. We 19 just have to complete that by August. The comparison of the 20- AP-600 design to the EPRI URD was a directive from the l
21 Commission to do that, although Westinghouse submitted 22 something in January, a comparison, we just need them to I 23 update it after they make all the final modifications to the 1 24 design.
25 Exemptions to the regulations, we have started a l (~~
ANN RILEY & ASSOCIATES, LTD.
Court Reporters l 1250 I Street, N.W., Suite 300 Washington, D.C. 20005 (202) 842-0034
--- -- - ---- j
l 7 l
1 list of them. Westinghouse needs to provide justification
) 2 for them. We suggested to Westinghouse that they wait and 3 see what we had in the SER so they know what exemptions are 4 needed.
5 DR. CARROLL: They do have the SER?
7 DR. KRESS: Are there a lot of exemptions?
8 MR. KENYON: I'm trying to remember. There's 9 probably six or seven. They are not unusual. Some of them 10 are either characteristic because of the special design of 11 the plant and mosc of them are things that we have seen in 12 the past on ABWR.
13 One of the topics we will be talking about today 14 will be the DBA analysis using a revised set of heat sink
, 15 consumptions. That is a chapter six item.
l 16 DR. KRESS: Is that on the containment shell?
17 MR. KENYON: It's on the containment shell.
i 18 Another item is the location of the fire 19 protection pumps. I think we have reached resolution on 20 that and we are just finalizing the documentation. I I
21 believe Westinghouse is going to move one of the pumps to 4 22 accommodate a concern.
, 23 DR. CARROLL: What's the concern, too close l'
24 together?
25 CHAIRMAN BARTON: In the turbine building, none of i
( ,/ ANN RILEY & ASSOCIATES, LTD.
Court Reporters 1250 I Street, N.W., Suite 300 Washington, D.C. 20005 (202) 842-0034 t
8 1 it was outside.
, {A) 2 MR. KENYON: ITAAC is of course a big issue. The 3 famous SER didn't really talk about that. We are putting 4 together an SER on it. We will be issuing that shortly and 5 we will be prepared to talk about that in the June 6 Subcommittee meeting.
7 DR. CARROLL: Do I understand correctly that 8 Westinghouse has revised the version of the certified design 9 material that I have on my desk at home?
l 10 MR. McINTYRE: Yes.
11 MR. QUAY: This is Ted Quay of staff. They have 12 just submitted on Monday a complete revision cf that. I J 13 also want to point out that your exemptions are given in l 14 Section one of ..e SER as well as the open and confirmatory
'V 15 items, so you have a complete listing in that chapter if you 16 want to go back and look at them. l l 17 DR. CARROLL: On certified design material, I'm j 18 going to toss out the book I have and I'll get a new one? I l
19 MR. QUAY: Completely.
20 DR. CARROLL: I can't help but comment that we 21 have to give Noel here a Guinness record or something for l 22
)
the amount of paper he sent us to get ready for this >
l 23 meeting.
24 MR. KENYON: We have an item concerning the 25 containment enclosure and building ventilation operation O
1 <
ANN RILEY & ASSOCIATES, LTD.
Court Reporters 1250 I Street, N.W., Suite 300 Washington, D.C. 20005 I (202) 842-0034
9 l' while moving fuel, We will be talking about that this week.
() 2 There is an~ issue on the QA program and again, we will be 3 talking about that this week when we get to it.
4 CHAIRMAN BARTON: Thank you, Tom. At this point, 5 I'll call upon Mr. Brian McIntyre,. Westinghouse, to start 6 the presentation.
7 MR. McINTYRE: Thank you, Mr. Barton. The one 8 thing that may be a surprise from what Tom said is we 9 weren't planning to discuse the containment analysis today, 10 I think Noel and I had talked about that, and we are going 11 to do that as part of the -- we have a PCS thermal 12 hydraulics subcommittee meeting that is on Tom's schedule.
13 We were going to discuss it with the thermal hydraulics 14 subcommittee at that time.
\_) 15 The first speaker today is Terry Schultz. He is 16 going to be talking about chapter six. I think it's l l
17 unfortunate that Dr. Wallis isn't here because at our last 18 ' t.ieeting, there were many, many questions which were deferred l
19 to Mr. Schulz to the point of people were wondering if there 1
20 was a Mr. Schulz, and I guess without any further ado, I I 21 feel he should have a drum roll, because this is one of the 22 more anticipated presentations that we have had, we will 23 bring you Mr. Schulz with chapter six and the description 24 of these engineered safety features of the AP-600.
I 25 DR. KRESS: That's really your name?
ANN RILEY & ASSOCIATES, LTD.
Court Reporters 1250 I Street, N.W., Suite 300 Washington, D.C. 20005 (202) 842-0034
10 1 MR. SCHULZ: It certainly is. Brian has been
() 2 feeding me comments from you from previous meetings and they i 3 kept coming. In fact, I even got a few more yesterday 4 before I came down. i 5 I hope that I can answer your questions. There is 6 a lot of material in chapter six. I've prepared some 7 presentations on these four areas. I think most of your 8 questions were in the passive core cooling system, a few of 9 them were in containment cooling and hydrogen. I hope to 10 speak to them as I go through these systems.
11 In the back of the handout are some PSID type 12 drawings, which I won't necessarily refer to except in 13 response to questions or possibly to talk about one or two 14 of the questions that you had, so they are more of a O 15 reference material.
16 Unless you have any other suggestions, my 17 intention was to go through these four pretty much in the 18 order that they are in the SSAR, starting with the passive 19 containment cooling system.
20 I know you have had some contact with this system, 21 and in particular with the discussion of the analysis of the 22 containment. I, of course, am going to be talking more !
23 about the aystem and how it works and what it is supposed to 24 do, how it is designed.
25 You see here the requirements that the system ANN RILEY & ASSOCIATES, LTD.
Court Reporters 1250 I Street, N.W., Suite 300 Washington, D.C. 20005 (202) 842-0034
(
11 1 performs in terms of limiting the peak pressure to less than
() 2 design pressure, considering large break LOCA's, steam line 3 breaks, reducing the containment pressure in the long term.
4 A footnote here, our ghost calculations do not actually 5 credit reduced leakage in the long term. He, wever, it is 6 still a functional requirement placed on the system.
7 The system also has to continue working 8 indefinitely. The system provides three days worth of water 9 in the initial elevated gravity tank. After three days, 10 actions are taken to continue water supply. There are two 11 basic ways of doing that. One of them is to use off site 12 equipment and water supplies. The other way of doing that 13 is to use equipment that is part of the passive core cooling 14 system, to extend the water on site supply from day three to 15 day seven,.and then after day seven, we only need to bring 16 water to the site and continue to use the on site pumps.
- 17. That's basically what the system does from a 18 containment cooling point of view. It has an ancillary 19 function of providing long term and fuel pit make up. Now, 20 in a normal plant operation, when you are not in a 21 re-fueling, this would only occur after seven days and the 22 water would be basically coming from off site and it goes up 23 to'the storage tank and then can drain down either to the 24 containment and/or to the spent fuel pit.
25 In some re-fueling nodes of operation, where PCS ANN RILEY & ASSOCIATES, LTD.
Court Reporters 1250 I Street, N.W., Suite 300 Washington, D.C. 20005 (202) 842-0034
12 1 water cooling is actually not required, and all or most of (Q,f 2 the core can be off loaded into the spent fuel pit. The PCS 3 water can be used earlier than seven days for spent fuel pit 4 make up.
5 DR. KRESS: To achieve the less than 45 psi 6 pressure, did I understand correctly that you had to invoke 7 two dimensional conduction in the containment shell?
8 MR. SCHULZ: You are getting into --
9 DR. KRESS: Is this too much -- is this for the 10 other meeting?
11 MR. SCHULZ: This is really for the other meeting.
12 The passive water containment cooling system plays a very 13 important role but is not the only featurn that limits the 14 peak pressure. Obviously, the size of the containment is a
'(-' 15 major factor in reducing or limiting the peak pressure. In 16 the longer term, 10, 20, 30 minutes, you can also get into 17 peaks and the performance of the PCS is important in that.
18 I'm not really going to talk about the details of the 19 performance. I'll talk a little bit about water flow rates, 20 but not specifically about the analysis.
21 DR. KRESS: How long does it take you to get down 22 to half of the peak pressure?
23 MR. SCHULZ: Let me defer that, if I could. One 24 of the things I want to talk about is water flow rates and l
25 how the system works, and I'll speak about reduced pressure.
() ANN RILEY & ASSOCIATES, LTD.
Court Reporters t 1250 I Street, N.W., Suite 300 l Washington, D.C. 20005 l (202) 842-0034 c_____________ ___-
13 1 DR. CARROLL: What's the volume of this r,
(] 2 containment? Refresh my memory.
3 MR. SCHULZ: It's about 1.8 million cubic feet.
4 DR. KRESS: It's bigger than standard.
5 DR. CARROLL: An 1,100 megawatt plant would be 6 more like 2.2?
7 MR. SCHULZ: It would be bigger than that. A y 8 typical two loop plant would be somewhat smaller than this, 9 so it's kind of in between. It does have a little bigger 10 reactor coolant system volume than a typical two loop plant, 11 which is one of the reasons why it doesn't need a --
12 DR. CARROLL: I said a four loop plant would be 13 2.2 or something like that.
,, 14 MR. SCHULZ: Yes, it's a relatively big
\ ,/ 15 containment for the megawatt rating.
16 One other function it provides is a seismic fire 17 water capability, and this water is actually stored in the 18 passive containment water storage tank and there's 19 connections to the fire protection system. I'll point those 20 out in the drawing.
21 What I'd like to do now is to put the sketch of 22 the system up and then talk about the word points on your 23 next two overheads.
24 DR. CARROLL: One more question about the l 25 functions. The three to seven day period, where you use on s
(' ') ANN RILEY & ASSOCIATES, LTD.
Court Reporters 1250 I Street, N.W., Suite 300 Washington, D.C. 20005
!- (202).842-0034
14 1 site pump and water supply, assumes that you have gotten on
() 2 site power back?
3 MR. SCHULZ: I can point out here, this tank, 4 ancillary water storage tank, and these pumps, were part of 5 the agreement that we reached with the staff on post- 72 6 hour actions. What we have provided is this equipment is 7 not safety related, but it is provided with some hazard 8 protection, seismic and high wind, including power supply.
9 Now, the power supply for this comes from two 10 small ancillary generators. These are very small. They are 11 not the main standby diesel generators that we have in the 12 plant. They are only sized to provide these pumps some I&C 13 capability and fans. They are very small capacity. They 14 are also seismic and wind protected.
15 DR. KRESS: Refresh my memory. How do you keep 16 that water from freezing in the wintertime?
17 MR. SCHULZ: There are heaters provided. In the 18 elevated tank, we have a heater that's in the circulation 19 circuit, and it maintains the tank at least 40 degrees F.
20 DR. KRESS: This is always in circulation?
21 MR. SCHULZ: If need be, yes. It doesn't 22 necessarily have to circulate all the time in warm weather, 23 but in cold weather, it will probably circulate 24 continuously.
25 DR. KRESS: In an accident, that takes the system
() ANN RILEY & ASSOCIATES, LTD.
Court Reporters 1250 I Street, N.W., Suite 300 Washington, D.C. 20005 (202) 842-0034
l 15 1 from a passive system to an active system? l (3
t' xY 2 MR. SCHULZ: That system doesn't have to function 3 during.an accident, so it's a feature that is required to 4 maintain a plant prior to the accident.
5 DR. KRESS: Prior to.
6 MR. SCHULZ: This tank is over a half million 7 gallons worth of water. There is a lot of heat capacity in 8 it. Also, during an accident, heat's being released from 9 the containment that's right under that tank. It tends to 10 minimize heat losses and minimize the cool down. The water l 11 that's in these pipes up to the isolation valves is in a i 12 room that's right under that tank. That room is also heated 13 to prevent that from freezing.
14 Down stream of these valves, the pipes are kept
/~N ks- ) 15 dry because of the isolation valves. There is also leakage 16 collection detection equipment that's located inside the 17 heated room, just in case there might be some leakage past 18 the valves, so that water cannot get into the pipes and I 19 possibly plug them.
20 There are two drains off the bottom of the )
l 21 annulus. These are redundant. Each one can take a maximum i 1
22 flow rate we can get from the gravity injection, and they 23 are designed so that even with no water evaporation, that 24 they can drain off the water and prevent flooding of the l 25 outside and possibly blocking of the air passage.
I
l 4 DR. KRESS: Cut down on algae growth.
5 MR. SCHULZ: That kind of thing. That's all. We 6 don't do anything to help PCS performance. It's basically 7 just for the algae.
i 8 DR. KRESS: Keep it clean, i 9 MR. SCHULZ: Keep it clean, right.
10 DR. CARROLL: Algae does grow.
11 DR. KRESS: I used to manage a swimming pool. I 12 know.
13 DR. CARROLL: Hydrogen peroxide. Where do those 14 drains go at the bottom?
15 MR. SCHULZ: They just go out into the yard. They 16 are above grade. There shouldn't be a problem with plugging 17 or backing up the drains.
18 DR. CARROLL: It is possible for a person to go 19 down those cooling passages and make inspections and stuff?
20 MR. SCHULZ: Yes. We have -- this is not in your 21 handout package, but it's in the SSAR. It shows a bit more 22 of a schematic -- not schematic, a real section of the 23 containment.
24 Between the concrete shield building and the steel 25 containment shell, is about three feet. There is about two (n)
ANN RILEY & ASSOCIATES, LTD.
Court Reporters 1250 I Street, N.W., Suite 300 Washington, D.C. 20005 (202) 842-0034
17 1 feet on the ot tside, and that's where access would be
() 2 gained. This actually depicts a personnel lift, so they can 3 be lowered up and down and control that. The baffle that 4 separates the concrete from the steel that provides about a 5 foot of clearance on the containment side to sort of 6 optimize the cooling performance, that baffle goes all the 7 way around the containment. The panels can be removed there l 8 to actually get access and look at the containment shell to 9 see what's going on on it, to see that it is in good 10 condition.
i l 11 DR. KRESS: How is that baffle supported? l 12 MR. SCHULZ: Supported from the containment.
13 There are little round pegs that are welded to the 14 containment and they come out about a foot and then there's l 15 a little frame that supports each panel. The panels are 16 bolted to that frame, so they can be removed and put back 17 on. That foot distance is basically constant. The 18 containment heats up and grows a little bit. It moves with 19 the containment.
i 20 CHAIRMAN BARTON: With the problems being l 21 experienced with containment coating adherence, et cetera, 22 and you classified the external coating system on this 23 containment as safety related, is it possible to do a 100 24 percent examination of the external surfaces of the 25 containment?
() ANN RILEY & ASSOCIATES, LTD.
Court Reporters ,
1250 I Street, N.W., Suite 300 Washington, D.C. 20005 (202) 842-0034 L
18 1 MR. SCHULZ: Where the passive containment cooling
() 2 system functions, yes. Basically, your interest in what is 3 basically above the operating deck, because the air flow 4 comes down and turns at about the operating deck and then 5 goes back up. All that is open to inspection 100 percent.
6 Now, you get into other issues as you get lower 7 than that. There's an area where's there penetrations.
8 That still is accessible, and eventually you get down into 9 the buried concrete part, which is not accessible.
10 Where the PCS water flows, it is 100 percent 11 inspectable. We also have a functional test that we will 12 not only do in the initial plant start up but will repeat 13 periodically during the life of the plant. You turn the 14 water on and measure the distribution, that you get enough 15 coverage to be consistent with the safety analysis. We are 16 able to run that kind of a functional test. It's not a heat 17 removal test, but it does show the weirs are working, that 18 coating isn't degraded somehow, that you don't get the film 19 formation and that kind of thing.
20 CHAIRMAN BARTON: Thank you.
21 DR. CARROLL: What sort of corrosion protection if 22 any do you have in the lower head of the containment vessel?
23 MR. McINTYRE: Terry, I think Richard is going to 24 talk about that during chapter three tomorrow.
25 MR. SCHULZ: Richard should be able to talk about
( ANN RILEY & ASSOCIATES, LTD.
Court Reporters 1250 I Street, N.W., Suite 300 Washington, D.C. 20005 (202) 842-0034
19 ;
1 that in terms of the thickness of the material, the coatings
( 2 and that whole situation.
3 DR. CARROLL: You don't have cathodic protection?
4 MR. SCHULZ: I don't know.
~5 DR., CARROLL; Okay. We will talk to Richard.
6 MR. SCHULZ: Yes. . When the system is required to 7 operate in an accident, containment pressure is the primary 8 signal used to actuate the water flow. The system sketch
'9 actually shows..the valving arrangement, so there are two 10 normally closed fail open air operated valves that would be 11; 'open by signals that initiates the water flow. The initial 12 water flow is relatively high, more than 400 gallons per 13 minute. That flow rate does two things. It fairly quickly i
14 flows out over the containment and gets the water over the (m) 15 . whole surface, and it takes a little bit of time to do that.
16 In addition, that water flow rate.is much more .
I 17 than is needed just to remove decayed heat, so it is capable 18 of reducing the. containment pressure, so within about three 19 hours, the pressure is substantially reduced. I don't think 20 it's quite half design yet. That's not achieved until more 21 like 24 hours2.777778e-4 days <br />0.00667 hours <br />3.968254e-5 weeks <br />9.132e-6 months <br /> where about half design pressure in 24 hours2.777778e-4 days <br />0.00667 hours <br />3.968254e-5 weeks <br />9.132e-6 months <br />.
22 DR. CARROLL: Contrast this to what happens with 23 non-safety related core spray.
24 MR. SCHULZ: Containment spray?
25 DR. CARROLL: I mean containment spray.
ANN RILEY & ASSOCIATES, LTD.
Court Reporters 1250 I Street, N.W., Suite 300 Washington, D.C. 20005 (202) 842-0034
20 1 MR. SCHULZ: In a typical PWR, we don't have
( 2 containment spray in AP-600 that is used in design basis 3 accidents. ItLis strictly severe accident capability.
4 DR. CARROLL: But you know how it functions.
5 MR. SCHULZ: You are asking the question for 6 AP-600, if we turned on the spray -- our spray system, what 7 would it do?
8 DR. CARROLL: And did not have the passive 9 containment cooling system.
10 MR. SCHULZ: Did not have. Did not have the 1
11 water? Our air flow is always open. As soon as you start l 12 heating the containment shell, you will get some heat 13 removal from it.
14 DR. KRESS: It's a function of your spray design.
\
15 MR. McINTYRE: I don't think we really know the 16 answer.
17 DR. KRESS: If you had a standard spray, it would 18 overwhelm the passive system and you wouldn't have to worry 19 about it, and it would come down, pressure within_like 20 minutes. I don't know what the spray system they have in 21 there would do because it has bigger drops and doesn't have 22 the pressure and the coverage.
23 MR. SCHULZ: It's probably more like one-half to 24 one-third the capacity of a normal safety related spray 25 system.
I
\
ANN RILEY & ASSOCIATES, LTD.
Court Reporters !
1250 I Street, N.W., Suite 300 i
Washington, D.C. 20005 (202) 842-0034 l
21 1 DR. KRESS: Are those valves counted as part of
() 2 the safety related equipment?
3 MR. SCHULZ: Absolutely, yes. All four of these 4 valves, the motor operated valves are normally open and tech 5 specs requires them to be open. They do get confirmatory 6 open signals. They are safety related in case they were 7 closed. They provide -- the reason they are motor valves is 8 so that if one of these air valves fails during normal 9 operation, a sinenoid burns out, an operator fails to open, 10 we can turn the water off to prevent it draining.
11 That still preserves the safety functioning of u..e 12 system.
13 There is a bucket that the water drains into 14 that's in the top of the containment. It's actually not 15 supported by the containment. It is supported by the shield 16 building. There are slots around that. That gives the 17 initial distribution from the center of the containment.
18 Then there are two sets of collection, 19 redistribution weirs that are arranged around the 20 containment, one of them on the head, actually, both of them 21 on the head, one out further, and this is used to ensure 22 minimum coverage, required minimum coverage. That's more 23 like 90 percent, to get these initial high flow rates and 24 that coverage is worse, is less at the lower flow rates 25 ultimately, in the range of 30, 35 percent at the low long ANN RILEY & ASSOCIATES, LTD.
Court Reporters 1250 I Street, N.W., Suite 300 Washington, D.C. 20005 (202) 842-0034
22
'l term flow rates.
2 DR. KRESS: You adjust the flow after what, 24 3 hours?
4' MR. SCHULZ: The flow adjusts automatically and 5 that's controlled by a series of stand pipes that are 6 located in the tank. For the first three hours, you get a 7 high flow rate, all of the stand pipes are passing water and 8 that flow rate is 440 gallons per minute. At about three 9 hours, the top stand pipe gets. passed by, so the flow drops 10 then to about 120 gallons per minute, and at that point in 11 . time, again, the pressure is substantially reduced.
12 After 30 hours3.472222e-4 days <br />0.00833 hours <br />4.960317e-5 weeks <br />1.1415e-5 months <br />, you are running on the other stand 13 pipes and at 30 hours3.472222e-4 days <br />0.00833 hours <br />4.960317e-5 weeks <br />1.1415e-5 months <br />, the second stand pipe gets passed by 14- and the flow drops to more like 72 gpm and at 72 hours8.333333e-4 days <br />0.02 hours <br />1.190476e-4 weeks <br />2.7396e-5 months <br />, the
'[
15 water flow is just above the bottom stand pipe.
16- DR. KRESS: Are those just open pipes?
17 MR. SCHULZ: It's more than that. There are 18 filters on them. There is anti-vortex formation necks on 19 them. They are engineered so that they don't tend to 20 collect any crude that might be in the water, that they 21 don't suck in air when the water level is a foot or two 22 above them, and the screens keep debris out of them.
l 23 DR. CATTON: The line connecting to the PCCWST, it 24 looks like a siphon.
I 25 MR. SCHULZ: Yes, except there's a siphon break at I ANN RILEY & ASSOCIATES, LTD.
l l Court Reporters 1250 I Street, N.W., Suite 300 j Washington, D.C. 20005 (202) 842-0034 E __ ____ _ _____ ---__- _ _ - _ - _ - - _ _ - - - - - I
23 l
1 the top.
() 2 DR. CATTON: What is its purpose, to fill the 3 pool?
4 MR. SCHULZ: The purpose is so that you can 5 circulate normally by taking water near the bottom of the 6 tank and bringing it down and circulating back through up to 7 the top. i 8 DR. CATTON: It's a water clean up system.
9 MR. SCHULZ: It's part of the water clean up 10 system. That's why the kind of strange looking arrangement. '
11 Because of the way this tank is constructed, we cannot have 12 any connections to the side or the top, because that's 13 really outside the plant, so all the penetrations in the I 14 tank are through the bottom of the tank.
15 If you want to have an overflow from the tank, it 16 is actually not shown on this sketch, but in the PSID in the 17 back of the package, the overflows come up through the 18 bottom of the tank and they go up above the normal water 19 level, so the water level increases and then it starts 20 pouring down and prevents overfilling of the tank.
21 This stand pipe over here is provided primarily 22 for the fire protection. That provides the seismic fire 23 protection capability. It goes up to the minimum water 24 level required for PCS operations. The water level is 25 normally maintained a foot or six inches above that stand
() ANN RILEY & ASSOCIATES, LTD.
Court Reporters 1250 I Street, N.W., Suite 300 Washington, D.C. 20005 (202) 842-0034
24 1 pipe and that provides the 10,000 gallons-that is required
/-
(N) 2 for the seismic fire protection capabilityi but the stand 3 ' pipe prevents the fire: system'from pooling off more than the 4 minimum that's required or assumed in the PCS analysis.
.5 DR. CARROLL: Why wouldn't you have put in 6 capability to pump'this drained water back up into the tank?
7 MR. SCHULZ: The idea all along has been that we
- 8. would design the water flow rates and the distribution so 9 that we don't have any water that collects in the bottom.
10 To really have a large steam line break, we think we will 11 evaporate essentially all the water.
12 If you look at very conservative analysis in terms 13 of heat transfer rates, we don't necessarily-evaporate all 14 the water. We also can have the vents that are less than 15 double ended ones, so the fact that we don't really have a 16 finely adjustable flow rate, we can have more water than we 17 need, so we need the capability of removing that water 18 without providing a problem, but if we have that less severe 19 event, we also don't need as much water.
20 CHAIRMAN BARTON: This water eventually ends up in 21 a pit someplace on the site?
22 MR. SCHULZ: Yes, it goes into storm drains.
23 After it leaves the end of this pipe, it will go into a L 24 normal water collection drainage system on the site, and 25 again, this is above grade, so it can't back up into the l l
ANN RILEY & ASSOCIATES, LTD.
Court Reporters 1250 I Street, N.W., Suite 300 Washington, D.C. 20005 (202) 842-0034 ,
1 i
_ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ - - - - _ _ - - - - - - - - - - - - -- - - ~ ~
25 1 system and prevent these drains from working.
2 CHAIRMAN BARTON: The level of your storage tank 3 is indicator controlled?
4 MR. SCHULZ: Absolutely. That's both to satisfy 5 tech' spec requirements, to show that we have the minimum 6 required water prior to the accident. It also is a 7 protection of a monitoring function so that during the 8 accident as the water is draining down, the operators are 9 kept apprised of that and they can 60 things to increase the 10 water level if they need to.
11 CHAIRMAN BARTON: Any way the instrument tubing on 12 that tank can freeze?
13 MR. SCHULZ: The filled system part of that is
'N 14 maintained with fluid that won't freeze. I think if we 1
15 looked at the actual PSID, there's a question both in the --
16 there's flow rate monitoring on each of the stand pipes so 17 the operators will know that each of the stand pipes is 18 working. They will also know when one gets uncovered and it 19 stops, as it is supposed to, and these systems are set up so 20 that there is a filled system here that is filled with fluid L 21 that won't freeze and the filled portion is kept very close 22 to the main process pipe which has flowing water through it 23 and is insulated, to try to prevent it from freezing.
24 DR. CARROLL: This is like Lycall or something 25 like that?
l O)
( ANN RILEY & ASSOCIATES, LTD.
Court Reporters 1250 I Street, N.W., Suite 300 Washington, D.C. 20005 (202) 842-0034
26 1 MR. SCHULZ: Yes.
() 2 DR. MILLER: All the instrumentation there is 3 safety related?
4 MR. SCHULZ: Not all of it, no.
I 5 DR. MILLER: It's not clear to me in reading the 6 SSAR which is and which isn't.
7 MR. SCHULZ: There is instrumentation, for 8 example, that's involved with leak detection, which is only 9 an issue prior to actuation. It's redundant but not safety 10 related. The level instrumentation is safety related and 11 that's actually --
12 DR. MILLER: I notice on one of them, that the 13 level instrumentation shares common taps that you show 14 through failure modes analysis, that doesn't matter?
15 MR. SCHULZ: Is that the core make up tank level 16 or this tank? I know we had a lot of discussion with the 17 staff on the core make up tank level instrumentation and
\
18 sharing of taps. j E19 DR. MILLER: Yes, it's on the core make up tank 20 level, yes. !
l 21 MR. SCHULZ: We can talk about that. l 22 DR. MILLER: That is now I assume a closed item? I l
23- It was an open item for a while.
24 MR. SCHULZ: Yes, we have reached agreement on the i
25 acceptability of that design and we did some things to
() ANN RILEY & ASSOCIATES, LTD.
Court Reporters 1250 I Street, N.W., Suite 300 Washington, D.C. 20005 (202) 842-0034
27 1
1 minimize the chance of an acceptable situation or to prevent '
() 2 unacceptable situations in the performance of that.
3 DR. KRESS: Why is there a connection to --
4 MR. SCHULZ: Two reasons. The original AP-600 5 design had a separate temporary type off site connection for 6 water make up post 72 hours8.333333e-4 days <br />0.02 hours <br />1.190476e-4 weeks <br />2.7396e-5 months <br /> and the passive containment ;
7 cooling had its own connections, so there were two 8 connections. They are actually physically separate on the 9 plant site because of the way piping was arranged.
10 When we got into the debate about 72 hours8.333333e-4 days <br />0.02 hours <br />1.190476e-4 weeks <br />2.7396e-5 months <br /> and on 11 site equipment and it was necessary for us to provide the 12 capability of going from three days to seven days with on 13 site equipment, it was easier for us to have one ancillary
,s 14 water storage tank and one set of pumps than two.
t 15 We provided that connected to the PCS and it put 16 water up into the tank and then from there, we could drain 17 back down to the spent fuel pit.
18 The other thing that did for us, thr- connection, 19 is if you have a re-fueling operation and you have say a 20 full core offload and new you have a large heat load in the 21 ' spent fuel pit but you have no heat load in the containment, 22 so you have this half a million gallons of water sitting on 23 top of the containment, which is a wonderful heat sink, 24- water make up supply, that we couldn't put into the spent 25 fuel pit.
A G ANN RILEY & ASSOCIATES, LTD Court Reporters 1250 I Street, N.W., Suite 300 Washington, D.C. 20005 (202) 842-0034
28
,-s 1 We provided this connection also for re-fueling (ji 2 operations, core offloads, where we dot need the water in 3 the containment and we have a tech spec on covering what the 4 heat load criteria is on when you can switch from PCS water 5 containment cooling to PCS water spent fuel pit. You can do 6 one or the other, not both, in the early time frame.
7 When the heat load in the containment is high, you 8 have to dedicate the PCS water to containment cooling, but 9 that also means you don't have a full core offload.
10 DR. KRESS: Does that mean that water might not be 11 available during shutdown and low power operation?
12 MR. SCHULZ: If the reactor is shut down a long 13 time or has some fuel transfer to spent fuel pit, I don't
_s 14 remember what the exact criteria is but if you have a normal 15 re-fueling outage -- you about reach this criteria where you 16 can put the PCS water into the spent fuel pit, so --
17 DR. KRESS: You do that by closing off the valves?
18 MR. SCHULZ: The way that is done is strictly 19 manual and it's done intentionally to be difficult to do 20 because this is a potential way of draining your PCS water 21 somewhere where we might not want it to go. We have in the 22 PSID, you actually have a valve here that's up in the room 23 right under the PCS water storage tank, and you have a 24 normally open leakage monitoring valve in a lower room that 25 is easier to get to, and also additional two isolation ANN RILEY & ASSOCIATES, LTD.
Court Reporters 1250 I Street, N.W., Suite 300 Washington, D.C. 20005 (202) 842-0034
l l
29 l 1 valves down there.
i
() 2 When you want to be able to use.that PCS water 3 make up, you have to go to the upper room and open this 4 valve.up and then r2 align these valves down here and you l 5 could actually then see what the flow rate would be and
(
! 6 adjust that flow rate with manual valves.
l 7 That make up capability is provided but it's 8 intentionally has mulciple valves in it that are normally 9 closed to prevent its inadvertent use.
l 10 DR.' CARROLL: You mentioned. third core offloads, 11 if something happens, since I left the business ten years 12 ago, it seems that the trend was everybody was doing tull 13 core offload because of hung up grid straps and problems
< 14 -like that.
t 15 CHAIRMAN BARTON: Now they are back to fuel 16 shuffle and make shorter outages.
17 MR. SCHULZ: The AP-600 is designed and the SSAR I 18 requirements on the spent fuel cooling system are consistent l 19 with full core offloading. We have the capability of
[- 20 running the plant that way and that's consistent with the l
l .21 plant design basis.
L
'22' The utilities we have talked to that have helped 23 us in this program actually have some different opinions on 24 how-they might run the system. Some of them would say 25 definitely third core offload. Some of them still are ANN RILEY & ASSOCIATES, LTD.
Court Reporters 1250 I Street, N.W., Suite 300 Washington, D.C. 20005 (202) 842-0034
t 30
- 1. thinking maybe full core offload. The plant is designed so
( 2 th'at it can be operated either way.
- 3. DR. CARROLL: A hung up grid strap can ruin your 4 day.
f l
5 MR. SCHULZ: Yes, certainly. Let me talk a little L 6 bit'about the potential of forming a vacuum in the 71 containment. I think that was one of the questions that was i
8 raised. We have done some analysis and there's actually a l 9 pressure versus time trace in the SSAR that shows the.
L 10 results of the most limiting condition.
! 11 We did try to think about all the different kind l 12 of things that could go wrong that possibly could lead to l 13 forming a vacuum, that included inadvertent-PCS water I
l 14 draining on the outside, pumping, the IRWST, out of the l f_)
l l
\-./ 15 containment as fast as you could possibly do, excessive fan 16 cooler kind of operacion, the HVAC system that is connected 17 to containment, if it's not connected properly possibly can 18 pull air out of the containment and tend to reduce the l- 19 pressure. Loss of all AC power when the plant is operating, 20 which would tend to reduce the heat input to the containment 21 and then allow the air cooling on the outside to cool the 22 containment down.
23 One thing that we did not consider, we thought 24 about, but we did not analyze, is inadvertent operation of I 25 containment spray, because of the design of that system, it ANN RILEY & ASSOCIATES, LTD.
Court Reporters l'
l 1250 I Street, N.W., Suite 300 Washington, D.C. 20005 (202) 842-0034
31 !
1 takes multiple mis-operation of valves to get that system on
'g3
() 2 line. That's not considered to be credible in design basis l
I 3 space.
4 DR. CARROLL: How about in Murphy Law space?
5 MR. SCHULZ: Well, in Murphy's Law space, you 6 would have presumably the chance that the operators would 7 see what's going on and do something about it. In this case, we looked at the analysis considering very extreme 8
9 conditions in a design basis sense. The worse limiting case 10 that we have found is the loss of AC power. We look at l
11 extreme environmental conditions in terms of the outside l 12 temperature, -40 degrees F., high winds going on. That 13 tends to maximize the cooling on the outside.
14 Prior to the accident, the internal conditions
/,_ s
'\s l 15 were relatively hot, assumed to be 120 degrees, which is 16 like the maximum internal temperature, and 100 percent 17 humidity. Both of those conditions maximize the potential 18 for a result in cool down and condensati of water to
- 19 reduce the containment pressure.
20 One interesting thing is the PCS water flow is not 21 assumed to operate and the reason for that is again, that
- 22 tank water is heated to 40 degrees F. If you turn it on, it !
l 23 actually mitigates the event. It was not assumed to l 24 operate.
25 The minimum vacuum or negative pressure that we I
(
! ) ANN RILEY & ASSOCIATES, LTD.
Court Reporters
(
1250 I Street, N.W., Suite 300 Washington, D.C. 20005 (202) 842-0034 L_____________ _ _ . _ . _ . _ _ _ _ .
32 1 calculated was about two PSID after about one hour. The f' s
() 2 pressure was still going down slowly at that point in time.
l 3 However, the event was assumed to be terminated by operator 4 action and he's got several things he can do at that point 5 in time in terms of opening HVAC vents from the containment, l
6 heating the containment up. He can even turn the PCS on at 7 that point.
8 DR. CARROLL: The normal protection against this 9 would be the purge system?
10 MR. SCHULZ: Yes, in terms of what the operator 11 can do to prevent this, The purge system is not normally 12 open. He can open it from the control room.
13 DR. CARROLL: It's not designed to automatically 14 open on low pressure?
s 15 MR. SCHULZ: That is also true. It does not 16 automatically open.
17 I'd like to move onto the hydrogen control system.
18 It provides three separate functions. It limits the build 19 up of hydrogen in a design basis accident. It limits the 1
20 build up of hydrogen in a severe accident. It monitors the !
21 hydrogen concentration in both of those events.
22 These three things are performed by three separate 23 subsystems, which I'll talk about separately here.
24 The hydrogen recombiner subsystem addresses the 25 DBA limitation. The hydrogen build up in this situation is
/*%
( ) ANN RILEY & ASSOCIATES, LTD.
Court Reporters 1250 I Street, N.W., Suite 300 Washington, D.C. 20005 (202) 842-0034
33 1 characterized by a slower build up. You have not
() 2 significantly damaged the core, maybe a large break LOCA 3 would be the limiting event, so you get a little bit of 4 hydrogen initially and then_a slow build up in the long 5 term, taking three or four weeks to reach four percent 6 hydrogen, if no protection was provided.
7 DR. CARROLL: This_ slow build up is the result of 8 radialysis?
9' MR. SCHULZ: Yes. You get a little spike as a 10 result of assumed, a minor LOCA clad interaction, but again 11 --
12 DR. CARROLL: Plus hydrogen in the coolant.
13 MR. SCHULZ: Plus hydrogen in the coolant. There 14 is also some chemical reactions, zinc, in the containment,.
15 where you get some hydrogen from also. This is the kind of 16 hydrogen build up you are talking about, the protection is 17 provided by PAR's. These are passive devices. There is no l
18 power required, no actuation required. We provide actually '
19 redundant PAR's, even though they are passive, in the 20 containment to deal with the bulk of the containment, which 21 is well mixed.
I 22 There are a couple of areas that may not be.so 23 well mixed and we provide partial sized PAR's to deal with 24 them. One is the IRWST and the other one is CVS equipment I L 25 room, to prevent --
I ANN RIL3Y & ASSOCIATES, LTD.
Court Reporters l 1250 I Street, N.W., Suite 300 l Washington, D.C. 20005 (202) 842-0034
34 1 DR. CATTON: Partial, that's half?
() 2 MR. SCHULZ: More a quarter. I think we have 3 actually. sized the CVS at a quarter. I don't know if we 4 have actually sized IRWST.
5 DR. CARROLL: These are the German design?
6 MR. SCHULZ': Yes, I think they have actually 7 tested. I don't know if they have actually developed the 8 material, but I know they did testing on it.
9 DR. CARROLL: We heard a presentation a number of 10 years ago on it. At that time, there was still some 11 questions about poisoning of the catalyst and things like 12 that.
13 MR. SCHULZ: We have had a lot of discussion about 14 that. A lot of the concern is really in the area when you
) 15 have a severe accident and we are not relying on these 16 working in a severe accident. We rely on the ignitors.
17 There still are some issues in design basis space and we 18 think those have been addressed by testing. T,at testing 19 will have to be basically repeated under our QA requirements 20 because it wasn't really done in that respect.
21 DR. POWERS: The testing that I have seen has been
- :22 interesting tests and what not, but I have not seen things 23 .that you usually see in connection with catalysts like i 24 looking for evidence of coking, surface occlusions and 25 ' things like that. It's been basically kind of interval ANN RILEY & ASSOCIATES, LTD.
Court Reporters 1250 I Street, N.W., Suite 300 Washington, D.C. 20005 (202) 842-0034 l j
35 1 testing. Is there more that I should be aware of on this q
Q 2 testing?
3 MR. SCHULZ: I can't answer that question.
4 DR. POWERS: Is coking a concern on these?
5 MR. SNODDERLY: Excuse me, Dr. Powers. My name is 6 Mike Snodderly, Containment Systems and Severe Accident
'7 Branch. We were responsible for providing the evaluation of 8 the hydrogen recombiner subsystem.
9 I would probably refer you to some of the EPRI 10 poison test reports and also there's some work that's been 11 done by the French, exposing it to -- that was more for ,
12 fission product poisons, but there has been testing done for 13 coke, for example. They developed a very heavy black soot i 14 by burning oil fires and different types of material 15 typically found in containment to develop a coke or soot, 16 and then they saw that it affected performance anywhere from 17 10 to 20 percent, 20 on the maximum and in most cases, it 18 was a 10 percent reduction in efficiency.
19 DR. CARROLL: Do you have test modules in the 20 PAR's that you can take out periodically?
21 MR. SNODDERLY: I think one thing to keep in mind 22 is the way the SSAR is written and the SER is written is it 23 talks about a type of a device that would be needed, and 24 there are two devices that the staff is aware of that could 25 meet that requirement currently today, so it's more of a ANN RILEY & ASSOCIATES, LTD.
Court Reporters 1250 I Street, N.W., Suite 300 Washington, D.C. 20005 (202) 842-0034
36 1 proof of principle, that you would need such a device, and
(
) 2 we are aware of those devices.
3 To answer your question, the two devices that we 4 are aware of are both made in Germany. They both contain 5 the catalyst within slotted cartridges, and the way the 6 surveillance's are now designed, you would remove a certain 7 percentEge of those cartridges, test those cartridges to a 8 known hydrogen concentration and verify a certain 9 temperature inside the catalyst.
10 Of course, if you saw failure or degradation, you 11 would increase the population size, if you didn't see a 12 problem. That's the way the surveillance is currently 13 designed.
14 DR. SEALE: My recollection is that the operation
's.- 15 of these recombiners gives you a rather interesting 16 environmental problem, namely a warm to hot plume exiting 17 the recombiner. That might be a mixing device that we might
! 18 think of inside containment.
19 DR. FONTANA: I think the tests have shown that 20 mixes above the recombiner mostly. j l
21 DR. SEALE: I understand that but my real concern l i
1 22 is does this pose any problem in terms of location, to 23 ensure the protection of the equipment within the 24 containment.
25 MR. SNODDERLY: I'll let Westinghouse address one I
I i
\
ANN RILEY & ASSOCIATES, LTD.
Court Reporters 1250 I Street, N.W., Suite 300 Washington, D.C. 20005 (202) 842-0034
37 1 aspect of that question, but I think I can address a certain O)
(, 2 aspect, and that would be that the testing that we have 3 done, the staff, and that Westinghouse has referenced, shows 4 that we would expect the PAR to start working somewhere less 5 than one percent hydrogen and keep the concentration well 6 below two percent.
7 Once you get up around three percent, four 8 percent, that's when you start seeing some really high 9 temperature plumes. Therefore, I think your concern is 10 addressed, but if Mr. Schulz could, maybe he could talk 11 about the location of equipment relative to the PAR.
12 I believe there's not a lot of equipment. The 13 exit plume would go mostly into the upper containment. I 14 don't think there is anything that would be influenced, but
(~)
(_ / 15 I'll let Mr. Schulz answer that. )
16 MR. SCHULZ: The two mai.1 PAR's are located above 17 the operating deck and there is nothing right above them 18 that is safety related at all. In fact, the IRWST vent is 19 also at the operating deck. It doesn't prevent a threat to 20 anything, any instrumentation or valves. All the valves are 21 pretty much below the operating deck.
22 DR. CARROLL: If we weren't playing this silly 23 design basis accident game, would we need this system?
24 Would the ignitors take care of all of our problems?
25 MR. SCHULZ: The ignitors are not nearly so t.
' ANN RILEY & ASSOCIATES, LTD.
Court Reporters 1250 I Street, N.W., Suite 300 Washington, D.C. 20005 (202) 842-0034
38 1 effective of keeping the concentration low, so the PAR's do
/
(3j 2 provide a nice capability that the ignitors don't do as good 3 a job at.
4 MR. SNODDERLY: If I could also add, I think if 5 you didn't have the requirements of 50.44 and only 50.34 (f),
6 yes, you can meet 50.34 (f) with just the ignitor system. I 7 guess I would refer you to some work that was done with the 8 CNRA and they came up with I guess -- there is some talk 9 where certain European designers wanted to use exclusively 10 recombiners and I think the conclusion of that work was that 11 the optimal system is a combination of recombiners and 12 ignitors. They really compliment one another very well, 13 where the ignitors are weak, the recombiners really 14 compliment them.
15 DR. CARROLL: Thank you.
16 DR. KRESS: How much hydrogen are we talking about 17 in DBA space?
18 MR. SCHULZ: One figure of measure there is the 19 time it takes to get to like four percent, which is three or i
20 four weeks, without any removal-of hydrogen, as compared to !
21 a severe accident where you have extensive damage of the i 22 core, you can get into the realm of -- I think our 23 calculations are without removal and good mixing, about 13 '
24 percent, and that happens in a handful of hours or 25 something, depending on your particular sequence. l l
() ANN RILEY & ASSOCIATES, LTD.
Court Reporters 1250 I Street, N.W., Suite 300 Washington, D.C. 20005 (202) 842-0034
39
{
1 DR. KRESS: The major source of that hydrogen is I
() 2. radialysis or is it water metal --
3 MR. SCHULZ: Radialysis.
4 CHAIRMAN BARTON: No , not in a severe' accident.
5 DR. KRESS: DBA.
6 MR. SCHULZ: Radialysis.
7 MR. SNODDERLY: Excuse me, Dr. Kress, in chapter 8 six of the SAR, there is an excellent figure that shows the 9 hydrogen, the anticipated hydrogen production versus time l
)
10 for design basis accident and then what the profile would be 11 if you had one PAR working and if you had both PAR's 12 working. I think that would answer your question, and to !
13 also add that the major contributor, I think you will see 14 from that figure, is radialysis, because a hydrogen 15 concentration goes up a little bit at the very early stage l 16 due to metal water reaction and then you see it slowly rise i 17 and that's radialysis.
18 DR. POWERS: As long as we are still on the PAR's, ,
19 I'd like to come back to the coking. Mike, you indicated i
20 that they had looked at soot formation and what not. Was 21 that in the French work?
22 MR. SNODDERLY: Sorry, that's the German -- I'm 23 sorry. The manufacturer's name has slipped from my mind 24 right now. It's testing done by the manufacturer, NIS.
25 That work was done I'd say in the early 1990 time ANN RILEY & ASSOCIATES, LTD.
i -()
Court Reporters 1250 I Street, N.W., Suite 300 Washington, D.C. 20005 (202) 842-0034
l l
l 40 1 frame, and we can get you a reference if you'd like, of what
() 2 documents the tests. Basically, it was a fairly simple 3 test. They just took a cartridge and then burned certain 4 material. I definitely remember there was cable material, 1
5 oil, welding fumes and some of the others. l 6 DR. CARROLL: How about spray painting?
l 7 MR. SNODDERLY: They didn't spray paint but they 8 did burn paint. They burned materials that had paint i
9 coatings. Like I said, the worse degradation, I can't l 10 remember what caused it, but caused a 20 percent reduction, 11 and in most cases, it was a 10 percent reduction in 12 efficiency, and then if you look at the margin that's 13 provided, you can handle -- first of all, as Westinghouse p
14 stated, they provided two PAR's, redundant, and you can even
- 1
(/ 15 handle up to a 30 to 40 percent reduction in efficiency if 16 you just have one PAR and still keep it below 3.5 percent 17 hydrogen in the containment.
1 18 DR. POWERS: The question I have is if it was done l 19 in Europe, they don't typically have sulfurous containing 20 cabling there, when they burned cable, was it European cable i
21 or American cable?
l 22 MR. SNODDERLY: I know that you had mentioned to 23 us in previous meetings a concern about sulfurous acid, 24 sulfuric acid, that could be seen in containment. There was 25 work done subsequent to that. EPRI was aware of the concern
) ANN RILEY & ASSOCIATES, LTD.
Court Reporters 1250 I Street, N.W., Suite 300 Washington, D.C. 20005 (202) 842-0034
41 1 you raised and they did expose it.
() 2 What they did was they took American cables and 3 they exposed them to a steam environment and fire. They 4 .didn't actually verify that they got sulfurous or sulfuric 5 acid, but they did burn American cables in an environment to 6 try and create sulfuric acid. We think it was created and 7 it showed a reduction, as I said, of less than 20 percent.
8 DR. POWERS: The problem you run into with these 9 testimonials is basically theJe recombiners have pladium or 10 platinum surface materials on them and there have been an 11 awful lot of catalytic chemists look at that material and 12 they find it does get poisoned by sulfur, and you have 13 sulfur in the system.
14- You do a test and you say, well, we didn't see it.
15 You are contradicting a huge body of literature, and you 16 have to.tell me something more about it.
17 DR. CARROLL: Didn't you say they did see a 18 reduction?
19 MR. SNODDERLY: They did see a reduction of less 20 .than 20 percent, but I understand what you are saying. If 21 there has to date -- there is another report that did
-22 address specifically sulfuric acid, but I don't know how 23 that test was conducted, and we can get you that report. To 24 ' address the concern that you brought, what was done was 25 cables were exposed to an environment of steam and fire to O
(jj ANN RILEY & ASSOCIATES, LTD.
Court Reporters 1250 I Street, N.W., Suite 300 Washington, D.C. 20005 (202) 842-0034 l
l 42 1 try and generate sulfuric acid, but it was not measured or
() 2 . determined if it was created, and they looked at how that 3 affected the efficiency of the PAR.
4 I believe there are other tests, and we can get 5 that for you.
6 DR. POWERS: Would we think the decomposition of 7 cabling on exposure to steam and fire would be different 8 than on exposure to radiation heat and steam?
9 MR. SNODDERLY: As you said, it would be difficult 10 to make that statement without performing the tests.
11 DR. POWERS: We know something about pyrolosis of 12 cables. We know there is a synergium between radiation and 13 temperature. There is a big body of information on that.
14 What I don't know is whether the chemical species being
(_e - 15 produced in those environments are different. That is 16 presumably a fire environment could be strongly oxidizing, 17 so you drive things all the way to the sulfurous, whereas a 18 lower temperature radiation heat environment could create 19 organic sulfur compounds that might behave differently. I l
20 honestly don't know.
21 MR. SCHULZ: The other thing you need to keep in 22 mind is we are not dealing with the severe accident.
23 DR. POWERS: No , no. I'm still working in DBA 24 space. In severe accidents, you have a different problem.
25 MR. SCHULZ: In DBA space, the conditions are not
() ANN RILEY & ASSOCIATES, LTD.
Court Reporters 1250 I Street, N.W., Suite 300 Washington, D.C. 20005 (202) 842-0034
1 43 '
1 that different in the longer term from normal operation.
(Oy 2 The radiation levels in the containment are not that i
3 elevated, the pressure and temperature come down. It's not 1 4 really that extreme. It's nothing like the environmental 5 qualification conditions that we put the cable through, 6 which are really more severe accident type conditions.
7 DR. SEALE: Could I ask, you have indicated that 8 the requirement at this point is more in the nature of a 9 black box requirement for a recombiner that has features 1
10 which are suggestive by the two German prototypes that are 11 available. Would it be inappropriate to carry forth Dr.
12 Powers' concerns about sulfuric or sulfur type poisonings of l 13 the catalyst in the SER so that concern can be addressed 14 when we know what the recombiner more clearly looks like?
15 MR. SNODDERLY: Yes, we could, but I'd like to say 16 or suggest two things before we maybe address that 17 possibility. The first would be that there is a report that 18 does address the effect of acids, and I can get that and 19 provide that to Dr. Powers and we can see if that addresses 20 his concern.
21 If that doesn't, I would probably say these 22 recombiners are supplied to meet the requirements of 50.44, 23 to meet the requirements of 50.44, the staff suggests or you 24 can use and Westinghouse has committed to use Reg Guide 1.7, 25 and the environment created by 1.7 I don't believe would t ANN RILEY & ASSOCIATES, LTD.
Court Reporters 1250 I Street, N.W., Suite 300 Washington, D.C. 20005 (202) 842-0034
44 create these-acid environments.
'l'
- .2 DR..SEALE
- I agree the sulfur problem is a rather 3 arcane one, but that doesn't mean it's not a real one.
4 MR. SNODDERLY: We will that report and provide 5 .that to Dr. Powers and-see if we still have a concern. .(
i 6 CHAIRMAN BARTON: Thank you. Moving right along. I 7 MR. SCHULZ: To wrap up the hydrogen control 8 system with this slide and get onto passive core cooling 1
9 system, we do have ignitors in the containment to deal with l 10 severe accident. Of course, the characteristic of hydrogen 11 in a severe accident is a lot more hydrogen and it appears 12 faster.
13 We have 64 ignitors that are located throughout 14 the containment and basically redundant groups. They are 15 distributed to try to have access to points of release from 16 the reactor coolant system. They are powered by off site i
17 power, on site power, which is the main diesels or non-1E 18 batteries. They are actuated manually either through the 19 control system or the diverst ctuation system can also a
20 ~ actuate the ignitors. l l
21 DR. CATTON: With respect to the placement, 22 Siemens has a set of rules for ignitor placement that is I 23 experimentally based. Do you have something similar?
I 24 MR. SCHULZ: Yes. There was a lot of discussion 25 with the staff on placement of ignitors. The SSAR has a i i
() ANN RILEY & ASSOCIATES, LTD.
Court Reporters 1250 I Street, N.W., Suite'300 Washington, D.C. 20005 (202) 842-0034
)
I 45 l 1- bunch of rules that includes looking for points of release
() 2 -but not putting them too close, put them in environments 3 where they can actually burn hydrogen. 1 I
4 There was an actual analysis done of containment 5 conditions. There was also expert panel type opinions.
6 DR. CATTON: This is all in the SSAR?
7 MR. SCHULZ: The criteria is in the SSAR. the 8 locations are in the SSAR.
9 DR. CATTON: Do you do anything special to control 10 the natural circulation so that you can get better l
11 performance from these ignitors? '
12 MR. SCHULZ: The AP-600 containment -- !
13 DR. CATTON: Sloped ce_ lings, no place that stuff 14 can get trapped, that sort of thing?
m/ 15 MR. SCHULZ: We do some of that, like the IRWST 16 has a flat ceiling. That's not actually sloped but it's 17 flat.
18 DR. CATTON: Vent in the top?
19 MR. SCHULZ- Then there are vents on the top of 20 that. There are rooms under the operating deck. The 21 operating deck has some significant openings through it, 22 both around the outside edge near the containment because it 23 doesn't obviously not connect to the containment steel, 24 containment shell, and there are also some other vents that 25 have actually been'placed in there to try to facilitate
() ANN RILEY & ASSOCIATES, LTD.
Court Reporters 1250 I Street, N.W., Suite 300 Washington, D.C. 20005 (202) 842-0034
46 1 circulation, generally down around the outside and up p) q 2 through the loop compartments. That plays a role in the 3 passive containment cooling system performance. It also 4 plays a role here in severe accident hydrogen mixing.
i 5 DR. CATTON: You use the circulation patterns in )
6 deciding where to put the ignitors?
7 MR, SCHULZ: That was a part of the whole picture.
8 DR. CATTON: The Siemens experimental program.also 9 showed you have to.be careful'about doorways, you go through 10 a contraction and what happens on the other side of it. If 11 the air flow is from one door to another, you get different 12 performance.if you put the ignitor here than over there.
13 MR. SCHULZ: I'm not sure I can specifically 14 answer that question about doorways.
k s/ 15 DR. CATTON: Will I get that answer if I look at 16 the SSAR?
17 MR. SCHULZ: I don't know. There is some 18 discussion about ignitor replacement, why things were in 19 like a room, but I don't think there is as much detail as 20 you may be asking for.
21 CHAIRMAN BARTON: Why don't we put that down as a 22 question from Dr. Catton.
23 MR. SNODDERLY: I think what you are going to 24 find, Dr. Catton, is each enclosed volume room was provided 25 with ignitor coverage, dual ignitor coverage. You won't
( () ANN RILEY & ASSOCIATES, LTD.
Court Reporters 1250 I Street, N.W., Suite 300 Washington, D.C. 20005 (202) 842-0034
47 1 find --
(A,) 2 DR. CATTON: Both at the entrance and exit of the 3 room? I 4 MR. SNODDERLY: That is what I was going to get 5 to. There were several criteria that were used and those 6 criteria are listed in a table in chapter six, and I think 7 if you go look at that, that would be helpful to you. It 8 played a major role in how the ignitors were located.
9 You will not find a criteria that states where the 10 ignitors should be placed relative to the doorway. You will 11 find criteria that talk about how the natural circulation 12 patterns were looked at for the area or enclosure or room, 13 so that the hydrogen ignitor is going to be located close to 14 or above the hydrogen source, so theoretically, you are
\-s/ 15 going to burn it before -- the idea was to burn it before it 16 gets to the doorway, so you wouldn't get the build up. !
17 You look at each enclosed volume and you say to 18 yourself where are the hydrogen sources in this room and you 19 locate the ignitor above and as close to as reasonable the 20 hydrogen source. Theoretically, you should be burning that 21 hydrogen before it wants to leave that room.
22 Now, one of the exceptions to that was that the 23 CVCS area, I'll call it, and although there is ignitor 24 coverage, that was one area that our consultant and people 25 that we used at the Office of Research who have had O)
( ANN RILEY & ASSOCIATES, LTD.
Court Reporters 1250 I Street, N.W., Suite 300 Washington, D.C. 20005 (202) 842-0034
i 48 1 experience in placing ignitors saw a potential for flame j e- -
2 l
acceleration, which I think is your concern, where flames l 3 are accelerated through a small opening such as a doorway.
l l 4 .There were some areas that we found susceptible to l
5 that. One of those.was the CVCS room and there, there is a i
l ~6 PAR located. I think that gets into where the two systems '
7 compliment one another and that the PAR should start working 8 much earlier before the ignitor and I think that gave us a 9 great deal of comfort, that that PAR was provided at that 10 location.
11 Other than that, I think if you look at that table i
l 12 in chapter six that discusses the criteria and how the 13 ignitors were located and if you also look -- Terry had the 14 drawing up a little earlier that showed the relative 15 openness of the containment above deck and how 16 compartmentalized the containment is below the 135 foot 17 elevation and the ignitors that were provided in those 18 compartments. I think it gives one a fairly good feeling i
19 that there is substantial coverage with this design and I I 20 would say it is at least equal to that provided for the ice ;
21 condensers and the BWR Mark III's.
22 CHAIRMAN BARTON: Why don't we have Dr. Catton 23 look at that and if there are any questions, we will get 24 back to the staff or Westinghouse with ACRS' questions. j 25 MR. SNODDERLY: I think that would be a good
() ANN RILEY & ASSOCIATES, LTD.
Court Reporters 1250 I Street, N.W., Suite 300 Washington, D.C. 20005 (202) 842-0034
49 1 approach. '
(}j r
2 MR. SCHULZ: The hydrogen monitoring system, we 3 provide monitors that are capable of monitoring the hydrogen 4 in the containment. They write a post-accident monitoring 5 function in a DBA situation.
6 In a severe accident, you also use these monitors 7 to operate the igniters. Sixteen monitors are provided.
8 Three of them are safety-related and powered by 1E supplies, 9 connected to the protection system. Thirteen of them are 10 non-1E and they're kind of subdivided further. Two of them 11 are connected to the diverse actuation system which then 12 allows the operators to use the DAS controls to run the 13 system, and 11 of them are connected to the control system.
14 DR. MILLER: I have a question on the monitors.
O
\_ l 15 One is why did you have three that are 1E and not four, !
16 since the basic approach was a two-out-of-four? And the l 17 other, are the DAS ones going to be different? You 18 specified the performance requirements of the monitors in l 19 there, in the SSAR. Are the ones from DAS going to be 20 different for diversity or not?
21 MR. SCHULZ: When you're talking post-accident 22 monitoring and you're not actuating something, when you talk 23 about two-out-of-four logic, you're talking about reactor 24 trip or ESF actuation.
25 These sensors don't actuate anything. So if you ANN RILEY & ASSOCIATES, LTD.
Court Reporters 1250 I Street, N.W., Suite 300 Washington, D.C. 20005 l (202) 842-0034 I
50 1 look at the PAMS requirements, the most you need of anything
() 2 is three to provide the unambiguous kind of input to the --
3 DR. MILLER: So it meets the post-accident 4 requirements.
5 MR. SCHULZ: Pardon me?
6 DR. MILLER: That. meets the post-accident, Reg.
7 Guide 197.
8 MR. SCHULZ: Yes. DAS is aimed at addressing 9 common mode failure of software and I&C hardware. It does 10 not have inherently any sensor diversity requirements. In 11 this case, they are not required to be different.
12 We have looked at sensor diversity in PRA space 13 and when that was done, we saw no coupling between sensor 14 problems and software I&C hardware problems.
15 DR. MILLER: No inherent common mode type problems 16 on hydrogen monitoring.
17 MR. SCHULZ: They have not been identified, for 18 example, in the PRA as being an important sensor type.
19 Basically, group things, pressure transmitters, high 20 pressure, low pressure, RTD type sensors and in PRA space, ,
i 21 looked at failures of those kinds of things. Special type 22 sensors were kind of grouped separately and tended not to be
'l 23 failing at the same time you would fail these other sensors.
24 And pressure sensors or DP sensors tend to be more 25 challenging from a PRA point of view because they use many,
() ANN RILEY & ASSOCIATES, LTD.
Court Reporters 1250 I Street, N.W., Suite 300 Washington, D.C. 20005 (202) 842-0034
51 1 many different places and different functions and you lose
() 2 more of your instrumentation with that kind of common mode 3 failure than something that would just take hydrogen out.
4 DR. CARROLL: What are the three locations 5 monitored that are 1E?
6 MR. SCHULZ: I don't know the answer to that 7 question. We could find out for you.
8 DR. CARROLL: And similarly, what are the two l 9 non-1E DAS?
10 MR. SCHULZ: I believe those are in the upper part 11 of the containment. DAS is generally assuming that things 12 are reasonably well mixed and you're kind of going for a l 13' limited, beyond design basis kind of situation there. {
14 The passive core cooling system performs a number 15 of very important functions. Emergency.RCS makeup and 16 boration, this is in non-LOCA kinds of sitrations. KP 17 injection, which, of course, is a post-LOCA type 18 requirement; eniergency decay heat removal, both in LOCA and 19 non-LOCA situations; and post-accident PH adjustment of the 20 containment water.
21 There are a number of PRA-related influences on l
l 22 the passive core cooling system.
23 DR. CATTON: Is this called PRHR by others?
24 MR. SCHULZ: No. You're talking about this or 25 this?
ANN RILEY & ASSOCIATES, LTD.
Court Reporters 1250 I Street. N.W., Suite 300 Washington, D.C. 20005 (202) 842-0034
52 1 DR. CATTON: The passive core cooling system.
(~h
\
y ) 2 MR. SCHULZ: The passive core cooling system l 3 includes the passive RHR and the core makeup tanks, IRWST.
4 It's actually a conglomeration of somewhat separate 5 functions.
6 They have a lot of interties to them, so we've put 7 them in one system.
8 The PRA has had a lot of influence on the passive 9 core cooling system design and that occurred at an inert 10 fashion going back into the 1980s, when we first started 11 working on this design, both affecting their redundancy and 12 the reliability diversity of the system.
13 DR. FONTANA: I notice, in looking at the picture 14 here, that the ADS-4 comes off the PRHR line. Does PRHR
(~h sl 15 become ineffective after ADS before it opens up?
16 MR. SCHULZ: It tends -- it may become 17 ineffective. The analysis -- typically, when fourth stage 18 opens, the RCS pressure is relatively low, except for maybe 19 in a DBI line break. But in a case of other events, where 20 the core makeup tanks drain in a more normal injection 21 fashion, by the time they reach the ADS stage four actuation 22 set point, which is relatively low in those tanks, the 23 pressure in the RCS is more like 50 pounds gauge or t
24 something like that, and the accumulators have emptied by !
25 that time.
l l
ANN RILEY & ASSOCIATES, LTD.
Court Reporters 1250 I Street, N.W., Suite 300 Washington, D.C. 20005 (202) 842-0034
53 1 And in the analysis models, the assumption is made
() 2 that when you put that nitrogen into the reactor coolant 3 system, that it may cause the passive RHR to stop working.
4 Now, I don't think we actually observed that at OSU, for 5~ example, and the passive RHR did continue working and did 6 continue removing heat with fourth stage open.
7 DR. FONTANA: It's not really' counting on it. You 8 don't need it.
9- MR. SCHULZ: We don't count on it. That's right. !
10 CHAIRMAN BARTON: Let me ask a question at this 11 point. Tom Kenyon, does the staff have any comments on this 12 chapter? Are'you going to have any comments on the FSER on 13 this chapter?
14 MR. HUFFMAN: This is Bill Huffman, of Projects.
15 We had one open item in this chapter that Tom briefly 16 touched on. I can give you a status at the end.
17 CHkIRMAN BARTON: I'm just trying to figure how 18' much time to give Westinghouse here, since we're really 19 running late. You've got till about ten after.
20 MR. HUFFMAN: It would only take a minute. ,
21- CHAIRMAN BARTON: Okay. You've got till about 12 22 after 10-to complete here.
23 MR. SCHULZ: Okay. Let me move on to talking
.24 about some portions of the system. This. portion of the 25- passive core cooling system includes the two accumulators,
() ANN RILEY & ASSOCIATES, LTD.
Court Reporters 1250 I Street, N.W., Suite 300 Washington, D.C. 20005 (202) 842-0034
54 1 the two core makeup tanks, the IRWST, and the associated
() 2 injection lines from them. The ADS valves which are shown 3 here are actually part.of the reactor coolant system, but I 4 will be talking about them because they work very closely 5 with the operation of the passive core cooling system.
6 DR. KRESS: I know that's a schematic, but are l
7 those relative volumes fairly accurate?
8 MR. SCHULZ: No.
9 DR. KRESS: They're not.
10 MR. SCHULZ: Each core makeup tank is about 2,000 11 cubic foot, the accumulators are the same size. The IRWST
- 12 is half a million gallons. It's quite a bit bigger than 13 everything else. So it's pictorially undersized in this 14 picture.
O l
k_s/ 15 DR. KRESS: The pressurizer and the core volume.
16 MR. SCHULZ: The pressurizer is actually smaller 17 than a core makeup tank. So it's like 1,600 cubic foot and 18 -those are 2,000 cubic feet.
19 DR. KRESS: And the pre-volume in the core is even 20 less than the pressurizer.
21 MR. SCHULZ: The whole reactor coolant system is 22 maybe 5,000-6,000 cubic feet, including lube piping, steam l i I 23 generators. Of course, most of it's in the reactor. The 24 amount of water stored in the core makeup tanks is about 25 that in the reactor coolant system, pounds wise. Now, the j
() ANN RILEY & ASSOCIATES, LTD.
Court Reporters I
1250 I Street, N.W., Suite 300 Washington, D.C. 20005 I (202) 842-0034
L 55 l i volume, since that's cold water, its density is higher. So
() 2 there's actually a little bit less volume in the two core 3 makeup tanks than the reactor coolant system by volume, but 4 by mass, they're almost the same.
5 A couple of points that I wanted to make here. I 6 think there's been some questions on water hammer, how we 7 designed for consideration, and it's been done over quite a 8 long period of time, i
9 Some of it was kind of common sense, in our minds. '
10 For example, both in the core makeup tank and the passive 11 RHR, which have two connections for the reactor coolant 12 system, both sort of an inlet and a return, we very 13 intentionally provided one barrier on the outlet side of the 14 device to -- and that ensures, for example, that the whole 15 system is pressurized.
16 So if this is the isolation, then the reactor 17 keeps this side pressurized and the cold legs keep this side 18 pressurized. So when you open the valves up, there is no 19 possibility for having a large pressure difference or an 20 unfilled system that could cause a water hammer on the 21 startup of the system.
22 The other thing that we do is we make sure that 23 the inlet line is hot. We do that by insulating it well and 24 routing it continuously upward to the high point. We've 25 done analysis that shows that internal circulation in these ANN RILEY & ASSOCIATES, LTD.
Court Reporters 1250 I Street, N.W., Suite 300 Washington, D.C. 20005 (202) 842-0034 c_ _ _ _ _ - _ - _ _ - _ _ _ _ _ _ _ _ _ _ . _ _ _ .
56 1 pipes keep the lines very close to the hot leg or cold leg
() 2 temperature, depending on what they're connected to.
3 Perhaps RHR Is actually connected to the hot leg.
4 And that does a couple of things. One, that 5- provides for effective initial recirculation of startup of 6 the system to have the hot water here and the cold water 7 over here. It also minimizes the potential for water hammer 8 in the case where there's a larger break and the cold leg 9 tends to void quicly while there is still water in that 10 line.
11 So steam trying to chase up that line will not 12 encounter cold water. So that in this line going up to the 13 top, you cannot get a cold water steam condensation effect.
14 We have also more recently done a systematic look-15 at AP600 using criteria developed by Peter Griffith at MIT 16 to look for water hammer situations and have not found any 17 meeting that criteria, and we provided a report to the staff 18 on that.
19 So we think that both in our initial design 20 assessment and more recent checks that we have addressed 21 that.
22 There was also some questions about low 23 differential pressure check valves. In particular, the 24 recirc check valve and the IRWST check valves, and are they 25 going to work in the conditions that they're exposed to, ANN RILEY & ASSOCIATES, LTD.
Court Reporters 1250 I Street, N.W., Suite 300 Washington, D.C. 20005 (202) 842-0034
57
~
1 will they work reliably.
() 2 With the addition of the squib valves that we have 3 with these check valves, we think we have'now put these 4 check valves in conditions that are within operating 5 experience and that are relatively favorable for these 6 valves.
7 For example, the recirc check valve basically sits 8 in a dry pipe. Now, the check or the squib valve provides a 9 leak-tight barrier between the water that's'in the IRWST and 10 the containment atmosphere. So ue don't expect any leakage 11 through that squib valve and this valve t.'ill sit in a dry 12 pipe.
13 We test that check valve by actuhlly going in and 14 you can see notes on the P&ID where we take a cover off of 15 the screen and we can place a test. device, mechanical test 16 device down the pipe and it's a short straight piece of pipe 17 through the check valve. We can actually push the check 18 valve open, make sure it's free, it operates freely, and we 19 do that at every refueling operation.
20 These check valves --
21 DR. CARROLL: These are all biased open?
22 MR. SCHULZ: No. These check valves that I'm 23 talking about here in terms of low DP are conventional 24 simple swing disk check valves.
25 DR. CARROLL: Do they have some sort of ANN RILEY & ASSOCIATES, LTD.
Court Reporters 1250 I Street, N.W., Suite 300 Washington, D.C. 20005 (202) 842-0034
58 1 non-intrusive position indication?
() 2 MR. SCHULZ: Yes, they do.
3 DR. CARROLL: How does that work? Magnetic?
4 MR. SCHULZ: We've talked to vendors about 5 existing technology and magnetic is one. We haven't 6 purchased.the valves and when we do, there may be other 7 things, but it will definitely be a non-intrusive type, 8 nothing connected to the moving parts of the valve.
9 DR. CARROLL: Okay.
10 MR. SCHULZ: The IRWST check valves was probably 11 the area of most concern in LSe original design, where we 12 had four check valves here and they were, at least as a 13 group, holding back RCS pressure from the atmospheric
/ __
14 pressure IRWST. So they were being held closed with high k~s\ l 15 DP, but they had to open with relatively low DP.
16 By putting, again, squib valves in here, we now 17 keep the DP off of those check valves. The DP will be j
)
i 18 across the squib valves. These valves will be emersed in l 19 the IRWST water, but that's cold water. There is no DP. So 20 we think it's a very good condition.
21- In addition to that, every refueling outage, we 22 will come down here and put a flow of air through test 23 connections that are specifically set up for these valves, 24 so that we can monitor their initial cracking open pressure 25 and the flow rate that fully opens those check valves.
ANN RILEY & ASSOCIATES, LTD.
Court Reporters 1250 I Street, N.W., Suite 300 Washington, D.C. 20005 (202) 842-0034
l 59 1 So every refueling outage, we will be testing l
() 2 these valves. Again, these valves are simple stainless 3 steel swing disk check valves. The accumulator check valves
- 4. are very similar pipe check valves to the IRWST check 5 valves, very typical of operating plants, again, simple 6 swing disk check valves.
7 Those valves will be tested every refueling outage 8 to fully open them with a small amount of injection flow 9 from the accumulator, where it will actually have the MOV 10 isolated, adjust the pressure in the accumulator so that 11 it's 40 pounds or so above the RCS, We'll open up the MOV 12 and we'll get enough flow from that to open the check valves 13 and verify that they opened by instrumentation on the check l 14 valve.
l
[\_ 15 But these are not considered to be an issue from a 16 low DP point of view, because you have a lot of pressure in l 17 the accumulator to'open them.
- 18 The biased open valves we have are in the CMT l 19 discharge. Thmt's the only place we have the biased open 20 check valves. They're a tilt disk type design, which is j 21 different from the swing disk valves used in the other check l 2.2 valves. They're waited so that they hang horizontal or 23 nearly horizontal in the pipe normally, and you can tell 24 that also by instrumentation on the valve, non-intrusive 25 instrumentation.
ANN RILEY & ASSOCIATES, LTD.
Court Reporters 1250 I Street, N.W., Suite 300 Washington, D.C. 20005 (202) 842-0034
60 '
1 So when the core makeup tank starts operation, t%
(a) 2 those valves are already open. So they don't have to 3 operate.
4 The accumulator flow, if it gets strong enough or, 5 in particular, if you have a cold leg break, you can get 6 some reverse flow back through the CMT. Now, in a smaller 7 break or inadvertent ADS type operation, when the 8 accumulators come on, it will tend to, because of the 9 pressure drop in this portion of the line at the high flow l
10 of the accumulators, will tend to develop enough pressure 1 11 here that the CMT flow will stop, but you won't get reverse 12 flow.
13 And so in that situation, these biased open check I 14 valves will stay open. So even though the core makeup tanks l 15 start injecting, then they stop, then they start up again 16 after the accumulators empty, these biased open check valves l
17 never close.
18 When they would close potentially is in a larger l
l 19 break LOCA in a cold leg or a break in the balance line, 20 where, when the accumulator comes on, it actually tends to 21 try to bypass the core and go out the break, and that's when 22 those check valves would open or close.
23 There's two of them put in there for failure 24 tolerance, just as there's redundant paths here to tolerate 25 failures.
/^\ ANN RILEY & ASSOCIATES, LTD.
( ,/
Court Reporters 1250 I Street, N.W., Suite 300 Washington, D.C. 20005 (202) 842-0034
61 1 After the accumulators empty, then those check O
( ,f 2 valves are expected to.open without failure consideration.
3 That's why you don't see four valves there.
4 DR. CARROLL: Tell me a little bit about your 5 basis for believing the squib valves are highly reliable.
6 What surveillance is done on them?
l 7 MR. SCHULZ: Squib valves are extremely simple )
8 devices. There's very few moving parts to them. Much 9 simpler than a motor operated valve or even an air operated 10 valve.
11 The technology of manufacturing, controlling the 12 quality of the propellant that creates the gas that causes 13 these things to work basically has its roots in military, 14 bullets, guns, whatever, and that has been developed to an 0
\sl 15 art and a science in terms of predicting the performance of 16 these things.
17 There is not a lot of experience with squib valves 18 of the size we're talking about, but the design is the same.
19 So there's a lot of experience, some nuclear, and a lot 20 non-nuclear, that says that these things are extremely 21 reliable.
22 Another factor in this reliability assessment is 23 when you go to the vendors who make the squib valves and j l
24 talk to them and say we're estimating in our PRA the 25 reliabilities of ten-to-the-minus-three, they kind of laugh l
i f ANN RILEY & ASSOCIATES, LTD. .
Court Reporters i 1250 I Street, N.W., Suite 300 l
Washington, D.C. 20005 l (202) 842-0034 l l
l
62 1 and snicker at you and say is that all you're assuming.
() 2 We guaranteed numbers that are better than that. We 3 designed to quote numbers that are better than that.
4 You go to a check valve vendor or a motor operated 5 valve vendor, they won't quote you a number at all.
6 DR. CARROLL: For good reason.
7 MR. SCHULZ: They don't know the reliability that 8 well. It's also sensitive to our operations and operating 9 conditions and the squib valves tend to be less so.
10 Surveillance is primarily done on a change-out of the 11 propellant and a sort of after-the-fact test to make sure it 12 would have worked, and this is part of the SME code, on the 13 sequence of that and how many you have to do.
14 DR. CARROLL: What's been the industry experience
(-) 15 with that? GE , of course, has been using squib valves for a 16 long time.
17 MR. SCHULZ: GE has used them in some 18 applications. They've had a couple of problems that related 19 to disassembly of valves and I think maybe also improper 20 connection of the circuit.
21 One of the things that we do as a requirement is 22 that anytime the circuit is disconnected and in particular 23 when you change the cartridge, we will check the continuity 24 of the circuit to make sure that it's been put back together l 25 properly.
r\
t ) ANN RILEY & ASSOCIATES, LTD.
Court Reporters 1250 I Street, N.W., Suite 300 Washington, D.C. 20005 (202) 842-0034
63 1 We try to ensure that the valve would not be --
2 if, after it's used, that the reassembly of the valve would 3 be-done without the possibility for leaving pieces in the 4 valve that would prevent its operation.
5 We have position indication devices that basically 6 tell you when the valve is open and if you don't take the 7 parts out after you've used it, it's going to continue to 8 stay open. So we've done some things that try to address 9 some of the problems that have occurred in these valves.
10 DR. CARROLL: What kind of valves are they as far
.11 as where the flow goes through? Are they gate valves?
12 MR. SCHULZ: No. I've actually got -- there's two 13 different types of squib valves we have in terms of what 14 they look like inside. I can get you a' copy of this if 15 you'd like.
l i
16 This is a fourth stage squib. This is the inlet 17 on the RCS pressure side. This is basically a flange joint i
18 and there is a nozzle which has an integral piece that 19 closes that nozzle off. So this is all sealed. You can't I 20 quite tell from this picture what's going on there, but 21 that's an engineered failure point right here.
22 And normally when the valve is sitting there, I 23 that's all metal, solid metal sealed. So there's no 24 possibility of leakage.
25 This piece here that eventually, when the valve
() ANN RILEY & ASSOCIATES, LTD.
Court Reporters 1250 I Street, N.W., Suite 300 Washington, D.C. 20005 (202) 842-0034
64 1 was opened, will flop down there, is captured in a pin here.
( ) 2 The actuator of the valve sits up here and the propellant is 3 stored up hare, and there's a couple igniters. Actually, in 4 ;the fourth stage, we have three igniters. Two of them are 5 safety-related'on different I&C divisions and a third one 6 comes from the diverse actuation system.
7 Any three of those can fire the valve --
8 DR. CARROLL: Any one?
9 MR. SCHULZ: Any one of them. Any one of the 10 three, that's right. Now, we take precautions back in the 11 actuation circuitry to make sure that a couple of things 12' have to happen before the valve gets a signal.
13 So there is parallel redundancy at the valve, but 14 you go back into the I&C system and we've got series 15 redundancy. So that any one of those signals is not subject 16 to a single failure. i 17 When the propellant is ignited, gas pressure is l 18 generated here and that builds up the pressure in this area !
19 and you basically have a hold-back pin that has a shear 20 point. When this pressure gets to be sufficient to break 21 that pin, then it drives the -- the gas pressure drives this 22 piston down and it impacts this closure, which then breaks 23 and the pressure from the RCS flops the disk out of the way.
24 Nothing leaves the valve in terms of parts and 25 pieces. There is a position sensor down here. So that when
()
1 ANN RILEY & ASSOCIATES, LTD.
Court Reporters 1250 I Street, N.W., Suite 300 Washington, D.C. 20005 1 (202) 842-0034 I
65 1 this thing falls'down, the operators can tell that the valve 2 is open.
3 DR. MILLER: Are we going to have further 4 discussion on the squib valve later?
5 CHAIRMAN BARTON: This also comes up in a PRA. So 6 we want to pick it up this afternoon.
7 DR. MILLER: Because I have something to say, a 8 couple questions on that.
9 MR. SCHULZ: We may actually have -- we're 10 supposed to talk about tech specs tomorrow. There is a long 11 time given to that.
12 DR. MILLER: It's in there, too.
I 13 MR. SCHULZ: What I'm just saying is that there 14 may be extra time then. I don't know what you want to do s_s 15 with that extra time.
16 DR. MILLER: I think with the squib valve, more l 17 the problem is in maintenance and installation than it is on 18 individual squib valves, which comes up in the PRA through 19 common cause failures. You're using them every place.
20 MR. SCHULZ: We've got three applications.
21 They're very important, though, absolutely.
22 We've talked about in-service testing, talked 23 about the water hammer, the low DP check valves. There was 24 some discussion about how the heck we came up with 2,000 25 cubic feet for the tank. It seems like a funny round 1
i ANN RILEY & ASSOCIATES, LTD.
Court Reporters 1250 I Street, N.W., Suite 300 Washington, D.C. 20005
- (202) 842-0034 1
l 1
66 1 number. It really goes back to some hand calculations we O
(_,/ 2 did a long time ago with this sort of a simplified sizing 3 criteria that basically said I want the accumulators to do 4 large break LOCA or refill, initial reflood, but we also 5 want the accumulators to remove decay heat for about 2.5 6 minutes.
7 And how did we pick 2.5 minutes? It was kind of 8 an arbitrary decision at the time. The core makeup tanks 9 are able to match decay heat from 2.5 minutes to 20.5 10 minutes.
11 How did we pick 20.5 minutes? Well, we thought by 12 that time we could get enough injection from the IRWST. So 13 that we looked at the IRWST at providing enough injection to 14 take away decay heat and. I guess, in all these cases, a C< 15 little bit of sensible heat in this various time-frames.
16 Now, the sizes have been then verified by all the 17 safety analysis we've done, both design basis and even more 18 cases run in the PRA thermal hydraulic or the PRA success 19 criteria calculations on literally hundreds of different 20 break locations, multiple failure sequences, looking at the 1
21 performance of these systems. ,
l 22 Let me jump to this picture here. This shows the l 23 conditions -- actually, in this case, the vertical scale is 24 actually fairly accurate; so not necessarily horizontal, but 25 the height of the pressurizer relative to the reactor in
() ANN RILEY & ASSOCIATES, LTD.
Court Reporters 1250 I Street, N.W., Suite 300 Washington, D.C. 20005 (202) 842-0034
l 67 1 IRWST is about right.
() 2 You see the plant here in a post-LOCA ADS 3 operation in the recirc mode. So the IRWST has drained down 4 and this is about the 107 level in the containment. You see 5 here a typical passive core cooling system valve room. This !
6 is all dry here. The water in the tank and the loop 7 compartments is sealed off, so those valves do not normally 8 flood when the containment floods.
9 Obviously, if you have a pipe break there, you can 10 flood one of the two rooms, but that's not really depicted 11 here.
l 12 What I wanted to kind of point out is a couple of I 13 things. One of them is the vents and overflows, and, again, 14 these are relatively accurate depictions.
04 15 These little slanted lines are intended to show i 16 louvers, and that's what we use to actually seal -- seal may 17 be to strong of a word. We impede interchange of air 18 between a containment and the IRWST. When you have steam 19 coming out either from passive RHR operation or, in this 20 case, sparger ADS operation, that will open those louvers 21 up.
22 Most of the vent area is through what we call 23 hooded vents that go straight up through the roof of the 24 tank and then make a 90-degree turn. What is not quite 25 accurate in this picture is that these things are relatively ANN RILEY & ASSOCIATES, LTD.
Court Reporters 1250 I Street, N.W., Suite 300 Washington, D.C. 20005 (202) 842-0034
68 1 close to the containment. There's a foot or so between the
[JI 2 back of the hood and the containment wall.
\
3 The vent comes out here and over here from the 4 IRWST. One of these vents will have a par in it. That's 5 not really shown here. Two of these vents will be set up so 6 that they can -- the louvers will open in the reverse l 7 direction. So in case of a large LOCA, large streamline 8 break, where the containment quickly pressurizes -- now, of 9 course, in that case, the IRWST will be fairly full and 10 there won't be a lot of air in it, but air would come back 11 in from the pressurizing containment and equalize pressure 12 in the IRWST.
13 The other thing I wanted to point out is the 14 recirculation operation, the screens and their performance (s,) 15 in AP600. You can see here, again, a relatively accurate 16 vertical picture of the bottom of the screen, which is a 17 couple feet off the floor, and that's unusual in PWRs. Most 18 of the time, the flood up level in a PWR is much lower. So 19 the screens are kind of pushed much closer to the floor of 20 the containment and they're not nearly so high and the water 21 level is a lot closer to the top of the screen.
22 AP600, because of the large IRWST volume and the 23 gravity injection lower flow rates, it takes, for a non-DBI 24 break, five 25 plus hours to get into this condition. The shortest time l
s I)~
ANN RILEY & ASSOCIATES, LTD.
Court Reporters 1250 I Street, N.W., Suite 300 l Washington, D.C. 20005 (202) 842-0034 l
t . _ _ _ _ __ ______ -__ _ - _
I 69 l 1 that we've been able to figure out is more like three hours ;
(D) 2 and that's where you're running the RNS pumps and they're 3 spilling into the containment and you just keep running them i 4 and that quickens the drain-down.
5 But in any case, you've got a lot of time for any 6 . debris generated by a LOCA to settle. In addition to t'+", I 7 we have, by restriction in the design, we're not allowing l 8 fibrous insulation to be used on pipes that could be L 9 impacted by the LOCA, and we have criteria in the SAR that 1
10 basically say you have to use metal reflective insulation on 11 any pipe that could be a LOCA or any pipe that's located 12 within certain criteria, which is like a 12-ID sphere around 13 the pipes and there's also a 45-ID -- this is without l 14 intervening structures, if you had a clear shot across the 15 containment, it can go that.far.
16 So any of those locations on even pipes that can't 17 be LOCAs, but could be impacted by LOCA jets, are not 18 allowed to use fibrous insulation. So we don't expect to 19 have any fibrous insulation in the water pool here.
20 The other thing that we have is that we are 21 allowing the use of non-safety-related coatings in the 22 containment. Now, we talked a bit about the containment 23 shell and that's a bit of a separate argument. The 24 containment shell, both inside and outside, that inorganic 25 zinc is required to be safety-related, and that's related to ANN RILEY & ASSOCIATES, LTD.
Court Reporters 1250 I Street, N.W., Suite 300 Washington, D.C. 20005 (202) 842-0034
70 1 passive containment system performance really.
O)
(
ss 2 The rest of the coatings inside containment are 3 allowed to be non-safety-related. Now, we do a couple of 4 things to deal with that issue and allow AP600 to function 5 properly. We take some steps to try to make the coatings 6 stay attached to the containment walls and floors by f 7 procuring coatings that are safety-related, qualified for 8 safety-related environments.
9 However, we don't place the safety-related 10 requirements on application and inspection. So as a result 11 of that, we also have to assume that those coatings can come 12 off into the water. And we have done analysis to show that 13 coatings cannot get onto the screens and one of the key r~%'
14 features in AP600 is something that doesn't show up here, l I
(
x_/ 15 but it's an umbrella plate that sticks out above the screens 16 about ten feet, in front of them and to the sides, and under i 17 that plate and in close proximity to the screen, if we use 18 coatings there, they have to be safety-related, but that's a 19 relatively small well defined part of the containment.
20 Anywhere else, we use non-safety related. So 21 basically that says that any coating that gets into the
, 22 water cannot get into the water right in front of the 1
23 screen. It would have to be ten feet away and with very 24 conservative settling calculations and our very low flow 25 rates, and we've got velocities in here that are probably an i
O) '
ANN RILEY & ASSOCIATES, LTD.
Court Reporters 1250 I Street, N.W., Suite 300 Washington, D.C. 20005 (202) 842-0034 l . _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ - _ -
71 1 order of magnitude less than current operating plans.
Il
( j 2 DR. CATTON: Is this described in the SAR?
3 MR. SCHULZ: Yes.
4 .DR. CATTON: Chapter 6 again?
5' MR. SCHULZ: Yes. 6.1 talks about coatings. 6.3, 6 you will find description of the screens, where they are 7 located specifically, where the plates are, and the 8 insulation criteria. So it's a bit in both of those i
9 places.
10 DR. CATTON: Where do you talk about the jet 11 impingement and things like that?
12 MR. SCHULZ: That's in 6.3. I think I'm going to 13 get yanked out of here pretty soon.
p, 14 CHAIRMAN BARTON: How much more time do you have?
15 Apparently you've still got a couple more slides to go.
16 MR. SCHULZ: I could say something about passive l 17 RHR operation, and I don't know if anybody has any questions 18 on the emergency habitability system. I didn't have a lot 19 of material on that.
20 So depending on what you wantrd me to try to l l
21 cover, I could probably wrap it up in ten minutes or less.
22 CHAIRMAN BARTON: Why don't we break now and come 23 back at 20 of 11? It's been quite a while you've been 24 going. Try to wrap yours up in another ten minutes or so.
25 We'll break now til'. 20 of 11. I
() ANN RILEY & ASSOCIATES, LTD.
Court Reporters i
1250 I Street, N.W., Suite 300 Washington, D4 C. 20005 (202) 842-0034 l l
C_ . _ _ _ _ _ . _ _ _ _ _ _ _ _ _ . _ _ _ . _ _ _ _ _ _ _ . . _ _ _ _ _ _ . _ _ _ . _ _ _ _ _ . _ _ _ _ _ _ _
)
72 1 [ Recess.]
() 2 CHAIRMAN BARTON: We're back in session. Terry,
'3 do you want to finish up? Can you finish up by five of 11?
4 We'll give the staff five minutes to discuss their open 5 item.
6 MR. SCHULZ: Sounds wonderful.
7 CHAIRMAN BARTON: Then we'll just go from there.
8 MR. SCHULZ: Let me move on to talk about passive l 9 RHR. The passive RHR is a very simple system mechanically.
10 A single heat exchanger that is connected to the reactor 11 coolant system, has an inlet on the hot leg and a discharge l 12 that's actually in the cold leg in the steam generator 13 channel head, which is effectively the pump suction. I
\
14 That is the lowest pressure point in the system, 15 so if the pumps are running or not running, the flow will i 16 always be in the same direction through the heat exchanger.
17 We don't rely on the pumps running, reactor coolant pumps 18 running, but if they do run, they actually assist the 19 operation of the system.
20 The system normally, like the core makeup tanks, 21 has a set of fail open air operated valves in the discharge 22 of the heat exchanger. Those are the only valves that are 23 normally closed. The inlet is open, so the system is always 24 pressurized. The initial conditions will, as in the core i
25 makeup tanks, this line will be hot up to the high point and ANN RILEY & ASSOCIATES, LTD.
Court Reporters 1250 I Street, N.W., Suite 300 Washington, D.C. 20005 (202) 842-0034
73 1 that is maintained that way because it's well insulated and (n,) 2 routed continuously up to the high point.
3 After.the high point, the piping does come down 4 kind of like is shown on the. schematic and that will quickly 5 become' cold water and the heat exchanger is,normally cold 6 sitting in - -of course, the tubes exposed to the RWST 7 water.
8 The heat exchanger itself actually is two flat 9 tube. sheets, channel head arrangements, which are connected 10 to the tank. wall. Inside the tank are 671 tubes, which have 11 a relatively short horizontal section, vertical section, and 12 another horizontal section back to the channel head.
13 One thing I didn't mention in the core makeup
,, 14 tanks that I think was a question, and that is what about 15 non-condensable gases and can they affect the heat 16 exchanger, that kind of thing. 4 17 We have a pipe stub that sits on the high point of 18 the system and the pipe stub has level sensors inserted into 19 it and if -- we don't expect this to happen, but if, during 20 normal operation, we get some non-condensable collecting up 21 there that will depress the water level and then set off the 22 level alarms and the operators will know about it, and they 23 have a means to vent that.
l 24 Now, the means is local manual valves. They 1
25 actually have to go in containment and open up a couple i
'( ) ANN RILEY & ASSOCIATES, LTD.
Court Reporters 1250 I Street, N.W., Suite 300 Washington, D.C. 20005 (202) 842-0034 E________________z_
74 1 valves. In this case, the vent goes into the IRWST to 2 capture the non-condensables and any water that might come 3 out before they close the valves.
4 ~We don't-expect this to occur, again. There is 5 hydrogen dissolved.in the water, but it's saturated or an
]
6 equilibrium at about 30 pounds gauge, and we're operating at 7 2,200 pounds gauge. So there's a tremendous amount of 8 over-pressure that will keep that hyrogen in solution.
9 That high point vent -- not vent, but the high 10- point collection indication also makes sure that in cases 11 where maintenance is done on the heat exchanger during 12 shutdown, that it is properly vented prior to operation, l 13 return to operation. So that in case they didn't vent the 14 system, you would clearly see that in this high point.
I d 15 We don't see an issue with post-accident operation 16 of the heat exchanger. Again, the hydrogen that's initially 17- there in the water is extremely -- a lot of over-pressure.
18 As I mentioned, when you get down to accumulator injection, 19 you're really into a situation where you've got a LOCA. And
~20 when the accumulator is empty and you might get a lot of 21 nitrogen into the system, we assume the passive RHR stops.
22 It actually probably doesn't, but we don't count on it l j
23 working after that.
24 The IWRST is, again, a large tank. So if you turn 25 on the passive RHR, initially the water in the tank is f l
ANN RILEY & ASSOCIATES, LTD. j Court Reporters 1250 I Street, N.W., Suite 300 Washington, D.C. 20005 (202) 842-0034 4
75 1 sub-cooled and it slowly heats up over about three hours.
() 2 It would then start to boil and steam would go into the 3 containment. And it doesn't show here, but on the other 4 picture I showed of the long-term cooling, there are large 5 vents off the top of the tank that will vent steam into the 6 containment.
In addition, there is a gutter that is also shown 8 on this picture here -- it gives you a pretty good 9 representation -- that would collect water, and this is 10 always around containment, this gutter is located, and it's 11 at the operating deck level or just below it, and that 12 gutter is then sloped so the water comes back to the IWRST 13 side and there are two pipe connections at separate 14 locations to return that's water to the IWRST.
Of 15 Now, that return is not normally functional.
16 Normally, there's a drain off the bottom, actually two of 17 them, that route, during normal operation, any condensate 18 collected down into the waste sump.
19 Following a passive RHR actuation signal, there is 20 redundant safety-related valves, again, fail closed air 21 operated valves, that isolate that normal drain so that any 22 water that's coming in here in a post-accident situation 23 back up and flow into the IWRST.
24 So in passive RHR operation, what that means is 25 that gutter will keep the IWRST essentially full, even if ANN RILEY & ASSOCIATES, LTD.
Court Reporters 1250 I Street, N.W., Suite 300 l Washington, D.C. 20005 l (202) 842-0034 !
l l
76 1 it's not a perfect return, it takes a long, long time to
() 2 degrade the level in the tank.
3 Actually, are there any questions on the passive 4 core cooling system? I think that's about all the 5 presentation. I could put up a slide on habitability 6 control system and talk briefly about that.
7 DR. CARROLL: Not specifically on the passive 8 system, but on EDS-4. It vents to compartments, but what 9 are those compartments?
10 MR. SCHULZ: The loop compartments is what we're
- 11. referring to.
12 DR. CARROLL: Steam generator.
13 MR. SCHULZ: Right. Steam generator compartments.
14 So this is a cut through the containment that actually is
(^%
(s / 15 below the loop level, so you don't really see them. These 16 are the two reactor coolant pumps that are below the steam 17 generator in the loops and fairly new, the concrete with the 18 reactor vessel is where the connection to the hot leo is 19 made on both sides and it comes up and there is a 20 flexibility and it discharges in the loop compartment in a 21 horizontal orientation.
1 22 DR. CARROLL: Has consideration been given to 23 minimizing the amount of insulation you blow off and things l 24 like that?
25 MR. SCHULZ: I mentioned the story on the
('~ ) ANN RILEY & ASSOCIATES, LTD.
Court Reporters 1250 I Street, N.W., Suite 300 l
W9.shington, D.C. 20005 (202) 842-0034
77 1 insulation. All the insulation in these loop compartments I
() 2 is going to be this metal reflective insulation and if you 3 blow it off, it sinks very rapidly, yes. And it can be 1
4- moved along the floor, but, again, our recirc screens, which '
5 are in -- one is in the loop compartment here and one of 6 them is out in this corridor.
7 DR. CATTON: It won't act like a kite and sort of 8 aerodynam*cally be carried alcng.
9 MR. SCHULZ: The flow rates that we have -- again, 10 when we start recirculation five hours after the LOCA 11 occurs.
12 DR. CATTON: And you assume it will all be 13 settled.
14 MR. SCHULZ: It will be settled in minutes, in 15 seconds. It settles relatively rapidly.
16 DR. CARROLL: Well, the big pieces do, but I've 17 ~ seen the results of some tests that have been run simulating 18 this and --
19 DR. CATTON: And even an inadvertent experiment.
20 DR. CARROLL: And you get some pretty fine stuff, )
i 21 too. l l
22 MR. SCHULZ: From the metal reflective insulation?
23 DR. CARROLL: Yes, sir.
24 MR. SCHULZ: The fibrous insulation is whole other 25 story. That stuff will stay in solution indefinitely. l
() ANN RILEY & ASSOCIATES, LTD.
Court Reporters J 1250 I Street, N.W., Suite 300 ;
Washington, D.C. 20005 l (202) 842-0034
I 78 1 There is only stainless steel in this and there is nothing 2 else.
3 DR. CARROLL: I understand. That's what was 4 tested. Do you remember it, Tom?
5 _DR. KRESS: Yes. You get some pretty small metal 6 parts that get transported around. I don't think they're 7 serious screen plugger up here though.
8 DR. CARROLL: I don't either.
9 DR. KRESS: The way the arrangement is and the 10 geometry.
11 DR. CATTON: I can't find the section in Chapter 6 12 that you referred to. Could you maybe have Brian or 13 somebody get me a more specific page number?
14 MR. SCHULZ: Yes.
15 DR. CATTON: I'm interested in the fibrous 16 insulation that's outside of the jet.
17 MR. SCHULZ: The section talks about the recirc 18 screens and the IWRST screens. So it'E in that sort of 19 equipment description section and there's about three or 20 four pages and the first part of it is criteria, and one of 21 those criteria relate to the insulation. I could point it 22 to you. l l
23 DR. CATTON: You have to pay a little bit more !
l 24 attention than just saying it's outside of the jet comb that I I
25 results from the break. There have been experiences where j i
ANN RILEY & ASSOCIATES, LTD.
Court Reporters 1250 I Street, N.W., Suite 300 i Washington, D.C. 20005 (202) 842-0034
)
79 1 if you have area changes and there is steam flow, you can
() 2 get significant shredding of this stuff and a distance.
3 M
.R. SCHULZ: This is all jacketed fibrous 4 insulation that we might use. It turns out that the only 5 place we're probably going to use it is on chilled water, 6 which is up against the containment wall.
7 DR. CATTON: If it's far away, it's probably all 8 right.
9 MR. SCHULZ: We're not going to use any in the 10 loop compartments, which means we're not going to have any 1
11 in the corridors going into the loop compartments.
12 DR. CATTON: Somewhere it says that and if I look 13 through this enough, I'll find it. It's 227.
14 CHAIRMAN BARTON: Go ahead, Terry.
O k ,/ m 15 MR. SCHULZ: I was just talking about the 16 discharge location. I guess I can move on.
17 MR. HUFFMAN: My name is Bill Huffman. I'm NRC 18 project staff for AP600. I thought that one of the 19 expectations of the committee was just to get an update of 20 the status of open items in any particular section you're 21 reviewing. 1 22 I noted Westinghouse did not specifically address I 23 the open item today because they felt it was part of the PCS 24 gothic aspect of your review and will be covered by i
25 subcommittee on June 11 and 12, I think. I just wanted to j
{
l
()
l ANN RILEY & ASSOCIATES, LTD.
Court Reporters l 1250 I Street, N.W., Suite 300 Washington, D.C. 20005 (202) 842-0034 i
I
80 1 let you'know that the staff does consider that issue
'( ) 2 technically resolved. It had to do with assumptions of heat 3 sinks inside containment for the containment pressure 4 analysis.
5 In the process of ITAC, the staff had requested i
6 that these heat sinks be specifically ITAC. Westinghouse,
! 7. on further consideration, decided they did not want to ITAC 8 certain small minor heat sinks such as stairways, gradings, 9 that kind of thing, and reperformed the analysis, not taking 10 credit for those heat sinks and the reason that was an open 11' item at the time of the FSAR was that the staff had not 12 reviewed that analysis.
13 They have reviewed the analysis and they do agree 14 with the conclusions of the analysis. There are some 15 documentation items that need to be completed to wrap up 16 this thing. So the staff co:,r'ders it now a confirmatory 17 item.
18 DR. POWERS: You can also now consider the 19 question of potential stratification of hydrogen in 20 containment as a closed issue.
21 MR. HUFFMAN: Stratification of hydrogen is not an 22 open issue. I feel reluctant to address it any further than 23 that.
l 24 DR. CARROLL: What would the ITAC say about these 25 heat sinks?
i O)
( ANN RILEY & ASSOCIATES, LTD.
Court Reporters 1250 I Street, N.W., Suite 300 Washington, D C. 20005 (202) 842-0034 r
E _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _
81 1 MR. HUFFMAN: That their presence would need to be
() 2 verified at the time of construction.
3 DR. CARROLL: But aren't you talking about 4 concrete?
5 MR. HUFFMAN: Those heat sinks, I think, will be i
6 verified. I think we were talking about minor heat sinks.
7 MR. McINTYRE: Jay, this is Brian McIntyre. What 8 they were looking for us to do was to have an ITAC on the 9 hand rails and the gradings and the thin steel and we
.10 thought that was not a winning proposition, but then not be 11 able to start your plan up because your hand rail was not 18 12 feet long, it was 17 feet six inches or something like that.
13- So what we did was backed off and only took credit 14 for things like the concrete and steel that's clearly there b)
'(/ 15 for structural reason and equipment. It's already there in 16 the ITAC for other reasons.
17 DR. CARROLL: Does it say anything about 18 additional insulating properties of this concrete wall or !
19 whatever or if you painted it?
20 MR. McINTYRE: Additional insulating properties. l 1
21 DR. CARROLL: Suppose you came back and painted 22 it.
1 23 MR. McINTYRE: It is painted. i l
24 CHAIRMAN EARTON: It has a seal on it.
l 25 f DR. CARROLL: I understand. But then for some i f
n/
( ANN RILEY & ASSOCIATES, LTD.
Court Reporters 1250 I Street, N.W., Suite 300 Washington, D.C. 20005 l (202) 842-0034
82 1 reason, you damaged it. You want to paint the whole thing
() 2 over again. Does that change anything or does that require 3 a reanalysis?
4 MR. McINTYRE: I don't have an answer for that. I 5 know that there is some discussion of what you can do with 6 this as far as how you maintain the paint and what the QA 7 program is. You can't have 30 layers of paint on.
8 DR. CARROLL: Okay.
9 DR. MILLER: I have one. For some time, the level 10 sensors on the core makeup tank were an issue. I went back 11 and looked at the FME and I couldn't find that. Is there a 12 way I could get the analysis of how you closed that issue?
13 MR. LEVIN: This is Al Levin from the staff. The
_ 14 initial design for the core makeup tank level sensors ks 15 involved the use of heated RTDs to sense the interface 16 location in the core makeup tank.
17 And there was a question that the staff raised 18 about the ability of those sensors to appropriately indicate 19 coverage versus uncoverage, and it was more a concern really
~
20 of actuating the system when it shouldn't be, rather than 21 not actuating the system when it should be, on the basis of 22 some of the testing that was done at Westinghouse as part of 23 the design certification test program.
24 It was determined by Westinghouse that those 25 heated RTTD sensors could not meet the performance criteria r
I ANN RILEY & ASSOCIATES, LTD.
Court Reporters ,
1250 I Street, N.W., Suite 300 l
Washington, D.C. 20005 (202) 842-0034
- _ _ _ _ _ _ _ _ _ _ _ _ _ _ i
l l 83 1 that had been established and they removed those from the
,\
2 design and replaced them with more conventional differential 3 pressure level sensors.
4 As far as the thermal issue, the thermal sensing 5 issue, with the original design, the removal of those, the 6 reference closed that issue.
7 DR. MILLER: I thought the issue was related to 8 the fact that they had common taps for the DP.
9 MR. LEVIN: That came up separately. That was 10 reviewed by the I&C branch and based on some additional 11 information that was provided by Westinghouse as to how it 12 was going to be configured, they determined that that was an 13 acceptable design.
14 DR. MILLER: I thought that was an open issue for 15 a while.
16 MR. LEVIN: It was an issue under discussion. I 17 don't know if it was ever reflected as an open issue.
18 MR. SHULZ: This is Terry SCHULZ from l
19 Westinghouse. There were official RE open item questions 20 asked and responses made on that issue of shear tap and 21 failure modes of that.
22 And there was some FMEA type work done, submitted 23 to the staff in connection with an RE response.
l 24 DR. MILLER: That issue didn't come through on the 25 PRA.
() ANN RILEY & ASSOCIATES, LTD.
Court Reporters 1250 I Street, N.W., Suite 300 Washington, D.C. 20005 (202) 842-0034 l
L__ _ _ _ _ _ _ _ _ _ _ _ _
84 1 MR. SCHULZ: It was dealt with in design basis
() 2 space. There were PRA considerations in the CMT level
'3 instrument type selection and we were actually-using 4 mechanical level switches, not level transmitters, and 5 that's actually driven by PRA considerations. ;
6 DR. MILLER: So the report you submitted, can I 7 get a copy of this?
8 MR. SCHULZ: It's in an RE response, so we could, 9 I guess, identify which one it was. Brian is making faces 10 at me, but maybe I can. Yes. We can tell you what response 11 that.was. It's a handful of pages kind of' thing.
12 CHAIRMAN BARTON: Moving to Chapter 14. Gene, 13 kind of put is in-perspective, where we are here, since it's 14 not 10:30 anymore. About 11 to 12, give you till 12:00, and 15 have the staff comments on this chapter, and then we'll 16 break for lunch.
17 MR. SCHULZ: I think I can do it faster than that.
18 I'll try to.
19 CHAIRMAN BARTON: While he's getting a battery, 20 let's find out if the staff has any comments.
I 21 MR. WILSON: This is Jerry Wilson for Projects..
22 We found the initial test program acceptable and on that 23 note, I would encourage Mr. Piplica to be brief in his 24 remarks.
25 DR. CARROLL: Wait a minute. I don't like that ANN RILEY & ASSOCIATES, LTD.
Court Reporters 1250 I Street, N.W., Suite 300 Washington, D.C. 20005 (202) 842-0034
85 1 philosophy, Jerry. We're doing'a review as well as you (O ,,r 2 guys.
3 MR. WILSON: I' wanted to be sure you had 4 sufficient time for questions.
5 DR. CARROLL: So ipso facto, it should be 6- acceptable to us.
7 MR. PIPLICA: My name is Gene (Poplica. I was the 8 chapter lead for chapter 14 and I'm going to discuss the 9 initial test program in an overview sense.
10 DR. CARROLL: What is yo'Ir background that makes 11 you qualified to put on an initial test program?
12 MR. PIPLICA: I've worked for Westinghouse for 33 13 years. I started out in the nuclear engineering department
.14 of the fuel division. I've done reload core designs. I 15 have worked on startup test programs in that time period.
16 Since then, I've --
17 DR. CARROLL: In the field?
18 MR. PIPLICA: Not in the field. Designing and i
19 writing programs to reduce data, core distributions and the 20 like, and think I was involved in the last eight years in.
21 the AP600 design certification testing program on thermal 22 hydraulic testing program. ;
23 The staff that supported me in the writing of this l 24 program included field people'. They provided the 25 information that I needed and the review of the initial test f ANN RILEY & ASSOCIATES, LTD.
Court Reporters 1250 I Street, N.W., Suite 300 Washington, D.C. 20005 (202) 842-0034 l
t.- .
86 1 program to be certain that it was done properly.
() 2 DR. MILLER: Those field people had physically 3 started up a reactor.
4 MR. PIPLICA: That's right. And Carl Dumsday was 5 the person that we most heavily relied on, as well as people 6 in the fuel division that did startup testing.
7 MR. McINTYRE: And, Gene, I think you also had 8 utility review.
9 MR. PIPLICA: That's right. We did have the 10 utility review of the startup test program.
11 DR. CARROLL: The same guys who liked the security 12 plan, right?
13 MR. PIPLICA: Yes, they participated, and others.
14 The test program was reviewed to NUREG-0800, the standard 15 review plan. It cites Reg Guide 1.7 for content and also 16 Reg Guide 1.68, among others. But Reg Guide 1.68, Rev. 2, 17 dated August '78, was really what we based the program on.
18 I wanted to point out that these reg guides are 20 19 years old and the AP600 was not a design in existence at 20 that time, and as a result, that led to substantial 21 interaction between the staff and Westinghouse to be certain 22 that the initial test program encompassed the tests that 23 needed to be performed for passive systems.
24 So we incorporated in this test program tests that 25 cover the new systems in the AP600 that weren't explicitly
) ANN RILEY & ASSOCIATES, LTD.
Court Reporters 1250 1 Street, N.W., Suite 300 Washington, D.C. 20005 (202) 842-0034
87 1
1 called out in Reg Guide 1. '
l () 2 The test program consists of three types of tests; 3 construction and installation, pre-operational, and startup. <
4 The ITAC performs during the first two phases of 5 the initial test-program and must be satisfactorily 6 completed prior to the initiation of fuel loading. Once the 7 ITAC have been approved and the fuel load begins, then the 8 tech specs take over for making sure the plant is safely 9 operated.
10 Types of construction tests are listed here. An 11 important point is that the testa are performed based on 12 information supplied by the vendors who supply that 13 equipment and that test abstracts are not provided in s 14 Chapter 14 for construction tests. We rely on the CO k- % 15 applicant to do that test program.
16 DR. CARROLL: And that'a okay by the staff.
17 MR. PIPLICA: Yes.
18 DR. CARROLL: As long as it's a COOL action item.
19 MR. PIPLICA: That's correct. The next phase of 20 tests, the pre-operational tests are quite important and 21 extensive. Now, I also gave you a listing of the table of I 22 contents for chapter 14 for the sole purpose that you could I 23 look at what systems are tested as part of the different 24 phases of the pre-operational and startup tests.
25 Terry talked about in-service testing of the ANN RILEY & ASSOCIATES, LTD.
Court Reporters 1250 I Street, N.W., Suite 300 Washington, D.C. 20005 (202) 842-0034
_______- ____-_- _. _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ - _ .___-_. . _ _ _ _ . _ _ _ _ . _ _ _ _ _ _ _ _ - _ _ _ _ _ _ _ _ _ _ _ ~
88 l 1 valves, for example. Pre-operational tests establish the l-() 2 baseline for that test program. That is the first time that l 3 the valves are.really put through any performance testing in 4 place in.the plant and that establishes the data and the l
l 5 baseline for going forward with in-service testing.
l 6 We test the equipment on the basis of each I
! 7 component and then we look at integrated system performance, 8 especially during the hot functional tests. Now, we don't 9 have in the test program a test that says hot functional.
10 But all the tests that are performed where heat is required 11 and provided by the reactor coolant pumps are part of the 12 hot functional testing program.
13 Now, an important point to make here is that the t
14 staff made sure that we incorporated in the testing program 15 that the principal design organizations, those people who l
16 are responsible for providing the equipment's and systems, l 17 are part of the development of the test specifications and 18 procedures that establish the performance and acceptance 19 criteria for the tests.
20 The staff wanted to be sure that those who knew l
21 what needed to be tested were involved in the' development of l 22 those programs by the COOL applicant. So that is 23 specifically stated in chapter 14.
24' And the test abstracts for the pre-operational 25 tests are in sections 14.2.9. Now, the startup tests begin ANN RILEY & ASSOCIATES, LTD.
Court Reporters 1250 I Street, N.W., Suite 300 Washington, D.C. 20005 (202) 842-0034
89 1 with fuel loading after pre-operational tests have been m
i I 2 successfully completed. The SAR, being a tier two document,
%/
3 can't talk about ITAC, but with that, the ITAC also have to 4 be successfully completed, and that's part of the rule, but 5 it's not in the SAR explicitly.
6 And we divided the startup tests into four broad 7 categories based on Reg Guide 1.68, Rev. 2, and they are 8 tests related to putting the fuel in the core, tests 9 performed after the fuel is loaded, but before criticality 10 -- for example, rod drop test, making sure that you have the 11 ability to shut the reactor down before you go critical.
12 Then we have the lower power testing related to 13 bringing the plant to initial criticality. These are the 14 low power physics tests, as they're c>mmonly known.
\_- 15 And then there are substantial tests after that 16 greater than five percent power to bring the plant to power l 17 and once this test program is completed successfully, the 18 plant is able to go into operation.
19 DR. CARROLL: A problem I think at least I have 20 had and I think a lot of utilities had is getting locked i
21 into completing the pre-op test program before you can load 22 fuel. Sometimes there are minor systems that have nothing 23 to do with floating fuel. There is still some work going l
24 on.
25 Have you taken that into account in developing
() ANN RILEY & ASSOCIATES, LTD.
Court Reporters 1250 I Street, N.W., Suite 300 Washington, D.C. 20005 (202) 842-0034 i
90 1 this?
(O j 2 MR. PIPLICA: Yes, we have. We talked to our 3 field people and we talked to utilities to try to find out 4 what types of things we could do and the thing that they 5 told us most was that there was always confusion on what was 6 required from a licensing basis for the test completion --
7 completing the program.
8 So what we tried to do is organize the 9 pre-operational test program in a manner that would make it 10 easier for them to figure out what the acceptance criteria 11 was and what order to perform the tests in, and I'll talk a ,
i 12 little hit about that in a slide after this one. I 13 DR. CARROLL: But the big problem is getting 14 construction completed on some of these. !
I_i (m) 15 CHAIRMAN BARTON: And then the design before that, !
16 too.
17 DR. CARROLL: And the design before that. You've 18 been on a wet drawing job, have you?
19 MR. PIPLICA: Okay. Startup tests make sure that 20 we can start up in a controlled and safe manner, including 21 shutdown systems. We want to verify the nuclear operating 22 characteristics and in some cases accident assumptions; for 23 example, rod drop time and coast down time for the reactor 24 coolant pureps. And it's very important that we obtain and 25 calibrate the data used for the control and protection
(#)
'~
ANN RILEY & ASSOCIATES, LTD.
Court Reporters 1250 I Street, N.W., Suite 300 Washington, D.C. 20005 (202) 842-0034 L___________________________ . _ _ _ _ _ _ _ _ _ _
91 l
1 systems in-the pre-op program and that we verify that the j e- -
( ,T) 2 plant is operating or can operate within the tech specs.
3 So there's a lot of tests that establish set 4 points, for example, a lot of points that look at these !
5 things. And just like the pre-op tests, the principal 6 design organizations have to be involved with the test I 7 specification and procedure development and section 14.2.10 8 describes these tests in detail.
9 DR. MILLER: This initial portion, was this unique 10 to this plan?
11 MR. PIPLICA: No. As a matter of fact, once you 12 load the fuel, you're pretty much the same. There were a 13 couple of unique aspects in the testing that we added as a 14 result of our interaction with the staff because of passive
(_) 15 systems.
16 For example, our primary method of removing decay 17 heat is the PRHR. So more emphasis is put on testing the 18 PRHR than the steam generator capability to do that, and I'm 19 going to specifically talk about that briefly.
20 Now, I want to read something out of the reg 21 guide. In the introduction it states, and I quote, "While 22 it is required that all structures, systems and components 23 important to safety be tested, it is not required that all 24 of them be tested to the same stringent requirements.
25 Specifically, criteria one of Appendix A of 10 CFR Part 50 f)
'~
ANN RILEY & ASSOCIATES, LTD.
Court Reporters 1250 I Street, N.W., Suite 300 Washington, D.C. 20005 (202) 842-0034
_ = _ _ _ _ _ - _ _ _ _ _ - _ _ _ _ -. . _ _ _
92 j 1 requires, in part, that structures, systems and components
() 2 important to safety be tested to quality standards .
3 commensurate with the importance of the functions to be 4 performed. A graded approach is inherent in the testing 5 requirements of criteria 11 of Appendix B to 10 CFR 50."
{
6 What this meant to-us, and we used this in the l 7 development of the program, is that you want to focus your 8 resources and your efforts on the systems that are most 9 important. That is not to say that systems of less i
10 importance need to be neglected. I 11 What it means is you want to develop your test 12 program so that you make sure that the safety systems are 13 tested extensively and thoroughly and to that end and to 14 help the organization of the program, we organized chapter 15 14 in this method.
16- We first picked the systems that performed 17 safety-related functions and described them in section 18 14.2.9.1 to a higher level of detail, the highest level of 19 !
detail that we could. '
20 For example, if you look in there and you look at !
21 the passive core cooling system tests, we have extensive 22 testing on this system, because it's of the highest 23 importance. Same thing with containment cooling, the 24 hydrogen control system and control room habitability. !
25 These are systems that Terry talked about earlier.
ANN RILEY & ASSOCIATES, LTD.
Court Reporters 1250 I Street, N.W., Suite 300 Washington, D.C. 20005 (202) 842-0034 l
- i l
i 93 l' -These are important safety functions.
() 2 -Then in section 2.9.2, we have systems that 3 perform defense-in-depth functions and then we go down and 4 as a result of --
I 5 DR. KRESS: How did you decide what a 6 defense-in-depth function was?
7 MR. PIPLICA: It's in the SAR. The SAR describes, 8 for each system, what its main functions are. The SAR l 9 states for this system it performs these following 10 safety-related functions. Then we refer to those SAR 11- sections to test those functions.
l 12 DR. KRESS: How did the SAR -- what criteria did 13 the SAR use to decide?
14 MR. PIPLICA: I'd defer that to Terry.
15 MR. SCHULZ: This whole concept of 16 defense-in-depth -- when Gene mentions defense-in-depth 17 here, it's talking about non-safety-related aspects of that 18 term. That term can be used and is used in the whole idea 19 of fuel activity containment, the fuel rods, the RCS, the 20 containment.
21 And in that sense, it's a safety-related 22 defense-in-depth. AP600 has another use of that term which 23 relates to non-safety-related systems, like the startup 24 feedwater and the diesels, that are not relied upon to 25 mitigate design basis accidents.
ANN RILEY & ASSOCIATES, LTD.
Court Reporters 1250 I Street, N.W., Suite 300 Washington, D.C. 20005 (202) 842-0034
94 1 DR. KRESS: It backs up a system that's relied on, 73
() 2 then you call it defense-in-depth.
3 MR. SCHULZ: We have our own criteria which is 4 very similar to that which basically says if a 5 defense-in-depth system in these non-safety terms is a 6 system that prevents or can avoid the need for passive 7 system operation, and that's a defense-in-depth system.
8 So like startup feedwater, if it works propelry 9 and the passive RHR is not needed, at least in most events, 10 the same is true of the CDS makeup for RCS leaks and small )
1 11 breaks, boration, that kind of thing, passive containment l 12 cooling versus fan coolers.
13 So it's the ability of a system to avoid the need 14 for a passive system operation. If it can do that, then m 15 we've labeled it as a defense-in-depth system.
16 DR. CARROLL: Now, the hydrogen igniters are a j 17 defense-in-depth system, are they not, or do you -- j 18 DR. KRESS: They're safety-related.
19 MR. SCHULZ: The igniters are not safety-related.
20 They're not a defense-in-depth system either, in this 21 context.
22 DR. CARROLL: So where do they end up in this?
23 MR. PIPLICA: They end up actually -- the hydrogen 24 control system performs safety-related functions, so the 25 PARS are tested there, and since that system performs a I ANN RILEY & ASSOCIATES, LTD.
Court Reporters 1250 I Street, N.W., Suite 300 Washington, D.C. 20005 (202) 842-0034
95 1 safety-related function, we also test the igniters at that
. 2 time.
3 DR. CARROLL: Because they back up --
4 MR. PIPLICA: Right. So we first look for the 5 safety-related functions and then we test the entire system, 6 even the functions that aren't safety-related are tested.
7 DR. KRESS: Now, is this'the priority?
8 MR. PIPLICA: This is a priority listing, yes.
9 DR. KRESS: It's an order going down.
10 MR. PIPLICA: It's an order. going down.
11 DR. SEALE: It's interesting, though, that service 12 water and chilled water systems are not safety-related.
13 They're defense-in-depth.
14 MR. PIPLICA: Yes, they are. There are 16 systems l O
( s/ 15 that have safety-related functions and there are 21 systems 16 that have the defense-in-depth functions in the initial test 17 program.
i 18 Then Reg Guide 1.68, Rev. 2 required us to add {
19 systems that process, store and control radioactive 20 material, solid, liquid and gaseous rad waste systems are 21 examples.
22- That's important from an operability or safe 23 operations standpoint. So it needs to be tested.
24 And then non-safety-related systems that are 25 require by Reg.' Guide 1.68, Rev. 2, were listed in the last
() ANN RILEY & ASSOCIATES, LTD.
Court Reporters 1250 I Street, N.W., Suite 300 i Washington, D.C. 20005 (202) 842-0034 l
l
96 1 section.
I 2 These are a lot of systems in the balance of 3 plant, turbine-related, for example. For the startup tests, 4 as I mentioned, we ordered them pretty much chronologically 5 as they would be performed, starting with fuel load, moving 6 to initial criticality, into the lower power physics tests, 7 and then the power ascension tests.
8 We based this test on what was designed in the --
9 what was in the Vogtle FSAR. We used that as a starting 10 point and we used the same type of format. Here we 11 describe the test objective, prerequisites, what method is 12 required for testing, and what is the performance criteria.
13 I didn't mention it on the previous slide, but 14 both in the objectives of the test and in the acceptance 15 criteria, we refer to the SAR, because the SAR describes in 16 most cases, when it talks about the system, it describes 17 what is important from a testing standpoint, from a 18 licensing standpoint.
19 Rather than reiterate that in chapter 14, we refer 20 back to the SAR section for that information.
21 CHAIRMAN BARTON: Why did you pick the Vogtle 22 units?
23 MR. PIPLICA: Because it was a plant that was most 24 recently put on-line.
25 DR. CARROLL: Some of your consultants came from ANN RILEY & ASSOCIATES, LTD.
Court Reporters 1250 I Street, N.W., Suite 300 Washington, D.C. 20005 (202) 842-0034
j 97 1 --
l
/~s
() 2 MR. PIPLICA: And they had experience in starting 3 that plant up.
4 DR. CARROLL: Whatever that utility is called 5 nowadays.
6 MR. PIPLICA: I don't know.
7 CHAIRMAN BARTON: Southern, is that the Southern 8 Company?
9 DR. MILLER: It may be different tomorrow.
10 MR. PIPLICA: One feature of the AP600 was that in 11 the ICE, that we've incorporated what we call special tests.
12 These are tests that further establish unique 13 phenomenological performance parameters that extend beyond j 14 those that were established in the design certification j
(_) 15 testing program.
16 And we expect that these parameters will not vary 17 from plant to plant and, as such, we developed, with the 18 staff's feedback, two categories of plants. The first plant 19 only and the first three plant tests. I 20 Originally, Westinghouse proposed about 21 tests )
21 that would be performed on the first plant only. But as a 22 result of interaction with the staff, that list was 23 substantially reduced, so that we now have six first plant 24 only tests and we have three first three plant tests.
I 25 And as the name states, we expect to perform the l
/% ANN RILEY & ASSOCIATES, LTD. T i'-"j j Court Reporters 1250 I Street, N.W., Suite 300 l Washington, D.C. 20005 !
(202) 842-0034 l
j
98
-1 .first plant only test on the first AP600 and on the first A
( ,) 2 three plant tests, a minimum of three plants will be used to 3- test.
4 DR. CARROLL: That jumped out at me when I read 5 it. I have a little trouble -- we've got some tests we're 6 going to do on every plant.
7 MR. PIPLICA: A big test program.
8 fDR . CARROLL: Big program. And then we've got 9- first time only. The first three, I mean, I just don't 10 understand why we need a first three category. It seems to 11 me you could neatly fit everything into the first one.
12 I sense a little bit of a compromise here.
13 MR. PIPLICA: Yes, I would agree with that. But I 14 tend to agree with what the staff asked for here because f
15 what we're doing here, the issue was will the program vary 36 from plant to plant and the issue was how well could we rely 17 on the ITAC program to be sure that the plants are built 18 identical for these systems and these -- by doing the tests 19 on the first three plants, we feel we can conclude that the 20 plants have been built identically such that the performance 21 of the plant for the phenomena that we're looking at does 22 not vary from plant to plant. ,
23 .DR. CARROLL: Give me an example of three plant 24 parameters that you're interested in.
25 'MR. PIPLICA: Okay. I'll skip over the first
() ANN RILEY & ASSOCIATES, LTD.
Court Reporters 1250 I Street, N.W., Suite 300 Washington, D.C. 20005 (202) 842-0034
99 1 plant only test, and here they are. There are three tests.
() 2 3
There are two core makeup tank tests. One involves natural circulation heat-up of the CMT. In the pre-op test program, 4 we will run the coolant pumps to bring the system up to 5 temperature and then we will open the valve and let the CMT 6 recirculate until it heats up.
7 Then once we get there, we'll close the valve, 8 rehear the system, so the whole system is hot, and then 9 allow it to begin to recirculate and transition to a drain 10 down mode, so that it drains -- this is a very important 11 injection feature of the plant.
12 DR. CARROLL: What is anybody concerned about 13 beyond the first plant?
14 MR. PIPLICA: The issue was to make sure that when 15 the plants are constructed, that they are constructed in 16 such a manner and that this phenomena of being able to heat 17 the tank and drain it will measure drain rates and heat-up 18 rates and temperature distributions in the CMT, will be the 19 same for plant to plant, and really provide additional 20 assurance that all the analysis that we've done for the 21 plant is still applicable.
22 That's really where it's coming from, because it's 23 a new concept, a new fatter.
24 There is also the issue of boron injection. These 25 2,000 cubic foot tanks are borate and this also tests the
() ANN RILEY & ASSOCIATES, LTD.
Court Reporters 1250 I Street, N.W., Suite 300 Washington, D.C. 20005 (202) 842-0034
100 1 injection feature, which is important from a criticality
-2 standpoint.
3 The other first three plant test is that of 4 ADS-123 lowdown. We'll test valve operation by initiating 5 the lowdown with the' ADS valves, but the real issue here is 6 sparger performance and hydrodynamic loading. If some of 7 you are aware of the thermal hydraulic testing that we did, 8 when we tested this system, we used a circular tank, a 9 larger circular tank, and in the plant it's an irregular 10 tank.
11 The tests were used to develop data to analyze 12 that tank to see how it responded to the lowdown. We will 13 test using accelerometers, pressure transducers, the 14 response of the IRWST to the sparger lowdown, and we will r
15 repeat that for.three plants. And if the test results are 16 the same, then you don't need to do it anymore, we won't 17 need to test that feature anymore. We will have the 18 assurance that thermal hydrodynamic loads are the same from 19 plant to plant to plant.
20 DR. KRESS: What will be your supply of --
21 MR. PIPLICA: This is a pre-operational test 22 performed during a hot functional and the reactor coolant 23 pumps will be-used to heat the plant to -- and we used non 24 trump to analyze 25 the lowdown.
() ANN RILEY & ASSOCIATES, LTD.
Court Reporters '
1250 I Street, N.W., Suite 300 Washington, D.C. 20005 (202) 842-0034
101 1 DR. CATTON: That makes me feel really good about
() 2 the results.
3 DR. KRESS: Will these tests include somehow 4 draining down the CMT to activate the ADS system? That line 5 of actuation's.
6 MR. PIPLIKA: The tests were performed for that 7 . function, but not this one. We're going to actuate it on --
8 we're going to wait until we get the temperature and make 9 sure we have as much stored energy as we can in the plant.
10 Another reason that we want to limit these to 11- three plants is these are very -- these tests consume 12 thermal fatigue loads. There is a usage factor applied to 13 the plant and by performing these blowdowns, you get l 14 substantial -- youknow, you're blowing down into the IWRST, 15 you're going to put energy into the containment. You're 16 also going to put a large usage factor on the plant isle, l
17 the nozzles and the like.
18 So we, Westinghouse, want to limit that because 19 we're using a $1.2 billion plant to blow down -- to do a 20- test of this nature and it's an extreme test, in our 1 21 opinion.
22 DR , KRESS: Was ADS-4 ever tested?
23 MR. PIPLICA: Yes. That's part -- that's tested 24 in every plant, i 25 DR. KRESS: That's on the part of every plant.
f 1m( ,/ ANN RILEY & ASSOCIATES, LTD.
j Court Reporters l 1250 I Street, N.W., Suite 300 i Washington, D.C. 20005 I
(202) 842-0034 c_______________________--___________ _ - _ _ _ _ - _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ - _ _ _ _ _ _ _ _ - - _ _ _ _ _ _ _ _ _ _ _ _ -
102 1 MR. PIPLICA: Yes. We have ADS-4 tested in every
'O plant.
3 'DR. CARROLL: I still haven't been convinced, for 4 example, ADS 1 through 3 require three tests. I mean, this 5 is a standardized plan, for goodness sakes.
6 MR. LEVIN: This 3s Allen Levin from the staff.
7 You're correct in that it was a compromise. These two 8 systems in particular rely on an integrated performance of a 9 number of key components and it was our feeling that
- 10. checking the individual components basically through ITAC 11 was not necessarily an adequate verification of the system 12 as a whole would perform as it was supposed to.
13 CHAIRMAN BARTON: But you would accept that on 14 plant four.
,O
\s / 15 MR. LEVIN: Well, yes, because three plants we l
16 felt was enough of a baseline that if Westinghouse could l 17 demonstrate that the plant could be built according to the 18 certified design, repetitively to that point, that that was 19 enough of an experience base to give us confidence that 20 these tests would -- at that point, would be extraneous.
21 Would not be necessary -- I shouldn't say extraneous. Would 22' not be necessary to have to demonstrate integrated system 23 performance.
24 If you look at the ones that we did agree to have 25 first plant only, those are basically tests where we felt
() ANN RILEY & ASSOCIATES, LTD.
Court Reporters 1250 I Street, N.W., Suite 300 Washington, D.C. 20005 (202) 842-0034
! 103 1 like even small variations -- like IRWST heat-up. It's an l
i /N .
( ) 2 enormous tank. Even if there are small variations in the 3 dimensions of the tank, we didn't feel that it was going to 4 affect that.much the way the tank would heat up and stratify 5 and one test in the first plant would be adequate to do 6 that.
7 But the CMT, on the other hand, you've got a lot 8 of different types communicating here and we've got to be 9 not only -- each one has got to be right, but they've got to 10 be put together right, too. So al the resistance is working 11 the way they're supposed to. It's a little bit more 12 complicated.
13 DR. CARROLL: Do you expect plant two to have 14 nominally at least identical piping system design to plant
's_) 15 one?
16 MR. PIPLICA: Yes. In this regard, yes. This is 17 tier one type stuf, it's a part of the rule. These are 18 safety systems and this is part of the verification of that.
19 These are ITAC type tests.
20 DR. CARROLL: They don't have to do the piping 21 design yet.
22 MR. SCHULZ: The detailed routing of lines and 23 number of elbows, placement of elbows, the exact slops, are 24 not ITAC.
25 There is a lot of ISAAC on the line resistance of f (s.
D i
ANN RILEY & ASSOCIATES, LTD.
Court Reporters 1250 I Street, N.W., Suite 300 Washington, D.C. 20005 (202) 842-0034
j 104 1 these key lines. In some cases, N-Max. There are ITACs on l () 2 some elevations, some key elevations, slopes in a few cases, 3 slopes usually we say that -- like on the passive RHR and 4 other core makeup tank lines from the RCS loops to the high 5 point. There are no downward slopes. The line is either 6 upward-sloping or minor horizontal type thing.
1 i
7 DR. CARROLL: May question is, will plant two have 8 a nominally dietnical design to plant or do they hat the 9 option of doing a new independent piping syste.. design.
10 MR. LEVIN: From Westinghouse's point of view, the 11 criterion is the resistance basically, like for CMT piping.
12 So hypothetically, they could re-route the piping, but they 13 would still have to demonstrate via ITAC that the 14 resistance, the overall resistance and whatever j
/"%
ks m 15 characteristics are ITAC meet the min / max or'the whatever is 16 applied in the ITAC as the design acceptance criterion.
17 And that's sort of another reason for the first 18 three plants rather than just the first plant only in this 19 regard, because if a slightly different routing is chosen, 20 it still meets the window of acceptability for resistance, 21 let's say, to be sure that the integrated system again I l
22 behaved the way it's supposed to. l 23 DR. CARROLL: Okay. That makes sense.
24 MR. PIPLICA: The first plant only test, there are 25 six of them, there's two here in natural circulation. RWC
') ANN RILEY & ASSOCIATES, LTD.
Court Reporters 1250 I Street, N.W., Suite 300 Washington, D.C. 20005 (202) 842-0034
105 1 heat-up, we added, primarily because of feedback we got from
() 2 the ACRS, this verifies the secondary side characteristics 3 on the outside of the PRHR as it's operating.
4 Pressurizer stratification is required by NRC 5 Bulletin 88-11 and is described in SAR section 3.9.3. It 6 measures temperature distribution and surge line 7 displacement. If you recall, we have that large serpentine 8 surge line. We want to verify our thermal analysis. So 9 that's only done one time.
10 Reactor vessel internal vibrations, that's 11 required by Reg Guide 1.2, and we agreed with the staff to 12 o chis testing for flow-induced vibration on the first 13 plant only to measure vibrational characteristics, and 14 that's normally what is done in new plant designs now.
- 15 From lessons learned at Three Mile Island, action 16 item I(g) (1) of NUREG-0737 requires natural circulation 17 testing of the steam generator, but in the AP600, the 18 passive RHR performs that function, so we test them both, 19 and I'm going to describe that test next.
20 Then Westinghouse wants to do a load follow 1 l
21 demonstration so that we can demonstrate the capability to 22 perform load follow with our'new gray rod design. So this L 23 was a test we wanted to do one time only.
1 24 So those are the first plant only tests that we're 25 agreed on.
l l
l
() ANN RILEY & ASSOCIATES, LTD.
Court Reporters j 1250 I Street, N.W., Suite 300 ;
Washington, D.C. 20005 l (202) 842-0034 !
I i
106 1 DR. CARROLL: Why do you commit to that to the NRC
( 2 then?
3 MR. PIPLICA: Part of Reg Guide 1.68 does talk 4 about load swings and load follow. We do have another test 5 for that purpose, but we want to demonstrate it to our 6 customers that it can be done.
I 7 DR. CARROLL: Okay. Anytime you make a needless 8 commitment to those guys over there, you're asking for i
! 9 trouble.
10 DR. MILLER: That's based on experience?
11 DR. CARROLL: Yes, based on experience.
12 MR. PIPLICA: Now, because of lessons learned from 13 Three Mile Island, this natural circulation test needs to be 14 performed and then the information from this test is used, 15 the data is used in simulators. You take the data, you put 16 it in a simulator and then you further train future 17 operators on how to manage the plant.
18 So this test performed at three percent power is 19 performed with the core loaded. We run the coolant pumps 20 and record the system temperatures around the system. We 21 trip the RCPs, maintaining power using the control rods, and 22 then we verify that natural circulation is obtained by 12 3 measuring the hot leg temperature in each loop and recording l 24 it as time goes on.
]
25 Now, during this test, the power was removed by
() ANN RILEY & ASSOCIATES, LTD.
Court Reporters 1250 I Street, N.W., Suite 300 Washington, D.C. 20005 (202) 842-0034
- _ _ _ __ _ _ - _ _ - _ _ _ _ _ _ _ _ _ _ _ _ _ _ _. _ _ a
107 1 dumping steam from the steam generators and the test is b) 4, 2 complete when you no longer have a change in the hot leg 3 temperature. So you're verifying the heat removal 4 capability of the steam generator.
5 This is what is done in today's plants. One time 6 for a particular steam generator design. Immediately 7 thereafter, following this test, we're going to initiate 8 flow through the PRHR heat exchanger by opening that valve 9 and we performed this identical test at OSU,-by the way, in 10 the pre-operational test and what happens is that heat 11 removal will gradually reduce through the steam generator as 12 the PRHR begins to remove the core power.
13 And when we do this test, we're going to get full 14 power conditions, flow, inlet and outlet temperatures, and
<O k/ 15 IRWST temperatures that will verify the performance of the l 16 PRHR, full power, full height, full pressure, full 17 temperature, the whole ball of wax. We'll have everything 18 we need.
19 And then the test is run to a limit of IRWST 20 heat-up of 150 degrees. We don't, again, want to put an 21 unnecessary thermal transient on the plant and we'll finish l
l 22 the test by closing the PRHR outlet valve and shutting the '
23 reactor down.
1 24 So I think this is a.very useful test and the data 25 that we're going to get from this will be very valuable to l
ANN RILEY & ASSOCIATES, LTD.
Court Reporters 1250 I Street, N.W., Suite 300 Washington, D.C. 20005 (202) 842-0034 L_.______._____________.__________._2__.___ _ _ . _ _ _ _ _ _ _ _ _ _ . . _ _ . __ _ . . . _ _ _ _ _ .
108 1 future ~ operators and simulators of the plant and it will
( 2 really give us a lot of information on how decay heat is 3 removed from this plant.
4 DR. CARROLL: I'm a little confused. You're going 5 to do this test in the five percent --
6 MR. PIPLICA: Three to five percent range to 7 . simulate the core' decay heat.
8 DR. CARROLL: How high has the reactor been in 9 power up to that point?
.10 MR. PIPLICA: It successfully completed the low 111 power physics testing.
12 DR. CARROLL: Okay. But there's not enough decay 13 heat to simulate --
14 MR. PIPLICA: Exactly.
15 DR. CARROLL: -- full power.
16 MR. PIPLICA: Right.
17 DR. CARROLL: So you'll leave the reactor.
18 MR. PIPLICA: Yes.
19 DR. MILLER: Up to two or three percent,_this 20 plant, you expect, will be exactly the same as the 21 Westinghouse PWR.
l 22 MR. PIPLICA: It's essentially -- yes -- a large 23 .two-loop plant.
24 .DR. MILLER: But going.from the physics testing to 25 this tecting, I would expect it's going to react ANN RILEY & ASSOCIATES, LTD.
Court Reporters 1250 I Street, N.W., Suite 300 Washington- D.C. 20005 ,
(202) 842-0034 -
i
109 1 differently.
()
2 MR. PIPLICA: Yes, because the PRHR will begin to 3 remove the heat, but --
4 DR. MILLER: I mean differently than a traditional 5
I plant. !
6 MR. PIPLICA: All we're doing is short-circuiting l 7 the steam generator and the PRHR -- so I don't think the 8 reactor coolant loop will perforra any differently. i 9 It's just going to demonstrate and show the 10 . operators how they could control the plant by removing decay i
11 heat through the PRHR.
12 DR MILLER: How about the steam generator part?
13 MR. PIPLICA: This is the same as a typical 14 existing operating plant, a Westinghouse operating plant, 15 This test was already performed.
16 DR. MILLER: Nothing new. Okay.
1 17 MR. PIPLICA: Yes.
18 DR. CARROLL: Each class of plant had --
19 MR. PIPLICA: Each class, yes. Each class of 20 plant had to perform this test.
21 DR. CARROLL: I know a lot about that one.
22 MR. PIPLICA: And my last overhead just discusses 23 what the combined license applicant has to do to perform the 24 test. The staff and I went through many iterations to make 25 sure that we had all these requirements covered,
() ANN RILEY & ASSOCIATES, LTD.
Court Reporters 1250 I Street, N.W., Suite 300 Washington, D.C. 20005 (202) 842-0034
110 1 organization and staffing, the test specification and
() 2 procedures, the administration procedures, evaluation of the 3 results, feedback to the responsible or principal design 4 organizations, the need to correct deficiencies and retest 5 if necessary, and, of course, the applicant has to perform 6 tests of their own on site-specific systems or systems that 7 were not included in the scope of responsibilities for the 8 design certification application.
9 So there is only -- section 14.4 is about a 10 page-and-a-half and it briefly describes what is required by 11 the applicant.
12 DR. FONTANA: I Pake it obviously you don't test 13 the passive containment tem and integral tests. You test 14 the water flow and al' ..at sort of --
O 15 MR. PIPLICA: Yes. We test the individual 16 features, air cooling, water distribution, water coverage, 17 flow rates for water on the outside of the containment 18 shell.
19 DR. FONTANA: Are you able to measure the heat 20 transfer from the inside of containment and the outside of 21 containment?
- 22 MR. PIPLICA
- We could, but we don't require that l
23 in the initial test program. The coating is verified.
l 24 DR. FONTANA: In a conventional plant, is the 25 spray system tested individually?
() ANN RILEY & ASSOCIATES, LTD.
Court Reporters 1250 I Street, N.W., Suite 300 Washington, D.C. 20005 (202) 842-0034
- ___ _ _ _ _ _ _ _ _ _ _ _ _ - _ _ = _ _ _ _ _ . __ _ ______ ________ ________________ _____________ _ _ ___________ _ _______ _
111 1 MR. PIPLICA: The containment spray inside, it 2 would be, yes.
3 DR. CARROLL: How are you going to turn it on?
4 DR. FONTANA: Well, you don't get everything wet, 5' but you test it.
6 MR. PIPLICA: You test it on an individual basis.
7 You make sure the valves work and -- i 8 DR. FONTANA: Yes, that's what I mean.
9 DR. CATTON: P'aw air through the nozzles, do 10 stuff like that.
11 DR. CARROLL: Actually, they did do a real test at 12 Zion.
13 CHAIPMAN BARTON: Somebody was coming up to sign
, 14 up for that.
\ 15 DR. CATTON: There was some purpose, because 16 there's a video and they have umbrellas and rain slickers on 17 and the whole bit.
18 MR. PIPLICA: Okay. 1 19 DR. CARROLL: I have one test that was a little
{
20 ambiguous to me. This is compressed and instrument air 21 system testing, 14280. I always worry about -- I mean, I 22 can lose the air supply totally in a variety of ways very 23 quickly. I can also lose it very slowly and the question 24 that always comes up is what sort of unwanted actions occur i
25 when that happens.
() ANN RILEY & ASSOCIATES, LTD.
Court Reporters J
1250 I Street, N.W., Suite 300 Washington, D.C. 20005 (202) 842-0034 i
L_______--_____-_____--
112 t
1 MR. PIPLICA: You mean how much do we rely on it-
'2 from a --
3 DR. CARROLL: I know you don't consider it 4 safety-related, but I'm not sure I know what that i
5 distinction really means.
6 MR. PIPLICA: Terry?
7 MR. SCHULZ: We have some air operated valves in 8 the plant for makeup tanks, passive RHR, passive containment 9 cooling, isolation valves. They are all fail safe in the 10 sense that whatever the safety-related function of the valve 11 is, either open or closed, depending on its function, it 1
12 doesn't rely on or require the air system. There are no air 13 bottles. We use the springs in the valve to push it to its 14 safe position.
15 DR. CARROLL: So in a slow decay or air pressure, l
16 you could end up with valves opening or closing sort of at 17 random. ,
4 18 MR, SCHULZ: Somewhere along the way, right. i 19 DR. CARROLL: 7ut do you do a test to understand l 20 what may happen there?
21 MR. SCHULZ: We understand what will happen in the i 22 end.
23 DR. CARROLL: I'm worried about the middle. I 24 MR. PERALTA: If I may interrupt. There is a test I
25 in chapter 14, 14.2.9.4.10, compressed and instrument air (G ,) ANN RILEY & ASSOCIATES, LTD.
Court Reporters 1250 I Street, N.W., Suite 300 Washington, D.C. 20005 (202) 842-0034
113 1 system testing. Testing is done and satisfies Reg Guide ,
I'h
( ,f 2 1.68 and Reg Guide 1.68.3 and one of the acceptance criteria 3 is to perform a test to verify the fail safe positioning of 4 all safety-related valves upon sudden or gradual loss of ,
5 instrument air pressure.
6 DR. CARROLL: So is it envisioned that they would i
7 do a series of gradual losses of pressure? '
8 MR. PERALTA: Right.
9 DR. CARROLL: How many?
10 MR. PERALTA: That will be decided once the tech i
11 specs and the --
12 MR. PIPLICA: Every valve would be tested.
13 DR. CARROLL: No. I'm saying that you could have
,, 14 an infinite variety of rate of pressures.
\s / 15 MR. PIPLICA: That's right. They would be tested 16 singularly and integrally all together. Whether or not 17 there are combinations in between, that would have to be 18 decided upon by the applicant when they rate the test 19 operating, test procedures and specifications and that's 20 reviewed by the staff at that time and that decision would 21 be made then.
22 DR. CARROLL: So has the staff taken a position on 23 that as to what would be acceptable?
24 MR. PERALTA: Well, the test abstract and the test 25 program is consistent with what we have proven in the past.
f C\ ANN RILEY & ASSOCIATES, LTD.
(_j Court Reporters 1250 I Street, N.W., Suite 300
! Washington, D.C. 20005 (202) 842-0034 L____________. . - _ _ _ _ _ _ _ _ _
114 1- It's similar to --
() 2 DR. CARROLL: That's why I'm asking the question.
3 People have gotten away with just having a sudden loss of 4 instrument air kind of test.
5 MR. PERALTA: We have,.in the SAR, both a sudden 6 and gradual loss.
7 DR. CARROLL: I'm not aware of anybody who has 8 ever done a gradual loss test.
9 MR. PERALTA: Watts Barre did recently,. I was 10 involved with that. It gets complicated. It depends how 11 the system is configured and how many valves are being l l
12 supplied by how many manifolds. So it's very site-specific. l 13 But a lot of the details will be, I guess, i 14 developed once the plant is built and we are involved with 15 the test-specs and test procedures. But I guess overall we j l
16 do expect, in principal, the test to be done. I 17 MR. PIPLICA: I'would envision, at that time, that 18 the principal design organization would have an impact on 19 this test by specifying what they thought was probably 20 events and how the system is made and what needs to be 21 tested, and then that, again, would be reviewed by the 22 staff.
23 DR. FONTANA: I understand the fuel is the same as 24 in conventional plants. Is the control rod configuration j 25 the same?
l g
ANN RILEY & ASSOCIATES, LTD.
Court Reporters 1250 I Street, N.W., Suite 300 Washington, D.C. 20005 (202) 842-0034
115 1 MR. PIPLICA: Yes, it is.
1
() 2 DR. FONTANA: How long does it take to do these 3 tests?
4 CHAIRMAN BARTON: Starting from where, 5 construction through 100 percent or where are you? What's )
i 6 the scope of your question? Pre-op testing? l 7 MR. PIPLICA: We've had some of our partners in 1
8 the certification program work on a construction and testing I 9 schedule, but I don't know the results of that.
10 CHAIRMAN BARTON: Because the construction tests 11 overlap during construction. That's why I said what's the 12 scope of your testing. You start pre-operational testing to l
13 power testing.
i 14 MR. PIPLICA: But there is a schedule.
i !
15 DR. CARROLL: Except pre-op testing, you do a few 16 systems early on, as soon as you can, like power and air and 17 things like that.
i 18 CHAIRMAN BARTON: Typically, it's a couple year l 19 program.
20 MR. PIPLICA: It's shorter in this case because
! 21 the entire plant is going to be built. )
1 22 CHAIRMAN BARTON: So you build them quicker these 23 days. The new ones you can build quicker.
24 MR. PIPLICA: That's the whole idea behind the 25 AP600 and that's why we've had people look at the testing ANN RILEY & ASSOCIATES, LTD.
Court Reporters l
1250 I Street, N.W., Suite 300 Washington, D.C. 20005 (202) 842-0034
l 115 1 schedule to be sure that it's integrated with the.
/9
( j 2' construction so that we can meet our first concrete core 3 load schedulo. I just don't know the details of it. I was 4 more worried about content rather than timing it when I ,
5 wrote this.
6 CHAIRMAN BARTON: Any other questions? Does the 7 staff have any questions?
8 MR. WILSON: No. As I said, we found this program 9 acceptable and are available for questions from the 10 committee.
11 CHAIRMAN BARTON: Any questions of the staff?
12 N response.]
(;o 13 CHAIRMAN BARTON: If not, we'll take a lunch 14 recess and come back at 1:00.
15 [Whereupon, at 11: 49 a.m. , the committee was 16 recessed, to reconvene at 1:00 p.m., this same day.] l l
17 j
i 18 '
19 20 i
21 22 l
l 23 24 l 25 l ANN RILEY & ASSOCIATES, LTD.
Court Reporters 1250 I Street, N.W., Suite 300 Washington, D.C. 20005 (202) 842-0034 1
- l. .
I 117 1 AFTERNOON SESSION 2 [1:02 p.m.]
! 3 CHAIRMAN BARTON: At this time, we'll get into the 4 PRA level two and three discussion in severe accidents.
5 Brian, you're the man here.
6 MR. McINTYRE: Thank you, John. The first part of 7 the presentation, the level two and three PRA, is going to 8 be made by Jim Scobel, and the second part, the external 9 reactor vessel cooling in severe accidents, is going to be i 10 by Bob Lutz. Mr. Scobel is the speaker.
11 MR. SCOBEL: My name, if you want to write it down 12 to remember it. Good afternoon. I'm going to talk to you 13 about the level two/three PRA and the work that we did for
( 14 that.
l 15 DR. KRESS: How can you do a level three? You
(
16 don't have a site, do you? l l 17 MR. SCOBEL: No, we don't have a site, but there 18 was generic data that was --
19 DR. KRESS: It's a generic site --
20 MR. SCOBEL: -- used to calculate the off-site 21 doses. As you know, for the PRA, we did internal initiating I 22 events at power and we also did internal initiating events 23 at shutdown, but for the shutdown work, we used the same !
24 phenomenology as the at power and we used linking for 25 systems. So that was kept consistent.
I
() ANN RILEY & ASSOCIATES, LTD.
Court Reporters .
1250 I Street, N.W., Suir.e 300 Washington, D.C. 20005 (202) 842-0034
l 1
118 1 So what I'm going to focus on for this
,m
() 2 presentation, just to keep it short, is what we did for at 3 power.
4 For level two PRA, the first thing that you need i 5 to do is you have to take all of the core damage frequency 6 and you have to group it into accident classes and you do 7 this by taking the sequences that have similar 8 ' characteristics and putting them into bins or accident 9 classes or plant damage states, or whatever you would like ;
10 to call them.
11 For this PRA, what we did was we took them and we 12 did them initiating event, timing of core damage, which 13 means did your core damage begin because of failure of 14 injection or failure of recirculation, and also what the RCS s
s_/ 15 pressure was at the time of the core damage.
16 By doing this, we came up with 11 accident 17 classes. Do you have the pointer? Here is one. Thank you.
18 Okay. And so we have these 11 accident classes 19 and probably the most important characteristic of the 20 accident classes was the reactor coolant system pressure.
21 As we go through the presentation, you'll see how that 22 applies, but basically we have a bunch of accident classes, 23 like the ATWS and the transients with no heat removal, that 24 would be like classic PMLB prime kind of sequence, that have l 25 pressures that are above the steam generator secondary l
' ANN RILEY & ASSOCIATES, LTD.
Court Reporters 1250 I Street, N.W., Suite 300 Washington, D.C. 20005 (202) 842-0034 f
L_____________-.__________________
119 1 relief pressure.
2 We have partial ADS accident sequences which get 3 you down below 150, but not low enough to be able to inject l 4 by gravity. And then there are a series of accident' classes 5 that have -- could potentially have gravity injection. They l- 6 are fully depressurized, but for some reason, you either 7 didn't get injection or you didn't get recirculation, which 8 is what took you into your core damage event.
S DR. CARROLL: But the initial condition for those-10 is full power.
11 MR. SCOBEL: The initial condition for these is l 12 full power. In fact, there is another set of similar-13 accident classes that start with LP for low power or 14 shutdown. But they correspond so that you can tell which
( 15 one is which. It's LP1A, LP3D.
16' DR. CATTON: Before you take that away, the last 17 meeting, when was that, Jay, 31 March?
i 18 DR. CARROLL: Yes.
19 DR. CATTON: A question came up about the 20 pressurizer and during a steam generator tube rupture, the 21 pressurizer level rises and falls, and what would happen if 22 it went solid and the safety stuck open and so forth. Is 23 that a consideration anywhere in the steam generator tube 24 sequences?
25 MR. SCOBEL: Yes, it is. Actually, in the --
ANN RILEY & ASSOCIATES, LTD.
Court Reporters 1250 I Street, N.W., Suite 300 Washington, D.C. 20005 (202) 842-0034
i 120
\
1 DR. CATTON: Could you tell me what probability #
() 2 was associated'with that? Conditional probability.
3 MR. SCOBEL: I can tell you the frequency of those 4 events is like ten-to-the-minus-eight.
5 DR. CATTON: For that steam generator tube l
6 rupture, I'd like to know just that little piece. Given 7 that you have a steam generator tube rupture, what's the 8 chances of overfilling the pressurizer?
9 MR. SCOBEL: I don't have that number.
! 10 DR. CATTON: Could you get it for me? The reason f 11 I ask is that your LOFTRAN calculations show several 12 different distributions of the heat load. One of them it
- l. 13 fills up. The data is pretty damn close to filled and then 14 the others are below. I'd almost say it's a 50/50 chance. )
l 15 You can get back to me with that.
16 MR. SCOBEL: Okay. I'm just thinking about that 17 particular -- the phenomenology of that particular -- !
18 DR. CATTON: It turns that into a different kind 19 of accident. Maybe it helps, I don't know. 3 20 MR. McINTYRE: Dr. Catton, the probability number I 21 you're looking at is something that's more of a level one i 22 and we are going to have a level one session -- Noel, help l
23 me here -- for a few hours at the next subcommittee meeting. I 24 We'll just do that then.
25 DR. CATTON: But do you understand why I'm asking?
I ANN RILEY & ASSOCIATES, LTD.
Court Reporters 1250 I Street, N.W., Suite 300 Washington, D.C. 20005 (202) 842-0034 i
r 121 1 It ties back into your V&V of LOFTRAN.
) 2 DR. KRESS: When you say injection, are you 3 takling about the core makeup tank? l 4 MR. SCOBEL: I'm talking about gravity injection, 5 IRWST. !
l 6 DR. KRESS: The IRWST.
7 MR. SCOBEL: That's right. The final gravity 8 injection that's the ultimate injection and then i
9 recirculation for the core. Every good level two PRA has ;
10 severe accident phenomena and these are the in-vessel l
11 phenomena that are considered explicitly in the level two 12 PRA for AP600. Of course, we have in-vessel hydrogen 13 generation, induced steam generator tube failure, in-vessel 14 steam explosion, in-vessel retention of core debris, which C-)s 15 is unique to the AP600 at this point, hydrogen detonation i
l 16 and deflagration.
17 We look at diffusion flames forming somewhere in 18 the containment where they can cause problems and long-term i 19 containment over-pressure from decay heat steaming.
20 DR. CARROLL: And the problem of hydrogen 21 diffusion flames cause containment --
22 MR. SCOBEL: It depends on where they're burning. ]
23 If they're burning up against the containment shell or 24 against penetration, then you can have an induced 25 containment failure or you also --
~s ANN RILEY & ASSOCIATES, LTD.
Court Reporters 1250 I Street, N.W., Suite 300 Washington, D.C. 20005 (202) 842-0034
122 ;
1 DR. SEALE: Or a stress, anyway. j 2 DR. SCOBEL: That's correct. And if they're 3 burning in other places, that can affect your equipment 1
4 survivability.
5 DR. FONTANA: When you look at hydrogen detonation 6 and deflagrations, first, in hydrogen detonations, maybe 7 first -- well, detonation to the deflagration, you have 8 accelerated flames. Do you look at that?
l 9 MR. SCOBEL: Yes, we do.
i 10 DR. FONTANA: And when you look at detonations, do 11 you look at shock focusing?
1 12 MR. SCOBEL: No. We say if you get a detonation, 13 you get containment failure, just to be conservative because 14 --
nV 15 DR. FONTANA: In effect, you're looking at 16 focusing, because that's how you would fail it.
17 MR. SCOBEL: Okay.
18 DR. CARROLL: Now, to get detonation, you need i 19 fairly high concentrations of hydrogen.
20 MR. SCOBEL: That's correct. l 21 DR. CARROLL: What did you do about the 22 possibility of hydrogen stratification?
23 MR. SCOBEL: We look at that and, in fact, what we 24 do is we look at time-frames for the hydrogen.
25 DR. CARROLL: Is there a report or something that
-t ANN RILEY & ASSOCIATES, LTD.
Court Reporters 1250 I Street, N.W., Suite 300 Washington, D.C. 20005 (202) 842-0034
L 123 i.
'1' ' covers this other than Bob Henry's?
() 2 :MR. SCOBEL: Chapter.41 of the PRA.
3 DR. CARROLL: Chapter 41.
L 4 MR. SCOBEL: All the hydrogen.
5 DR. CARROLL: I don't remember it saying very 6 much.
'7- MR,'SCOBEL: Basically, what it comes down to is 8 the stratification is mainly a stream stre.tification. The 1
l.
1
-9 hydrogen is well mixed in the containment because-you have I
10 stage four ABS open at the bottom of the containment that's 11 driving the flow around the --
12 DR. CARROLL: Is there somewhere that you have {
13 kind of documented all the rationalizations that-lead you to 14 it? I've read the EPRI report.
15 MR. SCOBEL: Appendix 6A of the SAR.
16 DR. CARROLL: Appendix 6A. And I just got chapter
.17 6.
18 DR. CATTON: You got them. Should have them.
19 MR. SCOBEL: But to answer your question, we do 20 have a higher probability of detonation in the long term due 1
21 to stratification of the steam and we treat that 22 conservatively by assuming dry air in some of the 23 compartments below the containment, where you actually have i 1 24 more of a probability of accelerating the flame. !
l 25 DR. CATTON: Well, the hydrogen tends to build up i
I ANN RILEY & ASSOCIATES, LTD.
Court Reporters 1250 I. Street, N.W., Suite 300 Washington, D.C. 20005
, (202) 842-0034 1
o 124 1- where the steam condenses out and some of the arguments-I've
() 2 heard are that it's condensing on the walls,'so it sweeps j 3- the hydrogen back in and mixes it up. But everybody has 4 ignored, in order to be. conservative, the cooling on the 5 top. Now, the cooling on the top, that ain't going to 6' happen, and most likely the cooling on the top is going to 7 be pretty good.
8 So I suspect that you're going to build up some 9 hydrogen up at the top of the containment. It's not going 10 to be swept down the walls. What's the chances of 11 overfillin -- i l
12 DR. POWERS: I don't_have an experimental basis )
1 13 that'says, okay, here's a containment that's been held with l 14 this kind of radiation flux at these kinds of temperatures 15 for this period of time and here's how much acid and yes I 16 did the calculation and, boy, they agree. I mean, I just ,
I 17 don't have that experimental base. I don't. Maybe you do.
18 MR. SCHULZ: I don't know that our calculation has 19 that either, but we have done a calculation. It is, I 20 think, more theoretical than experimentally based and it 21 does end up with us putting in two to three times the amount 22 of TSP than we otherwise would.
23 DR. CARROLL: I guess I perennially ask the I 24 question, based on the box of TSP I have on the shelf in my 25 workshop --
() ANN RILEY & ASSOCIATES, LTD.
Court Reporters 1250 I Street, N.W., Suite 300 l Wasl.'ngton, D.C. 20005 1202) 842-0034
125 1 DR. POWERS: Your solubility problem.
() 2 DR. CARROLL: Yes. I end up taking it and putting
-3 it on my anvil and hitting it with a hammer. I guess, as I 4 read it, you're going to take some of this out every outage 5 and see if it dissolves well, and that sort of thing, but 6 that's a real phenomena.
7 MR. SCHULZ: We assume in our dissolution 8 calculations for TSP that it's a solid block. .
9 DR. CARROLL: You do.
10 MR. SCHULZ: We do. We don't take any credit for 11 the fact that it originally was granular when it was loaded.
12 DR. CARROLL: So what's helping you is the f
13 temperature.
- 14. MR. SCHULZ: Yes.
i O' 15 MR. CARROLL: Water.
l I
16 MR. SCHULZ: The temperature, the water. It still 17 takes several hours to dissolve. The block that we're 18 talking about is relatively thin. It's about a 19 foot-and-a-half thin. The basket, because of the shape of 20 the containment, we don't have a lot of flat area in the 21 lower parts of the containment, so this TSP basket is l 22 basically hung on a wall, there are two of them. So they're 23 relatively tall, wide, and fairly thin.
t 24 We only take credit for water contact on one l 25 surface and when we do the calculation, we do assume it's a j l
ANN RILEY & ASSOCIATES, LTD.
Court Reporters l 1250 I Street, N.W., Suite 300 J l Washington, D.C. 20005 J. (202) 842-0034 L________________._____________________ . _ _ _ _ _ _ . _
126 1 solid block.
( ,j 2 MR. SCOBEL: Just a warning. If you ask a 3 question from a design basis space to me, I'm going to give 4 you the PRA answer, and that sometimes will make those guys 5 jump up. That's what happened in this case.
6 DR. POWERS: There's a problem that all modelers, 7 including myself, especially myself, that when we create 8 these models, we say they have a certain reality to it, and 9 here, especially when we're talking about the later term 10 sorts of things rather than confidence, the calculations, 11 where I think that phenomenologically isn't all in yet, and 12 I don't know, for instance, that the bounding case is put 13 the iodine all in the water and do the partitioning out of 14 it and ignore the absorption onto surfaces within the
()
V 15 containment.
16 It might be that your releases of iodine for your 17 later containment failure scenarios is worse when you put 18 things onto the surfaces, which I think you have not done.
19 I think conventionally all your iodine ends up in the water 20 down below.
21 MR. SCOBEL: That's correct.
22 DR. POWERS: Simply because we don't have the I 23 phenomenological evidence. That's the problem you run into.
24 MR. SCOBEL: We're not so subject of that in this 25 PRA because of these low frequencies. There is a reason for
( \ ANN RILEY & ASSOCIATES, LTD.
\~
Court Reporters 1250 I Street, N.W., Suite 300 Washington, D.C. 20005 (202) 842-0034
127 1 this, which is that these longer term failures are primarily
() 2 due to things like hydrogen detonation, or decay heat 3 steaming, pressurization of the containment, which were not
!' 4 particularly susceptible to, because of our igniters and 5 because of our passive containment cooling system.
L 6 So this configuration makes very good sense from 7 that standpoint and that also gets out of the woods as far i 8 as your question is concerned for long-term releases.
l 9 DR. CATTON: The MAAP code was the basis for l
I 10 working your way through these dominant event trees, I take 11 it.
12 MR. SCOBEL: No. For the containment event tree l 13 --
l
.(. 14 DR. CATTON: MAAP modeling dominant event tree
- l. : 15 path.
16 MR. SCOBEL: That's for developing the source i
17 term.
l 18 DR. CATTON: What did you do when you worked your i
i 19 way through these event trees to establish success criteria?
20 MR. SCOBEL: We used our phenomenological analyses 21 that were outside of codes, the large --
22 DR ., CATTON: You did not use MAAP.
23 MR. SCOBEL: No. We have MAAP analyses of 24 sequences for demonstrating the performance of the plant and 25 their sensitivity analysis.
I ANN RILEY & ASSOCIATES, LTD.
Court Reporters ,
1250 I Street, N.W., Suite 300 i Washington, D.C. 20005 (202) 842-0034 l
L -_____ _ - -_-_ _____-____-__ -__ __- ._ __ l
128 1 DR. CATTON: The reason I ask is because there was
() 2 two-loop plant and there were RELAP comparisons with MAAP 3 and the success criteria arrived at with MAAP were not 1 4 conservative.
5 MR. SCOBEL: Is this the German one?
6 DR. CATTON: No. This is Krsko, Slovenia.
7 MR. SCOBEL: Right. It was done by a German 8 consultant, though.
9 DR. CATTON: That was done by my student, who is l 1
10 in Slovenia, but that's a separate issue. !
l 11 MR. SCOBEL: Those were wrong.
12 DR. CATTON: His were wrong?
13 MR. SCOBEL: Yes.
14 DR. CATTON: Well, that's nice. l (ms/ 15 DR. MILLER: Did he do his Ph.D. on that?
16 DR. CATTON: No, that wasn't the subject of his 17 Ph.D. This was consulting work that he was doing.
18 MR. SCOBEL: I hope Rich is here to back me up.
19 The analyses that he did there were primarily for large 20 break and MAAP is well known limitation about modeling a 21 large break with MAAP, it doesn't have --
22 DR. CATTON: This wasn't done for a German outfit.
23 This was done -- Westinghouse did it, did the study of the 24 Krsko plant. And he was assigned to work with Westinghouse.
25 MR. SCOBEL: I didn't do that.
() ANN RILEY & ASSOCIATES, LTD.
Court Reporters 1250 I Street, N.W., Suite 300 Washington, D.C. 20005 (202) 842-0034
129 1 DR. CARROLL: Ivan uses sort of a scatter gun 2 appronch.
3 DR. CATTON: I just remember --
4 MR. SCOBEL: We might be talking about different 5 analyses.
6 DR. CATTON: I think we are.
7 MR, SCOBEL: Because Westinghouse --
8 DR. CATTON: Westinghouse was contracted to do the 9 study of the Krsko plant and then he was loaned to 10 Westinghouse.
11 MR. SCOBEL: Then that's not what I --
12 DR. CATTON: And it was done in Europe I think.
13 MR. SCOBEL: He's not the one. I'm not familiar 14 with those analyses. There is one that's constantly used to
/"'s
\%-) 15 say how bad MAAP's comparison is and it's all based on 16 large-break-LOCAs. And if you just do a double-ended 17 guillotine break with MAAP, you will blow down the primary 18 system due to flashing of steam and everything, but you i 19 don't have the -- like the bypass period that will 20 completely wash all the water out of the reactor vessel and 21 dry it out and then all the accumulator water will go around 22 the core barrel and go out.
l 23 That doesn't exist in MAAP and if you're modeling )
1 24 a large break in MAAP, despite doing a double-ended )
i 25 grillotine break, it doesn't work.
ANN RILEY & ASSOCIATES, LTD.
Court Reporters 1250 I Street, N.W., Suite 300 Washington, D.C. 20005 (202) 842-0034
130 1 If you wanted to do it to-trick -- to wee what
() 2 would happen,- you'have to trick the code by putting the 3- large break at the reactor vessel for a certain. period of 4- time.
5 DR.. POWERS: .You' answered the question when you
- 6. said you didn't do it.
7 CHAIRMAN BARTON: Let's move right along now.
8 MR- SCOBEL: . Yes, sir.
9 DR. CATTON: But I will ask the student.
10 MR. SCOBEL: And my apologies to your student.
11 DR. CATTON: I'll tell him who you are.
12 DR. POWERS: Where he lives, what kid of car he 13 drives and things like that.
14 MR. CARROLL: This is your Sicilian student.
15 MR. SCOBEL: I grew up in that neighborhood.
1 16 Okay. We used MAAP. We're back to'the MAAP thing. So for 17 each of of the release ccatrrgories in the MAAP run, we used 18 the largest source term, the one that would produce the 1
l 19 largest off-site doses to represent the entire release ;
i 20 category. l 21 For two release categories, we did use 22 contamination factors for things that MAAP doesn't model.
23 We don't have a model for the aux building and 24 we've got a DF of 3 based on some work-that was sponsored by '
25- DOE and also based on work that was sponsored by DOE, we l
() ANN RILEY & ASSOCIATES, LTD.
Court Reporters 1250 I Street, N.W., Suite 300 Washington, D.C. 20005 j (202) 842-0034 ;
l ,
131 1 used the secondary systems and a decontamination factor of
() 2 100 for the steam generator tube rupture.
\
3 DR. POWERS: Is that an experimental l
4 determination? !
i 5' MR. SCOBEL: No , it wasn't experimental. It was l 6 based on' work that was done at Pulstar. i 7 DR. POWERS: And so this is strictly a figment of 8 an analyst's-imagination.
- 9. MR. SCOBEL: That's correct.
I 10 DR. CATTON: Those numbers are a lot larger than 11 the ones that are used for the boiling water reactor 12 suppression pool.
l 13 MR. SCOBEL: This isn't due to suppression. i 14 Remember, this isn't due to water. This is due to the heat !
("h I sinks that are inside the -- all the tubes and the impaction
~
\ms/ 15 i
16 on the tubes and the heat sinks inside the steam generator i 17 secondary system,' that fission products can plate out on.
18 DR. POWERS: Did it consider restripping and 19 revaporization?
20 MR. SCOBEL: Yes, it did. There are later tests 21 that or later calculations that say that this-number is too 22 big for certain sequences. Those came out after this was 23 done. But we do have the. sensitivity analyses in the PRA 24 that don't include them. '
25 DR. POWERS: It's a very high decontamination
' ANN RILEY & ASSOCIATES, LTD.
Court Reporters 1250 I Street, N.W., Suite 300 Washington, D.C. 20005 (202) 842-0034 i 1
132 1 number. I think-Ivan's point, when he says it's higher than
() 2 the BAR suppression pool, is that we Ehink that that might .
3 be a bound on the decontamination you would expect in a 4 small closed system like this, that you get very high 5 numbers and they depend on an analysis in which you're 6 having to hypothesize what the properties are of the 7 particles and how they interact with surfaces.
8 Typically, you don't have things like bounds built
.9 into those models. Apparently, restripping and 10 revaporization have been taken into account. You also don't l
11 have electrostatic effects built into those models.
12 An awful lot of uncertainty to come up with a such 13 a big DF.
l 14 DR. CATTON: Wasn't the one in the suppression O)
\- 15 pool DF around ten?
16 DR. POWERS: I'm much more confident in 17 suppression pools than most, I guess. I give pretty high 18 numbers for them if they're sub-cooled and fairly deep. l 19 DR. CARROLL: I think 100 is the number that --
20 DR. CATTON: This is saturated.
21 DR. POWERS: I still have great faith in l
22 suppression pools. On the other hand, I know that hoops can 23 get you really royally ripped if you're pulling the wrong 24 kind of aerosol through them. l 25 MR. SCOBEL: The next two slides --
ANN RILEY & ASSOCIATES, LTD.
Court Reporters 1250 I Street, N.W., Suite 300 :
Washington, D.C. 20005 (202) 842-0034
1 133 1 DR. KRESS: When you're talking about source
() 2 terms, outside containment?
3 MR. SCOBEL: Outside containment, to the 4 environment. The next two clides -- and, really, your 5 slides you can look at better than I can put them on this 6 viewgraph. I showed noble gases and iodine, which is 7 particulate iodine, and cesium --
8 DR. KRESS: Since the noble gasses are hovering 9 around 80-90 percent, I would assume the initial release of 10 the fuel is hovering around 50 percent, and your releases 11 show something like at least an order or so magnitude lower 12 than that for iodine and cesium. Is that because you're 13 getting decontamination in the process somewhere? I was 14 comparing those two curvea.
15 MR. SCOBEL: Right. This number is highly 16 affected by the DF of 100. The bypass number. That's where 17 that 100 DF lives.
18 DR. KRESS: How about the CFE?
19 MR. SCOBEL: This has no DF at all.
20 DR. KRESS: That's what I don't understand.
21 MR. SCOBEL: Because this is to the environment.
22 So you have a release from the core to the containment.
23 DR. KRESS: It's natural attenuation in 24 containment.
25 MR. SCOBEL: There is some attenuation in ANN RILEY & ASSOCIATES, LTD.
Court Reporters 1250 I Street, N.W., Suite 300 Washington, D.C. 20005 (202) 842-0034 l
_ _ _ ____ ___ _ ___-_ _ _ _ _ _ _ _ _ _ - _ _ - _ - _ _ _ _ _ _ - - _ - - _ _ _ _ - - - _ - - - - --- _ A
134 1 containment and then you have the failure of the containment
( )~ 2 'which. flows down, the rest of it, close to ten percent.
l 3 So you have a certain release fraction of the 4- containment during that time. You have some deposition and 5 then you have the release over -- well, this was pretty much 6 instantaneous failure.
7 Since it is a steel containment and you can 8 postulate potential for bursting.
9 DR. POWERS: Again, to be clear, the CFI and the 10 CFL do consider iodine partitioning.
11 MR. SCOBEL: They do not.
12 DR. POWERS: On the late release category, do you 13 treat revaporization off the primary piping system 14 reinjecting material into the containment?
15 MR. SCOBEL: Yes.
16 DR. POWERS: Is that over by the time yo'uve 17 gotten to the CFL releases?
18 MR. SCOBEL: Probably, yes. I haven't looked at 19 that specifically for a long time.
20 For the consequence analyses, which is the level 21 three, we use the MACCS code. We primarily look at the site 22 boundary dose because we were trying to look at 25 rem at 23 the site boundary as our goal.
24 We also looked at a population dose. That was 25 primarily for doing the SAMDA analyses that we were required ANN RILEY & ASSOCIATES, LTD.
Court Reporters 1250 I Street, N.W., Suite 300 Washington, D.C. 20005 (202) 842-0034
l -
135 1 to do. And for these we considered no evacuation.
() 2 This answers the question that you asked at the
.3 beginning of the presentation. We used the generic site 4 population and meteorological data that specified in the URN I 5 and we used these source terms from the MAP analyses.
I 6 And in the end, we generated this curve. The dark l
j 7 line at the top is the total off-site dose risk of the 8 plant. This box represents the URN criteria, the 25 rem ac I 9 the. site boundary, the 24 hours2.777778e-4 days <br />0.00667 hours <br />3.968254e-5 weeks <br />9.132e-6 months <br />. You can see that in the --
10 for the lower releases, the probability is dominated by 11 intact containment, which drops off down around five -- this
! 12 is rem -- about five rem. From that point on, the large 13 release is dominated by bypass and early containment 14 failure.
I 15 Also, from this plot, you can see that at the 25 16 rem criteria, we have a frequency of 17 one-times-seven-minus-eight. It's actually 1.8-times 18 ten-to-the-minus-eight.
19 This slide then is a summary of core damage 20 frequency, large release frequency and the population risk 21 out to 72 hours8.333333e-4 days <br />0.02 hours <br />1.190476e-4 weeks <br />2.7396e-5 months <br />. It's not really changing much after that.
22 I included shutdown in here so that you can see --
23 at the beginning I told you I was going to focus on that 24 power end. This is where I'm throwing shutdown back in. So l
25 this is the total for the plant. So core damage frequency ANN RILEY & ASSOCIATES, LTD.
Court Reporters 1250 I Street, N.W., Suite 300 Washington, D.C. 20005 (202) 842-0034 ;
i l
! l
136 1 of'2.6-times ten-to-the-minus-seven and a large release 3Oj 2 frequency of 3.3-times ten-to-the-minus-eight.
3- So our CCFP, if you will, is a bit over ten I 4 percent.
5; DR. POWERS: Your shutdown entries up.there are 6 . annualized, right?
7' MR. SCOBEL: Yes. That's all per year. J 8 DR. CARROLL: And is there a great deal more J
'9' conservatism in the shutdown value than in that hour value?
10 MR..SCOBEL: From a. level two standpoint, yes, 11 because we used level two failure probabilities for the i
12 phenomenological events at shutdown, were, in fact, -- they 13 would be less. As far.as system failures, the system- l
-14 failures are right-because we used the fault tree linking. i 15 We used specific fault trees for shutdown. l 16 IM1. POWERS: When you looked at shutdown, some 17 substantial fraction of the events that involve some level i
18 of core damage and core heatup must be recovered.
19 ;MR . SCOBEL: That's true and that's not included 20 in the evaluation. These basically go to full core melt.
21 22 l 23 24 25.
1
() ANN RILEY & ASSOCIATES, LTD.
Court Reporters i 1250 I Street, N.W., Suite 300 L
Washington, D.C. 20005 (202) 842-0034 l
l
137 1 Also, in answer to that question, something I
() 2 should have stated at the beginning of this presentation, 3 the only accident management that we considered in this 4 presentation was to turn on the igniters and to flood the 5 reactor cavity if it needed to be flooded, so other recovery 6 actions like what you would be talking about there are not 7 included.
8 That is why we go to full core melt.
9 DR. CARROLL: And what tells the operator to flood 10 the cavity -- core exit, thermocouple?
11 MR. SCOBEL: Yes. If you have entered into FRC1, 12 which you do at 1200 degrees Fahrenheit, the core exit 13 thermocouple temperature and you step through FRC1, and at a 14 certain point you are not recovering systems. You are not 15 getting water into the core.
16 At that point the EOPs will tell you to flood the 17 cavity and prepare for severe accident by going to the CMG.
18 DR. CATTON: That's 1200 degrees F.?
19 MR. SCOBEL: That is at 1200 degrees F. That is 20 when you enter FRC1.
21 So -- for my conclusion slide for the PRA part of 22 the presentation, we talked about accident classes. We 23 talked about the containment event tree and the release 24 categories at the end, our source terms and our risks, so 25 our final conclusion is that we meet all safety goals with
() ANN RILEY & ASSOCIATES, LTD.
Court Reporters 1250 I Street, N.W., Suite 300 Washington, D.C. 20005 (202) 842-0034
I 138 1 margin.
() 2 3
DR. KRESS: You could meet those safety goals without having a containment at all, right?
l 4 Mk. SCOBEL: Yes.
l l 5 DR. KRESS: It was interesting to me that the core
! 6 damage freque ncy is about 10 to the minus 8 and the 7 additional containment failure probability is .1. I was i
8 just wondering about the balance there, if that was the 9 right -- if it is a depth balance that one likes to see.
l 10 Presumably you need a containment for this l
11 reactor, right?
12 DR. CARROLL: You don't need it for most reactors.
13 DR. CATTON: That's right. It's defense-in-depth.
14 DR. KRESS: If you want to make 10 to the minus 6, t
g i
g ,) 15 you do. You meet it for practically all of them.
16 DR. CARROLL: I don't know. I think if you do
'7
. your PRA's best estimate you probably don't.
18 DR. KRESS: All the PRAs I have seen says you do, 19 to get a 10 to the minus 6 -- here you don't. They have got 20 the core damage frequency way down, but you need a 21 containment for defense-in-depth.
I 22 MR. SCOBEL: We are providing defense-in-depth, 23 significantly.
24 DR. POWERS: But I think the question that is 25 legitimately raised ia that you do a calculation and you r
O ANN RILEY & ASSOCIATES, LTD.
Court Reporters 1250 I Street, N.W., Suite 300 l Washington, D.C. 20005 L (202) 842-0034
139 1 say, see, that core damage frequency is very, very low for a
() 2 3
plant that has never been built, there is no operational experience.
4 You are hanging your hat an awful lot on your 5 ability to calculate and to be predictive, which quite 6 frankly has a poor track record throughout history.
7 Analysts including myself and especially myself tend to be 8 overconfident in their ability to calculate things, so you 9 provide the containment as a ' defense-in-depth because 10 analysts, especially myself, are just very unreliable.
11 Now how much confidence do you really want to have 12 in your ability to protect yourself against people like 13 myself?
14 DR. CATTON: We probably have to wait until we 15 have had a few core melts.
16 You have to have an experimental database to judge 17 that.
18 DR. CARROLL: Might take awhile.
19 DR. POWERS: But I think Tom raises an interesting 20 question. Having acceded that the analysts are not 21 universally reliable, and that you want to have i
22 defense-in-depth, what is the appropriate mix to have that 23 ,
defense-in-depth, especially since you are going to have to 24 rely on another analyst to tell you what that CDF is --
25 CCDF -- the conditional containment failure probability.
[ ANN RILEY & ASSOCIATES, LTD.
\
Court Reporters 1250 I Street, N.W., Suite 300 ;
Washington, D.C. 20005 I (202) 842-0034
140 1 I think you raise an interesting and
()
A-2 thought-provoking question.
1 3 Assuredly, the .1 number is inconsistent with the 4 kinds of numbers we have looked -- hope to achieve in 5 existing reactors, but it is very consistent with what we 6 see out of the IPEs.
7 DR. KRESS: In fact, it's generally lower than a i
8 lot of the PWRs. !
9 DR. POWERS: It is certainly lower than the BWRs. i I
10 DR. KRESS: A lot lower than the BWRs. j 11 MR. SCOBEL: l In a paper that I wrote in which I ;
1 12 was promoting the ROAAM approach to evaluating l 13 defense-in-depth, the point of the paper basically is the ;
14 best thing is to rely on our engineered systems that allow C'
( )\ 15 you to not be subjected to phenomenological uncertainty and 16 to get the -- for example, primr s ry system depressurization, ,
17 allowing you to avoid induced steam generator tube ruptures. l 18 So when you are trying to evaluate all the 19 phenomena associated with induced steam generator tube 20 ruptures you could go crazy. You come out with a 21 probability number in the end that really doesn't mean 22 anything when really what you are trying to do is make sure l 23 that you don't have induced tube ruptures.
24 If you use the integrated ROAAM approach to force 25 the frequency of the induced tube rupture event below a 1
(Q j ANN RILEY & ASSOCIATES, LTD.
Court Reporters j
1250 I Street, N.W., Suite 300 l Washington, D.C. 20005 i (202) 842-0034 l
l
141 1 certain screening frequency that you have established for
() 2 your plant, then you don't have to worry about that 3 phenomenological uncertainty that you can't calculate very 4 well and that you are not reliable at.
5 DR. CATTON: I wish you wouldn't give the ROAAM 6 approach all the credit for that view. The Aerospace Vistas 1
i 7 did it from time zero and so do others. That's not ROAAM.
8 DR. KRESS: That's a good way to identify it.
9 though. If you say ROAAM we know what you mean.
10 MR. SCOBEL: It's the name that I am familiar with 11 that I am giving you.
12 CHAIRMAN BARTON: Any other questions for him?
13 [No response.]
14 CHAIRMAN BARTON: If not, we will break till 2:45
() 15 and then we will hear about severe accidents.
16 [ Recess.]
17 CHAIRMAN BARTON: Jim, are you ready to give your 18 final shot here?
19 .All right, Jim, you can proceed. Dr. Catton is 20 back here with us.
21 DR. CATTON: Waiting for me?
22 CHAIRMAN BARTON: Of course. He did this just for 23 you -- he is going to do it just for you.
24 DR. CATTON: Okay.
25 MR. SCOBEL: I don't have copies of these slides l
l
\1 O ANN RILEY & ASSOCIATES, LTD.
Court Reporters 1250 I Street, N.W., Suite 300 Washington, D.C. 20005 (202) 842-0034 i
=
142 l l
1 but Mr. Dudley took my paper copies and he is going to run some.
( 2 3 I made it, so I might as well put it up. This is 4 in-vessel retention of molten core debris. I always say IVR 5 and the NRC always says EVRC, and mine has less letters so I l
l 6 am going to use IVR.
I 7 You have seen this slide before 1 want to run 8 through this again. This is my 9 uniquely-suited-for-in-vessel-retention slide. I i
10 These factors are very important in the success of j 11 IVR and in fact I left one off that I probably should have l
12 put on. Like I said, the reactor coolant system is fully '
13 depressurized, which reduces the loads that the vessel has 14 to carry when it is in this state with molten core in the !
i 15 lower head.
16 The debris has a low power density which reduces l
17 the heat fluxes to the outside world. It is an oversize l
18 reactor vessel so you have more surface area per megawatt to 19 get heat out through the lower head and the reactor vessel i
20 is actually designed to be flooded on the outside. ;
21 There are no lower head penetrations to add a 22 vessel failure mode and the containment layout is designed 23 to flood the reactor cavity around the vessel even in design 24 basis accidents.
1 1
25 The IRWST provides a huge water source in the ANN RILEY & ASSOCIATES, LTD.
Court Reporters 1 1250 I Street, N.W., Suite 300 {'
Washington, D.C. 20005 (202) 842-0034
143 1 containment that can be drained into the lower part of the
() 2 3
containment to flood the vessel and we have this reflective insulation that allows the water in.
4 The one thing'I left off of here also is that the 5 in-vessel configuration of the core is very different than 6 conventional reactors. We have a reflector., a very, very 7 massive piece of stainless steel that the core sits inside.
8 The purpose is to reflect neutrons back into the core and 9 give it a flatter axial power profile, but this reflector 10 also creates a very strong thermal shield against the 11 relocation of core debris from inside the core region to the 12 lower head, and that comes into play strongly in the debris 13 configurations.
14 DR. SHACK: Just how massive is that reflector?
) 15 MR. SCOBEL: It's 40,000 kilograms.
16 DR. CATTON: How thick is it?
i 17 MR. SCOBEL: I have a slide -- it says it's about 18 13 centimeters thick.
19 DR. MILLER: That will solve all your problem with 20 vessels -- embrittlement?
21 MR. SCOBEL: Exactly.
2:2 DR. MILLER: No embrittlement problems.
23 MR. SCOBEL: That was one of the factors that was 24 taken into account in the reactor vessel failure initiating 25 event.
1
- ~ \ ANN RILEY & ASSOCIATES, LTD.
i I k-/ Court Reporters 1250 I Street, N.W., Suite 300 Washington, D.C.-20005 (202) 842-0034
'L44 1 DR. FONTANA: You are not likely to get a melt 2- path like TMI t o --
3 MR. SCOBEL: .That is part of my presentation.
4 DR. FONTANA: Oh, _sorry.
l
- 5. [ Laughter.)
6 DR. KRESS: How does that influence the core 7 relocation?
8 MR. SCOBEL: Strongly.
9 DR. KRESS: Is it,a heat sink?
10 MR. SCOBEL: It'is a heat. sink, yes -- it is a 11 heat sink and it is a barrier to sideward relocation.
12 DR. KRESS: See,.what-I am picturing is the heat 13 sink makes it take longer, but the core goes ahead and
-14 relocates eventually.
( 15 MR. SCOBEL: That's. correct.
16 DR. KRESS: And so it's just a little bit 17 longer --
18 MR. PIPLICA: That's correct, but it changes some 19' things with respect to our thinking, like for the TMI
- 20. relocation.
, 21 DR. KRESS: Oh, yes, it would change that l
22 thinking, I would imagine.
23 MR. SCOBEL: Exactly --
- 24. DR. CATTON: It would make the downward 25 penetration race.a little bit easier to win.
O ANN RILEY & ASSOCIATES, LTD.
Court Reporters 1250 I Street, N.W., Suite 300 Washington, D.C. 20005 (202) 842-0034 1
'145 1 MR'. SCOBEL: That may be true, but that is not
() .2-3 whatLends up happening.
Let me look at these slides. We can probably skip 4 . ahead to this. You probably know everything up to now.
5 Okay -- that is where we are skipping to.
6 Part of the IVR ROAAM and also not just the IVR 7 ROAAM but the in-vessel steam explosion ROAAM is'an 8 engineering evaluation of the relocation scenario.
9 This was-a rather detailed. study that was 10 performed by Argonne National Labs and.by UCSB together over 11 a long period of time. It was a very iterative process in 12 trying to determine how the core would get from the intact 13 configuration into the lower head and what the properties 14 'would be.
() 15 The important aspects to this were the formation 16- of a lower core blockage which blocks downward relocation, 17 the reflector -- which is this structure, like the former 18 but it is very, very' thick and very massive, the core 19 barrel, which is the outer ring that forms the boundary 20 between the in-core and the downcomer of the reactor vessel, 21 the formation of a molten pool in the upper corefregion and 22 the eventual failure through the reflector and a sideward 123 relocation from the downcomer into the lower head.
24 DR. POWERS: Why is it that we are going to have a 25 sideward relocation?
- /) %-
ANN RILEY & ASSOCIATES, LTD.
Court Reporters
.1250 I Street, N.W., Suite 300 Washington, D.C. 20005 (202) 842-0034
146 1 MR. SCOBEL: Because in the melt progression
()
' 2 analysis what they determined is that us you melt you agree 3 .that the metals in the core melt out first and the melting i
4 of the top of the core begins when the water level is still 5 up in the core region because you have the sort of oxidation 6 going on, and so the metals from above, particularly the 7 zirc and the control rods and gray rods, are dripping down 8 into the lower regions of the core and from the bottom of 9 the active core region inside the fuel rods there are solid 10 zirc plugs that fill up this bottom part above the core, 11 above the core support plate.
12 So what you eventually have is a very, very 13 massive heat sink that includes the core support plate and 14 the zirc plugs and the nozzles of the fuel assemblies which
() 15 freeze this metal into a solid mass at the bottom.
16 DR. POWERS: It sounds to me like somebody has 17 formulated a core degradation scenario based on looking at 18 an awful lot of tests with un-irradiated fuel, and once you i 19 start irradiating the fuel, won't you find a substantial 20 interaction that those metals with the fuel such that it
! 21 does not come dripping down there?
22 MR. SCOBEL: That the metal will remain in place? )
23 DR. POWERS: To remain in place it goes into 24 solution in the. fuel. l 25 DR. CATTON: There are as many scenarios like the i
- ) ANN RILEY & ASSOCIATES, LTD.
/' Court Reporters l 1250 I Street, N.W., Suite 300 Washington, D.C. 20005 (202) 842-0034
147 _
i 1 one you described as there are people who have looked at the
() 2 3
problem.
MR. SCOBEL: I haven't finished describing the 4 scenario yet'but --
5 DR. CATTON: The initial stages of the scenario -- l
! 6 it certainly isn't a scenario 1.nat describes Three Mile 7 Island.
8 MR. SCOBEL: Like I said, this isn't exactly like 9 Three Mile Island.
I i
10 DR. CATTON: It is? j 11 MR. SCOBEL: It's not.
12 DR. CATTON: Oh ---I didn't think so.
13 MR. SCOBEL: But I don't know the answer to your i 14 question other than to say that this relocation scenario was
() 15 part of the ROAAM peer review and this particular aspect 16 didn't really receive much in the way of questions. ;
i 17 DR. POWERS: Maybe that is a function of the I l
18 reviewers. I don't know. What I do know is that an awful 19 lot of our models of core degradation have been based on 20 experiments done with un-irradiated fuel, and what we find
)
21 indeed is there tends to be a lot of slippage of the metals 22 separating from the fuel when we have un-irradiated fuel.
23 As soon as we start going to irradiated fuel, we 24 find very different things occurring, and particularly we 25 find a. pronounced ~ tendency for you to get what is called O ANN RILEY & ASSOCIATES, LTD.
Court Reporters 1250 I Street, N.W., Suite 300 Washington, D.C. 20005 (202) 842-0034 l
148 1: fuel dissolution.into these metals, which means you are 2 forming a solution with them.
3 MR. SCOBEL: Okay, but I don't think that would 4 stop.them from refreezing at the bottom.
5 DR. POWERS: Well, what happens is that they 6 refreeze up in the core region.
7 MR. SCOBEL: But they would come down and 8 eventually refreeze at the tottom. f 9 DR. POWERS: Well, by the time they come down it 10 covers a large collection of molten material at that point.
-11 DR. CATTON: Or you could oxidize the clad a. bit 12 before, and you wind up with a huge pile of rubble before 13 any melting starts.
14 DR. POWERS: It seems to me that a much more 15 plausible scenario is in fact you get melting et slightly (
16 above the mid-plane and then you get collapse of everything 17 else, and to a rubble pile and that kind of heats 18 homogeneously.
19 DR. CATTON: And it doesn't take really high 20 temperatures to do that to zircalloy.
21 MR. SCOBEL: Okay.
22 DR. CATTON: Sounds like there's two parts to this
'23 story. The first part was with the in-vessel steam 24 explosion report. Is that correct?
l .25 MR. SCOBEL: The scenarios discussed in both the i
O ANN RILEY & ASSOCIATES, LTD.
Court Reporters 1250 I Street, N.W., Suite 300
{
Washington, D.C. 20005 (202) 842-0034 l
149 1 in-vessel retention and the in-vessel stream explosion
() 2 report and in fact Chapter 4 of the steam explosion report
! 3 is dedicated to this and I think is the better of the twa 4 write-ups.
5 DR. CATTON: I don't have that report. Noel, 6 could yc_ get me that report?
7 I just believed the in-vessel steam explosion 8 conclusion so I didn't bother to read the report.
9 DR. POWERS: That's because you have religion on 10 that subject, right, Ivan?
11 DR. CATTON: That's right.
12 See, once you make a pronouncement, you are stuck 13 with it.
14 [ Laughter. ]
( 15 DR. KRESS: You prescribed a scenario so far. You 16 haven't gone all the way yet. What is the purpose of the j 17 scenario? Are you going to end up with a given amount of ,
18 melt or --
19 MR. SCOBEL: To determine -- yes. To determine 20 what relocates into the lower head.
21 DR. KRESS: How much, and what, and when.
22 MR. SCOBEL: That's right.
23 DR. KRESS: And then that impacts on whether or 24 not the vessel is going to fail or something?
25 MR. SCOBEL: Well, it impacts on how you analyze i ANN RILEY & ASSOCIATES, LTD. l Court Reporters 1250 I Street, N.W., Suite 300 Washington, D.C. 20005 (202) 842-0034
150 1 it.
,-m) 2 DR. KRESS: How you analyze it.
(J l 3 MR. SCOBEL: Yes. l l
4 DR. KRESS: So you are setting the stage for the 5 analysis. Does that analysis configuration that you finally i 6 end up with depend heavily on this scenario or are there 7 many ways to get down there that end up in the same final 8 configuration?
9 MR. SCOBEL: I would say yes, it would end up in 10 the final configuration and the NRC and INEL would say that 11 there are several configurations along the way before it 12 ended up in the final configuration --
13 DR. KRESS: That need to be looked at as a .
14 transition period?
,~
k_ 15 MR. SCOBEL: That's correct.
16 DR. POWERS: I note that this report, which I take 17 it is the basis for this, was published in 1993, and --
18 MR. SCOBEL: Which version are you looking at?
19 DR POWERS: I think the version you gave us.
20 MR. SCOBEL: The final version came out in 21 October, '96.
22 DR. CATTON: Noel, I have the report.
23 DR. POWERS: About two years ago, which would have 24 been '95, so there was an opportunity for them to react to 25 it apparently, NRC hosted a session in which people from
[~ ANN RILEY & ASSOCIATES, LTD.
\- ) Court Reporters 1250 I Street, N.W., Suite 300 Washington, D.C. 20005 (202) 842-0034
j 151 1 France described some in-vessel experiments involving core l
() 2 3
degradation and the inability of all the codes, and there were a number of them, but I don't believe MAAP was one of 4 them that was included in there, inability of those codes to 5 model that core degradation.
6 It appeared that the difficult that had.been 7 encountered was there was more fuel melting in relocation 8 downward than what the codes predicted, substantially more.
9 The codes predicted one or two percent and.it was much more 10 like 20 or 30 percent and eventually 50 percent.
11 Was that factored into this?
12 MR. SCOBEL: Well, this analysis was not performed 13 with codes. This was a first principles analysis from 14 beginning to end of the relocation scenarios.
/~
's,),) 15 They looked at -- do you krow Jim Senicki by any 16 chance?
17 DR. POWERS: Yes, I do.
18 MR. SCOBEL: Senicki was heavily involved in this.
19 They looked at the melt relocation as the core became 20 uncovered and the water level was down and you had zirc 21 water reaction going on and the formation of a pool, and the 22 heat transfer that would be occurring to the side wall, and 23 an estimate of the timing for the failure of the wall, and 24 then an estimate of the relocation phenomena that would 25 ~ occur when the reflector failed.
t ANN RILEY & ASSOCIATES, LTD.
Court Reporters 1250 I Street, N.W., Suite 300 Washington, D.C. 20005 (202) 842-0034
7---.___
152 1 DR. POWERS: I guess I am still puzzled why that I
() 2 is different. All you are telling me is the particular 3 machine this model ran on involved a substantial amount of
! 4 wet-ware.
! 5 MR. SCOBEL: Yes -- just from my experience based 6 on you mentioned MAAP, you know, just based on my experience 7 I don't know SCDAP, RELAP or MELCOR very well but -- and it 8 is acknowledged in the ROAAM report -- MAAP can't model this 9 at all. There's no way that you could ever get MAAP to 10 model this scenario --
11 DR. FONTANA: I'm not sure you can with anything.
12 MR. SCOBEL: 2xactly.
13 DR, FONTANA: It so difficult and so many 14 different things going on. I think the approach you are m
15 taking, from what I hear, sounds pretty good. Maybe we 16 ought to let him go on for awhile and then --
l 17 DR. POWERS: Well, Mario, I guess the question is )
18 whether we care about what the experimental evidence is I
19 saying or not. j 1
20 DR. FONTANA: If that's the question, I wonder I 21 what you are raising here, because you are saying, as I 22 think I heard, is that fuel that is irradiated, the fuel in 23 a metal holds more together because there is more 24 interaction between them, you are saying, than in a I
25 situation where the metal runs off into the fuel -- is that
[~
N-ANN RILEY & ASSOCIATES, LTD.
Court Reporters 1250 I Street, N.W., Suite 300 Washington, D.C. 20005 (202) 842-0034
_ _ _ _ _ _ _ _ _______-_-__-________a
153 1 correct?
2 DR. POWERS:
{f That's certainly one thing.
3 DR. FONTANA: Yes. Now what did you do?
4 MR. SCOBEL: The metal was the first to melt 5 out -- you know, the things like control rods and whatnot, 6 and eventually ended up as a. frozen mass on the core support 7 plate and above.
8 DR. FONTANA: That's the control rods, but how 9 about the cladding?
10 MR. SCOBEL: The' cladding also would have 11 participated in that.
12 DR. CATTON: -- that significantly changes 13 everything else.
14 DR. FONTANA: I understand.
) 15 I would say let's go on and then kind of back up 16 and see what the effects -- DR. POWERS: I don't know.
17 Why is it useful to go on if we are going to ignore all the 18 materials interaction research that has ever been done on 19 clad and fuel at high temperatures?
20 DR. FONTANA: No, I wasn't going to ignore 21 anything. I was just saying what is the rest of his story 22 and then see how the issues that you are raising affect 23 that. Because I'think what you are going to say is that the
]
i 24 material comes down and creates a blockage before it runs l 25 through all the -- l l
i O ANN RILEY & ASSOCIATES, LTD.
Court Reporters 1250 I Street, N.W., Suite 300
]
i Washington, D.C. 20005 (202) 842-0034
154 1 MR. SCOBEL: That's correct.
2 DR. FONTANA: That's correct. And then so the 3 question is, if it comes down more, the fuel and the metal 4 are more coherent with each other, how much does that change 5 your basic model that the grid plate is going to get blocked 6 and go down all at once, is I think what you going to end up 7 with.
8 MR. SCOBEL: Well, no, what --
9 DR. FONTANA: No?
10 MR. SCOBEL: Well, yes and no. What ends up 11 happening is you form the blockage at the bottom and above 12 there you form a large circulating in-core molten pool that 13 --
14 DR. FONTANA: In your scenario, the above one is
() 15 largely molten ceramic?
16 MR. SCOBEL: That's correct.
17 DR. FONTANA: Okay. So one of the differences is 18 that it is more likely, more of the uranium -- more of the 19 uranium is likely to be in with the metal according to 20 Dana's perception. Is that going to change things much?
21 MR. SCOBEL: I --
22 DR. FONTANA: It's your modeling.
23 MR. SCOBEL: Let's find out as we go on. Think 24 about it, because it's -- you know, I am being presented 25 with this for the first time.
O ANN RILEY & ASSOCIATES, LTD.
Court Reporters
'250 I Street, N.W., Suite 300 Washington, D.C. 20005 (202) 842-0034
155 1 DR. POWERS: I guess I am --
2 DR. FONTANA: I am not saying ignore it.
3 DR. POWERS: I am going to have to really --
4 Mario, I have the right to pursue these things.
5 DR. FONTANA: Absolutely. I am not saying --
6 DR. POWERS: I have to understand why it is that 7 we have done an analysis that ignores all this materials 8 interaction phenomena. We have known that metallic 9 zirconium interacts with uranium dioxide, induces 10 liquefaction for a very long time. I don't know that it is 11 20 years, but maybe 15.
12 MR. SCOBEL: Well, the analysis includes the 13 dissolution of 5 percent uranium into the zirc that melts 14 -down to the core support plate. That's the materials k/.O 15 interaction that you are referring to. I don't know if you 16 find that satisfactory or not.
17 DR. POWERS: Well, I mean I see no reason, no 18 experimental basis to limit it to 5 percent.
19 MR. SCOBEL: Okay. The thing that I can say is 20 you direct your comments to Theofanous directly, you could 21 ask that they be included in the ROAAM report, you know, 22 addressed.
- 23. DR. POWERS: My experience is that directing L 24 comments to Theofanous only results in getting a review on 25 my ancestry and sexual habits.
1 ANN RILEY & ASSOCIATES, LTD.
Court Reporters 1250 I Street, N.W., Suite 300 Washington, D.C. 20005 (202) 842-0034
156 1 [ Laughter.]
() 2 DR. POWERS: Both of which I am intimately 3 familiar with.
4 MR. SCOBEL: So, the final result is that you have 5 a breakthrough eventually sideways through the reflector. I 6 should also state that the molten circulating pool is 7 basically a pool of molten uranium oxide with a steel that 8 is melted down from the reflector above on top. When you 9 break through here you have -- the break occurs at the top 10 of the oxide pool. The metal from the top fills up the gap 11 between the reflector and the core barrel, and then the 12 melting continues and you have a break of the core barrel 13 and you have the oxide and the metal from the top that is 14 remaining. The oxide first, metal comes down into the lower O 15 head.
16 DR. FONTANA: Okay. So now backing up to bring i
17 into consideration the points that Dana raised, you would l
18 have more oxide in with the metal, plus oxide by itself.
19 Will that change the scenario much in your opinion? I know 20 I am hitting you with a new question here.
21 MR. SCOBEL: Actually, I don't think so. That --
22 DR. FONTANA: I don't think so either.
23 MR. SCOBEL: I don't think it changes anything. I 24 am wondering what his effect is on -- what he thinks the 25 effect is on the lower blockage.
l
() ANN RILEY & ASSOCIATES, LTD.
Court Reporters 1250 I Street, N.W., Suite 300 Washington, D.C. 20005 (202) 842-0034 l
157 1 DR. KRESS: My problem is, if my answer depends r^s)
%.)
2 strongly on this scenario, I am really concerned. If the 3 final answer doesn't depend on this scenario, and that you 4 end up with a robust analysis that could get there anyway, 5 then I am not so concerned. So I have yet to see where the 6 scenario --
7 MR. SCOBEL: The only thing that the answer 8 depends on, based on the configurations proposed by INEL and l
9 the NRC, is the amount of metal that comes down in the first 10 relocation. I 11 DR. SEALE: Metal, what kind of metal?
12 MR. SCOBEL: Zirc and stainless steel.
13 DR. SEALE: In the first relocation, --
14 That's part of Dana's problem.
0 15 DR. FONTANA: Pardon?
16 DR. SEALE: I think that's part of Dana's problem.
17 DR. KRESS: Is it in the first relocation because 18 there is a transient problem with that melt --
19 MR. SCOBEL: No, it's actually part of a steady l
20 state problem. If you remember from the ROAAM Report, metal 21 layer, the thickness of the metal layer can determine how l 22 focused the heat flux is that is coming from the top of the 23 oxide layer.
l 24 DR. KRESS: So all this scenario is to determine 25 the thickness of that metal layer?
/
( ANN RILEY & ASSOCIATES, LTD.
\'
Court Reporters 1250 I Street, N.W., Suite 300 Washington, D.C. 20005 (202) 842-0034
158 1 MR. SCOBEL: That's correct.
() 2 DR. KRESS: Okay. So that's what we should be 3 thinking about. How does this scenario and possible 4 alternative scenarios affect the thickness?
5 MR. SCOBEL: And the location of a high-density 6 zirconium layer.
7 DR. FONTANA: It seems it would make the metal 8 layer a little bit thicker and have a higher generation rate 9 in it, so even more so it would go through the sides. I 10 think.
11 DR. KRESS: No , he is talking about after it gets 12 down there. The thickness of the layer after --
13 DR. FONTANA: Oh, by the way, is there still water 14 in the bottom head when this happens?
15 MR. SCOBEL: Yes.
16 DR. FONTANA: For a while.
17 MR. SCOBEL: Well, for a while. That's right.
18 DR. KRESS: Kind of give me an idea of how much 19 metal is in that final layer on top compared to the amount 20 of metal that might possibly be there if --
21 MR. SCOBEL: Well, in the final configuration, 22 it's the entire reflector and the lower core support plate, 23 and the nozzles from the -- and everything from the fuel 24 assemblies, because --
25 DR. KRESS: Zirconium in there?
ANN RILEY & ASSOCIATES, LTD.
Court Reporters 1250 I Street, N.W., Suite 300 Washington, D.C. 20005 (202) 842-0034
159 1 MR. SCOBEL: Sorry?
( f 2 DR. KRESS: Zirconium is in there?
3 MR. SCOBEL: Yes.
4 DR. KRESS: So basically it is a lot of metal.
5 MR. SCOBEL: It is a lot of metal. It's on the ;
6 order of 70,000 kilograms.
7 DR. KRESS: Compared to how much is in the active 8 core' region? i 9 MR. SCOBEL: There's no metal in the active --
10 well, there's zirc. But then you have to determine how much 11 of it oxidized.
12 DR. KRESS: Well, I am counting the core barrier, t 13 around the --
14 MR. SCOBEL: Oh, okay. Well, a lot of it is in 15 this active core region. Sixty -- or 40,000 kilograms is in 16 the reflector.
17 DR. CATTON: 40,000 kilograms of what?
18 MR. SCOBEL: Of stainless steel.
19 DR. CATTON: Now, the zirc alloy, when it' melts --
20- MR. SCOBEL: Yes.
21 DR. CATTON: Is the assumption made that it melts, !
22 runs down, then the fuel comes apart?
l 23 MR. SCOBEL: No, that doesn't -- I mean it is not j 24 like the fuel stands there stacked or'anything, that doesn't 25 matter to this.
() ANN RILEY & ASSOCIATES, LTD.
Court Reporters l 1250 I Street, N.W., Suite 300
- Washington, D.C. 20005 j (202) 842-0034
160 1 DR. CATTON: I think it matters with respect to 2 where the zirc alloy finally winds up. I mean if it all 3 falls down in this pile of molten zirc alloy and fuel 4 pellets, when it starts to heat up, the zirc may be in a 5 different location than the assumption that you are making.
6 And I guess the question is, how much of it gets into the 7 bottom and does it matter? And maybe you will hear that it 8 does matter in a little bit.
9 DR. FONTANA: I guess the bit question is does it 10 plug up that core support plate rather than flowing through?
11 Right.
12 DR. CATTON: I think if -- I mean zirc alloy is a 13 fairly high thermal conductivity. If it is molten and it 14 tries to flow through some of those passages, it probably 15 isn't going to make it. But I don't know how big the 16 passages are. It may even drip through.
17 DR. KRESS: I think the issue is you are going to 19 end up with some sort of configuration that has a certain 19 amount of fission product bearing molten UO2 and a certain 20 amount of metallic in there, which may have some fission 21 products in it, and you are going to analyze what that does 22 to the vessel. And the relative amounts of those things 23 down there will affect the answer. And somehow one has to 24 arrive at a guess for the relative amot'.nts. You can't put 25 everything up there down. You can't put none up there down.
() n.NN RILEY & ASSOCIATES, LTD.
Court Reporters 1250 I Street, N.W., Suite 300 Washington, D.C. 20005 (202) 842-0034
1 161 1 And you have to decide some way how much of each to put down l
() 2 there and where to put them. And I au not so sure I am
~3 comfortable with what I am hearing so far as a way to arrive 4 at a good guess as to the final configuration of what is 5 down there. )
6 DR. FONTANA: Well, in your model, won't the whole 7 shebang come down eventually?
8 MR. SCOBEL: Yes. Because the bottom -- the 9 support plate contacts the melt in the lower head. So the l l
10 support plate will melt and fail into -- and become part of '
11 the pool, and the reflector sits on the support plate. The 12 core barrel itself hangs from the top. So --
13 DR. CATTON: So you wind up with the whole mess in 14 there anyway.
t (m) 15 MR. SCOBEL: Yes.
16 DR. CATTON: So what does the scenario matter?
17 MR. SCOBEL: What it --
18 DR. CATTON: For the steam explosion it does.
I 19 Because it is the rate it gets there that matters.
1 20 MR. SCOBEL: Well, that's why it is much -- it is 21 addressed in much more detail in the steam explosion report l 22 than it is in the IVR Report. But what -- really, the point
, 23 that I am trying to make with respect to this is with 1
l 24 respect to the configurations that have been proposed that 25 failed the reactor vessel.
ANN RILEY & ASSOCIATES, LTD.
Court Reporters 1250 I Street, N.W., Suite 300 Washington, D.C. 20005 (202) 842-0034 1
162 1 DR. CATTON: I don't understand that.
() 2 MR. SCOBEL: Okay. INEL wrote a report about 3 evaluating the IVR ROAAM Report.
4 DR. CATTON: Yes, I read it.
5 MR. SCOBEL: Okay. And they proposed three 6 configurations.
7 DR. CATTON: Yes.
8 MR. SCOBEL: Which could -- which were not bounded 9 by the IVR Report. And basically this later on serves to 10 show that what they proposed, which was all the metal in 11 their melt was this -- this metal here, this small structure 12 that -- e.nd plus some of the control rod material. And when
\
13 you only include that much metal, you create a really, 14 really thin metal layer, and in the one-dimensional heat 15 transfer model, you end up with a large heat flux.
16 They are not including any metal that would come 17 down in the initial relocation from the reflector or the 18 core barrel, which is significant, and that was where I was 19 headed to.
20 DR. CATTON: A lot of speculation, on everybody's 21 part.
22 DR. SEALE: But in those models, isn't all the 23 uranium in the form of UO2?
24 MR. SCOBEL: Yes.
25 DR. SEALE: That's part of the question.
() ANN RILEY & ASSOCIATES, LTD.
Court Reporters 1250 I Street, N.W., Suite 300 Washington, D.C. 20005 (202) 842-0034
._ - - - - - - . . - _ . , - , - - _ - - - - - - - - - - , - - - - - - - - - - - - - - - - - - ' - ' - - - - ' - ' - - - - ' - - ' - - - - - - - - ' " - - " ' - " " - - ' ' ' ' ' - " - - - - - - " - " ' - - - ~ - - - - ' - ~ - - - - - - ' - - - ' - - - - - - - ' '
163 1- MR. SCOBEL: Well, it is -- in the 'RO1UU4 analysis
() 2 they look at decay heat in the metal layer. In the INEL 3 analysis, their base case has 13 percent of the decay heat 4 in the metal ~ layer. Do you --
5 DR. SEALE: I don't remember.
6 MR. SCOBEL: I think it's 13 percent.
7 DR. FONTANA: Well, they have -- one of the cases 8 I looked at, which I think was Case C, where the metal, some 9 metal is in the lower -- l 1
10 MR. SCOBEL: At the bottom, yes. :
11 DR. FONTANA: At the bottom. Which is the i
- 12. assumption where it dissolves -- '
13 MR. SCOBEL: Yes.
14 DR. SHACK: Let me just get this -- so you are
() 15 ! arguing then that the sequence doesn't really make any i 16 difference whether you get this molten pool and it breaks l 17 through on the side, and so you basically have a large slug 18 of molten _ fuel coming down first. You are saying you don't I 19 really care about that, what you really care about is i 20 whether the pool at the bottom consists of most of the metal 21 'up there and the fuel versus the INEL assumption that there i 22 is.very little metal. -
23 MR. SCOBEL: It is not even most of the metal.
24 It's just --
25 DR. SHACK: A large chunk.
ANN RILEY & ASSOCIATES, LTD.
Court Reporters 1250 I Street, N.W., Suite 300 Washington, D.C. 20005 (202) 842-0034
164 1 MR. SCOBEL: A lot of metal, that's right.
()
2 DR. SHACK: So the rest of it is all kind of 3 detail?
4 MR. SCOBEL: That's correct.
5 Dr. Fontana, you were asking a question about 6 Configuration C.
7 DR. FONTANA: Oh, I was saying that was one of the 8 configurations that is this report which has metal on the 9 bottom and ceramic on top.
10 MR. SCOBEL: That's correct.
11 DR. FONTANA: Which means it is a situation where 12 the fuel had to interaction with the metal to get denser, 13 didn't it?
14 MR. SCOBEL: That's right. The fuel -- it 15 basically has to be a very, very -- it has to be mostly zirc 16 and uranium and not a whole lot of steel to reach that kind 17 of density configuration.
18 DR. CATTON: So what is the zirc uranium metal 19 ratio?
i 20 MR. SCOBEL: In that lower?
21 MR. PALLA: This is Bob Palla with the staff.
22 What we had done as part of our review, and I can talk to 23 this a little later in my presentation, but we basically 24 looked at the peer review comments on the Theofanous report 25 and several of those reviewers, the peer reviewers, ANN RILEY & ASSOCIATES, LTD.
Court Reporters 1250 I Street, N.W., Suite 300 Washington, D.C. 20005 (202) 842-0034
l l
165 i 1 suggested that these interactions could occur, that a
) 2' .certain amount of uranium could be dissolved into the metal 3 layers carrying with it a lot of density which would sink 4 this molten pool from the top, the metal layer from the top 5 to the bottom.
6 And what we did is looked at those comments in the 7 support literature and basically take by assumption.the 8 fraction necessary-to make the layer heavier. And it was 9 not a large fraction, I want to say 5 percent, 5 atom 10- percent uranium was sufficient to cause the metal layer to 11 sink. And we carried with that metal layer a fraction of 12 .the decay heat which was, because of fission product source 13 term behavior, thought to be bound up in the tretals, plus, 14 you know, any decay heat from the uranium, that went with i 15 the metal layer to the bottom. And then a steady state 16 calculation was performed of the heat loads for that 17 configuration.
18 DR. SEALE: If you have excess uranium though, so 19 that it is even more dense -- if you have excess uranium, if l 20 you will, in this melange of material that sinks, so that it 21 is even more dense, and it comes in contact now with the 22 steel at the bottom of the vessel, what about the 23 interaction of that material with the steel? Is there any 24 special kind of chemical or metallurgical process going on 25 that might alloy and enhance the erosion rate that is
-( ANN RILEY & ASSOCIATES, LTD.
Court Reporters 1250 I Street, N.W., Suite 300 Washington, D.C. 20005 (202) 842-0034
t l 166 I
I 1 bringing the metal -- the steel into solution with the rest
/
(%) 2 of the material, thinning the structural residual layer of 3 steel in the process?
4 MR. PALLA: Well, yes, we think that there could 5 be a potential if you do get that kind of contact. We did 6 not look at that as part of the thermal analysis. We did 7 simply just the thermal analysis with this bottom metal 8 layer as '.ne boundary condition. There could be chemical 9 interactions. There was some concern about that that was 10 identified.
11 What we did with that Theofanous report, we looked 12 at it internally in the Office of Research, people familiar 13 with the RASPLAV program and work that we are doing on lower 14 head failure, Ali Behbhani, in particular, and several O
\_ / 15 others looked at this work and identified that chemical 16 interactions with the vessel are something that are not 17 fully understood. And their thinking was that the potential 18 for such interactions is probably going to be addressed as 19 part of RASPLAV. It is either separate effects testing or 20 RASPLAV testing. So our thinking is that that is the 21 appropriate place to lay these kinds of issues to rest.
l 22 DR. CATTON: Yeah, but AP600 is on the table.
23 What do'you do in the interim?
24 MR. PALLA: Well, we can -- I can talk to that 25 later. There's a number of arguments that we brought to l
l
() ANN RILEY & ASSOCIATES, LTD.
Court Reporters 1250 I Street, N.W., Suite 300 Washington, D.C. 20005 (202) 842-0034 s
167 I 1 bear on this, both probabalistic as well as the
() 2 deterministic calculations of ex-vessel loads. We {
L 3 considered all those things in making a determination as to l l l 4 whether we thought that this problem was adequately handled i 5 in the PRA. I 6 DR. KRESS: Did you need to invoke anything other 7 than thermal analyses for the vessel to fail in that 8 configuration?
l i 9 MR. PALLA: No, we did not.
10 DR. KRESS: The vessel failed anyway?
11 MR. PALLA: The vessel failed, in our calculation.
1 12 Now, it has -- there are limitations on our calculation.
i 13 But the critical heat flux was -- I believe for that l
14 Configuration C it was several times higher. I want to say O
(sl 15 about five times CHF. The CHF that we assumed in our model.
~!
l 16 Now, subsequent to our analysis, additional 17 testing was done at Penn State with insulation and, with the 18 conceptual insulation design, water is introduced at 19 essentially zero degrees, the bottom of the vessel. So 20 there is kind of like the most efficient cooling would be 21 occurring right at that lowest point, and the Penn State 22 tests show that the difference between tests with insulation 23 and without insulation, the greatest effect of the 24 insulation is on the CHF in the lower part.
25 DR. CATTON: It raises it.
('-) ANN RILEY & ASSOCIATES, LTD.
Court Reporters l
l J250 I Street, N.W., Suite 300 I Washington, D.C. 20005 j (202) 842-0034 '
l
- _ _ - _a
l j:
168 l' MR. PALLA: It-raises it. It raises it
'2 substantially.
3 DR. KRESS: It's the model's head that was l
4 completely --
5 MR. PALLA: It's a very small scale, but it 360 6 degrees, it the full, yes. Whereas Theofanous' test 7 configuration termed the OPU configuration is a slice in a 8 narrow channel basically. i 9 DR. KRESS: Penn State.
10 MR. PALLA: Penn State is the full hemisphere with l
11 insulation. Its real -- its only deficiency is really its j -12 scale and its inability to produce enough heat flux to reach
)
13 CHF. When they put the insulation on and ran the test, they l
-14 couldn't achieve CHF, it was outside the limits of their 15 ability to deliver heat with the heaters, it would have 16 burned up the heaters.
17 DR. KRESS: So what did you find out from the test l 1 18 if you couldn't reach CHF?
i 19 MR. PALLA: That CHF is greater than what they 20 were able to measure. They maxed-out. It is even higher.
21 DR. KRESS: So you put a bound on the CHF.
22 MR. PALLA: The CHF is something higher than what 23 they were able to measure. We don't know what it is, it is 24 something higher.
)
25 DR. CATTON: I heard you say five times. l ANN RILEY & ASSOCIATES, LTD.
Court Reporters 1250 I Street, N.W., Suite 300 Washington, D.C. 20005 (202) 842-0034
169 1 MR. PALLA: Oh. Well, our calculated CHF -- our
( 2 calculated ratio was -- we were measuring or predicting that 3
about five times CHF would be dumped out the lower bottom, 4 you know, bottom part of the vessel. What it looks like is 5 that -- and that's based on the -- I'll say the critical 6 heat flux without insulation. Now, if you took the tests 7 with insulation, you would have to adjust the CHF value 8 upward. That would reduce that ratio -- our ratio down. I 9
don't believe that the CHF ia going to be going up by a 10 factor of five by adding insulation, so it's very likely 11 that we would still predict something in excess of CHF at 12 the bottom.
13 DR. CATTON: It probably goes up right at the 14 stagnation point on the bottom, but I doubt that it's raised 15 very much higher up on the walls.
16 MR. PALLA: Well, okay. There is a report in 17 progress, I understand -- it's not out on the street yet 18 --that has that later data. The data was presented at the 19 CSARP meeting last week, and this is the first time I had 20 seen some of those details.
21 DR. CATTON: Do you have that paper?
22 MR. PALLA: I can -- we can probably check with 23 Research, and they might be able to -- if it's in a draft 24 enough form to --
25 DR. CATTON: Okay.
() ANN RILEY & ASSOCIATES, LTD.
Court Reporters 1250 I Street, N.W., Suite 300 Washington, D.C. 20005 (202) 842-0034
I
[ 170
'l MR. SCOBEL: Well, Bob Lutz has his presentation
() 2 of ex-vessel phenomena that has yet to be made.
3 DR. CATTON: We're now going to put the mess in 4 the bottom and somebody else is going to talk about it.
5 MR. SCOBEL: That's right.
6 DR. CATTON: And you're glad.
7 DR. KRESS: There was a question. The question 8 is, did you arrive at a probability of failure that you can 9- put into a PRA?
10 MR. SCOBEL: Based on ROAAM, it was zero.
11 DR. KRESS: Oh , okay. ROAAM rules it out.
I 12 MR. SCOBEL: That's right.
13 DR. KRESS: Or puts it below a level that you 14 don't care abr..ut.
l 15 MR. SCOBEL: That's correct.
16 DR. KRESS: So that was -- the ROAAM result is 17 that the probability is low and that you don't have to worry 18 about it.
l l
19 MR. SCOBEL: That's right.
20 DR. CARROLL: Provided.
21 DR. CATTON: We did similar calculations as part 22 of a look at vessel flooding for existing reactors, and we L
23 don't come out with zero, but there was no chimney.
24 DR. KRESS: There's no chimney in the pressure l 25 vessel --
I
()
ANN RILEY & AGSOCIATES, LTD.
Court Reporters l
! 1250 I Street, N.W., Suite 300 !
Washington, D.C. 20005 ;
(202) 842-0034 l
171 1 DR. CATTON: If the -- no, that's right, and the
() 2 heat load is different. There are a lot of differences.
3 DR. KRESS: We did the same thing.
4 DR. CATTON: I don't remember what the probability 5 was, but if the vessel was depressurized, it was pretty low.
6 But then we didn't consider things like the metal air or 7 anything else. We just took this fully mixed mess.
8 CHAIRMAN BARTON: Dana, is there any other thing 9 you want to pursue before you --
10 DR. POWERS: I think we have covered the central 11 issues.
12 CHAIRMAN BARTON: Okay.
13 DR. POWERS: I don't know how we address them. I 14 mean, there seems to be a controversy in the analyses here.
15 I mean, it seems to me --
16 DR. KRESS: I think I would like to see how the 17 probability was arrived at. You know, there has to be some 18 probability associated with different configurations, and 19 how those were assigned probabilities so that one could 20 decide the probability failure has escaped me at the moment.
21 So I might have -- I might have assigned a lot higher 22 probability to a configuration -- l 23 DR. POWERS: Well, it seems that we know something 24 about how the ROAAM process works, and they really come down 25 to the category I think they call physically unreasonable,
() ANN RILEY & ASSOCIATES, LTD.
Court Reporters 1250 I Street, N.W., Suite 300 Washington, D.C. 20005 (202) 842-0034
172 1 and it seems to me they have configured their core
() 2 . degradation scenario such that they arrived at a physical 3 unreasonable result. The only difficulty is that that core 4 degradation scenario seems not to be consistent with a wide 5 range of facts, experimental facts. The trouble is-we don't 6 have very many useful experiments that involve the link 7 scales of a reactor accident, and where we have done big 8 experiments, we've had to use unirradiated fuel. So now 9 you're -- you've got to link together small-scale 10 experiments with irradiated fuel to larger experiments with 11 unirradiated fuel to a reactor accident because they're all 12 big steps. It would seem prudent, then, to include a wide 13 range of scenarios, which does not seem to have been Cone in 14 this case.
15 Now, a couple of points are worth bearing in mind.
16 One is that if we have a large amount of steel such that we l l
17 can get the metal phase on the bottom, we probably do not j i
18 have an aggressive metallurgical interaction on the bottom.
l 19 On the other hand, if you have the metal phase on top, then l 20 you probably do have the conditions that lead to aggressive 21 metallurgical interactions, and so you can get the 22 possibility of sideways penetration, which are yet another 23 outcome.
1 24 I agree with you. Probabilities.are hard to i
25 assign when you haven't created a structure to allow you ANN RILEY & ASSOCIATES, LTD.
Court Reporters 1250 I Street,-N.W., Suite 300 :
Washington, D.C. 20005 (202) 842-0034
l l 173 i some flexibility in the configurations here; might have
() 2 impact on other things like the in-vessel steam explosion 3 and stuff like that.
4 DR. SEALE: The TMI core was what? Ninety days 5 over or 30 days over?
6 CHAIRMAN BARTON: It was more like 90. The 7 accident was like 90 days after commercial.
8 DR. SEALE: That's relatively fresh.
9 CHAIRMAN BARTON: Relatively fresh core, yes.
-10 DR. POWERS: I would say the biggest problem with 11 TMI as an experiment was that it was a very steam rich 12 experiment.
13 DR. CATTON: For a long time.
14 DR. POWERS: For a long time.
( 15 DR. CATTON: And oxidized the hell out of the 16 zirc.
17 DR. POWERS: Yes. You get into the heavy j 18 oxidation. Now, on the other hand, the configuration you 19 see in this strikes me as, it's very tenable, is that you 20 get some sort of a localized melting and then you get clads 21 in this rubble bed and a heat up of that rubble bed rather 22 than these continuous rod degradations, and in fact, the 23 rods around the core barrel region are relatively intact. I 24 mean, this is --
25 DR. CATTON: Well, there was water in them, water i ANN RILEY & ASSOCIATES, LTD.
Court Reporters i 1250 I Street, N.W., Suite 300 4 Washington, D.C. 20005 (202) 842-0034
174 1 flowing through it. They kept chilling it.
(~\
qj 2 DR. POWERS: Whatever the case --
3 DR. CATTON: You only need to -- when it's heated 4 up like that, you only need to run the steam through there 5 once or twice, thrice, and the clad gets really brittle and 6 starts falling off in little shards. We did that to our rig 7 at UCLA and we didn't have near the high temperatures that !
8 they have in the -- if you do, indeed, get to that state 9 with the reactor. We only went up to maybe 7, 800 degrees 10 F, and it was awesome what the steam did to the zirc. It 11 just -- it got all flaky and just fell out.
12 DR. POWERS: In some cases -- you may be being 13 fooled. You're sitting right on a phase transition at those 14 temperatures.
I \
\m / 15 DR. CATTON: That could be, but also, there were 16 the experiments at KFK where -- and there are photographs i
17 that are available where you'll see the clad -- oxidized 18 clad intact and all the fuel has fallen out.
19 DR. POWERS: Oh, yes, that happened.
20 DR. CATTON: So there's two scenarios, one where 21 the zirc gets down first in chunks of oxide, and another 22 where it stays up behind and the fuel gets out.
23 I don't know how you decide. The only thing I 24 think you can do is put it all in the bottom head and look I 25 at the range of variations that you can experience and try i
0%
l i
\-') ANN RILEY & ASSOCIATES, LTD.
Court Reporters 1250 I Street, N.W., Suite 300 Washington, D.C. 20005 (202) 842-0034
175 1 to attach some probability to it. If you can't attach a If') 2 probability to it, then you damn well better take the worst j
3 case or find a way to fix it.
4 DR. KRESS: Did you look at the chemical state of 5 __
6 DR. CATTON: Did we when that happened? No. We 7 were just upset.
8 DR. KRESS: Because the set temperature -- I can't 9 believe it's oxidized. It must be phase transition.
10 DR. POWERS: Yes, I'm sure Ivan was sitting right 11 on a phase transition and --
12 DR. CATTON: That often happens to me.
13 DR. KRESS: Yes, and that will tie it up.
14 DR. POWERS: And that is a huge volumetric change.
s_) 15 DR. KRESS: It doesn't make it an oxide. f 16 DR. CATTON: It ruined our day..
17 DR. KRESS: I can believe that, yes.
18 DR. POWERS: Phase transitions have a habit of 19 doing that.
20 CHAIRMAN BARTON: Can we move along?
21 DR. CATTON: You know, there's also -- Sol Levy, 22 as part of an exercise that we went through, did his own l
23 evaluation. All he was trying to arrive at is what it you 24 should put in the lower head and deal with, and he came up 25 with some ranges of metallic zirc, some ranges of steel and i
ANN RILEY & ASSOCIATES, LTD.
\~
Court Reporters 1250 I Street, N.W., Suite 300 Washington, D.C. 20005 (202) 842-0034 L-_________..._ - _ - - _ .
176
.~,
1 ranges -- and what he had done is he had taken every
. ( ,) 2' calculation that he could find and then made a few 3 reasonable assumptions about how the system would melt out 4 and he came to some kind of a -- what I thought was 5 reasonable, 55 or 60 percent of everything that was inside 6 wound up down in the lower head.
7 CHAIRMAN BARTON: Bob, are you ready?
8 MR. LUTZ: Yes, I'm ready.
9 CHAIRMAN BARTON: I think we're ready. We're not 10 going to get any readier by talking about this any more 11 right now.
12 DR. CATTON: You're the only guy here who has had 13 any experience -- real experience with these things.
14 CHAIRMAN BARTON: And it didn't go through the b
(m,/ 15 bottom of the head, so I guess --
16 DR. SHACK: Some things are better not talked 17 about.
18 CHAIRMAN BARTON: That's right.
19 [ Laughter.]
20 MR. LUTZ: What I want to talk about is the 21 analyses that we did of the ex-vessel severe accident i
! 22' phenomena and its impact on containment integrity. Let me 23 introduce this by saying within the PRA, and it's where 24 we've been for the last two and a half to three hours here.
25 The level 2 model assumes in-vessel retention for
() ANN RILEY & ASSOCIATES, LTD.
Court Reporters I
1250 I Street, N.W., Suite 300 2 84 b3
L 177 1 accident seguences where the RCS is depressurized and you
() 2: have a submerged vessel. If you don't meet either of these 3 conditions, the level 2 PRA assumes that.the reactor vessel 4 fails and a conditional probability of one that the 5 containment fails at the time of reactor vessel failure.
6 The question is whether this is a conservative 7 treatment or the degree to which it is a conservative 8 treatment of ex-vessel phenomena assuming that there is a l 9' conditional probability of one.
10 Go back and say with the present level 2 PRA 11 model, the way it's modeled with this conditional 12 probability of one, we find that the risk is acceptable and 13 less than the regulatory standards.
14 So the'next step, we got into discussions with the 15 staff during their review of in-vessel retention, and also 16 in terms of the requirements of SECY 93-087, we agreed that 17 while it was not necessary for the PRA in the modeling of 18 the PRA with the direction we were going with in-vessel 19 retention. We did agree to perform a limited number of 20 deterministic analyses of ex-vessel phenomena to identify i
21 potential containment vulnerabilities. I i
22 In particular, we looked at three things. We 23 looked at ex-vessel steam explosion, we looked at core j -24 concrete interactions, including the non-condensible gas i
25 generation, and we looked at direct containment heating.
]
i ANN RILEY'& ASSOCIATES, LTD.
Court Reporters 1250 I Street, N.W., Suite 300 i Washington, D.C. 20005 (202) 842-0034 1
I
i 178
,~
1 Come down and say we -- again, I would like to say (v) 2 we did a limited number of deterministic analyses in 3 selecting the criteria for what models and initial 4 conditions we were going to look at for these deterministic 5 analyses, what we wanted to do was having something that was 6 realistic but conservative for the assessment of ex-vessel 7 phenomena.
8 So what we did and what I'm going to present here 9 I would not call a best estimate, I would not call bounding; 10 it's somewhere in the middle.
i 11 I'm going to dump us straight back into the pot 12 that we were in about 15 minutes am,because the ex-vessel 13 phenomenon depends on the mode of reactor vessel failure and 14 the initial configuration of these core material that's A>
\_/ 15 within the reactor vessel, particularly for the ex-vessel 16 steam explosion and, as you'll see, to a degree for the core 17 concrete interaction work that we did, that the reactor 18 vessel mode in the modeling is important for assessing the 19 ex-vessel severe accident phenomena, the steam explosion and 20 the core concrete interaction.
21 What we did is we went back to the work that was 22 done at ANL and we took the initial configuration with the 23 metal layer overlaying the oxide core debris layer and all 24 of the core and reflector material in the bottom head at the 25 time of failure, the scenario that Jim presented earlier.
, t (Q,/ ANN RILEY & ASSOCIATES, LTD.
Court Reporters 1250 I Street, N.W., Suite 300 Washington, D.C. 20005 (202) 842-0034 .
l l
179 1 What we did from there is, again, in the Argone 2 work. Based on some analyses, they had hypothesized two 3 different reactor vessel failure modes that sort of looked 4 at you might say both ends of what could be possible. One 5 was a localized failure, which is a rather small opening 6 which progressively opens and sort of unzips, coming down 7 the vessel.
8 DR. CATTON: That's hard to imagine.
9 MR. LUTZ: It ablates as the -- ablates the 10 vessel, pours out. So it cuts it along a -- a longitudinal.
11 DR. CATTON: -- azimuthal line?
12 MR. LUTZ: Right.
13 And the other one is a zipper failure --
14 DR. CARROLL: That one is -- the second one is 15 believable. The first requires quite a bit of imagination.
16 MR. LUTZ: Well, we're looking at both sides, 17 what's the slow rate and what is the fast rate, and we'll 18 see when we get to core concrete interaction that there is 19 some significance to --
20 DR.. CARROLL: Which is the most damaging and most 21 headache?
22 MR. LUTZ: In terms of steam explosion, this one 23 obviously is. In terms of core concrete interaction, this l
24 one.
l 25 DR. CARROLL: That first one, there is a good l
i r*N
() ANN RILEY & ASSOCIATES, LTD.
Court Reporters l
1250 I Street, N.W., Suite 300 I Washington, D.C. 20005 (202) 842-0034 I
E . _ _ _ - . _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ - _ _ _ -. --. - _ _ _ _ _ _ _
180 1
1 picture in the Theofanous report of what's happening. It's l
?- 1
(%,) 2 spilling over the top and running down the side of the 3 thing, right? I 4 MR. LUTZ: Yeah. Basically, it creates a small 5 opening at the top and then pours out that. As the metal 6 comes -- or the level in the vessel is --
7 DR. CATTON: Oh, so it continues to erode the 1
8 bottom side of the whole --
9 MR. LUTZ: Uh-huh.
10 DR. CATTON: It's hard to imagine. The heat 11 transfer is pretty symmetric all the way around unless 12 something is trapping it. So I would expect, once it's 13 penetrated in one place, the rest won't be long behind. So 14 the hinge actually makes much more sense.
(__) 15 DR. POWERS: My experience, Ivan, having poured 16 hot metals on a few things, is that you get localized 17 failures and those failures grow, you don't tend to get 18 multiple failures.
19 DR. CATTON: I don't think I'm disagreeing with 20 that.
21 DR. POWERS: Yes, I think you are, and I'm saying 22 my experience is that they would get localized failures.
23 Now, again, I'll admit that my experiences don't 24 extend to masses bigger than about 500 kilograms.
25 DR. CATTON: If the pouring process cuts a hole, l
l l
! s I ANN RILEY & ASSOCIATES, LTD.
Court Reporters 1250 I Street, N.W., Suite 300 l
Washington, D.C. 20005 (202) 842-0034 l
< )
181 1 you're right, but if the stuff --
2 DR. POWERS: .I mean, even when you don't -- I hate 3 to use the -- the. language of the trade is when a crucible 4 containing a melt gets a failure, like ceramic fails or 5 something like that, or you eroded the ceramic, you tend to 6 get a localized failure and that failure grows by a stream 7 erosion very rapidly, and the drainage is so fast that you 8 just don' t. have an opportunity to get multiple f ailures.
9 But again, those experiences -- my own personal 10 experiences are limited to less than 500 kilograms. When
-11 you're talking about multiple tons here, that's a different
-12 story. But I have never heard of getting multiple failures.
13 Usually the drainage is so fast and the hole grows so 14 rapidly that you just don't get them. But --
(_/ 15 MR. LUTZ: If you bear with me for a second, 16 you'll find that by looking at both of them, I don't think 17 we have to question which one is more probable or not, 18 because we have carried both of these failures through and 19 looked at the impact of both of them. So just sort of hold 20 that thought until the end and see if it makes a difference 21 in your mind.
22 Just to throw some numbers around, and I'm not 1 23 going to spend a lot of time on them, but these are the 24 masses that were assumed inside the lower head at the time 25 of failure, and again, the failure was at the top or near ANN RILEY & ASSOCIATES, LTD.
Court Reporters 1250 I Street, N.W., Suite 300 Washington, D.C. 20005 (202) 842-0034 1
_ _ _ _ _ _ - - - - - - - - _ _ - - - - - _ -- I
182 1 the top of the metal layer.
l 2 These are the release rates for the hinge failure.
3 It was 30 to 40,000 pounds per second, with the whole thing 4 spilling out within ten seconds. For the localized failure, 5 it started at about eight pounds per second and 6 progressively increased to 940 pounds per second over quite 7 a considerable time period. If you look at that, that's 8 five or six hours, I believe.
9 DR. POWERS: I'll just point out the members that 10 this particular composition chosen here, the uranium would 11 be reduced to the stoichiometry right at the lower phase 12 boundary of uranium dioxide. In other words, it would be as 13 hypo-stoichiometric as it could possibly be. And the 14 activity of uranium metal in hypo-stoichiometric uranium 15 dioxide at the phase boundary is measured once kind of, and 16 that " kind of" measurement gave it an activity of about one; 17 in other words, it acts just like it's uranium metal.
18 DR. CATTON: Are you going to make the next leap 19 or are you going to wait?
20 DR. POWERS: I don't know the next leap -- I think 21 we passed the time of the next leap, --
22 DR. CATTON: Oh, okay.
23 DR. POWERS: -- that that would give you a high 24 enough activity in metal that this metal phase would, in 25 fact, be denser than the oxide phase, and this motal phase !
h ANN RILEY & ASSOCIATES, LTD.
Court Reporters 1250 I Street, N.W., Suite 300 Washington, D.C. 20005 (202) 842-0034
j i
l 183 l 1 as constituted there would have an aggressive metallurgic
() 2 interaction against any metal boundary that it encountered.
3 It would be very aggressive. You're almost at the optimal l
4 composition to get a heat of metallurgical interaction. l 5 Almost couldn't design it better.
6 DR. CATTON: So the vessel isn't going to last 7 very long.
8 DR. POWERS: Well, I don't know that. That's a 9 leap I can't take because as you pointed out, it's when the i l
10 heat in balances the heat out and the structure is then thin 11 enough and can it support, and that depends on doing the I 12 quantitative analysis, which I've not attempted to do. l 13 DR. CATTON: It seems to me someone --
14 DR. POWERS: My experience, experimental 15 experience with metallic reactions is the vessel doesn't 16 stand a chance.
17 MR. LUTZ: Let me address first the direct l
18 containment heating analysis that we did, this one sort of L '19 independent of reactor vessel failure mode.
1 20 Basically --
- 21 DR. POWERS
- I have to interrupt just to make sure I
22 everybody knows that not only was this work done at Sandia, l
23 but the authors and whatnot worked for me, reported directly l
t 24 to me. So I can't say a word. l 25 DR. CATTON: You can answer questions.
l l
ANN RILEY & ASSOCIATES, LTD.
Court Reporters 1250 I Street, N.W., Suite 300 Washington, D.C. 20005 (202) 842-0034
-_ _ _ _ _ - _ _ _ _ - _ _ - - _ _ _ - _ _ _ - -___-________-------.---_---_--------------------------.-------N
l 184 1 DR. POWERS: I don't even think I'll answer
() 2 questions.
l 3 [ Laughter.)
4 DR. CARROLL: Is it that shoddy a piece of work?
5 [ Laughter.)
6 DR. POWERS: I'm not going to say a word.
7 MR. LUTZ: We used the Sandia model to look at 8 direct containment heating. The model includes blowdown of 9 RCS gases, heat transfer from core debris to containment 10 atmosphere, oxidation of unreacted zirc, and hydrogen 11 combustion. We used a two-cell model, one to model the 12 containment area above the operating deck, which is the 13 primary area concerned for direct containment heating 14 because of the large volume, and then the steam generator
(~') i
\_) 15 loop compartments were sort of grouped together, the area 16 under the operating deck.
f 1
17 I have two number changes. When I made this slide 18 rp I took it off an earlier calculation version, and I 19 apologize for that. This agrees with what is in Appendix B 20 of the PRA. What we got in terms of a predicted pressure 21 increase was a 50 psi pressure increase due to direct 22 containment heating and if -- out of the model. If we 23 combine that with the PRA analyses using the MAAP code for 24 the highest containment pressure at vessel failure for those 25 sequences that went to vessel failure in MAAP, we find it ANN RILEY & ASSOCIATES, LTD.
Court Reporters 1250 I Street, N.W., Suite 300 Washington, D.C. 20005 (202) 842-0034
185 1 was 45 psig, so we are looking at a peak containment
._/
) 2 pressure from direct containment heating of something on the 3 order of 95 psig. That pressure is well below the point 4 where containment overpressure failures are significant.
5 DR. SEALE: How well is well?
6 MR. LUTZ: The 50 --
l 7 DR. SEALE: I would like to kind of have how much 8 I have got under my -- how much lee I have got under my --
9 MR. LUTZ: I believe if you look at the 10 containment fragility curve you are less than 10 to the 11 minus 3 probability of failure at this point.
12 DR. SLALE: Okay.
13 MR. LUTZ: The median is 132 psig or something 14 like that.
(,~
- s. 15 DR. SEALE: All right.
16 DR. POWERS: These containment fragility curves 17 are always a bit of a mystery to me because they are 18 calculated entities. I mean there is no experimental 19 movement up here. And it seems to me that there has been a 20 lot of work done to try to validate the calculational 21 vehicles to prepare these, these codes. And my experience 22 with those experiments is that they always find the models 23 fail at details in the -- in the model, I mean the 24 construction deals, that are below the level of resolution 25 of the codes used for the containment analyses. And though s
) ANN RILEY & ASSOCIATES, LTD.
Court Reporters 1250 I Street, N.W., Suite 300 Washington, D.C. 20005 (202) 842-0034 <
l l
j l
186 1 they get very high ratios between design pressures and
{
1
-2 failure pressures, it is always at construction details that (sL) I 3 -- where the failure occurs.
4 Now, people that do those things -- l l
5 DR. SEALE: The stress risers have a tendency to I 6 do that.
7 DR. POWERS: Yeah. The people that do those say, 8 yeah, but they are close to the bulk failures and whatnot.
9 But it's that step that always leaves you puzzled on what 10 you do with these containment failure probabilities, because 11 if the code is modeling the gross failure, and the failure 12 is occurring at the details, then it is the probability of 13 construction details that dominates everything. It is a 14 curiosity to me that -- I have never successfully resolved f~%
(s,) 15 it in my own mind.
16 DR. SEALE: Okay.
17 MR. LUTZ: The other data point that one can use 18 is the service level C value for the containment is 90 psig, 19 I believe, so it is just a couple of psig over the servica 20 level C value, which is tied to the ASME code.
21 So based on that, one could predict that direct 22 containment heating would not be predicted to challenge the 23 AP600 containment or present a significant challenge.
24 I might add that if you go back into the Appendix 25 B, we did some conservatism in modeling and we didn't model I ANN RILEY & ASSOCIATES, LTD.
' ~ -
Court Reporters 1250 I Street, N.W., Suite 300 Washington, D.C. 20005 (202) 842-0034
187 1 all of the flow paths out of the -- into the loop
.n
() 2 compartments. Let me just go through the three flow paths.
3 There is a permanent seal ring around the vessel joined to l l
4 the -- at the vessel flange. We assumed that that was !
5 completely lost, so it opened up that whole area, which 6 maximizes the direct pathway to the upper containment.
7 We did not take credit for the penetrations 8 through the biological shield where the coolant pipes go ;
9 through because we couldn't in our own minds say what was 10 going to happen to all the insulation and everything else 11 that is around the reactor vessel, so we just assumed that 12 -- we thought conservatively that those holes would be 13 plugged. And the only other pathway out is the ventilation 14 shaft that goes up to the loop compartments. So we were A i 6
(_ ,/ 15 looking at the ventilation shaft and the seal ring area.
16 DR. CATTON: Wouldn't that fancy insulation 17 chimney just get jammed into every hole?
18 MR. LUTZ: That's why we assumed that --
19 DR. CATTON: So you don't have any?
20 MR. LUTZ: -- there was no hole there. Basically, 21 we assumed that that's what happens.
22 DR. CATTON: All right.
23 MR. LUTZ: And these are the numbers we came up 24 with.
25 DR. CATTON: So what is the pathway to the upper (n)
ANN RILEY & ASSOCIATES, LTD.
Court Reporters 1250 I Street, N.W., Suite 300 Washington, D . C .. 20005 (202) 842-0034
188 1 deck?
() 2 MR. LUTZ: Through the gap between the flange and 3 the concrete.
4 DR. CATTON: But wouldn't that fill up with this 5 stuff?
6 MR. LUTZ: Well, --
7 DR. CATTON: If it does, what happens?
8 MR. LUTZ: If it does, then we don't have a 9 calculation, because the direct -- these values just go way 10 down. We tried to maximize the pressure that we would get 11 from direct containment.
12 DR. CATTON: Well, what would the pressure be in 13 the cavity if that happened?
14 MR. LUTZ: Well, the ventilation shaft is rather 15 large in terms of flow area, so it would relieve through 16 that into the loop compartments.
- 17. DR. CATTON: Somehow I am wondering if we are 18 talking about the same thing. You have got the cavity.
19 MR. LUTZ: Correct.
20 DR. CATTON: And you are feeding it all this hot 21 gas and nasty stuff.
22 MR. LUTZ: Correct.
23 DR. CATTON: And you jam all these fancy 24 insulation into all the crevices and cracks that the gas is 25 going to get out, so the pressure in the cavity is going to 1
i
()'-
ANN RILEY & ASSOCIATES, LTD.
Court Reporters i 1250 I Street, N.W., Suite 300 Washington, D.C. 20005 (202) 842-0034
(
1 i _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ . _ _ _ . _ _ _ _ _ _ _ _ _ _ _ _ _ . _ _ _ _ _ _ _ _ _ _ _ _ _ . _ _ _ _ _ . _ _ _ _ _ _ _ _ . . _ _ _ _ _ . _ _ _ _ _ _ _ . _ _ _ _ _ _ . _ _ _ _ _ _ _ _ _ _ _ _______________.__________;
189 1 go quite high. Have you done that kind of calculation?
L
() 2 MR. LUTZ: Let me draw a picture. When we get 3 into the steam explosion stuff, we can look at the other 4 side.
5 DR. CATTON: I am just -- DCH is what is on the 6 table now.
7 MR. LUTZ: Right. Right. There's sort of two 8 regions to the cavity. There's a flow path here, there's 1
9 obviously-a flow path here, and there's a large ventilation 10 shaft that goes up from there.
11 DR. CATTON: How big is the ventilation shaft?
l 12 Several feet?
13 MR. LUTZ: Five or six square feet. I don't know 14 the exact number. It's in the Appendix B.
15 DR. CATTON: Because the chimney that you have 16 built does a really good job of directing the flow. But it 17 doesn't look to me like it is going to be able to withstand l
18 a DCH at all. You are going to -- it is just going to move 19 ahead and plug everything in its way.
20 DR. FONTANA: I think part of the idea, you drew ,
i 21 that -- you drew that thing to the left there. I think that 22 comes a URD requirement that the debris come off the side l 23 and get trapped in that, so it wouldn't come up and plug up 24 that.
25 DR. CATTON: But, Mario, the URE is talking about
/"
( Jdni RILEY & ASSOCIATES, LTD.
Court Reporters 1250 I Street, N.W., Suite 300 i Washington, D.C. 20005 '
(202) 842-0034
190 1
the debris that comes out of the vessel and not the -- all h 2 these thin sheets of stainless steel that are used to 3 insulate it.
4 DR. FONTANA: Even so -- even so, you are going to 5 have some inertia as that thing takes a turn and tries to go 6 up.
7 DR. CATTON: It is going to be like an airfoil, 8 who knows where it is going to go.
\
9 DR. FONTANA: Yeah. But --
10 DR. CATTON: Well, give it a 50/50 probability 11 then. We are in the world of PRA.
j 12 DR. FONTANA: By the way, the vessel is not 13 depressurized under this scenario here, correct?
14 MR. LUTZ: That is correct.
15 DR. FONTANA: Yeah. Okay.
16 MR. LUTZ: But what we assumed in order to 17 maximize this pressure is that this pathway was open, this 18 pathway was plugged by the insulation and that this pathway 19 through the ventilation shaft is open.
20 DR. CATTON: I am -- I guess I could rephrase -- I 21 could phrase the question. What happens if the pressure 22 gets really high on that cavity? What kind of pressure can 23 it take?
24 MR. LUTZ: Well, if we look at the loadings from 25 steam explosions, we can --
lk JJRJ RILEY & ASSOCIATES, LTD.
Court Reporters 1250 I Street, N.W., Suite 300 Washington, D.C. 20005 (202) 842-0034
191 1- DR. CATTON: But those are impulsive, that's
() 2 3.
different. This'is not going to be an impulsive, it's going to be over a longer period of time.
4 MR. LUTZ: More a quasi-steady state.
5 DR. CATTON: A steam explosion is different.
6 MR. LUTZ: I am not --
is Jim here?
7 DR. CATTON: Are you going to lift the vessel up !
1 8 out of there or something? That was the concern the Germans I 9 had, was that you would lift the vessel with a DCH.
10 MR. SCOBEL: This is Jim Scobel from Westinghouse.
11 I think the insulation would be more likely to lift that the l
12 vessel.
13 DR. CATTON: Well, that's exactly the point. The 14 insulation lifts, it plugs the holes.
15 MR. SCOBEL: The insulation.that is plugging --
that hole, the ventilation hole is more like probably 50 16 17 square feet than this. It's a big manway kind of a hole.
18 Itds open and if you were to --
i 19 DR. CATTON: I gather you didn't take a look at 20 it.
21 MR. SCOBEL: We didn't look at direct containment 22 heating.
23 DR. CATTON: And --
24 MR. SCOBEL: And the blowdown forcing the l
25 insulation.into that hole. But if you are going to talk i
ANN RILEY & ASSOCIATES, LTD.
Court Reporters 1250 I Street, N.W., Suite 300 Washington, D.C. 20005 (202) 842-0034 i
192 1 about lifting the reactor vessel, I think the pressure is.
() 2 going to blow that insulation into the tunnel above rather 3 than lift the reactor vessel. It's going to balance the 4 forces, which one is stronger. But I would pick the vessel.
5 DR. CATTON: You are well trained in the ROAAM 6 method.
7 MR. SCOBEL: Thank you.
8 MR. LUTZ: To answer your question, no, we did not 9 look at the pressurization in the cavity.
10 DR. CATTON: I think it's a simple thing to do and 11 maybe you ought to do it just to make sure that --
12 DR. FONTANA: I did that a long time ago, but I 13 don't remember.
14 DR. CATTON: You looked at that?
O V 15 DR. FONTANA: Yeah. Some assumption on the 16 thickness of the concrete walls and the pressure blowing 17 them out too. I just can't remember.
18 DR. CATTON: Maybe you just fold this stuff up and 19 blow it right on through.
20 DR. FONTANA: I think that's probably what would 21 happen.
23 assume any water in a cavity or did you assume -- I guess 24 you assumed the IRWST did not discharge. Was there any 25 residual water from the RCS?
() ANN RILEY & ASSOCIATES, LTD.
Court Reporters 1250 I Street, N.W., Suite 300 Washington, D.C. 20005 (202) 842-0034 u-_---____ _ _ _ _ _ _ _ _ _ _ _ _ - _ - _ _ . _ _ . _
193 1 MR. LUTZ: No. No.
() 2 DR. CATTON: I don't think you will' lift the 3 vessel either, but it would just be interesting to know what 4 that pressure is if you plug up some of those holes. ;
5 DR. FONTANA: That problem goes away if you I l
6 depressurize.
l 7 DR. CATTON: Of course it does.
8 MR. LUTZ: That is correct. Direct containment 9 heating is --
10 DR. CATTON: DCH is gone.
11 DR. FONTANA: Yeah. So I mean what are the 12 probabilities of not depressurizing the vessel? Well, I am i 13 not asking for --
14 MR. LUTZ: If you go back and look at the level 15 one results, I think it's on the order of 6 percent of the 16 core damage sequences.
17 DR. FONTANA: Thanks.
18 DR. CARROLL: It's misleading to talk that way 19 because you also have to say what the core damage sequences 20 are doing that's playing it a little different, much lower.
21 MR. LUTZ: That's correct. If you go back to 22 Jim's slide and you look at the 3BE sequences that led to 23 containment failure, that's the ones we are talking about.
24 Let's move on to the ex-vessel steam explosion
( 25 work that was done. For ex-vessel steam explosions what we i
ANN RILEY & ASSOCIATES, LTD.
Court Reporters 1250 I Street, N.W., Suite 300 Washington, D.C. 20005 (202) 842-0034 L____-..-__-_ - - _ - - - - - ---
i 194 1 did is we took the TX's version 5 code that was developed at
() 2 _the University of Wisconsin and we applied that to the two
- 3 reactor failure modes that I described previously.
4 DR. CATTON: When you used the TX code, what 5 assumptions did you make about the particulate sizes?
6 MR. LUTZ: The stream coming out or the --
7 DR. CATTON: Yeah. The stream comes out, it 8 breaks into big pieces, and big pieces break into little 9 pieces, and those sizes determine what the loads are. Are 10 those built into the TX code or de you put them in as an 11 input parameter or what?
12 MR. LUTZ: I did not personally do the 13 calculation, and I don't know the answer to that question.
14 DR. CATTON: Okay. Well, I'll look at the report O
15 and see if there's an answer there.
16 MR. BASU: This is Sud Basu from Recuarch, if I 17 can answer that question. The fragmentation models are 28 built into TX. You start with an initial particle size, 19 that's something that you --
20 DR. CATTON: That's the initial fragmentation, 21 right? And then the shock fragments it into smaller and, 22 depending on what you choose, you get any answer you want.
23 MR. BASU: Correct.
24 DR. CATTON: Okay.
25 I just want to put TX in the right perspective.
l I
(/ ANN RILEY & ASSOCIATES, LTD.
Court Reporters 1250 I Street, N.W., Suite 300 Washington, D.C. 20005 (202) 842-0034
195 1 That whatever comes out of it in speculation. Now, there
( ) 2 may be some basis for the speculation, but it is speculation 3 nevertheless. And it is a one-dimensional code, at least it 4 used to be.
5 MR. LUTZ: It still is. It does model fuel 6 fragmentation, fuel co and inter-mixing, fuel co and energy 7 transfer, and explosive loads, where applicable. We did use l 8 both reactor vessel failure modes. And what the code 9 predicted were the folloaing explosive loads. And you can 10 see that the localized failure mode with the relatively 11 small, or slow pour rate, doesn't compare to the hinged 12 vessel failure mode in terms of the explosive loads. So 13 most of the resc of what we did, we focused on this hinged 14 failure mode where we got 24,700 psi, 71 psi seconds.
15 There were several sensitivity studies performed, 16 they are reported in Appendix B, on some of the initial 17 conditions for the TX code. We'didn't see any large b 18 variations from this base case for reasonable variations in ;
19 the model parameters, so we went ahead with this loading and 20 looked at the impact on the structure.
21 DR. CATTON: Is that 24,700 psi the impulse?
22 MR. LUTZ: Yes.
I
?. 3 DR. KRESS: 71 is the -- integrated? j 24 MR. LUTZ: I believe -- I believe so. And, again, l l
25 I am not an expert in this, so am hedging a bit.
1 l
()
l ANN RILEY & ASSOCIATES, LTD.
Court Reporters 1250 I Street, N.W., Suite 300 Washington, D.C. 20005 (202) 842-0034
196 1 DR. CATTON: So what you do when you evaluate the rh
( ) 2 impact is you take that shock pressure and run it over the 3 whole surface of the inside and see that nothing --
4 MR. LUTZ: That's correct.
5 MR. BASU: This is Sud Basu aga'n. I am not sure 6 if I heard you correctly. Did you say 24,700 psi is the 7 impulse?
8 DR. CATTON: No , it is not the impulse. The other 9 one.
10 MR. BASU: That's the peak pressure.
11 MR. LUTZ: That's --
12 DR. CATTON: But it's not, it's peak pressure. i 13 MR. BASU: That's the peak pressure. The second 14 number is the impulse load.
('m/) 15 DR. CATTON: Well, I guess I could figure it out.
l 16 DR. CARROLL: It's called long division.
17 DR. CATTON: That's hard.
l 18 MR. LUTZ: I believe it was like 4 or 6 ;
19 millisecond. So we turned that over to the structural 20 people and had them do a structural assessment of the 21 loadings. For the -- what they found for the hinged vessel 22 failure mode, they found that localized failure of the 23 cavity walls might occur, but the containment vessel, the 24 embedded liner within the concrete, is not predicted to fail 25 and that the strain, the vessel strain is less than 20 1
() -
ANN RILEY & ASSOCIATES, LTD.
Court Reporters 1250 I Street, N.W., Suite 300 Washington, D.C. 20005 (202) 842-0034
197 1 percent of its ultimate strain.
s.,
(V) 2 DR. CATTON: The shock doesn't see the metal 3 containment vessel, does it?
4 MR. LUTZ: No. No. There's concrete on top the 5 vessel and then concrete again.
6 DR. CATTON: And if you fail the cavity wall, that 7 doesn't cause -- it apparently doesn't cause a problem.
8 MR. LUTZ: Does not cause a problem.
9 DR. CARROLL: What does localized mean in that 10 context?
11 MR. LUTZ: The peaking loading and failure is the 12 wall that is between the reactor coolant -- or the reactor 13 vessels and the two sides of the reactor cavity. There was i 14 a wall.
('N
(_,) 15 DR. CATTON: Jay, you get a prescure pulse that 16 crawls up the wall. And if there's a weak spot, it would !
17 crack it or break it or something.
18 DR. CARROLL: I am just wondering what happens 19 with reinforced concrete.
20 DR. CATTON: It probably just cracks.
21 MR. LUTZ: It was determined that there wouldn't 22 be any gross failures that would cause other things within 23 the containment to change location. If that's where you are 24 leading with the question.
25 DR. CARROLL: Okay. Good.
(' '
ANN RILEY & ASSOCIATES, LTD.
Court Reporters 1250 I Street, N.W., Suite 300 Washington, D.C. 20005 (202) 842-0034
198 1 DR. POWERS: Did you look at or are there even ]
() 2 possibilities for generating missiles from these ex-vessel 3 steam explosions?
4 MR. LUTZ: We looked at the reactor vessel'itself.
5 Other than the insulation, there is nothing down in that 6 area. We don't have any core tubes or anything else down in 7 that area.
8 DR. POWERS: There is nothing to throw, so you 9 can't create much a missile with nothing.
10 MR. LUTZ: We did look at the reactor pressure 11 vessel response and found that for the hinged vessel failure 12 mode that the maximum width of the reactor vessel was less 13 than six feet, and that's within the height of the shield 14 walls that surround the reactor vessel, so they just sort of 15 pop up and down again.
16 We also looked at the impact of the movement of 17 the vessel on penetrations, and there's enough formation of l l
18 plastic hinges in the coolant pipes to isolate the vessel i 19 movement and prevent failure of containment penetration.
20 DR. CATTON: When you do this, what kind of a 21 shape do you give the pressure pulse?
22 MR. LUTZ: Triangular, I believe is what they i 23 used. I 24 DR. KRESS: Did the TX code use all this mass that ,
I 25 is on slide 5?
ANN RILEY & ASSOCIATES, LTD.
Court Reporters 1250 I Street, N.W., Suite 300 Washington, D.C. 20005 ;
(202) 842-0034 l
199 -
1 MR. LUTZ: Yes, sir. There was -- it is rate
)
( -) 2 dependent, if we look at slide 5. I 1
3 7R. KRESS: Yeah, it has got a rate at which it l 4 pours out. l I
5 MR. LUTZ: Right. And I am not sure exactly how 1 6 TX handles the rate dependency. But this amount was put in 7 over -- at this rate, over ten seconds. I am not familiar 8 with the inner workings of the code.
9 DR. KRESS: Well, it assumes a time for 10 triggering. It only uses the amount that pours in up to 11 that time of triggering. You don't know what that figure 12 assumption was?
13 MR. LUTZ: I don't recall right now. I believe it 14 is in the Appendix B material.
15 DR. CATTON: Is that built into the code too?
16 MR. LUTZ: The --
17 DR. CATTON: So when you use the TX code, you 18 don't make any assumptions?
19 MR. LUTZ: I didn't use the TX code.
20 MR. SNODDERLY: I believe, as Mr. Lutz said, it is 21 in Appendix B, but it's -- the trigger time was based on the 22 fall time, the time for the melt to go from the top to hit 23 the floor, and that was in the one second time frame.
24 They also modeled -- the pour rate was modeled as 25 226 jets of a .068 diameter. So they -- in other wordc, TX i
O t' ,/ ANN RILEY & ASSOCIATES, LTD.
Court Reporters 1250 I Street, N.W., Suite 300 Washington, D.C. 20005 (202) 842-0034 E----_-----------._------
( 200 1 you can't -- TX doesn't like huge hinged failure melts, so
() 2 you have to model it as a series of -- as a set of streams, 3 jets. The staff did some confirmatory analyses, and we gave l l l 4 that report, the Committee requested that report and we !
5 provided it a while ago just to summarize. We came out, we i
L 6 convened a panel of experts and based on -- we came up with l l '
7 a similar -- we found their assessment to be -- their melt j 8 mass configuration to be reasonable.
9 We disagreed with what -- we agreed on the failure 10 location, but we differed on the size. They assumed either I 11 a .06 meter or this, a very small-failure on the side or 12 this hinged failure. We ended up in the middle, we think 13 that it is going to be a side failure, but about .4 meters.
14 And then what we also found, we did calculations A
ksIs 15 with the one-dimensional TX code and the two-dimensional PM 16 Alpha S-PRO SIM code. That's Theofanous' code. Both codes 17 were developed under NRC sponsorship. What we found was 18 that PM Alpha S-PRO calculated nigher impulse loads than TX, 19 but they seemed to be in the same order of magnitude as !
20 those calculated by Westinghouse for the hinged failure.
21 We came to a similar conclusion as Westinghouse, 22 that we would expect the cavity walls to fail but we would 23 expect the impulse to accelerate or crack the concrete, j 24 separating between the liner, and that would cause a !
25 deflection of the liner but would not exceed the strain of l
() ANN RILEY & ASSOCIATES, LTD.
Court Reporters 1250 I Street, N.W., Suite 300 Washington, D.C. 20005 (202) 842-0034 l
\
-___-_n__._ _ . - - . - _ - _
201 1 the liner, so the liner stays intact but the cavity fails.
A
() 2 DR. KRESS: This looks like about a .1 percent 3 energy conversion.
4 Did you calculate that?
5 MR. SNODDERLY: I'd have to get back to you what 6 the actual energy conversion was. I don't have that right 7 on my fingertips right now.
8 DR. KRESS: It would be a neat thing -- variance 9 and sensitivity. Rather than start from known pour rates 10 and stuff, just use the energy conversion and vary it as a 11 sensitivity and see how much it takes to fail the 12 containment.
13 MR. SNODDERLY: Work backwards -- say, because --
14 you're right.
t
\_/ 15 We had the Structural Branch look at the 16 structural analyses provided by Westinghouse and they found 17 the model to be reasonable -- anyways, I mean I have it 18 right here.
19 DR. CATTON: There is no way that they shocked to 20 propagate above the deck?
21 MR. SNODDERLY: Oh, well, they also looked at what 22 that shock would do to the bottom of the vessel, but they 23 assumed that the hot legs are plastic and fail.
24 In other words, they don't assume any resistance )
25 there.
l'
('- ) ANN RILEY & ASSOCIATES, LTD.
Court Reporters 1250 I Street, N.W., Suite 300 Washington, D.C. 20005 (202) 842-0034
202 1 They just assume how high up -- because we were
() 2 concerned about could you impact the liner going up.
3 DR. CATTON: That wasn't the question. When I 4 think about it, it was a dumb question.
5 DR. CARROLL: What are the implications of 6 shirring off the hot legs at this point in the accident 7 progression?
8 DR. CATTON: Probably nothing.
9 MR. LUTZ: The core is out of the vessel at this 10 point.
11 DR. CATTON: You just have more steel in what is 12 left to attack the basemat.
13 MR. SNODDERLY: I can answer Dr. Kress's question.
14 What we -- which was --
15 DR. CATTON: While you are doing that, the 16 concrete that is around the vessel and so forth, what does 17 that look like relative to a existing PWR? Is there more 18 concrete, less concrete? About the same or what?
19 MR. LUTZ: I believe it is much thicker.
20 DR. CATTON: Thicker?
21 MR. LUTZ: Normally you are looking at a three 22 foot, three-five foot thick wall. This is -- I don't know 23 the dimensions but it is massive. It is massive.
24 DR. KRESS: It's pretty well protected.
25 DR. CATTON: We did these kind of calculations for
() ANN RILEY & ASSOCIATES, LTD.
Court Reporters 1250 I Street, N.W., Suite 300 Washington, D.C. 20005 (202) 842-0034
203 1 an existing PWR and just wrote off ex-vessel steam r~'T
( ) 2 explosion.
3 MR. SHERRY: What is the --
4 DR. CATTON: Took the whole core.
5 MR, SHERRY: Did you look at the quasi-steady 6 pressure just due to blowdown of the vessel prior to 7 clearing a vent path from the cavity?
8 MR. LUTZ: No, we didn't.
9 MR. SHERRY: It would seem to me that would be 10 quite high until you have established a vent pathway, if you 11 are failing the vessel, versus a large lower head failure 12 when a vessel is at high pressure -- and the cavity will 13 pressurize and the pressure won't decrease until you have 7
14 established a vent pathway.
"\s l 15 MR. LUTZ: But the ventilation shaft is there as a ,
4 16 vent pathway.
17 MR. SHERRY: Yes, but it's filled with water.
18 MR. LUTZ: No , it's not.
19 MR. SHERRY: It's not filled with water? Water is 20 surrounding the vessel, right? I mean I don't understand 21 what the --
22 DR. CATTON: There must be water if you have a 23 steam explosion.
24 MR. LUTZ: Okay, I see where you are going. I was 25 going back to the analysis we did on direct containment.
x (e
) ANN RILEY & ASSOCIATES, LTD.
Court Reporters 1250 I Street, N.W., Suite 300 Washington, D.C. 20005 (202) 842-0034
204 1 MR. SHERRY: No , this is a case where you now --
O t j 2 you are flooded.
3 MR. LUTZ: Yes, well, first of all, we 4 conservatively -- when we did the direct containment heating 5 analysis we.didn't take any credit.for the floor 6' penetrations for the' loops through the biological shield.--: l 7 actually six -- the penetrations.
8 Each of those is, I forget the numbers, but they 9 are not insignificant, so you have some vent path up to the ,
l 10 high level that wouldn't be flooded at that point, so you do l 11 have a relief path from the cavity.
12 MR. SHERRY: Between the location of the bottom of 13 the vessel and the first liquid gas interface, what is the 14 depth? How much water has to be vented? What depth of 15 water has to be vented before you actually have a -- a vent 16 pathway?
17 MR. LUTZ: It depends on what scenario you are 18 talking about. For the scenarios that we looked at for 19 reactor vessel failure, they were ones that basically the 20 water level was at the top of the lower head, that 21 elevation, at the time of vessel failure. I 22 MR. SHERRY: If you would have the IRWST discharge 23 it would be much greater than that?
l 24 MR. LUTZ: With the IRWST discharged, you would 25 essentially be full in that area.
l
() ANN RILEY & ASSOCIATES, LTD.
Court Reporters j
1250 I Street, N.W., Suite 300 i
Washington, D.C. 20005 ;
(202) 842-0034 1
205 1 MR. SHERRY: Is that right, Jim?
() 2 MR. SCOBEL: Pardon?
3 MR. SHERRY: With the IRWST completely discharged, 4 the cavity would be full?
5 MR. SCOBEL: Yes, the cavity would be full up to 6 the lip nozzles.
7 MR. SNODDERLY: This is Mike Snodderly again. If 8 I could, I would like to try to. attempt to answer Dr.
9 Kress's question about if we work backwards to determine how R10 big of a load we could handle.
11 Westinghouse -- well, the Staff estimated the 12 worst case loading based on our sensitivity studies was 13 about 650 kilopascal seconds and we estimated that that 14 resulted in about 5 percent elongation in the steel i 15 containment vessel.
16 The ultimate strain capacity of the containment 17 vessel is 20 percent, so you figure -- I don't know if you 18_ can do a linear extrapolation, but -- so that gives us some 19 idea of -- there appears to be a great deal of margin there 20 as far as the liner is concerned.
21 Looking at the report, I don't know -- I can't 22 readily -- there is not a table where we did energy 23 conversions but sure -- I can't promise I could find it, but i 24 'what I am saying -- I can attempt to try to get back and 25 give you some idea.
I- 2
(')
~
ANN RILEY & ASSOCIATES, LTD.
Court Reporters 1250 I Street, N.W., Suite 300 Washington, D.C. 20005 (202) 842-0034
206 1 DR. KRESS: The reason I asked that is that
() 2 3
experimentally with large steam explosions, about the most
'you ever observe is about 3 percent. If I were going to do 4 this, I would pick out a big mass and say 3 percent is about 5 as'high is it will get, and use that, tura it into a shock 6 wave -- it's far below the failure threshold.
7 You know, I can't look at these numbers, multiply 8 them by MC sub-P T and the Btus and then look at your pulse 9 width and get -- I am having a little trouble converting to 10 get a percent.
11 MR. SNODDERLY: And I understand your concern 12 because a lot of times when you look at fuel cooling 13 interaction studies it is an important parameter because 14 some studies go as high as 10 percent and some are as low as 15 .1 percent, and you have got to know am I looking at 16 something that is closer to the 10 percent range or am I at 17 the .1 percent, so are you being conservative or 18 non-conservative, and unfortunately I can only tell you that 19 we let Texas do it's job and it was the Staf f's developed 20 code.
21 Did you want to -- somebody else?
22 MR. BAJWA: Su Bajwa. Tom, the 3 percent 23 conversion you were referring to is for the illumine )
1 24- experiments.
25 DR. KRESS: Yes, it's a method actually observed. I 1
I
() ANN RILEY & ASSOCIATES, LTD.
Court Reporters 1250 I Street, N.W., Suite 300 Washington, D.C. 20005 (202) 842-0034 l
207 1 MR. BAJWA: With other simulants you get 1
) 2 percent, sometimes less than 1 percent --
3 DR. KRESS: Sometimes you get nothing, yes.
4 MR. BAJWA: So there is your range and when you t
5 are talking about the reactor materials, we haven't seen any I
6 evidence where it says that the conversion is anywhere even I i
7 close to a fraction of a percent.
8 DR. KRESS: I understand but I would like to know 9 what this percent is, just to calibrate it with respect to 10 what I have in mind for steam explosions.
11 MR. LUTZ: Again, I don't think we have the answer 12 or the Staff at this time.
13 Let me address core-concrete interactions is the 14 last thing that we looked at.
15 One of the things that we found when we first 16 started looking at core-concrete interactions is we sort of 17 questioned the uniform spreading assumption that is commonly 18 used in looking at core-concrete interactions.
19 We had two things that led us to believe that 20 uniform spreading was not realistic. One is the reactor 21 vessel failure mode with the metal coming out and followed 22 by the oxide, and second is the shape of the AP600 reactor 23 cavity region, which is sort of two regions that are 24 connected via a tunnel. !
25 What we did is Argonne has a code called l
f ANN RILEY & ASSOCIATES, LTD.
l Court Reporters 1250 I Street, N W., Suite 300 Washington, D.C. 20005 (202) 842-0034
208
~' l .MELTSPREAD that we used to look-at the spreading of the core 2 material as it comes out of the vessel and then used that as 3 an initial condition and used the MAAP models for 4 core-concrete interaction and for the containment.
5 Using what came out of MELTSPREAD is an initial 6 condition for these, so we sort of ran MAAP in a separate 7 _ effects mode where we fed it the input from the MELTSPREAD 8 code, j
9 DR. KRESS: The uniform spread would be the least 10 conservative, because it spreads it out over the biggest 11 area?
12 MR. LUTZ: That is correct. That is correct.
13 DR. KRESS: Using something else than that, you 14 are going more conservative?
15 MR. LUTZ: That is correct, or at least we think 16 we are -- I just said that.
17 So what did we find from the MELTSPREAD code? The 18 hinged vessel failure case, and if you look at Appendix B 19 there's some nice pictures in there, we found that there was 20 a relatively uniform spreading, at least the final 21 distribution, on the reactor cavity floor with an average 22' debris depth of about one and a half feet, but we had a very I 23 unequal distribution of oxidic and metallic melt in each of 24 the sides of the reactor cavity. !
i 25 The part of the cavity under the reactor vessel i
() ANN RILEY & ASSOCIATES, LTD.
Court Reporters 1250 I Street, N.W.,-Suite 300 Washington, D.C. 20005 (202) 842-0034
209 1 was primarily oxidic melt and the side under the reactor
() 2 coolant drain tank was principally metallic melt, so what 3 this told us is that when we went to the MAAP analysis we 4 sort of separated the two sides.
5 We said okay, this is the initial condition for 6 this side, this is the initial condition for the other side 7 of the reactor cavity --
8 DR. KRESS: This is a dry cavity?
9 MR. LUTZ: Correct -- and we let the MAAP 10 core-concrete interaction models determine the ablation 11 rates and the gas generation rates.
12 So as you would expect, the reactor vessel side of 13 the cavity eroded and the erosion rate was much greater than 14 the other side of the cavity that contained reactor coolant 15 drain tank -- so in effect what we are doing is 16 concentrating the erosion of the concrete into half the 17 cavity by using the results of the MELTSPREAD analysis.
18 Now if we go ov?r and look at the localized 19 reactor vessel failure case, what we find is a non-uniform 20 spreading of debris over the cavity floor. If you think 21 about this, this is a very slow pour rate out of the cavity 22 over a prolonged perioc, and it sort of builds up a pile 23 under the reactor cavity or under the reactor vessel and not 24 very much of it flows over into the other side of the 25 reactor cavity.
ANN RILEY & ASSOCIATES, LTD.
Court Reporters 1250 I Street, N.W., Suite 300 Washington, D.C. 20005 (202) 842-0034
210 1 So we have got about three foot of debris on the
() 2 reactor vessel side and about eight-tenths of a foot on the 3 other side of the reactor cavity.
4 We also found that we had the same unequal 5 distribution of oxidic and metallic core debris. What was 6 on the reactor vessel side again was primarily oxidic melt, 7
and what was on the reactor coolant drain tank side was 8 primarily metallic melt.
9 DR. CATTON: Does that make a big difference?
10 MR. LUTZ: Big difference in terms of?
11 DR. CATTON: Well, it's your bottom line, because 12 this is the result of a scenario that goes all the way back 13 to how the thing melted.
14 MR. LUTZ: That is correct. That is correct.
15 DR. CATTON: What would happen if you reversed 16 that? Would it change your bottom line very much or what 17 would happen?
18 MR. LUTZ: No, it really wouldn't change it at 19 all, I don't think. If we said that this side was primarily 20 oxidic and this side was primarily metallic --
21 DR. CATTON: Or if they were layered. Would any 22 of these things change your conclusions very much?
23 MR. LUTZ: My initial reaction would be no, but I 24 am --
25 DR. POWERS: My impression on the analyses that
() ANN RILEY & ASSOCIATES, LTD.
Court Reporters 1250 I Street, N.W., Suite 300 Washington, D.C. 20005 (202) 842-0034
211 1 they were trying to do was to get ultimate penetrations that
() 2 the pressurization were too low to threaten the vessel, in 3 which case the analyses are driven primarily by just the 4 integral of the decay heat curve.
5 But where the distributions of the metal and the 6 oxide make a difference is on the fission product release, 7 but for their purposes that was dictated to them.
8 DR. CATTON: Okay, so it really doesn't matter.
9 DR. POWERS: So it probably wouldn't make any 10 difference at all.
11 DR. CATTON: Well, that is kind of what I thought.
12 DR. POWERS: Typically what happens is melts are 13 basically melts except when you have a reactive metal like 9
14 zirconium, which will drive your temperatures up and create 15 very, very reducing conditions for you. Then you get a 16 certain class of fission products off. The more refractory 17 materials tend to come off, but since their source terms 18 dictate them to them, there is no modification that they 19 would be looking to make for that.
20 MR. PALLA: This is Bob Palla with the Staff.
l 21 I just wanted to all something so the committee is i 22 aware of it.
23 It may not change the bottom line results but one 24 thing that this distribution of melt does affect is the 25 thickness to the liner. The distance to the liner is
() ANN RILEY & ASSOCIATES, LTD.
Court Reporters 1250 I Street, N.W., Suite 300 Washington, D.C. 20005 (202) 842-0034 l
t
l 212 1 different on each side of the cavity.
() 2 In' fact -- not for the localized failure but for 3 the hinged failure mode, it's hypothesized that with the 4 amounts of mass that are involved, which is essentially the 5 whole core plus the core support plate and parts of the core {
l 6 barrel, there is enough melt release that it could actually 7 flow into the sump and so it is in this regard that it is to 8 your benefit to have debris that is mainly metallic so that l -9 it doesn't have the decay heat generation in that side, 10 because the sump is on the far side of the cavity where the 11 less, lower decay heat producing constituents end up, and so 12 that is the only difference but I don't think that that 13 would change the results regarding overall time to --
14 basemat melt-through really.
15 Basemat melt-through was our figure of merit and 16 our criteria rather than liner melt-through so I don't think 17 it would affect that.
18 4 DR. POWERS: That is always an interesting debate 19 on whether the liner melt-through means anything or not. I 20 don't know the answer to it. It's kind of hard to know 21 whether basemat penetration is all that important either.
22 MR. LUTZ: We sort of took the same view and we 23 used both of them as potential indicators of containment j 24 integrity just to see where we stand and I think one can i
25 make qualitative arguments to say that it is somewhere in i
() ANN RILEY & ASSOCIATES, LTD.
Court Reporters 1250 I Street, N.W., Suite 300 Washington, D.C. 20005 (202) 842-0034
l 213 l 1 between the two -- it's probably closer to this one but !
./
2- again they are qualitative, so we looked at the two 3 indicators and we also looked at -- since there is no 4 concrete type specified for the AP600 basemat. ;
5 We looked at the impact of basaltic and limestone 6 as we know one has a faster erosion rate than the other, but 7 the one generates more noncondensible gases, which has an 8 effect on. containment pressurization, so you get a trade-off 9 one way or the other.
10 DR. POWERS: What happens when you -- three to 11 seven or whatever number of days it is -- here it's ;
12 relevant -- in 29 days you penetrate to the basemat. Then )
13 what? I mean what happens?
14 MR. LUTZ: I don't know. I don't know.
y !
l
\_ 15 DR. POWERS: I mean one thing that happens, you 16 can fully well imagine happens, is that there's a lot of 17 water in there and radioactive material that presumably 18 flows into the ground, doesn't it?
19 Does anybody ever worry about --
20 DR. SEALE: It depends on what is under there.
21 CHAIRMAN BARTON: What is underneath the mat.
22 DR. SEALE: Sure. At TMI you had a red sandstone 23 that had purged water in it but the main water table was a 24 couple hundred feet below grade so it was just more basemat 25 for all intents and purposes.
l l
l ANN RILEY & ASSOCIATES, LTD.
I Court Reporters 1250 I Street, N.W., Suite 300 Washington, D.C. 20005 (202) 842-0034
214 1- DR. CARROLL: I am sure their sites over the
~
2 groundwater is fairly high --
3- CHAIRMAN BARTON: Depends on where you build it --
'4 DR. CARROLL:
-- with the inflow of water.
5 -DR. POWERS: You might be interested in some 6 experiments that I think.were done at KWU at which they.
]
7 looked at-what happened when you go from the transition from 8 concrete to earth. What they found was that the core debris 9- cindered the dirt ahead of it prior to melting'it, and so it 10 was just like what Bob said'-- it's just more basemat.
11 DR. SEALE: It looked like concrete'.
12- DR. POWERS: It just looked like more concrete.
13 [ Laughter. ]
14 DR. POWERS: So in the end the only thing that
-( 15 counted was the pressurization that was occurring because in
-16 those particular experiments they really couldn't'tell the-17 difference in when they quit eroding concrete and started 18 eroding dirt.
19 MR. LUTZ: Yes, that's a good point. ~j 20- DR. KRESS: If you can figure out a way to keep 21 the people that type this stuff up from putting.in the word I 22 " basement" instead of "basemat" let me know how you do it,
.23. .because I end up'with that on every one of mine also.
- 24 [ Laughter.] l l 25 MR. LUTZ
- That spell-checker will kill you.
ANN RILEY & ASSOCIATES, LTD.
Court Reporters 1250 I Street, N.W.,. Suite 300 Washington, D.C. 20005 (202) 842-0034
215 1 DR. KRESS: Doesn't help.
( 2 MR. LUTZ: The spell-checkers on these word 3 processors don't --
4 DR. POWERS: If you-go to the button that says l
5 " add to dictionary" -- they click that.
6 CHAIRMAN BARTON: Now that we are through with the 7 basement, where do we go from here?
8 DR. KRESS: Take that out and put in basemat 9 instead.
10 DR. FONTANA: There was no attempt here to look at 11- the effect of water, overlying water?
12 MR. LUTZ: No , there was not. I 13 DR. FONTANA: Okay -- this was just plain dry.
14 MR. LUTZ: This is just how long does it take to 15 'get to-the embedded liner, which is down to .8 feet and the 16 bottom edge of the basemat.
17 What we looked at was the hinged failure and the 18 localized failure modes, and these are the numbers that we 19 came up with.
20 In all cases the end of the basemat was at least 21 several days for the basaltic concrete'and tens of days for 22 the limestone concrete.
23 Now if you go --
24 DR. FONTANA: Do you know what the containment 25 pressure would be at those points?
l (A_/ ANN RILEY & ASSOCIATES, LTD.
Court Reporters 1250 I Street, N.W., Suite 300 Washington, D.C. 20005 (202) 842-0034 I
L 216 l
1 MR. LUTZ: That is just what I was going to grab (f 2 for.
i 3 DR. FONTANA: Sorry about that.
4 MR. LUTZ: Two slides later.
5 What we looked at was Service Level C containment 6 pressure. We never reached it in these three cases at the 7 time of basemat melt-through. In fact, here is the pressure 8 at the time, at'the 3.4 hours4.62963e-5 days <br />0.00111 hours <br />6.613757e-6 weeks <br />1.522e-6 months <br />, the 11 days and 28 days, or 9 whatever the numbers were on the previous slide. These are 30 the pressures.
11 We can see that the only case that we came up to 12 Service Level C was the localized failure mode with 13 limestone concrete that produces lots of noncondensible 14 gases.
15 If you would take Dana's one step further and say 16 that even after it goes through the basemat it retains --
17 you don't have pressure relief, then obviously this one 18 would -- this one isn't far behind this one.
19 DR. KRESS: Did you burn those noncondensible 20 gases?
21 MR. LUTZ: No. No. The containments at these 22 pressures would be steam inert.
23 DR. KRESS: Steam inerted, okay.
24 MR. POWELL: Excuse me. This is Bob Powell of the 25 Staff.
() ANN RILEY & ASSOCIATES, LTD.
Court Reporters 1250 I Street, N.W., Suite 300 Washington, D.C. 20005 (202) 842-0034
i 217 1 I just wanted to point out something to the
() 2 committee, and it has to do with Service Level C.
3 What has been done in the PRA and for 4 calculations -- I'll say for calculations of these ex-vessel 5 loads in comparison to the containment capabilities -- what 6 we normally do is use the fragility curves and as you have i i
7 heard before, some of these loads are like the DCH-related I 8 load is a load the value at which there is any significant 9 probability of failure. i 10 Another measure is Service Level C, and one could 11 say, well, if you are at Service Level C chere is very high 12 confidence that integrity would be maintained as well, and 13 so you could to some degree use Service Level C. and the 14 fragility curves interchangeably.
15 But what I wanted to point out is that there is a 16 disagreement between the Structural Engineering Branch and 17 Westinghouse with regard to what is actually Service Level C 18 for AP600, and it gets down to which sections of the code, I
19 ASME code, are being applied, so I think the statements here 20 are still true in the sense that where you will see Service 21 Level C here you should probably think of a pressure at 22 which the probability of over-pressure failure becomes 23 noticeable. j 24 I think our Service Level C value, the Staff l 25 value, would be 60-something -- 60 and a couple PSI, l
/~
ANN RILEY & ASSOCIATES, LTD.
Court Reporters 1250 I Street, N.W., Suite 300 Washington, D.C. 20005 (202) 842-0034
218 1 DR. FONTANA: Is the basis for Service Level C the
(.
() 2 beginning of yield, if you know?
3 MR. LUTZ: I am not an expert on that. I do not 4 know.
5 MR. SOBROWSKI: This is Joe Sobrowski of the 6 Staff.
7 When we talk about Chapter 3 tomorrow we will have 8 the Structural Engineering people here, so if you want to 9 talk about how we calculate at Service Level C, we can bring 10 it up then.
11 The only point that Bob was trying to make is just 12 to alert you to the fact that there is a disagreement for 13 what service level C is on the AP600, between what the staff 14 says. So if you look in our FSER --
s_-) 15 DR. CARROLL: Is that considered an open item? l l
16 MR. SURIS: No. And the reason that it is not 17 considered an open item is because even with the lower 18 pressure, our calculated 62 psig, and I think that's what it !
19 is, if you look in the ultimate capacity, we have a section 20 in the FSER called the ultimate capacity, and if you read 21 that, it talks about what we calculate and it's in Chapter
.22 19. If you read that, it talks about what we calculate for 23 service level C and why we find it acceptable. And, j 24 basically, the basis is -- it goes back to SECY 93087, and 1 25 for the more likely severe accident scenarios, you don't
/~
( }j ANN RILEY & ASSOCIATES, LTD.
Court Reporters 1250 I Street, N.W., Suite 300 Washington, D.C. 20005 (202) 842-0034 L _ ______-_____ . _ _ _ _ _ _
219
< 1 exceed service level C for I believe 24 hours2.777778e-4 days <br />0.00667 hours <br />3.968254e-5 weeks <br />9.132e-6 months <br />. And what we 2 say is for the more likely severe accident scenarios you do l 3 not exceed 62 psig, so, therefore, we find it acceptable on 1
4 that basis.
5 DR. CARROLL: Has either the staff or Westinghouse 6 submitted a code case to ASME to see who is right?
7 MR. SURIS: Well, actually, if you -- again, this 8 is something that we can ask the structural people tomorrow.
l 9 But if you look in Chapter 3 of the SSAR, there is a 10 discussion. Basically, it revolves around the factor-of 11 safety that you use on the equipment hatches. And if -- q l
12 Westinghouse quotes both values. If you use a factor of 4 13 safety -- and forgive me, these numbers may be incorrect, 14 but the ballpark is 2.5 and 1.67. If you use the more i 15 conservative factor of safety that the staff believes in, it 16 shows the equipment hatches have a vulnerability at about 62 17 pounds. If you use a different factor of safety on thobe 18 equipment hatches, you end up with around 90 pounds.
19 So there's -- the agreement that was reached with 20 the staff was basically, you can go ahead and put in those 21 -- you can go ahead and put in those different service level 22 C values, but our basis, and what believe is service level C l
l 23 for the AP600 is less. And that's -- that's the only point 24 of the discussion that we were trying to make sure that the 25 Committee was aware of.
1 ANN RILEY & ASSOCIATES, LTD.
Court Reporters 1250 I Street, N.W., Suite 300 Washington, D.C. 20005 (202) 842-0034 E _
220 1 MR. LUTZ: Core-concrete interaction. If you 2 look, I think you can conclude that the impact of 3 core-concrete interactions on containment integrity is not 4 really sensitive to the assumed mode of vessel failure. The 5 numbers are in the same order of magnitude for the hinged 6 failure or the localized failure.
7 There is a high probability that containment will 8 fail by basemat penetration as opposed to overpressurization 9 if core-concrete interactions occur and cannot be arrested.
10 And based on the times being significantly more than 24 11 hours1.273148e-4 days <br />0.00306 hours <br />1.818783e-5 weeks <br />4.1855e-6 months <br /> and not on the order of three days for the limiting 12 type of concrete, we have concluded that it is not necessary 13 to specify a concrete type for the AP600 basemat under the l 14 reactor vessel.
r 15 And the last -- I 16 DR. FONTANA: Do you remember if the utility 17 requirements document called for some kind of liner material 18 or any kind --
19 DR. SEALE: Liner?
20 DR. FONTANA: Not a fancy liner, but something 21 other than concrete.
22 DR. SEALE: You mean the basemat?
23 DR. FONTANA: In the basemat.
24 MR. McINTYRE: In the basemat?
25 DR. FONTANA: Yeah. The utility requirement I
ANN RILEY & ASSOCIATES, LTD.
Court Reporters 1250 I Street, N.W., Suite 300 Washington, D.C. 20005 (202) 842-0034
[
221 L
1 documents, is i't silent there? I don't remember if it made 2: - some -recommendation'.
.3 DR. CARROLL: I don't think it did.
4 MR. McINTYRE: I don't think it does. I can tell l 5 you for whatever it down there, that we would be in 6' compliance with it.
7 DR. CARROLL: You are thinking of some refractory 8 ceramic material or --
l 9 MR. McINTYRE: No. No.
10 DR. CARROLL: A core catcher maybe?
11 MR. McINTYRE: No. No. No.
12 DR. CARROLL: How come you guys don't have a core 13 catcher?
14 (Laughter.]
g-% !
(,) 15 CHAIRMAN BARTON: Your conclusions are?' !
16 MR. McINTYRE: Thank you.
17 DR. POWERS: I am going to just ask a question. I 18 Do you consider it, maybe not in your design activity, but 19 it would be a feasibility for a designer -- or for somebody 1
20 that bought one of your plants to use a HAC, a high I 21 aluminous cement for that concrete that is above the liner? l 22 MR. LUTZ: For what reason?
i 23 DR. POWERS: Well, a couple of reasons. One, it 24 is a lot easier to maintain, and whatnot. And it is, of i
25 course, virtually -- relative to ordinary concrete, it is
(%
() ANN RILEY & ASSOCIATES, LTD.
Court Reporters 1250 I Street, N.W., Suite 300 Washington, D.C. 20005 (202) 842-0034
h 222 1 immune to concrete attack and generates very little gas on
() 2 what modest attack does occur. And it has a lot of other 3 advantages for installation. It has some disadvantages, to 4 be sure, but it does have a lot of advantages, and its use 5 -- would there be any reason to prescribe that?
6 MR. LUTZ: Well, the way we stand right now, we do 7 not make a specification for concrete in that area. So, the 8 --
9 DR. POWERS: You could put anything you wanted to 10 in the design.
11 MR. LUTZ: The COL applicant is free to --
12 CHAIRMAN BARTON: And he will look at, Dana, what 13 is the price differential between his limestone or the thick 14 layer --
15 MR. LUTZ: Sure.
16 DR. POWERS: It turns out the price differentia!]
17 are about a factor of 3.
18 CHAIRMAN BARTON: About what?
19 DR. POWERS: A factor of 3. And so it makes it, 20 for a typical construction process, it makes it very 21 unattractive to use it. What you have to do is trade it off 22 against how easy is it to --
23 DR. SEALE: Maintain.
24 DR. POWERS: Maintain and whatnot. And how it has 25 to be poured in this, you know, with a liner coming up and ANN RILEY & ASSOCIATES, LTD.
Court Reporters 1250 I Street, N.W., Suite 300 Washington, D.C. 20005 (202) 842-0034
f 223 I
'l things like that. I was just curious. I like HAC, it's fun A 2 stuff to work with.
(_j 3 DR. KRESS: Is there any specification at all on 4 .the concrete?
I 5 MR. LUTZ: Just on strength, I believe. l 6 DR. KRESS: Strength and--
7 DR. POWERS: Do you have a 5,000 pound requirement )
i 8 for it? !
9 MR. McINTYRE: We are the wrong people to talk 10 about that.
)
11 - DR. POWERS: I'mean that's -- that's real good 12 concrete. I mean it is hard to get that.
13 CHAIRMAN BARTON: That's typical for this kind of j i
14 concrete, 5,000 pounds. l 15 DR. POWERS: Four thousand is required in that, 16 and even that -- I think they went overboard when they l l
17 required 4,000. !
18 MR. McINTYRE: Richard Orr can talk about that 19 tomorrow.
l 20 CHAIRMAN BARTON: We used 5,000 for it, j 21 DR. POWERS: I know people do, but it's -- it's 22 hard to work with that stuff.
23 CHAIRMAN BARTON: Back to the -- l 24 MR. LUTZ: Last slide. We look direct containment 25 heating, didn't see a challenge to containment integrity. ,
() ANN RILEY & ASSOCIATES, LTD.
Court Reporters 1250 I Street, N.W., Suite 300 l Washington, D.C. 20005 '
(202) 842-0034
f l
224 l
1- We looked at ex-vessel steam explosions, did not see a
() 2 challenge to containment integrity. We looked at 3 containment failure from overpressurization from 4 non-condensable gas from core-concrete interactions. We 5 found that if it. occurs at all, it is well beyond the 24 i 6 hour time frame. And no-through of the basemat is well I
7 beyond 24 hours2.777778e-4 days <br />0.00667 hours <br />3.968254e-5 weeks <br />9.132e-6 months <br />, even though no-through the embedded line 8 occurs before 24 hours2.777778e-4 days <br />0.00667 hours <br />3.968254e-5 weeks <br />9.132e-6 months <br />, we don't consider that a real 9 failure mode. Something closer to the end of the basemat, j l
10 which is at least several days, is a more credible criteria {
11 to use for failure. So from the ex-vessel side, based on l l 12 the analyses we have done, we don't see containment 13 challenges that are outside the safety goals that we had i
14 set. j O ~ 15 DR. KRESS: The only other challenge to
)
16 containment is hydrogen and you say that belongs over some 17 other -- it belongs with the PRA and not part of this 18 analysis, or what?
19 .MR. LUTZ: Which one? I'm sorry.
20 DR. KRESS: Hydrogen burns. I mean you don't 21 mention hydrogen on there at all.
22 MR. LUTZ: The hydrogen burns have been looked at, 23 and the amounts of hydrogen are rather substantial within 24 the PRA because they include some recovery sequences where 25 you have got a significant fraction of the core zirc 1
l
() ANN RILEY & ASSOCIATES, LTD.
Court Reporters 1250 I Street, N.W., Suite 300 Washington, D.C. 20005 (202) 842-0034
, 225 1
1 oxidized.
() 2 For all of the ex-vessel cases that we are looking 3 at here, I didn't look at each and every one of them, but 4 just based on experience, I would expect the containment to 5 be steam inerted in every case.
6 MR. SNODDERLY: Excuse me, Dr. Kress. Could I l 7 take a shot at it?
8 DR. KRESS: Yes, j 9 MR. SNODDERLY: Just for SECY 93087, it directs us 10 -- for the staff, that the position taken in 50.34 (f), to l 11 continue with that approach for hydrogen control for severe 12 accidents. And so we reviewed the igniter system that.they 13 supplied in response to 50.34 (f) and the fact that you use 14 the supply system to handle 100 percent metal-water reaction 15 with 10 percent average.
16 DR. KRESS: That's DBA space, right.
17 MR. SNODDERLY: Well, I think that's beyond 18 50.34 (f) . That's 100 percent. For DBA it would be you were 19 governed 50.44, which was say 5 percent metal-water 20 reaction.
21 DR. KRESS: And did that not --
22 MR. SNODDERLY: And that resulted in their 23 providing the passive autolytic recombiners. And for
- 24. 50.34 (f) , severe accidents, they supplied the igniters.
25 DR. KRESS: Yeah. Well, did that not result in ANN RILEY & ASSOCIATES, LTD.
Court Reporters 1250 I Street, N.W., Suite 300 Washington, D.C. 20005 (202) 842-0034
226 1- .any containment failure? That's -- I know the igniters are
() 2 there and it'results in burns. But I haven't seen the 3 resulting pressure.
4 , MR. SNODDERLY: Jim, could you talk about this?
5 MR. SCOBEL: The igniter, the burning at the 6 igniters occurs at a very low concentration in the 7 containment and essentially what you would have would be 8 diffusion flames inside the steam generator compartments 9 coming from the stage 4 ADS valves, because that is 10 primatily where the hydrogen would be coming out. And that 11 really doesn't have much of a pressurization effect on the 12 containment and, thermally, it is not -- it is in a location 13 where it is not a problem to the containment shell at all 14 because it is shielded by the steam generator dog houses.
O
\_ ,/ 15 So with the igniters operating, there's no real pressure 16 transient for the containment, and we look at also when the 17 igniters don't operate.
18 DR. KRESS: When you specify 100 percent of the !
19 hydrogen, do you specify a rate at which that comes out? l 20 MR. SNODDERLY: Jim can probably answer better 21 than I. But I do believe they -- I mean they assumed a rate 22 of release.
23 MR. SCOBEL: The rate of release had more of an l 24 effect on the detonation analysis and the diffusion flame 25 analysis that was done at the IRWST wall. The igniters are j i
()
1 ANN RILEY & ASSOCIATES, LTD.
Court Reporters 1250 I Street, N.W., Suite 300 Washington, D.C. 20005 (202) 842-0034
227 1 on before the hydrogen comes out, and the rate really isn't
() 2 going to have that strong of an effect, because you are just 3 ' going to start burning right there at the exit as it comes 4 out.
5 DR. KRESS: Well, I would assume the rate at which 6 hydrogen comes out fixes the rate at which the energy is 7 release, and then it is a race between loss of heat and 8 pressurization and some way -- somehow the rate seems to me 9 to be important without making the calculation myself. I 10 would have thought a rate -- it was a race between 11 pressurizing the system and getting rid of the heat that is 12 added. So 'I, you know, I don't know, I haven't done the 13 calculation.
14 MR. SCOBEL: When you look at a 100 percent burn, 15 like for a global burn, you are burning 100 percent of the 16 oxygen instantaneously. '
17 DR. KRESS: Well, that's --
18 MR. SCOBEL: That's the rate.
19 DR. CATTON: That's the number he is looking for.
20 What is that?
21 MR. SCOBEL: Okay. That is less than service 22 level C. Based on -- it's about 90 -- it's about 100 psig, 23 somewhere in that neighborhood. So it's -- you would have 24 to look at the report to get the exact number.
25 DR. KRESS: Okay.
ANN RILEY & ASSOCIATES, LTD.
Court Reporters 1250 I Street, N.W., Suite 300 Washington, D.C. 20005 (202) 842-0034
228 1 MR. SNODDERLY: And also, Jim, for equipment
( 2 survivability, you did some calculations with MAAP where you 3 artificially bumped up te get 100 percent metal-water 4 reaction and you had a release rate over time for those 5 loadings too.
6 MR. SCOBEL: That's true.
7 MR, SNODDERLY: So that's what I was thinking of 8 when I said that you had dci.a it.
9 MR. SCOBEL: Okay. Those -- actually, we didn't 10 do -- we had FAI do those calculations. I dun't know the 11 numbers off the top of my head, but they are in the Appendix 12 D of the PRA report if you want to look at that as well.
13 When you have a diffusion flame, when you have a 14 constant burning going on, really, the thing that you are 15 looking at more is temperature than pressurization.
16 MR. SNODDERLY: Dr. Kress, before I guess we get 17 going with the next presentation, I just -- the staff 18 estimates that if you have like a 500 kilopascal second 19 impulse load, we estimate that would be 14 kilojoules of 20 energy and tnen if you assume 9,000 kilograms of meltmass 21 participates, we estimated that that.would result in about 22 .16 percent energy conversion rate. So that's what I think 23 TX -- it's in that ballpark. Okay.
24 CHAIRMAN BARTON: Before we hear from the staff, 25 would the Subcommittee members like to take a break, or do
() ANN RILEY & ASSOCIATES, LTD.
Court Reporters 1250 I Street, N.W., Suite 300 Washington, D.C. 20005 (202) 842-0034
229 1 we just.go through and bang this out?
{A) 2 DR. KRESS: Let's bang out it.
3 CHAIRMAN BARTON: Let's bank it out.
4 Bob, you're on.
5 MRI PALLA: I'm in the Containment Systems and 6 Severe Accident Branch of NRR, and I had: reviewed the level 7 2 PRA and, in particular the supporting external reactor 8 vessel cooling analysis that was a key assumption in the 9 PRA.
10 I just wanted to bacically outline the 11 organization of the various sections of what we have done 12 and the logic that we used to conclude on this topic.
13 Basically, we are going based on an assumption -- on a 14 position that we had taken some time ago now, and that k- % 15 basically was we didn't want to put all our eggs in one 16 basket. We did not want to rely solely on external reactor 17 vessel cooling and' ignore completely the possibility that 18 you could fail. We knew that certain events, you would have 19 potentially not a flooded cavity, so we knew at the outset 20 there was a small probability of failure of the vessel, even 21 if external vessel cooling is 100 percent efficient.
22 And so what we decided to do is look at a 23 combination, a two-pronged approach. We would rely l
24' primarily on external reactor vessel cooling, but we would j i
a
! 25 do a limited evaluation of ex-vescel phenomena to assure ANN RILEY & ASSOCIATES, LTD.
Court Reporters 1250 I Street, N.W., Suite 300 Washington, D.C. 20005 (202) 842-0034 L
230 1 , that the vessel -- that the containment was not vulnerable 2 to early failure if the vessel, reactor vessel was in fact 3 to fail. So we did this in combination, this combined 4 approach.
5 We received several submittals. The key submittal 6 for external reactor vessel cooling is the DOE report 7 prepared by Professor Theofanus. There was a lower head 8 integrity study also prepared by Professor Theofanus that 9 addressed in-versel steam explosions, and there were two 10 other DOE reports that went along with the steam explosion 11 reports that -- I think they were also DOE studies, and they 12 addressed the propagation and the pre-mixing aspects of fuel 13 coolant interactions, in-vessel fuel coolant interactions.
14 There was an additional DOE report that addressed
,O
(-) 15 the spreading of melt in the cavity. This is an Argonne 16 study and this is the work that Bob Lutz described. It is a 17 study using the thermal code, the melt spread code and the 18 MAAP code to assess the integrity of the vessel floor, given 19 that the vessel -- that the reactor vessel fails.
20 In Appendix B is where you will find all of those 21 external -- the deterministic calculations and it was a key 22 part of our study. In numerous rounds of RAIs that evolved 23 over a five years plus period and, in fact, change -- as a 24 design change, some of these RAIs are changing. But this is 25 where you will find the key information in your review, that n
() ANN RILEY & ASSOCIATES, LTD.
Court Reporters 1250 I Street, N.W., Suite 300 Washington, D.C. 20005 (202) 842-0034
231 1 we looked at. And our technical reviews are broken into
() 2 :section in terms of phenomena.
3 External reactor vessel cooling is described in 4 the FSER, but there are also two key documents. I believe 5- the Committee has them, and if you don't, you should be 6 aware. There was an internal review of external vessel 7 cooling done within the Office of Research. That -- you 8 have got that already? Okay. It was a technical evaluation 9 report from INEL. The two key documents, if you need more 10 information, I have just identified the different sections 11 in the FSER.
12 I should point out-that for external vessel 13 cooling -- for external steam explosion -- external vessel
. 14 steam explosion, there is an ERI report that was done under D
\_, 15 sponsorship of the Office of Research, and that entered into 16 our evaluation as well.
17 I want to just jump to the conclusions that you 18 will find in the FSER. I am going to summarize what went 19 into the thinking here. To summarize, we believe that the 20 reactor pressure vessel will remain intact, given that the 21 coolant system is successfully depressurized and the cavity 22 is flooded, in accordance with the success criteria,, which 23 essentially says you dump the entire IRWST, it would bring 24 the water level up to the level of the nozzles of the l 25 vessel.
l ANN RILEY & ASSOCIATES, LTD.
Court Reporters l 1250 I Street, N.W., Suite 300 Washington, D.C. 20005 (202) 842-0034
232 1 We say that based on several studies that were f
i 2 done. Now, there was a study done, a calculation done with 3 the SCDAP code. And, I guess, Dana, this comes as close as 4 anything comes to addressing the kind of issues that you 5 talked about the complexity of the meltdown progression and 6 utetic formation, et cetera. Our view is that SCDAP 7 represents the embodiment of most of the knowledge that we 8 have of the melt progression process. And what was done was 9 INEL performed an AP600 specific calculation for a 3BE 10 sequence and looked at this model -- they used the detailed 11 lower _ head model, which nodalized it into -- I forget the 12 exact number of elements. And then they looked at the heat 13 transfer into the wall and out of the wall to the 14- surrounding water. They, within that lower head model, they 15 incorporated their view of the appropriate heat. transfer 16 correlations to use for the molten pool in the lower head.
17 And for the CHF correlation on the outside, they used the 18 available Penn State CHF data at that time.
19 Now, we compared that data with the data that was 20 available also from the OPU facility that Theofanus had l 21 developed, and we decided to use the Penn State data as a 22 representation of the CHF.
23 This calculation was a transient calculation in 24 nature. The SCDAP calculation predicted a series of 25 discrete debris relocations essentially following the same ANN RILEY & ASSOCIATES, LTD.
Court Reporters 1250 I Street, N.W., Suite 300 Washington, D.C. 20005 (202) 842-0034
233 1 sequence progression as was hypothesized or determined in lh 2 the DOE report, perhaps not based on codes but based on 3 qualitative assessments.
4 We, I would say for the large part, confirmed the h
5 general nature of the progression and the actual failure 6 locations where the debris would leave the in-core pool and 7 be poured into the lower head.
8 We observed, and it is an artifact of the code, N
9 these discrete relocations and we used the code and the 10 data, if you will, the information about the constituents of 11 these different pours, we used that to really structure some 12 of our thinking about the potential for these, for other 13 debris states that might form, but in essence, the code has 14 other -- SCDAP/RELAP will give you the relocations into the 15 lower head and then, as I mentioned, there was this lower 16 head model that was added to handle the heat transfer in the 17 lower head, but that model had some very serious 18 limitations.
19 It did not do any kind of this layering that we 20 have hypothesized, the metals on the top or on the bottom.
21 It basically mixed the debris up uniformly. It did have a 22 crust -- I mean it does have the models that on the 23 periphery and adjacent to the vessel you will have a crust 24 and you will have a molten inner pool where you will apply 25 your correlations to determine the heat transfer to the --
ANN RILEY & ASSOCIATES, LTD.
Court Reporters 1250 I Street, N.W., Suite 300 Washington, D.C. 20005 (202) 842-0034
234 1- towards the outside.
' (b) 2 But that-is one limitation. You know, it's a 3 transient calculation but it does not address the metal 4 layer on the top or on the bottom or -- it's basically a 5 mixed pool, but it-told us that if you did achieve basically 6 a uniformly-mixed pool in the long term -- actually, I'd say 7 all along the way -- it's a transient calculation - .we did 8 not show any failures, any heat fluxes that exceeded --
9 DR. KRESS: Did you calculate a rating depending 10 on how much debris was down there?
11 MR. PALLA: I think it did. I am not sure 12 geometrically how it does that, but --
13 DR. KRESS: Probably made it look like a finite 14 element.
15 MR. PALLA: I think it might have made it look 16 like a sphere in a sphere. It did something with the 17 geometry --
18 DR. KRESS: The rating number gets heat transfer 19 coefficient and gets heat flux, adjusts the crust thickness 20 according to that -- heat thrust -- picture a boundary at ,
21 the crust during the meltdown.
22 MR. PALLA: Precisely. It would do that. You 1 j 23 would have a crust and that crust thickness could change l 24 over time and -- the problem was when you get another I 25 relocation --
f-()- ANN RILEY & ASSOCIATES, LTD.
Court Reporters ]
1
[ 1250 I Street, N.W., Suite 300 I
Washington, D.C. 20005 i (202) 842-0034 L__________.__. _
l 235 )
i 1 DR. KRESS: -- you have to swell all that up and
() 2 do it again, I guess.
3 MR. PALLA: I think realistically you would .
4 actually quench and then you would have water in the lower 1
5 head when this happened. You would quench and then you 6 would have another relocation and another relocation.
7 It takes you until you reach the core support 8 plate before you have dried out the water that is down 9 there. There is water there up until the time of just the 10 event -- essentially the last of these five relocations, 11 which coincided with the time that it took to get to the 12 core support plate.
13 That is about when the water was dried out, but we 14 did not have that quench model operable. It was essentially !
15 -assumed that there was no quenching. This is an attempt to 16 represent it in a bounding and conservative fashion, so 17 there was no quenching assumed.
18 But that model, that calculation was one datapoint 19 that we used to say, yes, this looks like the best tool that 20 we have here for looking at melt progression, using that
- 21. data, putting it in the lower head, running a transient 22 calculation. Subject to the limitations it looks like it 23 will stay below critical heat flux, i
24 The second thing we looked at was the hypothesized 25 configuration that was proposed in the DOE report, the i
ANN RILEY & ASSOCIATES, LTD.
Court Reporters 1250 I Street, N.W., Suite 300 Washington, D.C. 20005 (202) 842-0034 i
_ ____-- - _ _ _ 4
236 l
1 metals on the top, and essentially all of the steel in the f~
(s) 2 vessel. You are putting in the core support plate and the l 3 reflector and the chunk of the core barrel, so huge amounts 4 of steel.
5 If we get there, it looks like things -- it is 6 pretty robust, there's margins to failure and I indicate 7 even with expanded treatment of uncertainties.
8 What we did is we were tasked by INEL to look very 9 carefully at the DOE report, at the peer review comments on 10 the report, and on how those peer review comments were 11 dispositioned. We also asked them -- we knew there would be 12 some residual concerns. We knew that not all of the peer 13 reviewers' comments were addressed. We knew that heat 14 transfer correlations were assumed to be a given without any
(_s) 15 consideration of uncertainties.
16 There wasn't any internal heat generation in the 17 metallic layer that you could expect to be there from 18 fission products, so we modified -- we started with the DOE 19 model but we added on certain parameters. We shifted 20 uncertainty distributions to account for earlier times to 21 core relocation for example was one shift.
22 We changed the massivity values in more 23 conservative directions in general, but even when all that 24 was done, if you reached the final bounding state, there's 25 margin of failure still.
(O) ANN RILEY & ASSOCIATES, LTD.
Court Reporters 1250 I Street, N.W., Suite 300 Washington, D.C. 20005 (202) 842-0034
237 1 The last point there is that in-vessel fuel h 2 coolant interactions are not expected to fail the reactor 3 pressure vessel also, based on the analyses and the caveats 4 for the scenarios that we looked at. The ones that we think s
5 are more likely we don't expect that the vessel will fail 6 from a steam explosion.
7 The second point I wanted to make was that we 8 looked as as part of the INEL study the potential to form
\
9 other debris beds, and if you did form it what would the 10 effect on the heat loads be?
11 We did not try to quantify any probabilities or 12 assign any probabilities for forming alternative debris 13 states, but we simply recognized that if debris is going to 14 be discharged into the lower plenum in a series of 15 relocations, if it quenches it takes up -- there is a 16 voiding that occurs, that essentially if you voided with a 17 50 percent void fraction you could basically fill the lower 18 plenum, lower head of the vessel up with about half of the 19 core, and then the balance of the core would come down on 20 top, so we thought there was a potential for loads to occur 21 in the interim time period and there is also a potential for 22 things to occur such as a sandwich steel layer.
23 We thought if you had these sequence of pours, if 24 there was quenching, if you fill the lower head with 25 quenched and articulated debris and then you put the I ANN RILEY & ASSOCIATES, LTD.
Court Reporters 1250 I Street, N.W., Suite 300 Washington, D.C. 20005 (202) 842-0034
238 1 balance of the core on top maybe you could have a steel ll 2 layer and maybe the core support plate in fact sandwiched in 3 between two heat-producing parts of the debris bed, in which 4 case you would heat the metal layer from the bottom and heat 5 it from the top.
6 So we looked more carefully at these alternative 7 configurations, treating them as hypothetical because there 8 wasn't really any evidence to say that they in fact formed.
\
9 What we found is that in all three cases we expect that you 10 could exceed critical heat flux.
11 Now I do want to point out some of the limitations 12 in this calculation. As I indicated, the SCDAP calculation 13 was a transient calculation, but our calculations for these 14 alternative debris states assumed that it was a steady state 15 situation. We basically took a snapshot of the situation 16 during a transient and said if we in fact had this condition 17 occur in the long-term and if this steady state solution is 18 valid, then these are the heat fluxes that we would get.
19 In fact, it has not been established that these 20 debris bed configurations would form at all, and if they did 21 that they would persist a long enough time that a steady 22 state solution is valid, so there are limitations there from 23 the point of view that it is really a steady state 24 characterization of what we know is a transient state.
25 Another limitation is -- and this is really ANN RILEY & ASSOCIATES, LTD.
Court Reporters 1250 I Street, N.W., Suite 300 Washington, D.C. 20005 (202) 842-0034
l 239
( 1 relevant for the sandwiched steel layer -- we only looked at r
() 2 one dimensional-heat conduction in the reactor vessel and if 3 the layer is very thin you obviously have the ability to 4 remove heat. The vessel is essentially thin in that area 5 and if it'is surrounded by water you could dissipate some 6 additional heat through the metal that is above and below 7 the elevation'of the metal layer.
l 8 Another conservatism that is important to realize 9 is that we had, we used the Penn State data for critical l 10 heat flux that was available at the time of the calculations 11 and this data did not include the effects of the insulation, 12 thermal insulation.
13 If you had the thermal insulation, it appears that O
14 critical heat flux is substantially higher, most noticeably k/ 15 in the lower elevations but even in upper elevations there 16 is a shift.
17 DR. CATTON: You measured it higher?
18 MR. PALLA: They saw higher CHFs even all the
- i l 19 way -- 90 degrees.
20 DR. CATTON: You need pretty high velocities to l 21 impact the CHF very much.
22 What are the velocities?
23 MR. PALLA: I am not sure what the velocities are. !
24 I know that they were observed visually. I don'c know that 25 they were measured in the tests.
l
( )' ANN RILEY & ASSOCIATES, LTD.
Court Reporters 1250 I Street, N.W., Suite 300 Washington, D.C. 20005 (202) 842-0034
1 l
240 1- DR. CATTON: See, right on the bottom it's
/%
( ,) 2 stagnation. That's very different. Up on the wall it's 3 just flow past the surface, and if you look at some of the 4 CHF data that you can find where there's flow you see that 5 the force convection curve just intersects the nucleic 6 convection and the CHF doesn't change a whole lot until you 7 really have high velocities, and you don't have high 8 velocities that were that high.
9 MR. PALLA: I thought that what might be happening 10 is by putting that vessel insulation -- it's about six 11 inches or nine inches from the surface of the vessel and it 12 is conical at the bottom -- water enters in at the bottom-13 and then flares out, goes around. It was my impression that 14 this could be entraining a lot of liquid and just sweeping 15 it in that narrow gap.
16 DR. CATTON: It's a function of the velocity and 17 you don't need any new experiments to evaluate that. There 18 is lots of data.
19 MR. PALLA: And I think the velocities obviously 20 must be increasing in order to increase the CHF that much.
21 It is noticeable. As I mentioned before, the Penn 22 State data is limited in a sense that the test vessel 23 couldn't be heated high enough to even measure CHF at 24 certain locations. It exceeded the capabili t y of the j 25 vessel, so it is something even higher than what was O ANN RILEY & ASSOCIATES, LTD.
( /
Court Reporters 1250 I Street, N.W., Suite 300 2 84 I b3 i
241 2 measured.
() 2 Okay. So there are limitations in what we did but 3 what we -- we're basically unable to conclude still that 4 these configurations can be excluded completely so it.is our 5 view that while we say that we expect the vessel to remain-6 intact we cannot conclude, we don't feel that we could say 7 that that would be the case for all scenarios. We just 8- think that the situation is far more complex, the ability to 9 fully model the formation of the debris bed starting with 10 the quenched debris and moving on through to the formation 11 of fully established natural circulation in a molten pool, 12 we just weren't in a position to buy that, especially in 13 view of what we had seen for these hypothetical debris bed 14 configurations.
O
\~ / 15 DR. CARROLL: Have you sat down with Theofanous 16 and asked him to respond to these concerns? He seems very 17 positive in his report that these things are physically 18 unreasonable.
19 MR. PALLA: Well, our interactions really are with 20 Westinghouse. We did provide the INEL report to 21 Westinghouse and we did not pursue these.
22 I think it would be hard for us to have been 23 convinced otherwise. I think there is a lot of modeling 24 that would be needed. I still think that there would be 25 residual uncertainties and I think it would be -- I think
() ANN RILEY & ASSOCIATES, LTD.
Court Reporters
- 1250 I Street, N.W., Suite 300
' Washington, D.C. 20005 (202) 842-0034 l
u________________ _ _ _ _ _ _ _ - . _ _
i i
242 ,
1 the first and the last bullet would'probably still stand
) 2 even if a. case is made that some of these configurations are
-3 not viable.. I 4 DR. POWERS: I think we have to be very 5 sympathetic with that point of view, that there would be 6 residual uncertainties concerning core degradation 7 processes.
8 You have outlined many of the things that cause 9 you pause about SCDAP analysis, the most recent of which is, i
10 yes, it's having troubles modeling some aspects -- core 11 degradation in experiments with irradiated fuel -- because 12 of the limitations of some of its chemical modeling and I
-13 phase diagram modeling.
14 I think if we had staff here from the RASPAV (n.
\ 15 . program they would be anxious to tell us that modeling some.
16 of these chemical interactions is proving to be quite 17 challenging even for dedicated codes.
18 MR. PALLA: It's hard to get the tests to come 19 off.
20 DR. POWERS: Well, I think they are getting tests.
21 They just don't know what to do with the results exactly.
22 So I think you have to be fairly suspect when !
23 someone tells you that the configuration of debris and how 24 it exactly gets down in the head is known with any kind of 25 positiveness and it makes a difference. '
A Q ANN RILEY & ASSOCIATES, LTD.
Court Reporters 1250 I Street, N.W., Suite 300 Washington, D.C. 20005 (202) 842-0034
243 1 My own feeling is that there are real strong
() 2 reasons to think that metallic oxide layer inversion is at 3 least possible, just based on the chemistry. Now whether it 4 actually occurs or not is something that, gee,.only an 5 experiment would tell you the difference one way or another.
6 MR. PALLA: I think our view really, and you can 7 see this in the FSER, is that basically we are looking to 8 the RASPAV experiments as really, you know, the place that 9 this could be answered and it may not answer it still.
i 10 In whatever answers are produced I suspect there 1
11 will still be uncertainties in this area, but we were just l
12 unable to preclude these from occurring. We think that 13 tests like RASPAV could give us more information to judge I i
14 that, and we are surely in no position to assign !
15 probabilities to it, so this last statement is really a 16 statement that I think we would be hard to defend anything 17 other than that.
18 DR. CATTON: RASPAV is not going to address the 19 scenario issues.
20 MR. PALLA: It is going to address things like 21 interactions --
22 DR. CATTON: It will address the chemistry but it 23 won't address some of the convection type problems.
24 MR. PALLA: It might address the questions about 25 what is going to be on the top, what is going to be on the
() ANN RILEY & ASSOCIATES, LTD.
Court Reporters 1250 I Street, N.W., Suite 300
! Washington, D.C. 20005 (202) 842-0034
244 1 bottom.
/'N
() 2 DR. CATTON: That depends which test you are 4
3 talking about. The one where they stick that big electrode l
4 down the electric fields and the currents are going to stir 5 that to a level? It's going to be mixed.
6 MR. PALLA: There's definite limitations, yes.
7 DR. CATTON: It's going to be mixed.
8 DR. POWERS: I think there's been some evolution 9 in the thinking and the design on the RASPAV total 10 experimental program.
11 I think the prototypic material type of l
12 experiments are more directed toward these chemical issues 13 Bob is talking about and less toward quantification of the I i
14 actual heat transfer. I think they are relying on the l 15 thermal hydraulics community and their simulants to give 16 them the heat transfer and to augment and change that based 17 on whatever chemistry they see.
18 Unfortunately, they are seeing some really i
19 complicated chemistry, but the experiment configuration is !
20 such that it is very difficult for them to address steel 21 oxide type melts. They are working primarily with 22 hypostroicimetric oxides -- they see stratification, by the 23 way, into two layers.
24 DR. KRESS: I think it was prudent of the Staff to 25 use a true defense-in-depth view of this issue and require l
( j ANN RILEY & ASSOCIATES, LTD.
Court Reporters 1250 I Street, N.W., Suite 300 Washington, D.C. 20005 (202) 842-0034
245 1 an additional look as if the vessel was penetrated, but what
() 2 I see is that that look that Westinghouse took showed that
- 3. the system was still robust, even though penetrated, and as 4 far as I could see it didn't require any additional --
1 5 didn't lead to any additional requirements.
l 6 MR. PALLA: That's right.
7 DR. KRESS: So all it is is a prudent look at 8 analysis and gives you a nicer feeling about the --
9 MR. PALLA: We are left with uncertainties. We 10 think it is going to stay intact but it might not --
11 DR. KRESS: If it does, you are still pretty well i 12 satisfied with the consequences.
l 13 MR. PALLA: If it does, you are done, but if it 1
14 doesn't then we say, well, what do the deterministic I
.O
\-s/ 15 analyses show us? Well, the deterministic analyses show us 16 that you are going to probably maintain containment l 17 . integrity still.
18 Even though the PRA conservatively assumed that 19 vessel failure equals containment failure, the deterministic 1
j 20 calculations say no, that the containment is very likely to 21 stay intact.
22 DR. POWERS: What I don't understand is that you 23 have core melt accidents and you presume you retain 24 in-vessel but maybe you don't. It changes that probability 25 as well.
~3 g) ANN RILEY & ASSOCIATES, LTD.
Court Reporters 1250 I Street, N.W., Suite 300 Washington, D.C. 20005 j (202) 842-0034 L __ _ _ _ _ _ _ _ - _ _ _ _ _ _ - _ .
246
- 1. Yes, maybe your containment is robust, but maybe
() 2 you are challenging it more frequently too.
3 What I am saying is that the results of the PRA at 4 least on the consequence side aren't transparently obvious 5 to me now, that's correct, and if the vessel is going to 6 fail every single time -- as opposed to never -- the number 7 has got to change.
8 DR. KRESS: That may be the --
9 DR. SEALE: --
the vulnerability.
10 DR. KRESS: We need to look at the PRA.--
11 MR. PALLA: Now which assumption are you speaking 12 of?
13 DR. POWERS: Now if I assume that when I core 14 debris down into the lower head I do not fail the vessel. I 15 get the current results.
16 MR. PALLA: That's right. >
17 DR. POWERS: If, on the other hand, I fail the 18 vessel every time I put core debris down there, I am going 19 to get some other result.
20 MR. PALLA: Then I guess the question is what --
21 now if you went to the PRA, which was conservative because 22 they had the latitude to be conservative -- they had no j 23 significant vessel failure frequency -- so you could say 24 when I fail my vessel I'll just assume my containment always 25 fails. If you take that approach and say now every time j 1
() ANN RILEY & ASSOCIATES, LTD.
Court Reporters j
l 1250 I Street, N.W., Suite 300 i Washington, D.C. 20005 (202) 842-0034 l
l
'247 1 that I relocate.into the lower head I am going to fail the
() 2 vessel'and then'I fail the vessel and I am going to fail the 3 containment, then you basically say my containment failure 4 frequency equals my core melt frequency.
5 That-is actually what we said in the second bullet 6 here. Because you have got such a low core damage frequency 7 in this plant, you could even do that. You can say that 8 'every time that I relocate into the lower head I am going to i i
9 not take any credit for external vessel cooling.. I am going j l
10 to fail the vessel and then I will even'be as conservative 11 as I was in the.P1UL and I will fail the containment and you l i
12 still will stay below the large release goal, thankfully !
13 because of the low frequencies that you went into thi's thing 14 with.
15 Now you could take an alternative approach and say let's scrap that idea about taking the containment out every 16.
17 time that I-fail the vessel. If you had to go back and 18 sharpen your pencil and then say external vessel cooling 19 doesn't always work -- I am going to fail some part of the 20 time -- then it is logical to look ex-vessel and to take
]
2 11 credit for some of the robustness that we realize we had !
22 not, and I think some of the reasons, and Jim Scobel could i
~23 correct me with regard to why some of these assumptions were 24 made, I think the major PRA update that was made in which 25 they went to the conservative assumptions about RPV failure ;
s
( ANN RILEY & ASSOCIATES, LTD.
Court-Reporters 1250 I Street, N.W., Suite 300 i Washington, D.C. 20005 (202) 842-0034
248 1 equals containment failure, high pressure core melt equals
() 2 3
steam general tube' rupture -- this was a major PRA revision and simplification that was made before the deterministic 4 calculations had been performed, so there really was not a 5 good basis for sharpening the pencil and there was enough 6 room to make the conservative assumptions and that was what 7 was done.
8 Then, subsequently, the deterministic calculations 9 were performed. The margins appear to be there and if one 10 were to go and redo this analysis you would want to take l 11 credit for those ex-vessel calculations and perhaps the 12 probability of vessel failure would be offset by now the 13 probability that the containment is able to survive that.
14 DR. CATTON: You get low numbers based on 15 calculations you know how to do, so you get to stay away 16 --or you put yourself in a position you don't have to worry 17 about a lot of the science fiction, and that's good.
18 MR. PALLA: So, anyway, we did use deterministic 19 calculations. We went into it with a recognition of the low 20 probabilities and the ability to even meet safety goals 21 given the worst-case assumptions, and our conclusion is 22 we'll accept the characterization of the external vessel 23 cooling based on a combination of those, and that's all I 24 need to say.
25 DR. CARROLL: The saga doesn't end here, however, l
ANN RILEY & ASSOCIATES, LTD.
Court Reporters 1250 I Street, N.W., Suite 300 Washington, D.C. 20005 (202) 842-0034 E--.-.----_ - - - - _- - - - - - - - - - - - . . - - - _ - - - - - _ - - - - - - _ - - - - - - - - - _ _ ----- _-- - _--------
m- i l'
249 i
1 on internal vessel cooling. I can imagine that a number of j
() 2 operating plants, whatever, for one reason or another may be 3 coming in and saying, hey, we would like to take credit for 4 this, and it's really important to continue looking at the 5 phenomena, because obviously they have a tougher problem.
6 It's closer to --
7- MR. PALLA: It's a good point. There are several 8 plants that have taken credit for this in their individual 9 plant examinations, Commonwealth. plants.
~ 10 In addition, each of the PWR owner's groups in 11 their severe accident management guidance, they identify 12 reactor cavity flooding as a severe accident management 13 strategy. Now, -- and I think what you say is exactly 14 correct.
15 The way we view AP600 is kind of the leading edge.
16 If this strategy doesn't work here, it won't work anywhere, j i
17 because this design is amenable to it for the reasons Jim 18 Scobel outlined. It's got the stand off - the insulation is 19 very supportive of the concept, as well as the power density; 20 and the clean lower head. You go to an operating plant with 21 close clearances and you're counting on ingression; it may 22 be quite misleading to have people going on the assumption 23 that this is going to work in operating plants when they've 24 got higher power densities, penetrations, tight insulation, 25 maybe can't even vent the steam if they can get it in there.
() ANN RILEY & ASSOCIATES, LTD.
Court Reporters 1250 I Street, N.W., Suite 300 Washington, D.C. 20005 (202) 842-0034
_ _ - - _ _ _ - - _ _ - - _ . - - - - _ - - ------.--_-_a
250 1 DR. CARROLL: Yes, but I would imagine that people
() 2 would change out their insulation to something compatible 3 with this concept. It's not hard to do.
4 MR. PALLA: I don't see that happening.
5 CHAIRMAN BARTON: I don't see people going through 6 that effort, Jay. I really don't.
7 MR. PALLA: I think that some plants -- and maybe 8 Bob Lutz could clarify if you're interested, but some plants 9 had considered whether they wanted to make their cavity 10 floodable, and I think they're deciding against that. So to 11 my knowledge, no utilities are moving to make physical plant 12 modifications in this area.
13 CHAIRMAN BARTON: No, I don't think you're going 14 to find anybody coming in willing to go through that change, O 15 no.
16 MR. PALLA: And then you have to weigh the 17 potential, am I buying into a steam explosion in ex-vessel.
18 CHAIRMAN BARTON: What are they really gaining by l 19 going through that effort?
20 MR. PALLA: Okay.
21 CHAIRMAN BARTON: Is that it?
l l 22 MR. PALLA: That's it. !
23 CHAIRMAN BARTON: Thank you.
24 Any questions of the staff or Westinghouse on the 25 material we covered today? We did cover a lot of material.
ANN RILEY & ASSOCIATES, LTD.
Court Reporters i 1250 I Street, N.W., Suite 300 l Washington, D.C. 20005 (202) 842-0034
251 1 If you've got questions or some input to a report, please
(( "%,J . 2 get it to me or Noel while it's fresh in your mind.
3 No other discussion or --
4 MR. McINTYRE: Before you gavel this to a close, 5 can we go over the Dudley list for the things that we should 6 have covered today and before some of the Westinghouse 7 people --
8 CHAIRMAN BARTON: We can do that if you're 9 prepared to do that.
10 MR. McINTYRE: Yes. I think it would be a good 11 idea before some people scoot out of here.
12 CHAIRMAN BARTON: Okay.
13 MR McINTYRE: Are the Westinghouse people ready?
14 Terry, would'you like to step to a microphone here, because 15 your name is next to several.
16 I'm just going to go down the list and I'll 17 identify them by the date and the question, and the 18 Westinghouse people knew that they should have somehow 19 addressad this, so we'll see if you guys think that they 20 have and make sure that they have.
21 The first one was from 1/11/95. Mr. Schulz was 22 requested to provide a SSAR reference on how the AP600 23 design improved the in-service test program.
24 MR. SCHULZ: This is chapter 3, right, that 25 three-nine-six talks about in-service testing. It lists
() ANN RILEY & ASSOCIATES, LTD.
Court Reporters 1250 I Street, N.W., Suite 300 Washington, D.C. 20005 (202) 842-0034
252 1 what is done on AP600. It does not explain why it is better l /^
( )s 2 than current plants, and that's typical in the SSARs. We l 3 don't talk typically about how the design is better or worse 4 than -- of course not worse, but any better than existing 5 design; we just explain what the design is and what is done.
6 So I did talk today about several aspects that I 7 thought where we had done improvements in terms of check 8 valve testing and things like that when I was talking about 9 the passive core cooling system.
10 So the only SSAR reference that can be provided is 11 396, which does address in-service testing.
12 DR. CARROLL: Moving on.
13 MR. McINTYRE: Phil is not here anymore, but is
,, 14 that a satisfactory response?
(_s/ 15 DR. SEALE: Well, other thinge are happening in 16 in-service testing right now, too. There's a Wr .ttinghouse 17 owners' group effort in that area involved in a pilot plant.-
18 and that may turn out to be influential in the long run.
19 MR. McINTYRE: The next one was a Carroll 20 Michelson question, requested -- it was also from that same 21 meeting -- requested design description of event, vacuum 22 breakers in the top of the IRWST.
23 MR. SCHULZ: I did try to talk about that when I 24 had the -- that sort of sectional view of the containment 25 and pointed out the vents in the louvers.
(e
'~'
s) ANN RILEY & ASSOCIATES, LTD.
Court Reporters 1250 I Street, N.W., Suite 300 Washington, D.C. 20005 (202) 842-0034
253 1 DR. SEALE: Yes.
() 2 MR. McINTYRE: Carroll Michelson, same meeting, 3 requested explanation of what effect the freezing of the 4 drain and/or supply lines would have on the containment 5' cooling systems.
6 DR. SEALE: Covered that.
7 DR. CARROLL: Covered that pretty good.
8 EMR. McINTYRE: The next one is Gene Piplica.
9 Verify that the initial test -- this was a Jay Carroll 10 question from 5/31/95 -- verify that the initial test 11 program for the instrument air system is more extensive than
.12 for previous designs. I don't think he covered more 13 extensive, but he did say that it is covered,. and we talked 14 about the slow leaks.
15 DR. CARROLL: Yes. That's what I had-in mind.
16 MR. McINTYRE: Okay.
17 DR. CARROLL: What happened to the one before 18- that?
19 MR. McINTYRE: That's a Richard Orr question.
s 20 DR.' CARROLL: Oh, okay.
21 MR. DUDLEY: The one even before that, request 22 explanation for the potential of -- I
- 23. MR. McINTYRE: That's Richard Orr tomorrow. .
1 24 MR. DUDLEY: All right. I thought we heard some i 25 of that today.
' () ANN RILEY & ASSOCIATES, LTD.
Court Reporters 1250 I Street, N.W., Suite 300 Washington, D.C. 20005 (202) 842-0034
254 1 MR. McINTTRE: The potential for creating a vacuum
() 2 in containment?
3 DR. CARROLL: Yes, I did, too.
4 CHAIRMAN BARTON: If we have the expert tomorrow, 5 we'll hear it tomorrow.
6 MR. McINTYRE: The next one is 7/19/96, requested 7 additional information on the effects of a partially filled 8 cavity on containment failure probability.
9 DR. FONTANA: That's answered, essentially.
10 MR. McINTYRE: Down at the bottom, at least on 11 this page, is a 2/3/98 question from Jay Carroll requesting 12 additional information regarding the check valve testing 13 program.
14 DR. CARROLL: I'm happy. I'm happy.
O
(, / 15 MR. McINTYRE: Do you think you need that report, 16 Jay?
17 DR. CARROLL: No. l 18 MR. McINTYRE: The next one is a 2/3/98 question 19 f rom the subcommittee, and it was a Terry Schulz question, 20 or we gave it to Terry. Westinghouse agreed to determine 21 why containment hydrogen levels are displayed in the MCR but 'q 22 are not alarmed. {
23 MR. SCHULZ: Okay. I didn't really speak about l 24 that. They are alarmed. The SSAR mentions that. They have ;
25 high alarms. 1 I
1 J
ANN RILEY & ASSOCIATES, LTD.
Court Reporters i 1250 I Street, N.W., Suite 300 l Washington, D.C. 10005 i (202) 842-0034 ;
I
255 1- DR. CARROLL: Didn't we find at least what we
() 2 reviewed, that there was no indication that they were?
3 MR. SNODDERLY: The confusion may be in. Chapter 4 18, the list of minimum inventory controls and displays, it 5 is not listed as being alarmed, but it does say, as Terry 6 said, in the SSAR and in our SER that it is alarmed, but it
.7 didn't -- it's not included in that minimum inventory 8 listing. The staff found that acceptable for the reason 9 that there is no action associated with hydrogen 10 concentration except for accident evaluation. There's no 11 action that igniters go on because of high thermocouple 12 ten.perature and recombiners are self-initiating.
13 MR. McINTYRE: Mr. Scobel has suggested I back up.
14 There was one 7/19/96 where the Subcommittee requested (O
m/ 15 analyais of steam generator tube response during severe 16 accident temperature and pressure conditions.
17 MR. SCOBEL: This is Jim Scobel from Westinghouse.
18 That goes back to the assumption in the PRA that high 19 pressure core damage results in steam generator tube l 20 failure.
21 DR. SEALE: And you made that a one, didn't you?
22 MR. SCOBEL: That was a one.
23' DR. SEALE: Threw in the dice.
24 DR. POWERS: But the additional complication of 25 'the DF of 100 on the secondary side of that, I mean leaves
() ANN RILEY & ASSOCIATES, LTD.
Court Reporters ;
1250 I Street, N.W., Suite 300 )
, Washington, D.C. 20005 l (202) 842-0034 i L-____________-_________- _ _ - - - _ _ _ ._ _ _ _ _ _. . _ _ _ _ _ _ _
256 1 that question clearly alive, doesn't it? I mean what I
() 2 don't understand is how we have a confidence of a DF of 100.
3 MR. SCOBEL: The PRA also provides a sensitivity 4 if there is no DF for that case.
5 DR. POWERS: You're right. You did mention that.
6 MR. McINTYRE: ?/3/98 Subcommittee, Westinghouse 7 agreed to determine whether the containment hydrogen level 8 indicators are diverse actuation devices.
l 9 MR. SCHULZ: Th19 is Terry Schulz. I spoke to the l
10 fact that the design does not use diverse sensors in 11 monitors, hydrogen monitors.
12 DR. CARROLL: But they are not actuating anything, 13 are they?
14 MR. SCHULZ: That is correct. I am not quite sure 15 we understood the real question there.
16 DR. MILLER: I asked the question again today. I l
l 17 don't think I asked it on 2/3/98, but I asked it today. And 18 the response, I thought was appropriate. They don't actuate 19 anything, it's to be diverse.
20 MR. McINTYRE: So we are happy with -- satisfied?
21 DR. MILLER: I am.
22 CHAIRMAN BARTON: Okay. Next Brian.
23 MR. McINTYRE: And I think those were all the ones 24 that we were going to try to cover today, so thank you.
25 CHAIRMAN BARTON: Brian, I think --
() ANN RILEY & ASSOCIATES, LTD.
Court Reporters 1250 I Street, N.W., Suite 300 Washington, D.C. 20005 (202) 842-0034 L
257 1 DR. CARROLL: You knocked off Carroll, 4/1/98,
() 2 safety valve.
3 MR. McINTYRE: We decided it was a level 1 PRA 4 question.
5 DR. CARROLL: I think we have heard enough today 6 to be --
7 MR. McINTTRE: Okay.
8 CHAIRMAN BARTON: You're happy with that one?
9 Boy, you're easy today.
10 MR. McINTYRE: Thank you. Thank you. Thank you.
11 CHAIRMAN BARTON: You are really easy today.
12 DR. CATTON: Well, I'll wait to hear the answer.
13 MR. SNODDERLY: Excuse me, Brian, I think the 14 vacuum, the potential for containment vacuum was covered 15 today. I think what you would hear in Chapter 3 is that 16 whether the 3 psid -- the ability of the containment to make 17 3 psid, the potential for a vacuum was that -- was what you 18 heard today, was that it is the maximum vacuum that was 19 created was 2 psid from a very minus 40 degree day, and 120 20 degrees F inside containment.
21 DR. SEALE: Yeah.
22 MR. SNODDERLY: So I think if that doesn't answer 23 your question, you should say so now, because I don't think 24 you are going to hear any -- Chapter 3 is just going to say, 25 yes, the design can tolerate 3 psid differential pressure.
() ANN RILEY & ASSOCIATES, LTD.
Court Reporters 1250 I Street, N.W., Suite 300 Washington, D.C. 20005 (202) 842-0034
258l 1 MR. McINTYRE: So you're happy? Okay.
(7
/
j 2 DR. KRESS: How cold does it get in Pittsburgh?
3 MR. McINTYRE: Actually, it was minus 40 a couple 4 of years ago.
5 DR. SHACK: Forty-eight mile an hour wind.
6 MR. McINTYRE: I was up on my roof shoveling snow 7 when it certainly felt like it, yes.
8 DR. SHACK: How cold does it get in Shanghai?
9 DR. CARROLL: I don't know. The month I spent 10 back there, I think it was below minus 40 every day.
I 11 CHAIRMAN BARTON: That was because of humidity. l 12 [ Laughter.)
13 CHAIRMAN BARTON: Any other discussion or 14 questions?
O)
's 15 MR. McINTYRE: No. That's from the list for 16 today. Thank you.
17 CHAIRMAN BARTON: No other comments. Then we will i 1
18 recess till 8:30 tomorrow morning.
19 [Whereupon, at'5:43 p.m., the meeting was l 20 recessed, to reconvene at 8:30 a.m., Thursday, May 14, 21 1998.]
22 .
23 24 25
/T I 3 ANN RILEY & ASSOCIATES, LTD.
Court Reporters 1250 I Street, N.W., Suite 300 Washington, D.C. 20005 (202) 842-0034
REPORTER'S CERTIFICATE This is to certify that the attached proceedings j () before the United States Nuclear Regulatory Commission in the matter of:
NAME OF PROCEEDING: SUBCOMMITTEE ON ADVANCED REACTOR DESIGNS CASE NUMBER:
PLACE OF PROCEEDING: Rockville, MD were held as herein appears, and that this is the original transcript thereof for the file of the United States Nuclear Regulatory Commission taken by me and thereafter reduced to typewriting by me or under the direction of the court reporting company, and that the transcript is a true and accurate record of the foregoing proceedings.
(D 11
( JM trW hnHundley ( ;
Official Reporter Ann Riley & Associates, Ltd.
O
s s dr e e
6 r e
r tu at t ui gm f em aoc ea S
b g peF t
r o Si u n a g t
c as t
e h y i n
r en gM e Ct fe lz e9 e8 R i u n9 h g1 i
n s5 Ra c n o oD e r e 1 AS S E 3 L s 1
eo3 may S d t t 1 T eM t
c t may i
s e a Se r S y
mRM o 0 e Cd e 0 e yc r
o a n 6 n ivs Pig d
vd A A An E
o
)
4
(
6 m
t e
s y
S l
o r
t n
) o
) 4 C 2 2 y 2 6 i l
t
( i 6
( b m a m t e i b
t t
e s )
a s y 3 y S 6 H
S l ( y g o c n t r m n i n e e g
l o o t s r o C y e C n S m t e g E 6 n g n e o i l
m r o R n m d y o o o
E i a H C R T d t n e l e o t
n r o o P t n C e C t
r n
A e s e m
n e C o
H m i v
s i
a i v
s n C
t P s n s i a
s a o a P C P M 0 i c
p 0 o
- 6 T - - - -
P A =
~
~
o$
s A
B D -
M r o y -
E l
f p
) p T i g
y u
s S s p e g
l p r Y 5 k a p u
yt pa e -
S l
4 a s -
(
e r pw n u e r G e l
g n t st e i
s a r i
t ef s a i
N e mn w a o w f
I d r o s t
L n t ei t c g n&
op w& p u O h a gu n ndi i
t c m
& pe r e
O t o relos pmk a on ra up t
sk l mupm a O C l s
e eint an oc ooe i t t
l u
pei t
w n
T t obr er e n ea cetraba eisl t p i
o N e e u m mr pr r n s u t
i n e t
c E ul i s sintoo op s n ou f t e
s s mpat emn e
r s a r aes oet s n o
r M eat r s opun a r
e s u e p
p N spten n c ost u s s e r
ot n s mimt eo n s e y i I
s f A i tc eAnlar o
, t u ai t s ydamr a
f o
T e nmCiuteh sf un a c t yf ao d 7t e c
N i
FiOt nla n gwd e 7 r g e no s u r
n e o taL c c os3 y r s O
t deogs e l U 3Al f
n o o s
C t a c r e oedssae t r
o s e
i t c l a c d f d ci nm E etsL uDi3A v - - - i v us Ri m d o o V y-t L i
R e
P r
P r Fi e r
e S I
S f e -
h a
S S - - - -
t O - 3 0 A P
8 9
/
2 1
/
5 S
- L T
llllIlll l '
u em s
s r e mi eu n r s g di ps el
.e aa M
st r en p ssw ui c oc w o m E i f
s eh sl os i
T r o e lata n ru S / ci e u ao i slp r
Y n a
iS pd n t h ei0 e/
wsf S e i
r A pr a d D 3 ni r
dnl y m oir t
G t
e ro s yt aio d a i
t as t ul u e N a as pt m b d er as g
I wS p p i
t r
et L eM 3 t
og s rr P
s ra ap)mdr 27 i
i t
d O i t
a a
t r p gt n ed r l O
i i i e a ~ ei C
i n v sl y
f r s2 e o y 4h s r i
t o/wt u s el a f
b 4~ a y u b hgin i
v c l a o r T nl ai e t gd( ch t
s uronsoo
- N p v a 0l l eee t 3d d i
t E n g
od e o m 0 road r s al t
l el a n ag i
t 0, t M
i s n st yau eg ar ea t
1 3 cl n
o os wa mt oioi nat r i t i t
n c v r
N Di eaoep epni e slwpd y5 ;el c gosl
- l f l r
- I c e of o A md t pd t au e raidn e 2 ni au2 d c e h i r i t
ovf en T t smta a u st awir t
c r 1 i t ooaz t yepn w~b u r ps nis N
l St s wps e a aeeoh t r l go t e seep a
- f O
- i t
yr i r
d y l slooewt et eHF t
e -
sk C t a er f
viaa v l
i l v
ef aa ic d u omr woo E
l eiAFViSW-R s s s s
t r
eBTCI V y a a a I t P e - - - P - - W- - -
S f a
S S - - -
A P -
~
E
~
O e
M r u
E n T l
i a
T i o S e f e
S t
c e
W i p r u
C p Y t o C d n
s s
S r 7 p -
P t
a e r
o s y p d4t G r as p ep l
p r
e v
N z yu aae t
au o I
L h dk a
a r
pc s t n
h r e O t i
fom ei s ms m wl O y am gp gi n e i n
a s
C l
i p g s t n
O p0 u0 2 u p u o .
T n g
s 0,7 Tk S c c
s N i s
r 06 e0 ;
Wb a t n
E e t a4;pyCs s e v
D wk mpCe e M
l r
m d napu u ppd p N e t e t s mo i
v g I t s
y nglegf a e er or r n
A ef p r i l
o T Si l r- ast r n alaST o N d oyo e otet r t cf st awgWey s c O l a
e ms ar r eefi00,C e0 r r Auta no l
C r nt t Ry- e o aac i8Cei r g NWWm1 Pf A t
E t en s A V
i I
f a
sLo - - S e
R P -
S n o
S N - - - 5 O A 8 9
/
2 1
P *
/
S L
T I l ,
l y~
=
M -
E A T -
S _
U o;;,
Y S
G v nAw N
I L - ,3_T g
O
~ ,
m O ,"'
m*
C - e s
, n T ",ii' ,l*
a a
N VI E
M N1i- e. I i! u
- ,T l ,-
r'N~
N _N I
\
A \
Q T -
N ~t /'.
i O
C /,
E S,'/
- V
/ ! .j.*,
I
,vI S
S m'* s A ~ o n
a P
e r r T s eu v o N n o eh s
E i
t y 1 I
id t s r di d s
- S n o i ni ete
_ N c w umls f a
A l at hp h ed k i
_ t R n n e ev m%asmpn 0 o O T t n me 8 0 ; 0. i 41 d e2 t
c n e a e
M ol ,
m ;~ o m i r b d FF r U i n n n vi d e0 0 u et s r a
_ U t a i oerec m42 u 1 s a us re nl oagt ;
s p
- C or ct nei r nt ot Cs s s n
- t s or e o A f o clo omyn pi Aao o l i n n py V nm b l t a si i
- t ndi o ude i
- t u r al a r sond i T olemh t p si o nto amt t
o Loicdnl al c pien na d n i s r i
- N eopomwt r i n o n a o mm i
E deta umcn sA iu t
eCe t c n et er r n r st cpay m S ge M n aT f sl l ai n e v mxt cit e n N o Sd egst e a E reEInPlsu ne t
t I
A CSWi sC R aur os no l
gt n x e v tn PI FPLC iE RE T e t
i m
N v i
E - - - - - - L O
O C *
- l i
g M n E
i i
x T m S d o
Y o g m S A s
e o B d o r
L s D i
v t O t n
e g o r
ps
)
n e
R d i n
w t R t t n n m T
i c nA eeip N
c a l l
o o
eP m( mv u q
e es nT e O
f i r gr aS nWS l
s e on e n e t C t n
v e t r
n gr ai n oR cI C V
es o or b N d c dr am n n n E
i c& t n yn c ed r dd odi ii c s G aA eh eeieee s reB mgf eoo mr n c qaaa ut t t O A eD t sd r
y y utay pi i
t e r
c ooo c c R B vt l l l D s op es Sh d n a n o sRR l
D - l t
s s t bd i
ueb u c c aA PP otaiR AAt Y oog n S a t
mn /ouP u cd ee t t d d H l ppi l
r o r l
er er eiaa ez o z oot n yt ae/zii
- t T dt r r e n n o t
i n i bt gr v e n eie r s r r i s s N d oo ccm mfaops swl u aa o
l t t E
i v
on n n c sl o aof p p eee es wS pp oee M P r
ggg R deoCs own n l
N sr ooo r r niSP eNTOOz I nddd oy y y e v gr o i l
A i
i t
tc HHH oP U T n u
dy r - - - - - -
N F - - - H - -
O C * *
~
~
O E t s
n s s r
e t n e M d i
i o
ht o
E c c p s r T a e s i e
r
)
S e Aw S e r
e a
e e D op Y
t t
v e
l e a (
S s r
n b E se 1
g e E -
i r
L n g 1
- t n e ott O i w
o d o
r n
o n
e n a b
R l l
o y n r
d i
c e
r aE T f h
l a
o c a 2 1
N r l
o e n e
i r t
) ,o n
O t n
o e w o
r e S 'n L
C t v ,
c o p
pS e P sr n eA s (
o N e g
r a
t iDs t
s E sn O E o r
e n
n r o,So o p
1 ns e
G d yn t
n s r epeL m Aro l o nE O h d egomu s r
wP e Bt n eDo cd e1 r
R mteotr nniag i r pb ooy t sd yivt s a er D eladyigt n ot ey r e1 w t
s yr eh nowis a l
S>p opiood1 r o Y S yf ect
- f l
u t
u t i v ), p n oS s H b e
- t o poi t gn o on a f
S sr noipMi oi gr e r
uauf n y gt t T S sd d r
dib m nic/
n n gr eP e ( t t N r ni l
yed dd t
i r ou nr a E a o
o nb ouh 4 bugre ee i
E t t
i mfStor s1 b i
t n ea n c M I n st 6 ge sasi s rr oc r i
t awtu o enMioed MgAon aE t
ni r d
N niF e v eDAPA z nr oPMo 31 ed m I
gr o i l
A oP i
U t gy oH 6
- T dy r - - - - -
dy r - - 1 - -
N H - H 0 O
O
- - 1 8
9
/
21 C * *
/
S I
T ll 1 1 '
M E
T lar l
o S p u vt o no Y k e
mc S a e r H m t p A G /
n a
et n R P
N I i o h e ym r a o ci n t
L a r f y
O b o e a d n t i t
s O S Cn o ot eo r c r
e C Ri t c ndd i
v E y c ye e/ t d c e cd ay R
j ded nei ne niclei t e
t c r l O
i i
vag yg a y b ole r
ee t
et r
t a C P r r mfamof ea s el i E st yESEP e
sR nf n V i oa o I
t c S - - - - N -
S n u
S F - -
A P
- T k
s M a Cr h n e h o S l
c0 ) i t
r ea0 Ac C er gu e1 r Re j R r s ,o P n
( i f a se )n f o l
r r ek oppa g e )y n ni d t io l
f oe i l
vvi e
s r Sl l ooa or r a t
n i e oCim leR( f p a c p gi z r
o d b ys g deST (n u s
i t
a y
a av ne3 os r
eI f t
e l
i a l ;
l b eWiRlar t e pS a vt aS / iu p
ot s n
a avD deni er cd e
u nytnA t
oso uit Sadva e n eo cd e r e pn ah O r f pwuomn t
- d l l
i Vr no )
l p C guf e Aarnmo r l
ak g yd b B t
n nb e r p Drst eno malai r e r gd e
h psu ( ogcR o b ua u ek ne Aa n et n " f o w e P r 8t i
rl S k pa Clul odio U fo/nd 3 en C a m om Oms L u e o vfr E p ca R nt cd pM ut e eon r K s ea nl odt a i ocv i pkd ero s t k pv iac Sf AoSC r D A m aoin vt r nu ot uf r ci n i
Dn o AiwtopA u mmd o M Pymr c aepag t d ect,eensc Tk as c s i S pe t
o oc pmki njTmga coa n aes i
t anMnta ufac t
u t i
C e samm eya mid t s yC i s m kt a R
k anno o- k t ne on u r t
t afe o ovo t af Moawu NCTA Mawur STAP Tawooc STCFA 0 S e r S 0 V o X 6 C C P O - - - - - - - - - - - -
- P A * * *
)
P M 2 U
E U F K C0 A C M A1(
E R
O [
C
, ]_ ]0
- {/
, [ ( '
O l l 4"
C
/ '.
1
'c.
,~ ~ / ', -
- g ~ l 'c,
- j L v
[ i Ut v F
E A RC L N)
C N2 M
O Cr I
O V 1 E A U:
D(
S) ,
T T
T R2 E
N Gr -
E R o MT A T
N S
N P1 S A
i S( W T R Y
TN i O ,
C T
S PE MEr N2
)
R O EL T
C S I uRo A S E
SC E V S
Si ( R R
HX E
RH P
{
V I R E
0 4
n A
(
4 I
n=: S"o D
A^
)
2 r L H SS S ].
[
1#
2
- "' Z I
R U . :
[
'( RP HM RU NP S
' a' S T 1
S P R
@' 3- m" E o A P
A R P
o t M O
P 0
3 1)2 Sr EO g- C f
G 0 A1 (
TS 6
P A
l
I O l AP600 LONG TERM COOLING IRWST INJECTION 1
ADS 1/2/3
-c>o--
[l CONTAINMENT SHELL
-r O ~ -
ce"<-
PZR SPARCER
/ WASTE SUMP (j IRWST ADS 4 AW
<V -
R
- r :CP -
1
} ._ !
SPILL PXS VALVE I ROOM ;
._ )
l O
4 I
i i
e AP600 LONG TERM COOLING RECIRCULATION ADS 1/2/3
-O<
- CONTAINMENT l SHELL l
l
-r '
G CUTTER
%l. "
PZR ->4-- I SPARGER
/ WASTE SUMP j
So
() IRWST
(~
p, <
__, RCP s
' : y_
RECtRC ,,
PXS VALVE l Y w
n e
d i
c s ls c r u
e a o s A h s i e C s d 3 t t
n y O nk n e t L- ) i n e i
s e n n si v n f a o eisl e
a r s n pot ob a ASl t n g e RD d o n l
i PAi n n i r SPakhn f uCC( st ea r
a n i ogh s o dRRs f n d a dnt otulvetam ei Xou L w o a yh oahl t
Hrht o u A d s l
o et l
t v s s R d t
V cit e t
uG t u
l bar ewn aidd e hec ReR hS O O s l l
a l ida r i
cd vi nor o v Plb M
l aa lo vd cu i
pr p o Ti E md r e t r ael d nt cr a r e Tg ,
t dS n
R oe nf o c r u
eu t sSn n aW n i R
r o
sT pS w eWn anWl nepR oo W/o l T mr f
gFo nSc;o c F m A eeu mCo P I
l f a nt Suac E sa pmd t t h n rCnin e
ce gR o d e oct H ywnr aon xh n i t
em d p/
t A
Sd ef r eivdt r Ewah hi uai t at n e uT Y teerf dl ea t
ag ncwt ca c ot l
ekMM A a yr wt opc i
ps t
eix Hlo ewal on BcC db a
C det ui t Rc ot af ocXcH l i t
n s om E ea esmr f
osa m H ed a e v a er f F - e Rt yh emRi s dd i evd o D not o pow at u e ecoHs efan ouRa r t eo et F r eu v
0 tuNTWA r
iSOFAPP s S PFA 0 a s t a C 6 S - - - - P - - - - - - R - -
O P A * *
- 1 ;
M e. _
P C
A N M.p E E T G
) R S w _
\ :
O F
R L O E T
C S S
" A E E V R
T S
W e R
1 E
M R
O C e b
E T
S Y T R H
R H
L S P" X
H R s o
H R E
R Z
I R
P U
S S
O E
R [
0 P 0
6 ,
P ,
A _
O s r
- l e
l
- i h
Y c T
I l d
e L o o
I n c n B e i o r o A
i k
a t
a a i t
a T
i t
n d d i d
I i aste r
Cr a B
l r
i e n a 2 s u Aers A h hi eo gd m yVdh H pH u l
s l 2
e i h f ) pl c7 r So I uax M sf r o f
nNo oNr h i
csmen mr r o i
O l t
e g n k eyt i i s h ot n e5 F 6 O a b w e2 f r i
f O 8 i
/e r t n d s7 ou R st r l i
iem r r e
,s 3 e ns a sn Eitdioeh ef i et L l oa l awsy vap r t O o olahsp cc o s r
Moloes Sr s
R / r n e agr se r t Aps ve i i t
l T sh nc if sf ea ( s e oo niser a n oiw l
l ml u N f h t
et npt C h ai o cd o on t o O
t i t
,l e sbioh y t ce a r C C Aci t si l Csi anp r A Vi d asf nh V Hwraoac t l A d aui t c8 m N Hy s
c 1
i c n t i s Ved Ht ae a/e 1 t t -
Resism sc Rf eihi ak l
I A t agadin t Cr y eai t
no M CasmiHmns M- ut Mt pmtao r r
nt o t od a ef emton x i
l a ou uee v aou aa 0 mNA ARH iSCAMM s
0 r o s 6 a N P O - - - 1 P 8 N
2 A
1
/
5 S
- L T
9 n, . ..
O p i .i i.
,l l I I ! I!!
Ll-n!i :
- i; Ili; I!
i l r i!:
, ,i:ii. i![I i: i ;
ll!!'!
'i!.,!!Ii i!!! ij i 'Il
!!:Fj i i
- i
!!!li
.i.i ll!
Isn! lil!!l!!,I;!bl! jjlll!!p!,i..ll8
!i
!'.!!lllii l!i q
- ll . i s
- 2. .
c--------i l " I
' e
!', l .'. J qr, rii;r. j !
I c. .Q!
HH I l !
___..__.._, ! 'ji $' !
r .
i i 7 9 'li. L _ _ ___J . _g ns l b" f - --
l
,a
,, i . ..
g-g l !: Iyp.
h a- i-I ! _g' 4
-W 'N 8'U8 8 f ~l. 2 . I j (___~___ _____ ~__ __ __ _ _ _b>
- 1TT i e e, g 2, ;,-6.E, si - '",itg_$.!
1
.- i, ,
!lji.e! , 1 e,. -- e _ .. ; is i- -
. i.e '
It --3 ,la !i
$! b d,,e' i k hlh'S i
i'ilLJ e
n a! _s ip: : : ::
i 6 iO I
~ ~'T_Ii8 ~"~l ib' b i'![i
.Its = Pi ~_~!!!!!!!!b
..g i ; ; j (---------g;,
~5 [1, I
= : ,. l hi i,
- 6
- 1
.. _.. = se eki es
!)
- i. - . .
sE4 l )
P'as i g-Ep na I
$Y l
- _ _ _ _ - _ _ _ _ _ _ - _ . _ _ _ _ - - _ _ _ _ - . _ _ _ _ . -_-_-____._____-______-_-_-_________.__-________________________-___-_____a
'b.
l l
O i ,, i i l' il il. i I lll!
l'Iil' :I!II
!,! l X' il #ill
?!
- v. lii!!11 in f ( ilff i
, , +
i i' l ,
@ f k3!j
!' , / ,
l ;I l
! i Ja!Q' !
a ! ,
'l
, I i
g
, s
!4 bg >
s J ,,j!
, l p,
s, VT, F(T- , .
. . . n" ... ,,
V !!
l -
~
h, s
[- ,
O ?
w .a IIl lll
~
q ']O N q
,u ... F
, i i ti g _ _ _ _ _-_4g l dL--
i **' g/ !f I l
. ; : i I' !
! l llli
~ '
E .. . i g
j,, l'i ll O e A 4 r
I
6 4
O
, w_4 i's 'l b i[it.
,7dil
<. s 3: ,;
.J 0 5
,, if. ggt 1 13 m['S$E ,,,, ;SI '
.T ~. . A.. J .
! c li h is
- hh4 SE6.'jh. , b. hgi ,na 3 j /y'/
q- " 'eli d ps, 7 . .
ri:
a i h,.,. ,(pig;,g[ip IM. ,l Up ,e,f
<,. ..p ,. e, , . ,, .
.. ,p.
ti!r,r4.. ;. g.q', ci .vi , i. ti_ . . p;;; 4 ,'..t g .. . -3 8 I,
,g /,
Y b
i ,..
h r
- 4r, i ., , .- amii 3 g
'e s ,,1 .. , ,. , .. ,,,
Ef9-i '
- (i$.;:
\i
'l'l.
in
-w
-p :
t -
.. f f g.
Si as i' I I #i i il II; !
I!!I i$i 8l
.d
'Fii
" si !';l!j:li 5
hl il $
f i, n !!i!,If!!ili!j'lijall l
\'n.,J 'ln. !
,1,,y-
!!! !!li. j ni!. :ii;,llI li,,.!!.!.y,!Elii! n s.,
y
.b . i! !' 'ca
<,i,!
!! E!!i i
,g, . i.. . ........... ...
3q ,
i i,
.. s N # 8_ .
m((1, I
's. \ 1 ""' s- I"'
@~i li6 8 Y- -
l I I '!,5*
Ilt g .b
.<g/ l i!
.i i ga. I'.i .
l I
- 3 l
A .
,r - 4i. @,.. . p..r. a.
se ,< mzpe, Ci I r, , ,, i '
, l hgml;p,q a if; '%t'iM'ggghiSp i 4 },
g i kiijfie .- ell Si;:;' .
'$[l i,! 3 ' g,i '8 '
x j:FFFF,p?,h:
o.
! O
o il;1 i i .-
,ii i
, l l ,. .
< ii:1, n.19 e 1' l l :i! g i lji 1 O +-d' si il,in i
3ll! li EP % -
i, ~' : , !!!!!!!!lj,lii h!!
~
h.~7..,.,s
,j , ,
.nir!!l
.a . . .a !l!! ,i.al i,
G g'....
2,.4. .-
l
~
[ h- '
gI il I!j !l
'i ' i ji' dIk_hsij li i , ,,
1,;I!.!
lji,
- 1l 'se i
i j!l!.!!li.ilil:li
,j il i i I, i
.i
.l1
!!'l li ! !.
i
- ,* i; it i, i i ~!
i i ini ih in
- i l
,, g !;1;!ll
- 1 i, i Il li;; i li!
_ W. ia s? \%e's I
$ li!!!i il!!:li1'uh; !jj!!nn i
l I
i.
i.:.ili! nil !i
$.L . ..
G .
G-... ,
- , sg
. .. i.
- i. g .pigi i il
- i. . ,4 J ii =
s, i
O je q-an,_
a:
3;,
i.
1i q$,;j;i,.j j[-[i % -
-<i '<> m.
s_ , :
n a,Fli,$ <j 7-.
.D. .ni ' 91 si [;ll
~j i i + 1 i < ,
i i i
I . .. ( 3
[j,'
i ;5l, lii'i i i ,
- j,iii[i bu.! i ~ i i . r s.
B,8.9.i-T .
i i . ..i 'i 1 ,i .,
'a\
'8, "
!!! i: iu 3 q
s.
,y
.d;,.
q tg i
- ,r,,gl. (
in g.3 !.i;,
l' E Uf l
l aif p! !
T--4i Q _i 9 h.,-.n ti 1(%.
i i
m, i
'f 'r i
~f7 6 ,'4'~%41, 'cN .
- lg. e l
e n't # i il j!, , l' ' '-BW , e d =i i-Eu_, e
. .n O
- 1i- 'ji v. :.!i lN[,,%
I '
s i
[h. !' ,,1 -['M
..[y;. gim ,.
I'" g ,
- e. iTh.,v:
, h .Mi ,
M
_j! ! .. . [
, , 4' g .i di .83 k$dr'fi gib M r- :8 .A$r-$ +: @4.';. .Ng
> 3-m 1
Qi y< ! :ll; pY
++ ,'li 1 ,9mi ,f M: 'I e
,+i . +s 7p A
' !y,*a !v, h 1, t I e
., L ,7 fg ';q kMm !rt=h--
i :11 -- ---- q
~
g g
ti :4 :
9- p t@
h L li " li ,
d' i( .-4 +8 .l.,
r i E
,a;.w:w-Yi'g; p e c, e.Ow++h. n %g i , e, il l i.
u!
a I
k+ t$' . .r m E ,, g g, i. , 6:
- yh e +sg ,e +E i' 'i ll
.. . i i, b + 1. i .,UE
.,; y
' '2/, ! !
- 44.W. ! !!l!
!Il I ses y- I g i ".. ; -
i 11 ji
.h{' l !!! !!
- V5j"N*..;F e
o "hf"$43 m$e u?;
, 'j!!0li!j' al i
!l ik , n! i!
l,
- i
, !!!i ,ai n !! iin rin.,;wi.,,,.,b.s,M
.s
,1 s ,e e
0 4
s
- i ,
hll! !
il 11 i !!'i';!iil$,'li! i ij;!!uii!:[ll i: ,j.::!i $l1!!!j!
nu.!!!
I ni d
i.i,d, li!11i11111 i d, !11: nii, I, Idli 1,
. ...,i
. . . . . . . . . . 11 .I . .
M. 3, !
i 6i y .e ..
is is!
'r- if-l , , 'ih,di 1 ui 3,s. sd s l '. .k'l Nij l
' l ' ~ +lini 4:ini i s A
i ! w'li!
t S IO- 3t i
'7- 'd, IT -
$6-I i ii g -
i, .a.
1.
O s
' h t V h)) .
' 0Es . .
7T 'gi' l's i ..,i 88
'E! '!'E ! 'E ! Mi 4 4' 4' 6-i i
i i l ik k .E6 l
i bus bp i,ia i
bui baib a
' h s'
(
6'
- ;'i
'd l a 'j i i i i '(i* i i
i,
_6 i
(! s 'l @} '@' a
{l i-i ). li
+M 3ji ii 2 ' ;PH
- i. I .).nAiwae 'i; l
i-a';e ggi 1l0 b
! ! l/
i
g .g.j -
t fi l ti ipg ?!Jis{--!;l'8 L$;rg s'
- [G) i
- s .; i
- 'i
. . . s I
- 3; ..A6f l
i 'g, : '. lign: .ri s! p!' i i @) :
N a
h' 'Tp, *I El g n.
, s g
ei -
1 I
t) T,s. ,!g%vn .
E .ss il i
$'t _
i he s's h 4 h
~
4 li li li O
li 1 1 L L
, . . . i;. . s.n . ._. _.'yi ,.._.._ .-. ,. . _ . . _ . . _ . . _ . . _ . . _ . . _ . . _ . . _ . . . - . . - . ,
C .
u -
i i
. .ei 1, I, .
! r- MI I- I ,* *
.,n, i
i j , 4, .'
l
- ,_-_i - - ,
y c,m ..., ;
! m 6 _ ;
! !IbA !$" r, l i
,- . !.i . r i
- ?, " i
-T: i l1! ini; ! !
e!! ,
i !
! ,i! ir = ! g .
- iti.g Pi I g ,----- J $ 1 l!!!l$ l IU !
f 'l dIb'------l
!!!! .b!!!
A-Si :q,q ,
r::. .!!illll!!.!.!
l A' 3I l 11, . O t...._.._.._.._....._.._.._ . . _ . . _ . . _ . . _ . . . . . - . . _ ._.._.._.._.._.._.._i 1i; ;
a gji 3,
i
. .itoj E I I
' III' ~!
P-*:*r- :' T 3
i
! ' ' 1l;I,
!Ill ll'Ii i'h l!)
lh:1 !9:3 I l'l !!!i !!!
- ' ' . , it isli n: si s l.i i=!
ci, l
' ' ,( ITC i is II$.I I !Ii j i
! ,!. n:: 'i s
'I i!"i'!!ailei5!'f!!!
l D
j s ,
LlJ ) gl
> a ,1,
l' .#'
l 5 5 1
g .
yi se si ;
I i I- 1 I ! ! !
31 El "I Hi
- ..[.......... ........) I
...}.......... . . . . . . . . . . , .
...}..........
, = _ .
...}...........
,=
. .. . . . . . . fr '
.l : =I
' l ;
!l l l I l l i . i ! .ii ! ! , i i ! ,, 1 i !n r .! l,li' e i .I e ! ! ll,1 i ! I e ! Il,i l
Ie! ! ll i ! Ici o+- 1 Il, ! C9- I !i. ! -
cc) i li, ! C9- i ll. !
i
! ., a 4CJi j , e! i l,lIi 4Ci
! , 4CJ! i i lil i l'c i ! 1,1Ii 'Iol ! ei" ! ll, ili I e ! !'
-! Oh !IF!i l ' OA I IF i ' i c3& t il, : , i C9- i !F i, i
4
.s 4 C J .i: . . 4CJ.i .13 :. ., s -4CJ!,:ll ,i , i 4CJ!:ll- ,
i i
ll,l 1 CN- i li' i ,llE9-ie i
i ll,1ie iei i If i l1 i C9-Ig j' !
i!Fi !
ol,I-i #- I e !'
i if i
! ,, 4CJj i ,, , 4CJli j ,, , 4CJj ! ,, , 4CJi
! ll,! ! Ici ! Ill i ie ! ll1 i i c ! ll1 i io
! . CN- Ilii
! CDA i li,' !l i CM- i li.'!l i C}J- i li,'i!
4 CJi, u.....................
u......................
4CDi l
. ' CJi !
4CJ! ,
u..................... w.......................
3 f
d L.h W
t
f A
1 N
E AP600 Ex-vessel Severe Accident Topics presented at ACRS Subcommittee Meeting by R. J. Lutz, Jr..
Westinghouse Electric Co.
May 13,1998 O
AP600 Ex-Vessel Severe Accident Topics Westinghouse Philosophy
+
The Level 2 PRA models in-vessel retention of the core debris for accident sequences where it is appropriate:
- depressurized RCS
- submerged reactor vessel
+
The level 2 PRA assumes that if either of these conditions is not met, the containment fails at the time of reactor vessel failure
- This is a conservative treatment of ex vessel phenomena
- The resultant containment failure probability is acceptable
- The risk is acceptable and less than regulatory standards I
AP600 PRA Appendut B 2 l
/
A A i O
AP600 Ex-Vessel Severe Accident Topics Westinghouse Philosophy
. To facilitate NRC staff review of the Level 2 PRA and In-vessel Retention and to meet the requiremetus of SECY-93-087, Westinghouse agreed to perform a limited number of deterministic analyres of ex-vessel phenomena to identify potential containment vulnerabilities
- ex-vessel steam explosions
- core concrete interactions, including noncondensible gas generation f - direct containment heating
. The criteria for selecting the models and initial conditions was to present a realistic, but conservative, assessment of the ex-vessel phenomena i
l AP600 PRA Appenda B 3 O
l AP600 Ex-Vessel Severe Accident Tol.ics Reactor Vessel Failure
- The reactor vessel failure mode / modeling can be important for assessing ex-vessel severe accident phenomena
- The reactor vessel failure mode chosen is based on analyses of the in-vessel core configuration at the time reactor vessel failure would be expected
- metal layer overlaying the oxidic core debris
- all core and reflector material in the bottom head at the time of failure
- Two reactor venel failure modes were chosen for analysis of the ex-vessel severe accident phenomena:
- a localized failure which progressively opens in a longitudinal direction
- a rapid failure that occurs in the latitudinal direction and results in hinging of the reactor vessel bottom head until it touches the cavity floor ,
AP600 PRA Appenda B 4 O
I IL
i 4
/^'a V
l AP600 Ex-Vessel Severe Accident Topics Reactor Vessel Failure Some important characteristics of the core debris transport at vessel failure are:
- melt mass inside the lower head:
.UO2: 167200 lbm ZrO2: 18700 lbm
'SS 15400 lbm Zr 28400 lbm
- release rate and release duration:
hinged failure 33250 to 41200 lbm/sec over 10 sec, localized failure 8.4 to 940 lbm'sec over 17525 sec.
- melttemperature: 2940 K (equivalent to 145 K superheat)
AP600 PRA Appenda B 5
'g
%)
l l
AP600 Ex-Vessel Severe Accident Topics i Direct Containment Heating
=
The potential for Direct Containment Heating (DCH) to challenge containment integrity was assessed using the screening model developed by Pilch et. aL st Sandia (SAND-912407C) 1
- The Pilch model includes containment pressurization from:
- blowdown of RCS gases a
heat transfer from core debris to containment atmosphere oxidation of unreacted zirconium ;
a hydrogen combustion
- The Pitch model is a 2-cell model i
+
upper containment above the operating deck
. steam generator / loop compartments below the operating deck i AP600 PRA Appendu B (
! . %)
3
O i
AP60 Ex-Vessel Severe Accident Topics Direct Containment Heating 60
- The Pilch model predicts a containment pressure increase of # psi due to direct containment heating.
- The highest containment pressure at vessel failure predicted in the Level 2 analyses is 45 psig. g
- The peak containment pressure from DCH would therefore be 9t psig, which is below the point at which containment overpressure failuries are signficant.
- Direct containment heating is not pndicted to result in a challenge to the AP600 containment AP600 PRA Appenda B 7 O
AP600 Ex-Vessel Severe Accident Topics Ex-Vessel Steam Explosions
- The potential for Es Vessel Steam Esplosions (EVSE) to challenge containment integrity was assessed using the TEXAS-V code models developed at the Univenity of Wisconsin at Madison
- The TEXAS-V code models:
a fuel fragmentation
- fuel - coolant intermixing
- fuel . coolant energy transfer explosive loads where applicable
- The TEXAS-V code represents a conservative assessment of the potential for steam explosions
- Steam explosion explosive loads were calculated for both reactor vessel failure modes described previously AP600 PRA Appenda B 8 O
O AP600 Ex-Vessel Severe Accident Topics Ex-Vessel Steam Explosions The peak steam esplosion loads were predicted to be:
- Ilinged failure mode- 24,700 psi 71 psi-seconds
- Localized failure mode: 87 psi 3 psi-secor.ds a
Several sensitivity studies were perfonned; the results confirmed that there are no large variations in the base-case analyses due to reasonable uncertainties in the model parameters AP600 PRA Appenda B 9 O
AP600 Ex-Vessel Severe Accident Topics Ex-Vessel Steam Explosions A structural assessment was done to evaluate the impact of these loadings on the critainment stnictures.
- For the hinged vessel failure mode, the following results were found:
localized failure ofcavity walls may occur containment vessel is not predicted to fail
- containment vessel strain is less than 20% ofits ultimate strain
- An assessment of the reactor pressure vessel response was also performed with the following results
- For the hinged vessel failure mode, the maximum liR of the reactor vessel is less than 6 feet a
this does not exceed the height of the biological shield walls
- the formation of plastic hinges in the coolant pipes isolate vessel movement and prevents failure ofcontainment penetrations AP600 PRA Appenda B 10 O.
r 5
L__-_______-_- _ _ _
O AP600 Ex-Vessel Severe Accident Topics Ex-Vessel Steam Explosions
- It is concluded that there is no significant vulnerability of the AP600 containment design for the perspective of ex-vessel steam explosions
- tSe cavity walls may be 6.mged, but the containment shell and embedded liner is predicted to rema:r: atact
- the reactor vessel may be infted upward, but a the reactor vessel will not contact the containment shell
+ the lines connect:d to the reactor vessel will not result in failures of the containment penetrations AP600 PRA Appendus B !!
O AP600 Ex-Vessel Severe Accident Topics Core - Concrete Interactions a Uniform spreading of core debris in the reactor cavity was found not to be realistic due to:
- the reactor vessel failure modes assumed
- the shape of the AP600 reactor cavity region
- The potential for core concrete interactions to challenge containment integrity was assessed using a combination of the MELTSPREAD and MAAP4 codes a realistic core debris spreadmg analysis was performed for each vessel failure mode by ANL using the MELTSPREAD code the resultant configuration was used as initial conditions for analysis of core concrete iriteractions using the MAAP4 code models AP600 PRA Appendus B 12 j
l 6
I O
AP600 Ex-Vessel Severe Accident Topics Core - Concrete Interactions
. The MELTSPREAD analyses produced the following results:
- Hinged reactor vessel failure case
- relatively uniform spreading of debris over the reactor cavity floor
- average debris depth of 1.5 feet unequal distribution of oxidic and metallic core debris components in the reactor cavity
- reactor vessel side of cavity consisted of 85 to 90?i oxidic melt
- RCDT side ofcavity consisted of 75 to 8596 metallic melt
- leads to the conclusion that core concrete interactions will not be
, uniform on both sides of the cavity AP600 PRA Appenda B 13 O
AP600 Ex-Vessel Severe Accident Topics Core - Concrete Interactions
- The MELTSPREAD analyses (Continued)
- Localized reactor vessel failure case
+ non-uniform spreading of debris over the reactor cavity floor
- debris depth on reactor vessel side of 3 feet
- debris depth on RCDT side of 0.8 feet
- unequal distnbution of oxidic and metallic core debris components in the reactor cavity
- reactor vessel side ofcavity consisted of 70 to 8096 oxidic melt
- RCDT side of cavity consisted of 80 to 90. metallic melt
+ leads to the conclusion that core concrete interactions will not be uniform on both sides of the cavity AP600 PRA Appenda B 14 O
7
V
! O AP600 Ex-Vessel Severe Accident Topics Core - Concrete Interactions
- The MAAP 4 core concrete interaction model =ss used to assess the impact of on containment integrity
. Loss of containment integrity due to core concrete interactions is not well defined. Two bounding indicators of potentialloss integrity were used:
- the time that the embedded liner is penetrated by the core debris
- the time that the bottom of the basemat is penetrated by the core debris
+ Both basaltic and common limestone-sand concrete were evalusted AP600 PRA Appendus B 15 O
AP600 Ex-Vessel Severe Accident Topics Core - Concrete Interactions
- The MAAP analyses of core concrete interactions produced tha following results Basaltic Limestone Concrete Concrete Hinned Vessel Failure Mode Melt-Through of Embedded Liner 7.5 hours5.787037e-5 days <br />0.00139 hours <br />8.267196e-6 weeks <br />1.9025e-6 months <br /> 15 hours1.736111e-4 days <br />0.00417 hours <br />2.480159e-5 weeks <br />5.7075e-6 months <br /> Melt-Through of Basement 3.4 days 28.9 days Localized Vessel Failure Mode Melt Through of Embedded Liner 8.9 hours1.041667e-4 days <br />0.0025 hours <br />1.488095e-5 weeks <br />3.4245e-6 months <br /> 16.4 hours4.62963e-5 days <br />0.00111 hours <br />6.613757e-6 weeks <br />1.522e-6 months <br /> Melt-Through of Basement 3 4 days 11.0 days AP600 PRA Appendut B 16 O
8
. . . . _a
(__
l II L
O l
l AP600 Ex-Vessel Severe Accident Topics Core - Concrete Interactions The MAAP4 code was also used to assess the knpact of containment i
pressurization from core concrete lateractions on the containment integrity
- the indicator of a containment integrity challenge was chosen to be the Service Level C pressure of 90 psig
)
AP600 PRA Appenda B 17
/~
.t AP600 Ex-Vessel Severe Accident Topics Core - Concrete Interactions
. The MAAP analyses of core concrete lateractions produced the following results Basaltic Limestone j Concrete Concrete Ilinned Vessel Failure Mode Service Level C Preseure N/A N/A Ctmt Pressure at Basemat Melt-Through 29 psig 86 psig Localized Vessel Failure Mode Service Level C Pressure N/A 11 Days l
Ctmt Pressure at Basemat Melt-Through 43 psig % psig AP600 PRA Appenda B 18 9
f I O AP600 Ex-Vessel Severe Accident Topics Core - Concrete Interactions
- Based on these analyses, it was concluded that:
- the impact of core concrete interactions on containment integrity is not sensitive to the assumed mode of reactor vessil failure
- there is a high probability that the containment will fait by basemat penetration, as opposed to overpressurization, if core concrete interactions occur and are not arrested
- it is not necessary to specify a concrete type for the AP600 basemat under the reactor vessel AP600 PRA Appendut B 19 O
AP600 Ex-Vessel Severe Accident Topics Overall Conclusions
- Based on the limited set of detennlairtic analyses of ex-vessel severe accident phenomena for AP600, it was concluded that the goal of pentecting the containment fission product boundaries during the first 24 hours2.777778e-4 days <br />0.00667 hours <br />3.968254e-5 weeks <br />9.132e-6 months <br /> is met
- containment failure due to direct containment heating is not predicted to occur
- containment failure due to ex-vessel steam explosions are not predicted to occur
- containment failure due to overpressurization from non-non-condensable gases from core concrete interactions will not occur in the first 24 hours2.777778e-4 days <br />0.00667 hours <br />3.968254e-5 weeks <br />9.132e-6 months <br /> and is not expected to occur at all
- while melt-through of the embedded liner due to core concrete interactions is predicted to occur before 24 hours2.777778e-4 days <br />0.00667 hours <br />3.968254e-5 weeks <br />9.132e-6 months <br />, this is a very pessimistic criteria for l containment failure; melt-through of the bottom of the basemat (an i optimistic criteria) takes at least several days to occur l
AP600 PRA Appendus B 20 O
l 10 l
1 O
NRR STAFF PRESENTATION TO THE ACRS
SUBJECT:
External Reactor Vessel Cooling l
l DATE: May 13,1998 l
O l PRESENTER: Robert L. Palla l
t TITLE /ORG: Sr. Reactor Engineer Containment Systems & Severe Accident Br Division of Systems Safety and Analysis Office of Nuclear Reactor Regulation i
TELEPHONE: 415-1095 l l
SUBCOMMITTEE: Westinghouse Standard Plant Designs I
l
I EXTERNAL REACTOR VESSEL COOLING g
. Staff position provided in SECY-96-128, and supported by ACRS, Commission, and Westinghouse use a balanced approach to address the adequacy of the AP600 design for ex-vessel events
- primary reliance on in-vesselretention of the core, complemented by
- limited analytical evaluation of ex-vesselphenomena
. Key submittals made by Westinghouse
- In-Vessel Coolability and Retention of a Core Melt (DOEllD-10460)
- Lower Head Integrity Under In-Vessel Steam Explosion Loads (DOEllD-10541), and companion reports on propagation and pre-mixing
- Analysis of Melt Spreading in an AP600-Like Caviiy (DOEllD-10523)
- Deterministic evaluations of ex-vessel phenomena (PRA, Appendix B)
- Responses to several rounds of RAls g'
. Technical reviews performed by NRC and contractor staff, and i documented in FSER and supporting reports j
- ERVC - 19.2.3.3.1, RES in ternal report, INEEL/ EXT-97-OO 779 IVSE - 19.2.3.3.5.1
- EVSE - 19.2.3.3.5.2, ERl/NRC 95-211 DCH - 19.2.3.3.4 CCI - 19.2.0.3.3 l
l l
l O1 1
1
1 O CONCLUSIONS i
AP600 RPV is expected to remain intact, given successful RCS depressurization and reactor cavity flooding transient heat fluxes below CHF in SCDAP analysis (non-stratified) steady-state heat fluxes below CHF for " final bounding state" (FIBS) even with expanded treatment of uncertainties in-vessel FCI not expected to fail RPV
. Staff-sponsored calculations for alternate debris configurations indicate that FIBS may not bound all heat loads stratified intermediate state intermediate state with " sandwiched" steel layer metallic / oxidic layer inversion -- metals on bottom O
e in view of potential for hypothetical debris configurations to exceed CHF, staff believes that RPV failure cannot be ruled out for all possible core melt scenarios O
C CONCLUSIONS (continued) g
. Deterministic analyses for AP600 indicate that RPV failure will not result in early containment failure
. Probabilistic analyses indicate that containment failure frequency will remain below large release goal even if credit for ERVC is eliminated and RPV breach is assumed to lead to containment failure
. Staff accepts Westinghouse's characterization of ERVC in the PRA based on:
margin to failure for configurations believed to be more likely deterministic and probabilistic analyses of impact of RPV failure O
O
$@/ 3. Design of Stivetures, Components, Equipment, and Systems s
O 4 4 4 DETAIL s
[ # a, ser.r-v
, h a., w -o.
- a. or+ Nm DETAIL EL ir-s-
- n. er+ g , ,,,,
Et so -s-Figure 3.8.2-8 (Sheet 1 of 2)
Location of Containment Seal O Revision: 22
@ W85tingh0Use 3.8-131 April 6,1998
i s
nr ob .
ie t
e e t
nD=
e r 4 o Ro=
l e C-n, s e s t n, el Vo_
I nM-f -
o--
o -
n i
o t
n t
e e
R n l o e i t
s s l a
e s u 0V=
0 n- - n o l elet i n
o i s
n O6I r i
t e s s a e a r v Po-Ad f
e=
i z
r u
o c
y s s tr e e e t v v n u i
t c
e o ro e r
t s t o l f
i u= s i
st p y e S e n ct cd a r p e a a at "y r l l
y e dr e r e e n l e
u d r h e d n
ed e r e
q S wel e m oi zb nTi i
n C a wiaS f r
U R p s r d ot we ol n WR v o o oRl V l
l f
u l o of l n cI "
O eeeeeeee
y t
i n l e
l i
o s b i t
e s a a c e c v o
l o f a
o "t l
e r u e d l e r M=Cl i
s n a t l e f eMle l l l
a n ak A == s sr e mwre a e
t
'n A =- eo fo et l s x we e w s VC s s ef o ved o i O
O- - i s
Re ",f a vo oi s r e I
na l y
e n s r l
o Re= 06 n a sl ful an ep s n t x y e s V-I 4i t o g d n
t a a af e
v 0 ni r o el l a o s 1
t e el e al h
u s e r
De I
n mat c c d
n d
u
/ Rig r i u u l E eit n h r t r c d c s o n On Da et muma 2 r
o c
O o oe
o m
a e
t s s r
s i n d n
d n
a p
i o a a m t
p d s i ar s h ee e c m
u- t r i h a g h e wnr r
s i w
o s wf t o yo e n
A S Cl t
a e
l - RVp l
m a w p= d Pp h s t
a e
i=
c eRu si t y r z
- idr s w at we n u eu nh cy i
r s go r ot ai t P s e u i t af r l i
e n ap u b r mit l ut n s ta pb n e uo s e n o e d sciv nw o ee e e
i o )
r r t o
a r e e c
n e
l mf e c o.(r cl S n e e n-o wt o s d y i-t a
n o o b
d t
o c o
i t
a eg n e w l u a i o o e l ak c t l o a l
R:_ o l
v o cf 2 el o pl o de g b M n e l
el t
e ei r x o
A r r i
r d e o r r a mra y A e rc to l
l c b r a n e wi O g e e e i
t r p ein R n owf e l
o pd i i
m el r c u s o eeeoee
g l
e s
l e
n s s ) o s e
e Ot a i
v V P c
)
mto o Olo) ) o g C eS t
pt T A gF r B t o b oin n
o I
( (
s Oi n(
r r i g d r e ei r t
P ut e t
a .
y y a a a Od ta l l p
I l
o,,= Ct n s A e g e ei l
t a det a x e a- i n
mi n t moh s
ed a g g e
m r
i mg t n s tn i i n z e ( n u y a - t a
i m
g p od i
l l h n ml b a u c
ui c x T i t i e
at s c i r i r
c a
m st n t
e jf eimmmm g eee 4
4 r ll .
l -
e-s s
e V-l o
o p
o y l
a T=s s t
i i
v s
t e
'n o d- n o
i s
m i
n e
a-o i
t u s n o
e r
c r
sb u c L =- et s
i r
ni o c e t eh yt a p oi tr s e a fa 2 si t cl l yi d r et a2 l l
e n r e d e a m at u rr n ta a m:
r n p oi o c a n c tad u y
i ploh p
u pi o c a r
r p
e t n t
ni d nd ct e e x o el cl a m m i
iie v ts o c h i t
r e c p -
u t u ter
.T i sh n
f i
sd f
f i i r
z g a d e
r el o x mv e e aaaaaaaa s
o e
- e c
- f a
a- r u
l xi-uret l
s e
F i r s s
t C e ae v er u-=-
l n
o oHi-l F a
m; n
s n
o ala c := i oi i
t s
i t m= t a o i
rh e=
r l
s sl u p CT t
s ih n a el t t e u U s s m i
P e v a z
L U ma o e
s a n ii s re o g n yi t i
t n o a i i
r a te l r r r aC u ne _ i g t ei r e c Aru_ l f
n o se l
i t e ol i a_ c s c
u r
u aF_
r l e e s s d
l i
a t
ulam uy r s r e s
f of g
st s n ct _
e s e n uc r ru vla r ct s i
t o
a i
i t
t t d i r m SS e na a l ei l n h mpi d s e c r e sn i k e e e F a mht rc oc e H wmuaaC O o e e
o 0
5 .
s 0 t
l <. n u
s F H i o
e= q0 C3 -
s u y R=_x q1 x cl l l
/
n a oi c M x1 a c ys o A _s m <eeh p r r
A _m_ i s u si a a" l
t l
O_ _ yf af o te el "e s i s
R l
,z n y i
i rb R at i l
i vlu i a
n V
t eb i
i a o s a s f s I ab n l b ro s e se ea r
o m o en p sr F n vu o e eo
o x
r E
f o h g
a u l
i t
o r
i s nh et .
swt y o n7 Pi9o9 l
a= ",en1 t n- 9 t c o 7 ee A-L 7RD 0
0l eg ,
E =, -
7 s n E 9 ed si N
TVo o I X -
l
/
EInF L0l e E0 6 s E P e s N
I AV o e
h d c en l
i n h a so sit w s ear s n v u n i
o -
ngi o t
u i i t
s f a
o n f
i b r s i r o u y sc s s t g s S l
a_
i i
s y d i
eS f nB n t cB oI l
a u cI cF o A n p a n uF sg ey en t b L i a
E_ B y i S n h
t mr nd e
E C a n s t o
u ed n
m l a
N .U s t n
s a u o en 3 bo f
I f o o n nt i
o e w ol a mist o eie s et n o n i
vir e r e p
e e o gt e oe vr r r R r c ar pa o ee e o
o s o n t i
o h c
t a a r o r
u g
p p
i f
a n_
o t
n e t e
C_ ad i
s t e
n s a
i O t e r yi f
t a t d n g a a n.
r= i n 5 P te u q
t e=:
=
r u A L
s s o t d
E a l
n A d e Rd d l
e p / e o s L o P t a o s E l A u h e E eSDl vB a el i
l v eI C vk N
i dF S el i f
a I
o e eeee
r e e r
y a o l
c e e d h i t x f
} l A o a
t r
o oC nx n f p od i
o t i
t o= p f c r a n e
ei Mlup t
e a, o f o r=
u t
n o )n n O e(
eA gA ap vA g =,
s R Mno i
i r o f ei t n=
o_
ya c n Ai s a c oiAs u
C_ Oe l
l of d al e s erg r e r fe re s ro t
i mlaids i i t
ep nr t n s at a te n l i
h l
oe T (p n i oim c v o o ee e
o n
e e
t w M e A b A A O n d
e R o B=n h c n i
t a
r o == i i
d u
i t w e g rmd a i r f n e n o a o u-g, "
s d i
s C i r n n f
n ==aye o c h a
o l s y t C r t l
l t aeiyi c n t e e alp r e
mle xt r f
f nd e od i i
t p ~
x o e o i
h To nr n o e e e l
i n
r e
m r w o e o p
t w t
o o y C
l b mt r n a c
n= ei u oi n pd e e o
i_
h t o e c r v d
s o
l t r o i a z Mv i
t r n n Ai n e o u e
oi Aly r
u i
g_ i sd r eOt s l
i a
f el v R o f n_
o i v s o n s n ci m
i t
n o of r
C . r d e
i d
e i
r z e a c s mre y mr a u el i l
cg a
l oiy e na r
n d tak c al l i l
a ar s eot ehl a i
t e u n mb l
a ct ee o a l
Mu c u mm o e o o
- e r
y- u l
g a
i t
i- n e v f R
i d r i-t i- V ol u oiaf s- I l f
n n e e- i e
v l r
i ed o g
n g s o o S=-
A- e sl ef Rn t
a vty
- i Pn m n a v s=
i i t
s r c e F f o<
L F D!
t e r
E n u RCG ul E L
=O N o a cf i
X L E R F R cR I
EL N mm WaV o I
I n m o
e e l ,
O e
b 1
o M
N O
w O
^
l O -
O i l CO CD O
H N l:
4
'h O
9 L.-________.___.__..__..__._
i O .
-,ew :
l ., ,
- I
, ut z t h
lesseA laJJeg eJog I
O c
~
u y u o --
im= 9
- 8'"'
! : !! ! , s ! f! [l! : !
/ , --
awu '? , s unansJ s ,
g a ,
s i s ,
l a
' / .
/ * , s
- -l ;
310Hs/ ' 3 '
s NCE '
>i
/ 'M30
/ ,' ..
{
3N1W '
s 8
, s uolo3umi I, j '
l s
y , ,
M/ , d M '
El aI
,.f e ,
s s i
] EjI O ' i ' '
o O
t n
e m
ego en t ai t an a r al ON S t o
Mgu TO NG c
emR r
e N EI i Dr go a
r OE S t TTTDE I
c rotc e
jPa AI w e TMR e orR r
Pt o r NMO i
E OT v
e r
en a cea 0 2
R go el 1 EC C S
na Rcu ait 1 0
O A 0 n 5 RB E 6 o y MdazfoN i 1 4
PUS R P
A n e t cedi r n o ao f
)
1 F D e 8 K jnisc e 0 FS E h t
9 9 . oavf i 3
(
o 1 J s
rt PSDO i f N
3 a O SA A s u 1 m : N E
REV t a
y a o L L RHD t S M h
T T E NTA I
T T S
S R R R'
- E E E T T T T C N N N E : E E E J E S S S B T E E E O U S
A D
R P
R P
R P
~
4 9 9 9 9 81 1
r 9 r 2 9 e 9 e68 1 b 9 b99 t 1 m99 s m e e 1 1 vyygp uet n
u oaaue J NMMAS 0
0 6 )
P n R s A o Ee )
i t Sd A E a Do D c (
CS S i f t r/gR F
(
U i t
o l pi nC r a O e v O c ets A o H n Re/ n r p
G g nTo i p
N i
s ons A e
i I t os T
i d 0 at m n S 0 u n g ts6l em i ipa s E
mAEeC vmo e W b uhteypo l D l
st ept a n
erf s o uR aSERF i
u f S RSEn ont f h oarSeE FSo gt i F n
'e n aDDch c
o nsi l
u i
t i eeal i i
s sl epuuvb c d p ssd ue e
O h c
WaIssAPD I S / ///
)\ l '
e ee O t e ee t t t
i t t ii m mm e m e mm oo t
e o c t e cc t b t bb S muu
- i i R mu S SS m n m nn C o cig o gg A bs cii bss H ue uee T SD SDD -
I i
cro ioo cr r _
W l l u
a c t
a e u a
t t cc e _
r aa S r e t e ee t e -
d d G yR t i yRR t i
N Hd m Hdd m I l e m l ee m O T ac o acc o E mn r a C mnn r aa C E ev l l evv l l
M hd u hdd u TA F TAA F G
N I
M 88 O 88 99 99 99 C 99 8
1 1 8 9
P 1 1
, , 9 28,9 8
U 25 9 1 9 1 1 9 1 1 1 - - ,
- - 1 1 3
, 1 77 -
3 1 1 6 8 1 1 yy e ee y y aa n nnl l u uuu u MM J JJJ J O
s m
- t e
- s y
s S
S m C
I d
n t e
s P a y t
S O n e d T m t e
a p
S l
_ ' s i u e r R-K t e
E q
y
- E p a s t
e
_ E h c
t n
f a
e S-W g n n o
n o
S i w p s N O I H l l
o m o
e r
u s
n f
o T f o C t a o t e
i n
F e s F m t
a e h e a c e s m O t r y c
- i r f t
_ n u t e g i c n t n a o t c o a e e S s u f
a r e p
r u d T r
U r S P S s i
c y m t s
T e S d t s l c r
_ t f e e a A A o A i n o r e T i c
n y e t
a T e p
n g
e n
l a h t
i l
a r
e l
u g
S o i s
i g
i t
i c
e u v e e t e n n T Q S R n D E I
a , , 4 6 7 9 2 i
c 3 6 1 1 1 1 2 f r r r r r r r i
n t e t e t e t e t e t e t e
g p p p p p p p i
s a a a a a a a o h h h h h h h n C C C C C C C O -
r s
i a
n l i
x i
o u t
a a t :
n d e w ~
n s o
s e
l e s a l m
e r b eee s r u t p y uf s y
d s e e l aode s h t nl t s a tcta d g
i l
i t e
n s t n nd a i
r e e ear l Y u u m nmr e
r R d s s y nig i
y A s n
i c
t i tat n a t e
e i
r f M o i f g n n vcga i O M i
t s
i c e o c coonins -
U e u
e p
t i
n r n malor n S q s l l
e f ooleuo i sc o n r s s nh sl f s h e o e
w u c t s n niwaeo i ot antysnst s s n l r
i t coiee n i e e pii f t av m l
a d
o o mta miau cl eiqoa urt t t i n w t t e
e d ad spn nc r l
e t n easSeoat ey a gel mr b v ofe knc r l a a r l
i a
p e
t a c ni i nt at o
v r m0 osnihl dn i na a p e0 rl eu f f s6Btaciet aPBee upx g
f a
f a bALhT b oer e t t S S - - - - - - - -
O - -
.