ML20202B340

Transcript of ACRS Advanced Reactor Designs Subcommittee Meeting on 980204 in Rockville,Md.Pp 98-367.Reporters Certificate & Assessment Process Encl

ML20202B340
Person / Time
Issue date:	02/04/1998
From:	Advisory Committee on Reactor Safeguards
To:
References
ACRS-T-3025, ASB-300-129, NUDOCS 9802110191
Download: ML20202B340 (282)

Text

,

=

. ; ..; : .  :.&i.Q,Q.k p f-p; ; ~n..  ; ; . ; .;, i . :% y *l. .l g. ;m .. 'l7.:g .

sq ..;N: "p;.K.;[%!,;;n;f  :..

.. . n , ; . ? . e::;;i':a

.,%. :9.c: a , % . x.s ;g.!g:c,6.Q.

9 414 :;.

,:;syq;& :p1.gw . xm:k;.: .,- y.Q,,3

, ".: .:;m.

,, y ,r ;w..;

? ,, y : % . l b.\ y.,7:.y ya

,q.- . ..~;s. y u'.L,g z.v y-o.. ..g. .: A ;(...,. g.:

..t.  :

.. c;.7 .

s 4, ;. . \

. ' . ;. *:. _ ,. . ; a :.,A );m;; y. ...y , q:.

. L(:lv. Q;rp4-.,;< ' :- ',;t,. %e.- *. . . . .'. , .,; , ';. 9 ' f.

1. (l'! / r. y pp. 8. .

.w : :.'.$. g':1 . . :: .. ; : :. ';gv.: \ % s.,. '.m,x . e,;- >?,,Cf ;y %;:

L T  :: < - . q .. . . . . -

. p.

a _

.. r  ; .

l ,

, (

m ,u m, a m x y m . mn . . m %g_

4.l.:. a .L ,

k w' w' ';k&.s f,y p;?.k f:g).,::.j::y,

j. n'

wwwe'.,

s. c.p Q' .  ; 3,9;s .

' y:.,,y,' ~

y b;'.'y;

?y,:fl.m

. . Q .;f Q .1 i

q:.g ;n g p ;. m c m.:.,; e. v n ; q ..;

. *k.(:,:y.J e m'j* m.kn:Qi 5

f.  : YY? '

l :k: Y. 'f. ; o:: f': .

.hh k: . h .

.f. . ',f.

.: . .: s: < > ; , <

.- .. . . I Q . c t:

.c f  ;,

f ., y q , ,;y (, .;g(p,9.[.

, i '. 2 l Y 3 y s 7'. * // ',q$ .8'['

,; f , . . . i,T.[ . . .

.,'[.; Td4 -

'l'

'[(. ] .1 ].. h .

?: . N . if 15.jly. VfQ%,,$.1,P:i,jN"1.,&',U.7.c,V'@q.Ar,l$NQW.hQQ%M -% Q.3 fpikW. ,a

.ll.f;;}..}&@:.Q:,:.a;%q$..}...Y.,&:s.,lhn'

~ ~. 7 . .y D&. y: Q y Q[

o ke

%4 ' a y,p gRaf.  : . ,b,~;.LA.g:,.ag&,:;ny.9

?;Qe. j p y. y.xp:N. >;:. -

y ; . Q. c a. ,7Q,)Q::&..s.ll(W$..

, ,3; ni:; r. .m.%,..g..( @ g .

p

. ygp&,f.t yyy:. &gQ> s...

^

' 3 n:1.sW: e,.n . W W.&q . ,ill N:f?lv.N< i.d, .o sC Q (&.a/.c~M,Q, , O'y.. s

. 4 v... . .s . ., f. . ,g.4 . j c .. .. .. ...t ...y ss, y .. .. . ~. m 4 sl,. . ' \; f y;'f 'I  :

" h.!

h. *,,.f A'. , fl:f? -

.; 'l -

,, NY :f::y!; D.h..lQd;;f. :l;& k,h'k. 8.:f ' .h

.:...W; , ':.fL:

VtfRQ &;;Qfh::d ., ' % 1;Qp % . %M D:ky-Q,td >'. . &;::.d:.

. . . s . . c ! : : p;. z 'Hf f $.

., ,,Y:)@.5.fl'F:

, ':.. , w ; . ,es

.' :Q. -

c.d.:

• \

3, 9

.m . . . . h :? , . u.

t* . v. x .,, .

- .h.. :h . , .

}f  ; v - ;:,

l y,. . y.

..3 s v:::- :.,.

. s,; .Q;;l'.. O,. *; ?

. . ,a . . , s > 4'. f;

.. e :- x ,L, w/ %r  ;, ?.h s y !, .; f.: h '~4.'. f6 R , Q s: ;, f.. .., t . : : &, :.n,i ,[ llo?. l '

.[ , ? .; \" ~ ' W":i:y ,Q" ' u.::'

an;;a%y? o n.:.);';& lQ+.R:

.m J. qv f few ; .. - s. : .+

.". ;y; +y ..[;,,, .c.& :7:/:y:;p,y'i . . <s; e ,em,:1m

!;y R . %.

a q.

4.&-<w . . i ,:

. . .; , ,y ;p  ; e. k.. :.y.:e .: a ,. .W,g%;-

m lt,;g,g::;M.

e../ ,x a ;,,. , .it z ;;y' . T. :..

W n.y0:l e; W..q$ $ K:

.N. , .

Y0 k,'h M" e /n

u 2 ../ H :p ;g, N, p! }g;f.. gy q,Nly. ,r}slQ,t v M.E&&, ' f f.,p,b . llf.Q 4 '

' * ;; f..~w ,. f h  ;,.,..

hp . Q M . 3..:n M' ;; Q.;q;..c 4 W JAqsJ:g t;..f  : s Q ..';;- % .[;'

. ~..a .~ .c .. r .

.# h.MN't.k:h,.ve Y.g i N # ' : . Y . ;.p.e. w u ~.9 7h. N] i ,Y hl / * ;,

,Ie,.[k;J.'ii,";Y) S. ;

l? 'f:f.f.ff'

'h '.  : h- i: .' W.  ::hYYl i:'  ? e

?.k f.$: k h.,.f'. k.\,f; h .g, .

khb::g$hk'['.[:,,.hck h ggy,.z,% f k 'j.h .%

k.'[Flh

. Y.') . {IWh;h9[dhyd %

i i YhlffkY.k V

' ' ~ '

h i '?. L ~W~ '

~4 *

.n -

+r

,;:%:,.:4

+<i:4n

.$  ; c! 'q - * )^ }H %A .

f ~. ' 1 *b h , h}t - b 'jD

h. 1

-g h h' ' ' t k.IN}ikk

I i I y.w h:[ 7.,

6 W .up:  :.l

• d ., ,

. .g, g%

y e g s; 3

m s .

( ,

g y az I

&$ k 3 a y-y~

k N

t x

",'P'

.~ c y 3n >

.,  ;.3 n m

g.

[~ 4 r- r -

1:

p*

' La  ;

• Q  ;

w . u .- c .-

? . v J,ah ,=

$ 9802110191 980204 Y- 1

• PDR ACRS f N q M@W * /

E T-3025 PDR rJ  ?

y',Oh .. MM4%

Mdb . .

enegrumpensM N6

i CM XA _

Adf5F3625 OFFICIALLTRANSCRIPT OF PROCEEDINGS.

NUCLEAR REGULATORY COMMISSION ~

ADVISORY 'COMMITTEE ON REACTOR SAFEGUARDSf

Title:

ADVANCED REACTOR DESIGNS-SUBCOMMITTEE MEETING-ra 4 ica s Docket No.: , m _ ,_3; . g : :,: en s:x: s xn ..s I 41- .-

ANvc-Work Order No.:. ASB-300-129-.

~

~

LOCATION: b Rockville, Maryland

}

- DATE: Wednesday, February 4,199 PAGES: 98 - 367 ACRS The 0057~ leil 0":'.8_"O05t1630T'~it:ee bU,bn ANN RILEY & ASSOCIATES, LTD.

- A -

1250 i Street, NW, Suite 300 l _- Washington, D.C. 20005 (202) 842 0034 mm ,x - IllIIIll11il11 t

)- i r ,,;,

rs l .

DISCLAIMER UNITED STATES NUCLEAR REGULATORY COMMISSION'S ADVISORY COMMITTEE ON REACTOR SAFEGUARDS FEBRUARY 4, 1998 The contents of this transcript of.the-proceading of the United States Nuclear-Regulatory Commission Advisory

( ,/ Committee on Reactor Safeguards, taken on February 4, 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.

. o

_ 9 8 '

1 UNITED STATES NUCLEAR REGULATORY COlittiSSION 2- ADVISORY _ COMMITTEE ON REACTOR SAFEGUARDS 3 ***

4 ADVANCED REACTOR DESIGNS S SUBCOMMITTEE MEETING 6-7 U.S. Nuclear Regulatory Commission 8 Two White Flint North

. 9_ 11545 Rockville Pike 10 Rockville, Maryland 20852-2738 11 12 Wednesday, February 4, 1998 13-14 The Committee met pursuant to notice at 8:31 a.m.

L ) 15 16 MEMBERS PRESENT:

17 JOHN BARTON, ACRS Chairman

-ROBERT SEALE, ACRS Member 19 -GEORGE APOSTOLAKIS, ACRS Member 20 -THOMAS KRESS, ACRS Member

-21 DON MILLER, ACRS Member 22 DANA POWERS, ACRS Member 23 ROBERT SEALE, ACRS Member.

24 ROBERT UHRIG, ACRS Member 2 5 -- GRAHAM WALLIS, ACRS Member ANN RILEY & ASSOCIATES, LTD.

Court Reporters 1250 I Street, N.W., Suite 300 Washington, D.C. 20005 (202) 842-0034 i

---a

99 1 PROCEEDING-S

() 2 3:

(8:31 a.m.]

CHAIRMAN BARTON: Good morning.

4 The meeting will now come to order. This is the 5 second day of a two-day meeting of the ACRS Subcommittee on 6 -Advanced Reactor Designs. I am John Barton, Chairman of the 7 Subcommittee.

8 ACRS members in attendance are George Apostolakis, 9 who is expected to join us, Thomas Kress, Don Miller, Dana 10 Powers, Robert Seale, Robert Uhrig, and Graham Wallis. . We 11 also have in attendance ACRS consultant Jay Carroll.

12 The purpose of this meeting is to hold discussions 13 with' representatives of the NRC staff and Westinghouse 14 Electric Corporation to gather information concerning the

) 15- AP600 advanced reactor design. Presentations will include

.16 items related to the test and analysis program and the 17 Standard Safety Evaluation Report chapters 1, 4, 5, 7, 8

-18 11, 13, and 18. The Subcommittee will gather information, '

19_ analyze-relevant issues and facts, and formulate proposed 20 positions.and actions as appropriate for deliberation by the

.' 21 full Committee.

22 Noel Dudley is the Cognizant ACRS Staff Engineer 23 for this meeting.

24- The rules for participation in today's meeting 25 have been announced as part of the notice of this meeting ANN RILEY & ASSOCIATES, LTD.

O- Court Reporters 1250 I Street, N.W., Suite 300 Washington, D.C. 20005 (202) 842-0034 j

. - -... - - -... . . - - .- -.- - .- - - . . .~ - _ - . . -

100

~

Li- previously published in the Federal Register on January 14, O 2 .1998.

3 A transcript of the meeting'is being kept and will'  ;

4 be made available as stated in the Federal Register notice.

5- It is requested that the speakers first identify themselves 6 and speak with sufficient clarity and volume so that they' -

7 can be readily heard.

8 We have received no written comments or. requests ~

9 'for time to make oral. statements from members of.the public.

_10 We will proceed with the meeting, and I-now call '

11 upon Mr. Brian McIntyre of Westinghouse Electric Company:-to 12 - begin;

-13 MR. McINTYRE:: Thank you, Mr. Barton, 14 .The first presentation-today is on. Chapter 8,-

O '

V -15. Electrical-Power,-and that will be-done by Tom-Hayes. The:

16- _ slides for this are in the package that was handed out:

yesterday.

18 MR HAYES: Good morning,-gentlemen. My name-is-19 Tom Hayes, Westinghouse AP600. electrical engineer.

.20 I'm here to give you a relatively brief overview 21 of the AP600; power systems -- electrical power systems---

22_ plant service electrical systems, if you will -- and because 23 of the amount of time available, it's going to be a very

24- broad-brush overview, and certainly, you're welcome to-ask 25 questions of more-detail wherever you have an interest.

O ANN RILEY & ASSOCIATES, LTD.

Court Reporters 1250 I Street, N.W., Suite 300 Washington, D.C. 20005 o--

(202) 842-0034

101 1- LI'd 'like to. start out by stating the obvious. The I_ ( ) 2. AP600 electric systems are essentially what you would expect 3 to find in any conventional nuclear plant except for the 4 places where they're different. I have a point there.

5 The reason is the components are not significantly 6 different -- a transformer is a transformer.is a 7 transformer, has been that way for many years.

8- What's different about AP600 is more as a result 9 of the passive nature of the plant --

some of the pedigrees 10 are different, some of the importance is different, but the 11 actual system architecture _is not significantly different 12 from what you would expect just normally, and for this talk, 13 because of the time available, I'd like to focus.on the 14 differences.

() 15 Following the guidelines of the-EPRI utility 16 requirements document, the AP600 electric systems follow-a 17 three-tiered approach to the design of the systems.

18L Tier one is, in general, what, for this body, is 19 probably considered unimportant loado. I always have-to 20 bite my tongue as I say that, because the man that owns the 21 plant considers them very important. If they are not there, 22- he can't make power. But from a safety point of view, 23 thay're relatively unimportant, the loads associated with 24 running the balance of plant.

-25 Skipping down to tier three, that's your standard AUN RILEY & ASSOCIATES, LTD.

Court Reporters 1250 I Street, N.W., Suite 300 Washington, D.C. 20005 (202) 842-0034

) _ . . _. .-

(;

102 1 _ safety systems Class 1E that you're used to seeing in a

() -2 power plant, and we have, in the middle tier, called tier-3- two, which is probably:where we spent most of our time-in 4 the design of-AP600,-it'r-that great gray area in_between-5 -what is traditionally non-safety and safety -- lots of 6' discussions with staff on exactly _what tier two means, 7 several gradations of what-tier two means, 8 Dealing with different -- I don't want to call

9. them interest groups but different disciplines in'the staff, 10 you get different approaches as to what tier two really 11- means, but-it's something that's half-safety, if you will.

12 The term defense-in-depth has been used. The term 13- regulatory treatment of non-safety systems has been used.

14 All of those features impact tier two to a_certain extent, 15 a.nd you'll find'that there's not a closed form answer to 16 what do we do_with tier two. It depends on whac part of 17 tier two you're talking about.

18 Now, specifically what the power systems are that 19 arefin AP600:

L20 The off-site power system, which is essentially 21 outside of the design certification -- it's an interface to c 22 .the plant, but it's not part of the design certification.

23 On-site power systems -- we have your standard AC 24 delivery systems, 60 hertz, 4160 volts, 480' volts, standard 25 inclear power plant or maybe you would prefer -- the analogy ANN RILEY & ASSOCIATES, LTD.

Court Reporters 1250 I Street, N.W., Suite 300 Washington, D.C. 20005 (202) 842-0034 m

103 l 1 is maybe better_-- a high-grade industrial plant.

'2 We have main generator, which is very 3 conventional, and;in the DC and UPS systems, we have:one that's Class-1E and one that's non-Class 1-E,

~

4 and again, 5 that's pretty standard for a nuclear power plant.

'6 Going down that list of systems, what are the 7 differences?

8 Off-site power systems -- AP600 has only one 9- .off-site power connection. For the actual design of the electric systems, that did not affect us.

.11 We actually penetrated the fence in two places, 12 but the licensing basis is.that those two connections that 13 go through the fence out-into tim transformer switchyard do 14 .not:have to be independent off-site.

O

(,,/ 115 Passive-nature of the plant makes it such that the 16'- assumptions.that were written into GDC-17 that=say AC power 17 is necessary for the safety of a plant are not. valid for 18 -AP600, 19- = MR. CARROLL: So, if this was a, quote, 20 " conventional" nuclear power plant, there'would be two 21 off-site.. connections.

22- MR. HAYES: Absolutely. Two independent, right, i 23- We do not require that.

24 -

It's one of-these places where the-technical 25 answer is real easy: It's a safe plant, so you don't need i

i ANN RILEY & ASSOCIATES, LTD.

Court Reporters 1250 I Street, N.W., Suite 300 Washington, D.C. 20005 (202) 842-0034

9 104 1 'it . . The leg 11 answer gets a lot harder, since GDC-17 is

() 2

-3 law; we have to get exemption from it. But I believe_we have resolved all of the problems associated with that.

4 MR. CARROLL: Now,-how would it complicate life 5 for somebody who bought one of these and-wanted to have two 6 off-site sources of power?

7- MR. HAYES: Well, as I mentioned a little while 8- ago, we've penetrated-the fence twice. We have a preferred a 9 power supply and what we call a maintenance' power supply.

10 So, if your site has tw independent sources, we have 4

11 provisions to use them.

12 MR. CARROLL: Okay.

13 DR. POWERS: In-making this design decision, did 14 you bear in mind questions that have arisen over the last

() 15 four or.five years concerning the_ reliability of the U.S._

16 electrical grid?

17 MR. HAYES: As far as just the stability of the-18 grid itself and is the grid going to=be there? Yes, we did.

19 DR. POWERS: What was your general conclusion jl 20 about the stability of the grid?

21 MR.1 HAYES: It's not yet a problem.

22 DR. POWERS: That's true. You're .at yet building 23 a plant either.

24 MR. HAYiS: I'll give you two answers to that.

25 One is the sale pitch answer, which says the grid

(-

[\~- ANN RILEY &. ASSOCIATES, LTD.

Court Reporters 1250 I Street, N.W., Suite 300 Washington, D.C. 20005 (202) 842-0034 1/. Jr c- ___.J

\

105-1 'is going to be as reliable as it's gcing to be and this

((

2 plant is much less -- is much safer with an unstable grid 3 than a conventional plant, because it doesn't need the grid.

-4 So, if I take the marketing approach, I'll tell you that, s- 5- for us, it doesn't matter.

6 On the other side, yes, we have worked through-

-7 some of the potential problems that t.here may be as a result 8 of grid instabilities and designed the protection logic, the 9- grid disconnect logic, if you will, ;uch that_the plant, 10 when and if-it's required to separate from the grid, it does 11 so gracefully.

12 This plant has its own generator circuit-breaker, 13 so there's not a bus transfer problem. - If the grid goes 14 away, the plant will reduce down to house loads, and all the

() 15- loads in the plant will continue to be served by the main 16 _ generator if it had been on-line at the time the grid went 17 down, i

18 If the-plant trips and causes.the grid to have a 19- problem, which is really the more likely scenario and 20 probably the one that-gets the -- the scenario that gets the 21- .most attention, we have prevented the generator from 22 tripping off for at least 15 seconds following a turbine 23 trip to keep the voltage up.

24 The reason we need to keep t.he voltage up is just 25 to ensure that the RCPs stay turning as long as they can, t ANN RILEY & ASSOCIATES, LTD.

Court Reporters 1250 I Street, N.W., Suite 300 Washington, D.C. 20005 (202) 842-0034

.. . _ _ _ - _ _ ____ l

,.__.m _ . _._.._._._..___m..-. ._ _ __ _ _ . ._.. _._ _

106 1 cr.pturing the inertia-from the turbine to get the natural 2- circulation going.

3 Once we get natural circulation going, we don't 4- need the grid, and at that point, you're talking about a 5' . grid that h'as gone down and it didn't matter if you had two 6 off-site connections or not, the grid was gone,. and we

7 _believe our passive systems will kick in - well, we know 8- our passive systems will kick in and will bring the plant-9- down safely.

110 Now, bre'we more susceptible to grid problems than 11 1 aLconventional plant? I don't think so. The PRA results.

12 say we're not.  !

13 When we go a little deeper in here, you'll find we 14 do have two diesels that will start._ They are not safety.

15 That will be one of the' differences. But between'the l 1

16 passive systems and the fact that we do-have two diesels, we 17 believe, if anything, we're safer following a grid 18 disturbance.

19 DR. KRESS: Have you tested that inertial tapping 20 of_the turbo-generator to see if it works well?

21 14R . HAYES: Not specifically.

22 DR. POWERS: They tried that at Chernobyl. It 23 didn't work out too good.

24 DR. KRESS: It might have worked pretty good, 23: DR. POWERS: Actually, the idea worked better than

(

.f*}% ANN RILEY & ASSOCIATES, LTD.

Court Reporters 1250 I Street, N.W., Suite 300 Washington, D.C. 20005 (202) 842-0034

107 i

li you might suppose in_the RBMK, but RBMK'is a long ways from

() 2 here._

l 3 MR. RAYES: Not specifically for AP600, though.

4 MR. CARROLL: I think it's well-established that

.5 --

6 MR. HAYES: Again, not a difference from a 7 conventional plant.

8 MR. CARROLL: People are wondering,'wouldn't it do

-9 it forever,-then, instead of just_15 seconds.

10 MR. RAYES: Sooner or later, it gets too slow.

11 MR. CARROLL: Well, you'd also have some heating 12 problems in the generator.

13 MR. HAYES: Absolutely. And it's not that we trip 14 after 15 seconds. We block it for 15 seconds.

15 After the 15 seconds, I don't know whether it's 16 going to trip or not. It will be_a function of things like 17 what's the generator protection look like, and we have not 18 established the. exact parameters of the generator protection 19 system.

20 DR. SEALE: I would point out-that'there is an 21 order of magnitude more difficult and detailed questions in 22 this area that would have to be addressed for a plant built 23 in many parts of the world, because there is no externe; 24 grid to speak of, certainly no reliability of an external 25 grid to speak of.

() ANN RILEY & ASSOCIATES, LTD.

Court Reporters 1250 I Street,'N.W., Suite 300 Washington, D.C. 20005 (202) 842-0034

_ _ --J

108 1 -So, from a technical point of view, this problem i 1

() 2 has got to be solved in a lot more detail with a lot more 3 serious interruptions and so forth than anything we're 4 talking about here.

i 5 MR. HAYES: But again, I'll remind you, we're much 6 less censitive to grid disturbances than a conventional 7 plant.

8 DR. SEALE: I didn't say it was any harder.

9 MR. HAYES: The grid is less important.

10 DR. SEALE: It's got a lot more hair on it.

11 MR. HAYES: You got that right.

1' DR. POWERS: Who gets -- who goes out and gets 13 this exemption from GDC-17, you or.the guy that wants to l

14 build the plant?

l O)

(_, 15 MR. HAYES: No , it will be part of the licensing 16 basis, t.d when staff gets up, you can ask them where it 17- stands.

18 My understanding, as a non-licensing person, is 19 that we write a letter requesting it, which we have done,

! 20 and then it comes inside these walls and I lose track of it, 21 quite frankly.

22 MR. HUFFMAN: This is Bill Huffman for the staff.

l 23 They have requested the exemption. The technical 24 group has reviewed the acceptability of the exemption, and 25 the exemption will be part of the rule for this plant.

i l

1 l

() ANN RILEY & ASSOCIATES, LTD.

Court Reporters 1250 I Street, N.W., Suite 300 Washington, D.C. 20005 (202) 842-0034 l

t

t 109 1 DR. POWERS: We've gotten the sale pitch argument;

() 2 3

we've gotten vague references to PRA analyses. What all accompanied this letter? I mean surely the letter didn't 4 say, dear sir, we want an exemption from 17, thank you very 5 much. There must have been some defense of the position.

6 Was that a probabilistic defense?

7 MR. HAYES: The letter itself really was about 8 what you just described, but the defense is the SSAR, all 9 20-some chapters of it, and I'm sure that was used in the 10 evaluation.

11 DR. POWERS: So, it was not a probabilistic 12 argument.

13 MR. HAYES: No. No , it was a passive plant 14 argument. But it was backed up with some -- I mean we 15 submitted a full PRA. So, buried in there, you can find 16 some answers to those kinds of questions.

17 DR. POWERS: Yes, that's the problsm, it's buried.

18 Is it an explicit treatment or is it an implicit treatment?

19 MR. HAYES: There is an explicit treatment of 20 station blackout conditions.

21 DR. POWERS: So, it's in the station blackout 22 section.

23 MR. HAYES: Yes.

24 DR. POWERS: Okay.

25 MR. HAYES: Okay. I got ahead of myself a little I ANN RILEY & ASSOCIATES, LTD.

k Court Reporters 1250 I Street, N.W., Suite 300 Washington, D.C. 20005 (202) 842-0034

110 1- 'while ago. .This is basically the-approach we've taken for

)- 2 how we used the fact that we-have a main generator 3 circuit-breaker and not a second grid connection, sequence 4 of events, depending on.what you're losing, how you 5 gracefully bring the plant down.

6 MR. CARROLL: Our former colleague, Charlie Wylie, 7- would definitely approve of all of that. 1 8 MR. HAYES: Bring him in.

9 Okay. .More differences.

10 Main AC power system is not safety-grade.

'll Terminology-wise, we've had some confusion in some .

12 documentation of the plant. We like to say we have no 13 safety-grade AC, which is sort of true. We do have i 14 battery-fed inverters that provide safety-grade UPS to 15 instrumentation cabinets.

16 So, if : slip up and say we have no safety-grade 17 AC, I'm not trying to confuse you. That is kind of the way 18 we ---we're talking heavy power. We don't haveLany 19 safety-grade heavy pow. -

We like to talk in terms of our 20 1E stuff is all4DC, but it's also DC-fed inverters for some s21 limited I&C-type loads.

22 The main AC power system has one relatively small 23- 1E function associated with it, so it's not totally clean.

24 .It's required from a safety analysis point of view that the 25 RCP breakers open under some event scenarios. That becomes

(\' .

ANN RILEY & ASSOCIATES, LTD.

Court Reporters 1250 I Street, N.W., Suite 300 Washington, D.C. 20005 (202) 842-0034

_ . . ,.___-_____._.-_m__ . __ _ __ _. _ , _ _ _ . _ _ . . . . . _ _ _ _ _ _ _..m.._

l .- 111 1 ' 1E . .- So, the-RCP circu't-breakers _are-1E, but that's-the 2 -onlpfAC_ power system _ device -- devices that-are 1E.

3- Otherwisa, the AC power system is divided into two -

\

4' load groups. Everything important in the plant there is

-5 redundancy for. It's fed by two' load groups, but I want to

. 6 '- caution that.I not mislead you. They are-not separated, 7- they'are not independent, simply:because.it's not needed for 8 this plant.

9 However, there'.s generally redundancy,~not.

. 10 necesstirlly single-failure in the pure sense, but wt'll 11 survive most single failures simply because of the 12 -redundancy.

13: The-diesel generators are non-safety. We believe L . 14 they're every bit as good as Class 1E diesels, but they--just l

' f)

V - 15 don'.t have_all the paperwork or the tech specs.- So, we may

-16 have a minor disagreement that we can agree-to disagree on 17 _. between us ar, d the staf f.

- 18 MR CARROLL: One thing that jumped out at me in l 19 going over this one more time was that you've_put the j .; 20 radiators for the diesels up on the roof of the --

21 MR. HAYES: And they-are not missile protected up.

22 there.

23. MR. CARROLL: Yes.

24 MR. HAYES: No,. sir,_they are not.

25 MR. CARROLL: Why wouldn't_you spend a few extra ANN RILEY & ASSOCIATES, LTD.

Court Reporters 1250 I Street, N.W., Suite 300 Washington, D.C. 20005 (202) 842-0034 s-

112 1 bucks and;make tnat problem go away. *

,-s

( 2 MR. HAldS: Because we don't believe missiles are

.3 -- I mean if there is an event that has missiles, we don't 4 believe it will take out both sets of radiators, although it i

P 5 {-could. We believe, in that highly unlikely event, the plant 6 is still safe.

7 =For the event that --

I mean I'll be honest with 8 you -- for the event that generates a missile that big, the 9 whole building is going to fall down. It's not a seismic 10 building either.

11 DR. FOWERS: Wind doesn't get them, wind-driven 12 missile?

13 MR. HAYES: The building has no high-level wind 14 criteria It's just your standard design building, bl

(_/ 15 So, if you start talking hurricane-grade winds or 16 tornado-grade winds, which are -- you really have to get up 17 in the 200-mile category before you start getting much of.a 18 missile -- the side's coming off of that building.  !

19 MR. CARROLL: I don't know. I think I would have 20 spent a few extra bucks and taken care of those things.

21 DR. UHRIG: T:e generators are sized so that only 22 one is required to carry the load?

23 MR. RAYES: Only one is required to carry the --

24 what we call defense-in-depth loads. I almost said the word 25 " essential," but that's not right.

ANN RILEY & ASSOCIATES, LTD.

Court Reporters 1250 I Street, N.W., Suite 300 Washington, D.C. 20005 (202) 842-0034

113 1 The loads that are there to minimize challenges to if )- 2' the passivo systems can be-satisfied by one diesel, that's 3 correct. Each diesel has an identical set of loads hanging 4- on it, purely' identical, for all practical purposes identical.

6 Either one starting gets you one group of fully 7 redundant things that are needed to keep the passive systems 8 from doing ugly things or unfortunate things, minimizes 9 challenges.

10 The intent here is that, 99 percent of the time, 11 both diesels will start, and following the start of both 12 diesels, you now have twice as much stuff as you need. All

-13 the HVAC kicked on; you only needed half of it. All of the 14 pumps kicked on; you only needed half of them.

() 15 You will start turning off the things that you 16- don't need two of and bringing on some lesser important but 17 still relatively important loads, but that'1 going to be a 18 manual operator-directed thing.

19 DR. UHRIG:- Not automatic.

20 MR. HAYES: Right. The automatic is load up those 21 loads required to minimize challenges to the safety systems 22 and you'll get two of everything most of the time, and if 23 you have a single failure, you get half of them.

24- DR. UHRIG: Manual loading up is part of the 25- emergency operating procedures?

ANN RILEY & ASSOCIATES, LTD.

-Court Reporters 1250 I Street, N.W., Suite 300 Washington, D.C. 20005 (202) 842-0034

.- - - . - . - - - - - - - .~ _-. - - .- .. . . ..... - . - -,

)

d

114-  !

1 MR. HAYES: We_ don't have procedures yet, but it  !

() 2 will be,.yes.

DR. SEALE:

l 3 -I ' nt curious . Our colleague, Dr.

4- Wallis, raised a question that I'd like to pase along at j 5 this point.

f -6 In.the operation of the' passive' systems, as I 7 understand it, the intent is that the operators are deaf, 8 dumb, blind, and mute and are not expected to intervene in

] 9 any way.

10 MR. HAYES: That's a bit of an exaggeration.

11 .DR.-SEALE: You have these systems -- I've 12 forgotten the terminology used for them but the systems that 13 supposedly will provide the first line of defense against an 14' initiator of some sort so as to preclude the need for-the

.()

4 15 passive systems to become engaged.

16_ In that case, there is apparently a fairly active 17 role:for.the operators to play, potentially anyway.

. 18 Now, I'm sure you've gone through and convinced 19 -yourself that,-if-these operators do the things you expect 20 _them to do, that when they get through-with all of this, 21 they've got a plant that's being protected by these 22 non-passive first line of defense systems, and the L23 transition,=if there is some failure along the way of those, i-24 :to reliance on the passive. systems will be smooth, seamless.

25 Presumably, the only thing that's necessary is for l ANN RILEY & ASSOCIATES, LTD.

Court Reporters 1250~I Street, N.W., Suite 300 Washington, D.C. 20005 (202) 842-0034 1

115 P

1 the air-operated valves to actuate to give you the alignment

() 2 that you need for the passive system to work.

3- MR. CARROLLt Plus some DC.

4. MR. RAYES: DC MOVs are in there, also.

5 DR. SEALE: Yes.

6 .MR.-HAYES: " Smooth" may not be a_ word I would-7 have used.

M DR. SEALE: So, it's not seamless. Okay.

9 Anyway, have you gone through and asked yourself-10 how can the operator ucrew this up? --

11 MR. HAYES: Certainly.

12 DR. SEALE: -- and make it not work?

13 MR.. HAYES: And the answer to that is in the 14 safety systems, and the initial conditions assumed for the

() 15 safety system, the automatic part, the operator, at lease to 16 stay within tech specs, will not be allowed to get out of 17 that arena. We've made it very ditticult for him to get out 18 of the assumed initial conditions.

19 In general, che longer he waits to screw up, the 20- cooler the plant's going to be anyway. So, if he screws up 21- in the first 30 seconds following some major transient, he 22 can't have run the plaric too far past the initial 23 ' assumptions.

24- ; The initial v .aptions for the safety analysis, 25  ! which rely heavily on - safety systems, the second tier in

~ i l

.( . ANN RILEY & ASSOCIATES, LTD.

-b

Court Reporters 1250 I Street, N.W., Suite 300 Washington, D.C. 20005 j (202) 842-0034

116 1 your terminology --

() 2 DR. SEALE: The t,assive systems.

3 MR. HAYES: -- the passive systems -- we believe-4 we have done a thorough job in muxing sure we have valid 5 initial condition assumptions to include the things the 6 operator can.do to screw up.

't In general, all of the things he could do to screw 8 up are usually down the line and could throw the plant into 9 sotaething that would be heading towards an outside .of an 10 analyzed condition but.less severe than what we did analyze, 11 just squeaked out in a different direction.

12 But the tocus has been on, no matter what the 13 operator does, the passive systems will ultimately kick in 14 and do the correct thing,

) 15 MR. CARROLL: But he can defeat those if he put 16 his mind to it.

-17 MR HAYES: Not easily.

18 .MR. CARROLL: Why is that?

19 MR. HAYES: This is more of an I&C question, which' 20 is later on today, but I'll give you the quick overview.

, 21 From the control room, the things he can do in the

.22 way of blocks and permissives are limited.

23 We have actually built in logic into the

, 24 protection system, what are the things he's allowed -- what 25 are the actions he's allowed to block, reset, provide i

i .

h ANN RILEY & ASSOCIATES, LTD.

Court Reporters 1250 I Street, N.W., Suite 300 Washington, D.C. 20005 (202) 842-0034

117 1 permissives for, and they are keyed to plant conditions.

() 2 So, he's not allowed to block a safety action if

.3 the plant is in an arena _where that safety action is likely 4 to be needed.

5 Yes, he can always go down to my battery rooms and 6 start taking a pick-ax towards the batteries, but we're not 7 talking about malicious here. We're talking about a 8 well-intentioned -- the well-intentioned operator is going 9 to, as a minimum, have to leave the control room'and start 10 going and doing outside-of-normal things to --

11 MR. CARROLL: So, there's not a switch in the-12 control room for motor-operator valves that he could close 13 the valve that's supposed to be open?

14 MR. HAYES: There are, but they don't work all the

) 15 time. They only work when the plant has met certain 16 conditions.

17 MR. CARROLL: Okay.

18 MR. McINTYRE: This is Brian McIntyre from 19 Westinghouse.

20 I think, when we talk about Chapter 6, when Terry 21 Schulz is here, he'll present something which are called 22 systems event matrices, and it looks at the various levels 23 of defense.

24 We've presented those, I think, one other time, 25 and you can see that there's not like a passive system on, O ANN RILEY & ASSOCIATES,-LTD.

Court Reporters 1250 I Street, N.W., Suite 300 Washington, D.C. 20005 (202) 842-0034

118 1 passive system off type switch, but there are various --

2- actually, we think, vtith tihe passive systems, that we have 3 more diversity and redundancy as far as systems and options 4 go.for a transient.

5 So, if something doesn't_ work or an operator does 6 -- I_ don't know what term was being used here -- defeat it, 7 there are other things that will_ occur in the plant.

8 MR. CARROLL: Okay. I have always believed 9 .there's two sides to that question. Taking the operator's 10 options away may be a mistake in some cases.

11 It's like locking valves. I've never believed in 12 locking valves, because sometimes, even though it's obvious 13 that that valve'should never be opened, sometimes we find 14 circumstances where it ought to be, then we run around 15 looking for a key.

.16 MR. HAYES: And Brian's right. The right chapter 17' to address.this in-is 6, slightly in 7. You may want to

-18 bring it up in the Chapter 7 discussion later on today.

19 But in general, we've taken it a-step away from 20 the way you would like to see us go. We have more things 21 defeated, less places the operator can get in trouble in 22 this plant than what you would normally expect.

23 The other thing that's slightly different about

24 -AP600 as a result of some of the concerns for post-72 hours, 25 three days after loss of all AC,.what are we going to do'for 1.t.H RILEY & ASSOCIATES, LTD.

Court Reporters 1250 I Street, N.W., Suite 300 Washington, D.C. 20005 (202) 842-0034

119 l

{ 1 this plant, our batteries are only good for three days. l e

() 2 3

We have added on-site small generators --

" ancillary" is our word for small -- 25KW each -- that are i i

. 4 there purely for post-accident monitoring, so the governor >

5 can rest easy at night when we've gone through one of these ,

i 1

1

-6 events where we have nothing and won't have rnything for a i 7 while, i

8 MR. CARROLL: They're not big enough to run the '

9 battery chargers, though.

10- MR HAYES: No.

11 MR. CARROLL: Okay. ,

.12 - MR HAYES: Batteries will not recharge from them.

t 13 MR. CARROLL: But if the stand-by diesels are --

14 MR. HAYES: Oh, absolutely,

) 15 MR CARROLL: -- you're in great shape.

16' MR. HAYES:

Absolutely. We're talking 4,000 KW

!- 17 here; we're talking 25 KW here. *

18. MR. CARROLL: So, the thinking is that, in that

[ 19 72-hour period, you'd be able to bring in some skid-mounted 1

1

-20 -diesels? .

21 MR. HAYES: There's two answers to that.

22 The official licensing position is you don't have 23' to bring in anything, you don't need them, because the 24 plant's safe.

25 There are some extremists of us that says you e .

I-ANN RILEY & ASSOCIATES, LTD.

Court Reporters i- 1250.I Street, N.W., Suite 300 l Washington, D.C. 20005 (202) 842-0034

. . , . , . . _ . - . _ , . - . . , _ , . . . . , . __ . , . . _ . . . . . _ . _ . . _ . , _ . _ _ . _ . . . - , _ . _ _ . , , . . _ _ _ ~ - . , . . .

120 1 don't even need these, because the plant is safe. You don't

() 2 need to know it's safe. If you believe the psychologists 3 among us, yes, you do need to know it's safe, so we have to 4 put them in there.

5 That's all these generators do. They provide 6 lights and a little bit of ventilation in the control room, 7 and they provide minimal displays for critical plant 8 parameters. That's it. That's all you need.

9 Now, in reality, you're putting in skid-mounted

\

10 generators, absolutely, but we had to go through the 11 exercise of showing what if you were Turkey Point in the 12 middle of Hurricane Andrew and there were better uses for 13 big skid-mounted generators that just might happen to be 14 available from the National Guard? Perhaps a hospital would 15 rather have that trailer-mounted generator than you.

16 MR. CARROLL: I hope they have their own and that 17 they work.

18 MR. HAYES: So do I.

19 DR. MILLER: In your SSAR, you say -- it's stated 20 that the batteries themselves will shut down the plant and 21 maintain it in a safe shutdown condition, period.

22 MR. HAYES: Well, the " maintain" is really not a 23 function of the batteries. Once it sends those valves to 24 their safe position --

25 DR. MILLER: Right.

ANN RILEY & ASSOCIATES, LTD.

Court Repcrters 1250 I Street, N.W., Suite 300 Washington, D.C. 20005 (202) 842-0034

i l

l 121 1 MR. HAYES: --

the batteries are done. So, the (b 2 " maintain" is really less -- yes, it's stated that way, but i 3- the intent there is, once the batteries send those valves to 4 their safe position --

, 5- DR. MILLER: --

the passive system will take it. l 6 MR. HAYES: -- passive systems kick in, the-valves 7 are aligned. There's even some argument that we ought to be i 8 going out there disconnecting batteries just to make sure

9 they don't go the other way.  ;

10 DR. MILLERS So, from that point of view, that's i 11 where you take the position that you don't need the i 12 ancillary diesel.

1 13 MR. HAYES: Right, f 14 DR, MILLER: In fact, the ancillary diesels are

( 15 not hooked to any valves. They couldn't make a valve move 16 if they wanted to.

4-17 They're only hooked to the post-accident 18 monitoring system, the control room lights and fans, and i 19 some pumps that we have for putting water back up on top of 1

l 20 the containment for cooling. But that's all outside i 21 containment. That's just providing more water.

22 MR. CARROLL: How about control room habitability?

23 MR. HAYES: We have-ventilation fans that are 24 powered from these. It's outside ventilation, no 25 air-conditioning.

ANN RILEY & ASSOCIATES, LTD.

Court _ Reporters 1250 I Street, N.W., Suite 300 Washington, D.C. 20005 <

(202) 842-0034

122 1 MR. CARROLL: Why do I remember, a couple of years

() 2 ago, a big controversy about how many air bottles you needed 3 and all that stuff?

4 MR. HAYES: That's pre-72 hours. The air bottles 5 are gone at 72 hours8.333333e-4 days <br />0.02 hours <br />1.190476e-4 weeks <br />2.7396e-5 months <br />.

6 MR. CARROLL: Okay. And then what happens?

7 MR. HAYES: We open the doors and turn on the fans 8 and bring in outside air.

9 MR. CARROLL: Is the outside air filtered?

10 MR. HAYES: No. ,

11 MR. CARROLL: So, the operators, at that point, 12 are in respiratory protection?

13 MR. HAYES: No.

14 MR. CARROLL: Why?

15 MR. HAYES: The Jim Grover analysis shows that 16 it's not needed.

17 MR. DUKE: This is Brian Duke, Westinghouse.

18 Even with the LOCA source term, the dose at that 19 point is below --

20 MR. HAYES: It's below the GDC-19 requirements.

21- MR. DUKE: We can meet the dose limits.

H22. MR. HAYES: We can meet the GDC-19 requirements ,

23 without supplemental filtration at that point.

24 MR. CARROLL: Okay. This is after 72 hours8.333333e-4 days <br />0.02 hours <br />1.190476e-4 weeks <br />2.7396e-5 months <br />.

25 MR. VIJUK: That's right.  :

ANN RILEY & ASSOCIATES, LTD.

Court Reporters 1250 I Street, N.W., Suite 300 Washington, D.C. 20005 (202) 842-0034

4 123 b 1 MR. CARROLL: And the staff has bought that?

2 MR. VIJUK: That's my understanding, 3 MR. HAYES: I'm not sure the right staff is in 4 hera right now.

5 MR.-HUFFMAN: This is Bill Huffman of the staff.

6 Yes, the staff has done their confirmatory 7 analysis, and-they agree, it's within the 5 rem.

' 8 MR. CARROLL: Okay.

1 9 MR. HAYES: Okay. Just to reinforce the diesels a 10 little bit, they're not safety-grade, they don't need to bei 11 4,000 KW each, so they're pretty good size. They're more or

12 less the size you'd expect to find at a nuclear plant. They 1- l

13 are there, and I've stated here, they're provided for i 14 defense-in-depth and investment prctection.

15 'Again, getting back to the earlier question, the '

16 defense-in-depth loads are the ones that come up first, the i

17 ones that are there to provide -- to prevent challenges to 18 the passive systems.

4

, 19 Then, if you get both diesels started, you can put 20 on some of the loads that are there to make the owner a 21 little happier and keep his expenses down.

^ 22 Diesels require -- start time requirement is two i 23 . minutes, not the 10 to 20 seconds you're used to seeing, 24 which makes the diesel manufacturers a. lot happier.

25 -I bring this auxiliary system bullet -- I bring up

, ANN RILEY & ASSOCIATES, LTD.

Court Reporters 1250 I Street, N.W., Suite 300 Washington, D.C. 20005 (202) 842-0034

. __ ._ - . 1. _ ._ __ -- - - . _ . -- - - ---

124

-1 because, since this is a non-1E diesel, we are taking the

() 2 approach of' buying as many of the auxiliary systems --

3 cooling water, what have you -- as part of just the diesel 4 engine _ package.

S' Therefore, we have minimized the amount of 6 auxiliary system description that we've put in the-SSAR,

-7 _

which is somewhat different from just a SSAR point of view 8 than cony ..tional plant, and'there are some Chapter 9 issues 9 that end up getting addressed in Chapter 8, because we've 10 -taken the position that, because fuel is the only one that's 11 AP600-specific, we're going to talk about the other 12 auxiliary systems in Chapter 8.

13 So, when we have the -- when you have the Chapter 14 9 review, you may end up with some Chapter 8 discussions

() 15 again.

16 Basically, good. commercial standards is what we're 17- going to use for these diesels, rather than IEEE 18 umpty-scratch for 1E diesels.

19 DR. POWERS: Don, in what series of IEEE standards 20 does umpty-scratch appear?

21 MR. RAYES: We're not using them, was the point.

22 'That's why.I don't know them. If we were using them, I 23 .could probably quote them.

24 DR. MILLER: He's not using umpty scratch or 323 25 or 344.

O-ANN RILEY & ASSOCIATES, LTD.

Court Reporters 1250 I Street, N.W., Suite'300 Washington, D.C. 20005  !

(202) 842-0034

_a

125 1 MR. HAYES: Yes, all of those.

() 2 Ancillary diesels -- we pretty much talked about 3 these.- These basically are small construction-site-type 4 diesels, very rugged. There have been some issues about can 5 these things survive a seismic event, and our crude answer 6 is you can kick them off the back of a pick-up truck, you 7 know.- Sure, they can survive seismic. They're in a 8 seismically-protected area of the plant, unlike the bigger 9 diesels. So, they will be protected against missiles; they 10 will be protected against all the GDC natural phenomena.

11 DR. MILLER: But they haven't been officially 12- qualified.

13. MR. HAYES:

No, they have not, and we've, at the 14 moment, taken the position that we could qualify them but we

) 15 don't see a reason to, and we believe staff has accepted 16 that. Some of the things about them do need to be -- the 17 missile protection is part of the building design criteria 18 that needs to be addressed. But because of-their rugged 19 nature and the easefof replacing them -- we recognize that 12 0 we've got three days to figure out we have a problem before-21- we even need these things. They're small. Every Aaron

~22 Rents has one, or something similar.that can be used. So, 23 you're really getting'down into the mud if you want to start 24 gold-plating these things.

125 Okay. Let's talk about'the Class 1E things just O ANN RILEY & ASSOCIATES, LTD.

1250 I Street, N.W., Suite Court Reporters 300 Washington, D.C. 20005 (202) 842-0034

! 126 l 1 briefly.

2 MR. CARROLL: Have you made'it easy to hook up 3 these ancillary diesels?

4 MR. HAYES: Absolutely. The only question is 5 -- going-to be, do you have to have a licensed electrician, or 6 can I come in and do it and still not violate the law?

7 MR. CARROLL: Is that also true of the bigger 8 ones?

9 MR. HAYES: No. The bigger ones are going to be a 10 good bit harder to replace.- They're 4,000 KW. You don't 11 just bring one of those in in the plant manager's pick-up 12 truck.

-13 MR. CARROLL: No , I understand that, but once I 14- got-it there, have you_put in terminal blocks where it's 15 easy to --

16 MR. HAYES: No.

17 MR. CARROLL: Okay.

18- MR.- HAYES: No . - It would require an electrician 19 going in there and actually disconnecting and reconnecting 20 --

21 MR. CARROLL: Okay. Running cable and --

22 MR. HAYES: Right.

23 MR. CARROLL: Okay.

24 MR. HAYES: Class 1E system -- slightly different 25 from what conventional plants are in that -- I'm speaking as ANN RILEY & ASSOCIATES, LTD.

Court Reporters 1250 I Street, N.W., Suite 300 Washington, D.C. 20005 (202) 842-0034

127 1 the electrical man. I provide four completely independent

() -2 divisions.

3 Now, there has been a lot of concern during the 4 licensing process that 19t's not get confused that this is 5 such that we have four-way redundancy. We don't. I do from 6= the electrical point of view.- I have four completely 7- independent systems, but the loads I feed are not four-way 8 redundant, which has led to some -- we've had some 9 discussions'back and forth on what's the appropriate tech 10 spec limitations, therefore, on one division out, two 11 divisions out, that sort of thing.

12 Standard tech specs were written basically 13 assuming two divisions, 100-percent redundant. If you have 14 one division out, you have no problem, you simply can't take

() 15 another single failure, but if you get a second failure, 16 you're dead in the water, because you've got nothing left.

'17 In our case, when we have one out, we still have 18 three active ones. If we lose one more, we still have two.

19 So, we're not nearly in as bad-a shape.

20 But on the other hand, we also -- we cannot claim, 21 when we only have one, that we have a plant we can properly 22 operate. I believe we've -- we have submitted tech specs 23 that are a little bit different from the standard in this 24 one place. Otherwise, our electric systems follow the 25 ' standard tech spec, and I believe we've come up with some

() -- - ANN RILEY & ASSOCIATES, LTD.

Court Reporters 1250 I Street, N.W., Suite 300 Washington, D.C. 20005 (202) 842-0034 1

-- . . . _b

128 1 reasonable approaches to that, but it's a little bit unique

() 2 for AP600 in this specific area.

3 From the point of view of fire analysis, we've 4 taken an approach as far as fire separation, which is not 5- quite the same-thing as electrical separation, fire area-6 separation ---we've taken the approach that we keep A and C 7 together and B and D together and keep them fire area 8 separated.

9 We've designed our loads such that, for at least 10 non-accident conditions, you can always shut the plant down 11 with either the combination of A and C;only or the 12 combination of B and D only, and that's been our basic 13 approach to fire issues.

14 Another difference, if you're familiar at all-with 15 magnitudes in conventional nuclear plants, our batteries are 16 -a lot bigger. There's a lot of lead in this plant.

17 We have -- in the Class 1E system,'we have four 18 24-hour batteries, one for each division, and two 72-hour

-19 batteries for monitoring purposes, and one spare battery, 20 So, thero's 49ven Class 1E batteries. Each 2 battery is rated at 6 900 ampere hours, technically two sets 22 of 2,400-ampere-hour cells, two 2,400-ampere-hour strings.

23 Why did we-do the two times-2,400? Becauce 24 there's not a qualified 4,800-ampere-hour battery out there.

25 Nobody's ever needed one that big to be qualified. We were O -

ANN RILEY & ASSOCIATES, LTD.

Court Reporters 1250-I Street, N.W., Suite 300 Washington, D.C. 20005 (202) 842-0034

129 1 trying to use previously-qualified equipment.

()

2 The battery manufacturers tell us they probably

-3 could qualify.their 4,800-ampere-hour ones, but we.have 4 chosen to try to stick with proven technology here. Doesn't 5 seem-like it would be a big advantage to use a different set 6 of cells.

7 DR. MILLER: In my understanding, reading the 8 FSAR, you need three to safely shut down the plant.- Is that 9 the way it reads?

10 MR. HAYES: Tua statement is, with three, we can 11 safely shut down the plant under all accident conditions.

12 DR. MILLER: Right.

13 MR. HAYES: With two, we can shut down the plant 14 safety under most conditions.

() 15 DR. MILLER: That's not in the FSAR, then.

16 MR. HAYES: The FSAR is a legal document. I just 17 wanted to clarify what the position is.

18 DR. MILLER: If I understandc in reading it 19 further, if you were down to, say, two, you have a spare 20 that you could put in. Is that right?

21 MR. HAYES: There's a spare battery and charger 22 .that can be -- it's a key interlock switch, twist-lock plug, 23 pre-installed cables, pre-installed but disconnected cables, 24 that can be used to replace a failed battery or charger, 25' that's correct.

ANN RILEY & ASSOCIATES, LTD.

O. -

Court Reporters 1250 I Street, N.W., Suite 300 Washington, D.C. 20005 (202) 842-0034

i 130 1 DR. MILLER: And that can be done fairly quickly.

() L2 MR. MAYES: Right. Right.

3 DR. MILLER: So,-in a way, you have another 4 redundant system.

5- MR. HAYES: Absolutely. Absolutely, for the 6 common failure thing.

7 Now, that does not cover directly the failure of 8 the inverter, which is also relatively important in any of 9 these divisions, but we have a regulating transformer that 10 gives you an alternate supply as long as you didn't need-the 11 batteries, which is relatively conventional plant sort of --

12- DR. MILLER: Now, are the inverters fairly 13 conventional?

14 MR. HAYES: Absolutely.

()

15 MR. CARROLL: So, those are static?

16 MR. HAYES: 'Yes, they're static. They're 17 three phase, which is a little bit different from a lot of 18 them. Most of them are single-phase. But the technology --

19 it's really three single-phase ones hooked together and 20 common marching orders to keep them in sync.

2 11 DR. MILLER: But they would be larger than -- is 22 that right? -- typically found in a plant.

23 MR. HAYES: Not significantly. The reason the 24 batteries are larger is-the times, not the KWs.

25 DR. MILLER: Okay.

O -

ANN RILEY & ASSOCIATES, LTD.

-Court-Reporture 1250 I Street, N.W., Suite 300 Washington, D.C. 20005 (202) 842-0034

_ - _ - _ - _ _ _ 1

131 1 MR. KAYES: The batteries -- I mean even our short

() 2 -- short time for these batteries is 24 hours2.777778e-4 days <br />0.00667 hours <br />3.968254e-5 weeks <br />9.132e-6 months <br />; conventional 3 plant is two hours, four hours for some plants.

4 But you find that -- and one of the things that 5 has driven some of our traditional - "traditionala is not

, 6 the right word -- some of our designers that have been l 7 around for a while are used to be the big motor loads --

8 big, brief motor loads being a real driving factor on how 9 you size your batteries.

10 That's not true for this plant. These batteries 11 don't even see those big DC MOVs. They're only on for a 12 minuts. Who-cares how big they are? They've got to run the 13 I&C cabinets for 24 hours2.777778e-4 days <br />0.00667 hours <br />3.968254e-5 weeks <br />9.132e-6 months <br />. That's where the capacity is 14 required.

() 15 As long as you don't try to put that big MOV load 16 right at the end of the battery discharge, you don't even 17 see it in your sizing _ calculation._ I.mean it's there 18 because IEEE says that's the proper way to do it, but if you 19 took.it out, you wouldn't get a different answer.

20 MR. CARROLL: The 24 hours2.777778e-4 days <br />0.00667 hours <br />3.968254e-5 weeks <br />9.132e-6 months <br /> is based on the fact 21 that they have done their thing and put the plant in a ,sfe 22 condition --

23 MR. RAYES: Yes, sir.

-24 MR, CARROLL: --

for the 72 hours8.333333e-4 days <br />0.02 hours <br />1.190476e-4 weeks <br />2.7396e-5 months <br />.

25 MR. HAYES: Right, They put the plant in a safe O _ _ _ _

ANN RILEY & ASSOCIATES, LTD.

Court Reporters-1250 I Street, N.W., Suite 300-Washington, D.C. 20005 (202) 842-0034

132 l

1 condition forever.

() 2 MR. CARROLL: Forever. Okay.

3 DR._ MILLER: And the 72 hours8.333333e-4 days <br />0.02 hours <br />1.190476e-4 weeks <br />2.7396e-5 months <br /> is required for the 4 post-accident monitoring.

5- MR. HAYES: Yes.

6 How, again -- I guess it will be in Chapter 6 --

7 there is one thing that we are -- that's a little *lt of a 8 twist on that, and that is the ADS, the de-pressurization 9- system -- we really don't want that thing to go off if we 10 can help it, and the way the system d e - again, the dumb 11 operator issue is we've got to get those valves open before 12 24 hours2.777778e-4 days <br />0.00667 hours <br />3.968254e-5 weeks <br />9.132e-6 months <br /> runs out. In fact, right now, we're saying 22 1/2  !

13 hours is the place to get them so you don't'get that spike 14 right at the end of the battery discharge that you may not

( 15 have the capacity for.

16 We are proposing procedures that says, at that-17 point in time, if the operator believes that ADS is still 18 not needed and that_the only reason he would go into ADS is 19 simply he's about to run out of batteries, he will provide 20 procedures to disconnect those batteries at 22 hours2.546296e-4 days <br />0.00611 hours <br />3.637566e-5 weeks <br />8.371e-6 months <br />, let 21 the plant go -- it's not safe yet, 100 percent'-- I mean 22 it's safe, but~it's not 100 percent past all the things you 23 might want to do -- then there ' sill be~ operator action-24 required to go in and reconnect the battery if he ever does 25 need those ADS valves.

ANN RILEY & ASSOCIATES, LTD.

Court Reporters-1250 I Street, N.W., Suite 300 Washington, D.C. 20005 (202) 842-0034-IM' E

. _ _ _ _ . _ . ~..m ~.-_.-___.___._..____.-_m _ _ _. ._.___ _ ._._.

i

}

133  !

t 1 MR. CARROLL: He can't manually -- locally operate  !

2 the MOVs?

3 MR. HAYES: Well, most of them are squib valves; >

4 they're explosive valves.  !

5 MR. CARROLL: Oh, yes, yes, yes.

6- MR. HAYES: This is,a matter of throwing a few l.

7 switches in the I&C room, and he's sitting there with 8 nothing else to do. I mean the plant is sitting there safe.

9 Talk to the governor is about all he'll have time to -- all 10 hetil be needing to do at that point.

11 So, that's a little bit of a quirk,-but from a l

12 pure dumb operator, he can walk away, he left' town, by 24 13 hours the ADS valves have_ opened, and they will

, 14 automatically open if Jun does not take action to stop them.

()

1

15. MR. CARROLL: Okay, b 16 MR. HAYES: Non-Class 1E is fairly standard .:

17 .non-Class 1E system. It.is a two-hour requirement on--the 18- batteries.

l.

19- Again, they're exactly the same batteries. We 20 only have one string in each battery for the non-1E 21 2,400-ampere-hour cells. So, every cell in the plant is the 22 same cell.

23 Now,-these don't have to have the'same paperwork 24 the others do,_but they're going to come off the same

25 production line, we believe.

l L AKN RILEY-& ASSOCIATES, LTD.

t _ Court Reporters L

1250 I Street, N.W., Suite 300 Washington, D.C. 20005

.(202)'842-0034 '

134 Two load groups, separation not required.

1 Again,

() 2 the hookup for the spare battery is such that, if you lose 3 l one of these, you can use your Class 1E spare battery to 4 replace one of these if you don't need it for one of your

! 5 CicJa 1E systems, 6 Now, given that we have this tier two. there's

? k t>en a lot of concern with the staff, how do you know it's 8 there, what's to prevent the operator from running this 9 plant with both diesels completely dismantled, his non-1E 10 system taken apart?

11 Tier three, of course, i covered by tech specs, i 12 j ur,: like you would in any standard plant, and for tier two 13 -- and this is not really my subject for today, but I'd like 14 to point out that we have, in addition to tech specs, in our

) 15 SSAR, short-term availability controls which are a -- we 16 call them tech spec-like. They don't carry the impact of 17 tech specs. They tend to have a little bit longer limited 18 condition of operation type r.tatement. But they do cover 19 some of the more important tier two electrical systems. ,

20 Stand-by diesels are in there, the off-site power 21 system is in there for some plant conditions, the ancillary 22 diesels ari in there, and the non-1E DC and UPS, at least as 23 far as its function of supplying power to the diverse 24 actuation system, is in there.

25 So, again, here we're into some of this -- tier ANN RILEY & ASSOCIATES, LTD.

Court Reporters 1250 I Street, N.W., Suite 300 Washington, D.C. 20005 (202) 842-0034

l 135 1 two has got various shades of gray. This is not everything

() 2 that's in tier two.

3 So, in tier two, we have some things that are  !

4 deemed tier-two important. They're covered by short-term e

j 5 availability controls which are similar to tech specs but-6 not quite the same thing as tech specs, not the same ,

7 pedigree as tecl. specs, and not the same implication as tech 8 specs.

~

9 Are there anymore questions? That's the end of i 10 what I had prepared.

i 11 DR. MILLER:

Everybody has agreed to that.

12 MR. HAYES: To the best of my knowledge, yes, 13 DR. MILLER: I had a question unrelated -- related 14 to what you've said but unrelated. This would go back to-() 15 Charlie Wylie, coo, a little bit.

?

16 In the area of lightening protection, I realize,

. 17 in the SSAR, it says that's the responsibility of the 18 licensee.

19 MR. HAYES: Right.

! 20 DR. MILLER: But you then specify a standard which 21- appears to be, based on my limited research here, out of 22 date.

23 MR. HAYES: NFPA-78?

24' DR. MILLER: NFPA has issued one later. It's 95, 25 I believe.

IdR4'RILEY & ASSOCIATES, LTD.

~ Court Reporters 1250 I Street, N.W., Suite 300 Washington, D.C. 20005 (202) 842-0034

l t

136 1 MR. RAYES: Ninety-two is the one I'm aware of.

() 2 It's interesting you should ask that. The revision of-the 3 SSAR that's coming out now has-fixed the NFPA reference to 4 780-92, I believe is the number.

5 DR. MILLER:- I have NFPA-780. I have the date 6 '95. That could be --

7- MR. RAYES: Well, I could be remembering wrong, 8E -too. But yes, we have changed that to-780 with some year.

9 Now, there .uty be -- you and I seem to have a 10 little confusion as to what's the correct year.

11 DR. MILLER: It doesn't really matter which year 12 it is.

13 MR. RAYES: That was a mistake on our part in the 14 SSAR in that the design documents for the plant _have -- were

() 15 changed to 780 two years ago time-frame, and we neglected to 16 get the SSAR updated.

17 But it is in the process of being updated right 18 now. I just signed off the markup of it like right before I 19 left yesterday.

20 DR. MILLER: The reason I raised that question is 21 I went to a workshop which made me have enough knowledge to 22 be dangerous, but anyhow, which said there are some subtle R23 - changes in there that affect the digital I&C aspects of-a 24 plant.

25 MR. RAYES: Yoa and I have the same-dangerous O ~

ANN RILEY & ASSOCIATES, LTD.

Court Reporters 1250 I Street, N.W., Suite 300 Washington, D.C. 20005 (202) 842-0034-

137 1 position.

() 2 DR. MILLER: So, you've taken care of that.

3 EMR. HAYES: I believe we have, that's right, but 4 not in anything you've seen, you're correct.

5 DR. MILLER: There was a study done by the NRC --

6 or Brookhaven,. actually, on behalf of the NRC -- that 7 indicated that a majority of the non-software-based failures 8 of digital I&C are a result of-lightning.

~9. MR. HAYES: Right.

10 DR. MILLER: So, the NRC has in process a 11 regulatory guide for lightning protection which this group 12 will see sometime in the summer.

13 MR. HAYES: And how we got stuck with the screw-up 14 -- I'll call it a screw-up since we didn't get the change in 15 there when we should have -- the SSAR should have been 16 changed when the documentat'.on-was, but we missed it.

17 In genera *., AP600 is designed to standards in 18 effect as of -- I believe it's September 1990. It's 19 sometime-in 1990.

-20. DR. MILLER: -Right.

21 MR. HAYES: Except where there's some good reason 22 to change that. And our customer in first-of-a-kind 23 engineering, the utilities, requested that change,

-24 apparently for the same reasons, is my understanding, and 25 so, we have made that change.

O ANN RILEY & ASSOCIATES, LTD.

Court Reporters 1250 I Street,-N.W., Suite 300 Washington, D.C. 20005 (202) 842-0034

.-__ ---. -- --- ----~- ------~

138

1 since it was driven by the utility side, we were 2 apparently a little bit sloppy in keeping the SSAR up-to I 3 -date, but it's in the process of being fixed even as we i

4 speak.

5 D AILLER: Another question, maybe more for the i 6 staff -- I was just curious curious whether the lightning

7 protection and a grounding system would meet the

! 8 requirements of a reg , quide which is now out on the street j 9 for comment ou EMI RFI, and also, would it meet the 10 requirements of the reg guide in process for lightning i

11 protection? That's a question you can't answer. It's 12 probably more of a question for the staff.

13 DR. SEALE: I believe you've got an answer.

a' 14 MR. TREHAN: Narinder Trehan from electrical 15 engineering branch.-

4 16 I am the person who has been talking to Research 17 regarding the reg guide.- It would meet that requirement.

18 DR. MILLER: Okay. So, the reg guide uses that 19 later guide by NFPA as a basis, then?

20 MR. TREHAN: Right.

21~ DR. MILLER: And-does it use also the-IEC 22 standards which maybe are similar or not?

23 MR. TREHAN: I believe it would use -- we would 24 use the latest guide available.

25 DR. MILLER: Okay.

ANN RILEY & ASSOCIATES, LTD.

O- Court Reporters 1250 I Street, N.W., Suite 300 Washington, D.C. 20005 (202) 842-0034

139 1 MR. TREHAN: I have been reviewing this I&C

() 2 3

computer system and electrical.

-- Oak Ridge, I believe, and There are two labs involved BNL. Each is handling a j 4 different issue, and we are looking into it.

i It will come 5 to us, electrical engineering branch, and we'll look into 6 it.

7 MR. CARROLL: Let's see. Yesterday, I was trying 8 to remember an insue with DC motor-operated valves, and I 9 guess I have jogged my memory and John helped me jog it, and 10 I guess the issue was, some years ago, there was a problem

+

11 in a number of plants who had not sized their cabling from 12 the motor control center to the valves large enough and 13 having all kinds of grief with the valves because they 14 weren't able to put the amperage through the cables. Are

' ,/ ^

(_,h / 15 you aware of that issue, and have you dealt with it 16 appropriately in the design?

17 MR. HAYES: Yes.

18 MR. CARROLL: Yes, yes?

19 MR. HAYES: Yes, we have. Yes, we're aware of it, 20 and yes, we have. From the point of view of -- the valve 21 vendors have eeen heavily involved, and if you really get 22 down to the fine print, you've got have them-involved and l 23 they have got to specify what's required for their valve.

, 24 MR. CARROLL: At the valve.

125 MR. HAYES: At the valve, that's correct.

ANN RILEY & ASSOCIATES, LTD.

\- Court Reporters 1250 I Street, N.W., Suite 300 Washington, D.C. 20005 (202) 842-0034

E 140  !

! 1 MR. CARROLL: But then some designer has to get it '

() 2 3

there. Okay.

MR. HAYES: Yes, we have addressed that.

l 4 MR. CARROLL: I'm happy.  !

5 CRAIRMAN BARTON: Any other questions from the 6 committee or the staff? Any comments from the staff on this

] 7 chapter?

! 8 DR. MILLER: I had another question for the staff 9 on the area of -- back to these inverters. .

10 Now, again, my understanding, inverters can be 11 generators of EMI RFI? das the staff or somebody looked at I

! 12 that? Now you've clarified one thing. The inverters are 6

13 not -- I thcught they u.. ;ht be larger. Have we looked at 14 those as EMI RFI generators? Do they meet the reg guide

.() 15 requirements, the requirements of our SER in that area?

16 MR. TREHAN: This is Narinder Trehan from

) 17 electrical engineering branch.

18 We did look into'the EMI and RFI and the harmonics 19 effect on the inverters, and we were given to understand that the harmonics would not affect the inverter operation, 20 21 EMI and RFI will not affect. .

22 MR. CARROLL: No , that's the wrong question.

23 DR. MILLER: I'm not certain I got --

24 MR. CARROLL: Are the inverters EMI RFI generators 25 --

O- . ANN RILEY & ASSOCIATES, LTD.

%/ Court Reporters 1250 I Street, N.W., Suite 300 Washington, D.C. 20005 (202) 842-0034

.-. _ - . - - - - - . . . . . . . _ - .-- ---.-~.- - -

141 E i  ;

1 DR. MILLER: Generators.  !

() 2 MR. CARROLL: -- that could cause some problems 1

, 3 with --

4 MR. TREHAN: That's what I really meant.

5 Harmonics generated by the inverters will not affect it.

i  :

6 MR HAYES:- They areLEMI RFI generators. That's a l-  ? fact. ,

i 8 DR. MILLER: Yes.- t 9 MR. HAYES: However, they're heavily filtered, 10 heavily limited, and on the other side, our-I&C cabinets are .

11 designed to be heavily immune to such things.

12 DR. MILLER:

I-understand that,-yes.

13 MR. HAYES

So, we have looked at it. We believe 14 we're well within what a required. >

15 DR. MILLER: The SER that endorsed the guideline -

16- by EPRI -- of course, the EPRI guideline specifies limits on

17 your genera *. ion of EMI RFI, My question is, do those 18' inverters meet those guidelines on generation of EMI RFI?

19- MR. HAYES: I believe they do.

20- MR. CARROLL: While we're on that subject -- I t

21' guess _this is an I&C question. Forget it'.

22 MR McINTYRE: Next we'll move to Chapter 13, .

23 ' Conduct of' Operations.

24 MR. CARROLL. Minus security.

25 MR. McINTYRE: Minus cecurity, excluding security.

' () ANN RILEY & ASSOCIATES, LTD.

Court Reporters 1250 I Street, N.W., Suite 300-Washington, D.C. 20005 (202) 842-0034

,-  :- _ - -. . . - , _ - - a - .- -..-.-.. -. - .. .- -.- -. .- ---

1 l-l 142 1 MR. CARROLL: We sure get into some strange 2 permutations.

3 Mk. McINTYRE: There was a question raised 4 yesterday on the availability of the security plan for the 5 members of the committee to review.

6 I understand the staff will provide-and place the 7 security design documents in here, locked up in our document 8 room, and-they will be available_to_ACRS members who are 9 here.

-10 MR. CARROLL: Who's the keeper of the keys?

11 Ethel?

12 MR. KERCH: Good morning, gentlemen. My name is 13 Steve Kerch, and I'm an engineer within our Man-Machine 14 Design Group, and I have been the lead for Chapter 18, Human

) -15 Factors Engineering, and also provided significant input to 16 Chapter 13, Conduct of Operations.

17 MR. CARROLL: There.isn't much in Chapteri13.

18 MR KERCH: Yes, exactly. You just stole a little 19 bit of my thunder.

20 There are six subsections in Chapter 13. About 21 three of the six subsections are also covered byfthe human 22 factors engineering program, Chapter 18, and of course, the 23_ -sixth' subsection is security, which'we're not covering here 24 this morning.

25 'The first five sections we're saying are the

() ANN RILEY & ASSOCIATES, LTD.

Court Reporters 1250=I Street,- N.W.,-Suite 300 Washington, D.C. 20005 (202) 842-0034

--_ -- -- - {

L 143' 1 responsibility of the combined license _ applicant.

() 2 The first section is the organizational structure-3 of the applicare, and again, we're saying that's a COL 4 responsibility.

5 Howeser, there-is a connection to the human 6 factors engineering program and the human system interface 7 design in that, whatever structure the COL adopts, they've 8- got to consider the human-system interface that the plant 9- staff will be-using to operate and maintain the plant.

10 The second subsection is on training. 'Again, 11 we're saying;that training program development is the 12 responsibility of the COL.

13 We do provide a WCAP document that's referenced by 14 Chapter 13 and Chapter 18, 15 This document provides designers input to two 16 things- to the design and implementation of the training 17 program for the operations personnel who will participate as 18 -test subjects during the verification and validation of the 191 human factors engineering program and the integrated 20 -human-system interface design.

21 This WCAP also provides input to the process used 22 to develop the specification of the role of the AP600 23 operator and the passing of this role and training insights 24 from the designer to the COL applicant.

25 DR. SEALE: So, this is the document, to use the

() ,

ANN RILEY & ASSOCIATES, LTD.

Court Reporters

-A 1250 I Street, N.W., Suite 300 Washington, D.C. 20005 (202) 842-0034

l 144 1 jargon of the training community, that has the designers

() 2 input into'the job-task analysis that is the basis for the 3 development of the training program for licensed. operators 4 and, I guess, unlicensed operators, as well, for equipment.

5 maintenance and so

• orth, which is the respousibility of the 6- combined license holder, and I assume the NRC then would 1 7 require that this be a program that meets the INPO training 8 requirements for our plant operators.

9. MR. KERCH: .Yes, sir.

The WCAP does state that a 10 training program would be developed in accordance with'the 11' systematic approach to training that INPO has adopted and 12 used in the industry and would use that approach to develop 13 a training program for the V&V personnel, and then any 14 insights that we learn from that, we_also provide --

() 15 DR. SEALE: Graft it ato the full program.

16 MR. KERCH: -- into the-full recommended program 17 that the COL would have to develop.

18 DR. MILLER: If-I were to go to the WCAP, which I  !

19 did not do, would I see a significant difference between

'20' what's in there and what I might find in a current operating H21 plant? I raise.that partly because I've gone to the WCAPs 22 in I&C. The truth is I don't see a whole lot of difference.

23 MR. KERCH: No , there would not be a significant 24 difference.

25 We tried to fat:or a few more things into the O- ANN RILEY & ASSOCIATES, LTD.

Court Reporters 1250 I Street, N.W., Suite 300 Washington, D.C. 20005 (202) 842-0034

145 1 program or define a few more things or recommend a few more r

( 2 things into the program, such as -- one of the things we're 3 going to have to do for these operators is train them in the l 4 use of a different interface.

5 DR. MILLER: It's a different interface, and it's 6 a passive plant. So, there's obviously a substantial 7 difference between them.

8 MR. KERCH: That's correct. And we would have to 9 stress and emphasize the differences between a passive plant 10 and a conventional plant.

11 DR. MILLER: But that's not in the WCAP?

12 MR. KERCH: It is. It is explicitly mentioned in 13 the WCAP as designers input to the program, but other than 14 that, there are no significant differences. I mean it takes

/~

(_)T 15 a systematic approach to training, it talks about the use of 16 the plant simulator, and so on.

17 DR. MILLER: So, 80 percent was pulled out of the 18 files and 20 percent was new.

19 MR. KERCH: That's correct.

20 Emergency planning is also the responsibility of 21 the COL applicant, including the communication interfaces.

22 However, we do, in the SSAR, provide several items.

23 Certainly the location of the Technical Support 24 Center, the TSC, the Operational Support Center, are al]

25 provided by subsection 1.2.5 of the SSAR.

l ANN RILEY & ASSOCIATES, LTD.

\/ Court Reporters 1250 I Street, N.W., Suite 300 Washington, D.C. 20005 (202) 842-0034 l

146 1 -The HVAC systems are defined and~provided for the

. () 2 main control room, the TSC, and the annex building, and

-3 that'sifound in-9.4 of the SSAR.--

4 There are high-level requirements for the TSC and-5 .the Operational Support Center provided by section 18.8 of 6 the SSAR.

7 MR. CARROLL: What's a high-level requirement?

8 MR. KERCH: Things like the number-of square feet-9 that are required that come out of the NUREG, things like 10 communications will be provided between the TSC and the 11 control room and the OSC.

12 DR -SEALE: This may be where George's seven-foot 13 phone cord shows up.

14 MR. CARROLL: Right.

15 MR.-KERCH: And-then' identification of plant l 16 _- variables provided for. interface to the. emergency planning-areas-are provided by section 7.5 of the'SSAR.

18 Operational review'- 1we simply state-that's:the--

19 combined license applicant's-responsibility.

20 And plant procedures, 13.5 of the SSAR -- again, 21 it's the responsibility of the COL.

22 -

However, we as designers have provided input to 23 the COL applicant for the-development-of plant procedures, l 24 which includes -- and it's in this WCAP document, 14690 --

25 includes _information on the AP600 emergency response ANN RILEY & ASSOCIATES, LTD.

Court Reporters 1250 I Street, N.W., Suite 300 Washington, D.C. 20005 (202) 842-0034

147 1 guidelines and information on the computerized procedures Og 2 system.

3 So, this WCAP also provides things like, you know, 4 what the COL needs to consider when they develop their plant 5 procedures, what guidelines, what accepted industry 6 standards to be used while developing their plant 7 procedures.

8 DR. UHRIG: Do you provide a draft set of 9 procedures?

10 MR. KERCH: We certainly prc: sde the emergency 11 response guidelines, which are then developed into emergency 12 operating procedures, and we -- not as part of the design 13 certification, but certainly we will provide --

14 DR. UHRIG: As part of your service to your

/D

(~) 15 customer.

16 MR. KERCH: As part of service to the customer, we 17 will provide start-up and shut-down procedures.

18 DR. UHRIG: They have to tailor them, then, to 19 their particular site, if there's anything unique about 20 their site.

21 MR. KERCH: Well, yes, that's correct. If there's 22 anything unique about the site characteristics, then they 23 would have to be tailored to the site. That is correct.

24 DR. UHRIG: Okay, 25 MR. KERCH: And that is Chapter 13.

['" ') ANN RILEY & ASSOCIATFS, LTD.

Court Reporters 1250 I Street, N.W., Suite 300 Washington, D.C. 20005 (202) 842-0034

148 1 DR. SEALE: Is there any part of-the ITAACs or g () 2 those kin'ds of implementation documentation that goes beyond 3 the site certification that covers the scope of the 4 procedures?

5 MR. KERCH: In Chapter 18, procedures are also 6 covered as part of the HFE program, but we do not have --

7 because we say that procedure development is the 8 responsibilit'j of the COL, there is not a specific ITAAC on 9 procedure development within the certified design material 10 document.

11 MR. CARROLL: The staff will have that requirement 12 in the certification.

11 3 DR. SEALE: Yes.

14 MR. McINTYRE: This is Brian McIntyre.

()

As Steve 15 pointed out, there is a COL information item. So, that's 16 the hook that the staff knows that the person building-the 17 plant has to come up with this and that they will get to 18 review it at that time. '

19 MR. CARROLL: I guess this is a better question 20 for the staff than for Westinghouse. What we've heard on 21 Chapter 13, these five. sections, looks to me to be very 22 similar to what was approved on ABWR and 80-Plus. Is that 23 perception correct?

24 MR. KENYON: Hang on a second. Mr. Wilson can 25 probably answer that question.

\ '

ANN RILEY & ASSOCIATES, LTD.

i Court Reporters 1250 I Street, N.W., Suite 300 l Washington, D.C. 20005 j (202) 842-0034 I

J

149 1 MR.! WILSON:- This is Jerry Wilson, NRR. Could you

() 2 re-ask the question, please?

3 MR. CARROLL: .In looking at these first five sections of Chapter 13, Jerry, they look very familiar to 5 me. They-look like what the staff had approved for ABWRiand 6 80-Plus. Is my perception correct?

7 MR.. WILSON: Well, in terms of the responsibility 8- of plant procedures for the COL, yes, that's correct.

9' MR. CARROLL: No, I'm talking-about all of the 10 sections we looked at, 13.1 through 13.5.

11 MR KENYON:- Mr. Carroll, this is Tom Kenyon from 12 the staff.

13- That's correct, these are the same, but I thought 14 you had a specific question concerning-ITAAC.

() 15 MR. CARROLL: No, that was somebody else.. That 16; was Bob Seale.

17 MR. KENYON:- Okay. You're correct. These are 18 essentially the same as what as in the two evolutionary 19- plants, and as Mr.-McIntyre mentioned, you know, the hook, 20 as he calls it, is the fact that it's referenced as-~a COL 21 action item.

22 MR. CARROLL: Unay.

23 CHAIRMAN BARTON: Any other questions or comments

-241 on Chapter 13?

25 (:No response.)

ANN RILEY & ASSOCIATES, LTD.

Court Reporters 1250-I Street, N.W., Suite 300 Washington, D.C. 20005 (202) 842-0034

150

(_ 1 CHAIRMAN BARTON: Thank you.

() 2- If not, we'll recess at this time, take a 3 15-minute break. We'll come back at 10:05. It looks like 4 Chapter 18 is going to take a while to go through, so we'll 5 take our recess right now.

6 Thank you.

7 [ Recess.]

8 CHAIRMAN BARTON: We'll reconvene the meeting and 9 start the presentation on Chapter 18, Human Factors.

10 MR. CARROLL:- Did I understand correctly that the 11 staff has no outstanding items on Chapters 8 and 13?-

12 CHAIRMAN BARTON: That's correct.

13 MR. CARROLL: Okay.

14 DR. SEALE: What about this chapter?

O 15

(,,/ -MR. HUFFMAN: This is Bill Huffman. No, we do not 16 have any outstanding items.

17 MR. CARROLL: And while we're on it, how about-l 18 Chapter 77 19 MR. HUFFMAN: Again, this is Bill Huffman. No, 20- -nothing in 7.

21 MR. CARROLL: Thank you.

22 MR. KERCH: Okay. Chapter 18 is Human Factors 23 Engineering. Again, my name is Steve Kerch, a member of the 24 Man-Machine Design Group and the lead Westinghouse author on 25 Chapter 18.

O ANN RILEY & ASSOCIATES, LTD.

Court Reporters 1250 I Street, N.W., Suite 300 Washington, D.C. 20005

-(202) 842-0034

i 151 1 Chapter 18 is organized and parallels the

() 2 organization of NUREG-0711 and complies with the criteria of 3 NUREG-0711, which is the human factors engineering program 4 review model. In addition, it complies with the minimum 5 inventory criteria which is found in subsection 18.12.3 of 6 the AP600 draft safety evaluation report.  ;

l 7 What I'm going to do is provide an overview of the

8. AP600 human factors engineering program and process. After 9 that, I'll go through each of the 10 elements and minimum 10 inventory in a little bic more detail.

11 MR. CARROLL: Now, has NUREG-0711 been updated or 12 revised since it was used for ABW and System 80-Plus?

13 MR. KERCH: I believe I'm correct in saying that 14 what was used for ABWR and System 80-Plus was the draft

) 15- version of the NUREG. So, the formal Revision 0 has come 16 out since then. 'It came out in mid '94, I believe.

17 MR. CARROLL: All right.

18 DR. SEALE: Were there any requirements in--the 19 utilities requirement document that covered human factors 20 considerations?

21 MR. CARROLL: Oh,-yes.

22 MR. KERCH: Chapter 10.

23 MR. CARROLL: Extensive.

24 DR. SEALE: Are you going to tell us anything 25 about those?

ANN RILEY & ASSOCIATES, LTD.

Court Reporters 1250 I Street, N.W., Suite 300 Washington, D.C. 20005 (202) 842-0034

152 1 MR. KERCH: I'm not going to go into detail -- I

() 2 didn't plan to_go into detail on those other than to say that they were used as input and assumptions and constraints 4 in some sense to our human factors engineering program.

5 DR,'SEALE: And_you complied.with the 6 Jrequirements.

7 MR. KERCH: Yes.

8. DR. SEALE: Okay, 9 MR. KERCH: And we complied with the requirements 10 -of the URD.

11 What I show here is a block diagram, simplified 12 block diagram, that provides an overview of the AP600 human 13 factors engineering program consisting of 10 elements and 14 minimum inventory, and what you see here are nine of the 10 15 elements _on this block. diagram, 16 The first element is human factors engineering 17 program management, which I'll talk more about in a few 18 minutes but basically defines the procedures and processes 19' used in the AP600 design, including the human factors 20 program, but on this figure you see element two, which is 21 the operating experience review, in the upper left-hand 22 corner,-operating experience review, which is an input to 23 the program, and it is a specific input not only to the 24 human-system interface design itself but, really, it's a 25 review of operating experience across the plant.

( ANN RILEY & ASSOCIATES, LTD, Court Reporters 1250 I Street, N.W., Suite 300 Washington, D.C. 20005 (202) 842-0034

L_

153 g 1 It covers topics and systems other than the~ main 2 control room and the human-system interface -- an example, 3 cooling systems, electrical syatem issues, and so on.

4 -Element three is --

5 MR. CARROLL: Now; that's_a staffing assumption.

6 :That's staffing for the plant or staffing for the human-7 factors team?

8 MR. KERCH: Staffing is one of the elements of the 9 program, and what we used here and what we're trying to show 10 here is that we-assumed certain staffing levels for the main 11 control room as input to our human-system interface design,-

12 and you will see later in my presentation that staffing is 13 _the responsibility of the combined license applicant.

14 -However, we have made certain assumptions and so_on.

(/ 15.- -MR. CARROLL: _ Okay, I got you.

16' MR. KERCH: -Other elements -- element three is our

17. functional requirements analysis and allocation element --

18 again, provides input to the human-system interface design, 19 and the human-system interface design is shown inside the large-dashed box here. That is the one-element that is:

21 dedicated to the actual design of the operation and-control 22 centers, such as the main control room, technical support 23 center,- and the associated human-system interface.

24 The. function requirements analysis and allocation 25 basically defines the level of__ automation for AP600 ANN RILEY & ASSOCIATES, LTD.

Court Reporters 1250 I Street, N.W., Suite 300 Washington, D.C. 20005 (202) 842-0034

154 1 documents and provides the rationale for it and, therefore,

() 2 is an input _to'our interface design.

3 MR. CARROLL: I. guess,-going back to-what I raised 4 before, don't I remember that one of the elements described 5- who comprised the human factors team? v 6 MR. KERCH: Yes.

7 MR. CARROLL: Where do I-see that?

-8 MR.-KERCH: That is element one, the program 191 -management element. I will address that in a few minutes.

10 MR. CARROLL: Okay. I don't see it here, but --

11 RMR . KERCH: Yes, sir. Element one is not shown 12 here, again because it's sort of the administrative element 13- that defines-the processes-and procedures that guides the entire plant design, including the human factors engineering 15 program. I 16 MR. CARROLL: Okay.

17 MR. KERCH: Element four is the task analysis, 18 again providing input to'the program and, specifically,-the 19 HSI'in the system interface design.

20 Element five is the staffing, which I've spoke of.

21 Element six is the integration of human.

22 -reliability analysis with the human factors engineering 23 program.

24- And again, each of these I'll talk more about in a 25 few minutes.

ANN RILEY & ASSOCIATES, LTD.

Os Court Reporters 1250 I Street, N.W., Suite-300 Washington, D.C. 20005 (202) 842-0034

155 1 Element seven is the human-system interface design

( 2 itself. So, what you see inside the dashed box is a block 3 diagram of that process,.where we go through a functional 4' design, functionally design our_ operation and control 5 centers -- that is, the main control room, the TSC 6 inter-mode shut-down-facility, and the local control i

7 Stations.

8 We also provide, from there, a functional design 9 of our various human-system interface resources -- alarm 10 system, plant information displays, and so on.

11 14R . CARROLL: Okay. So, this program does include 12 local control stations.

13- MR. KERCH: Local control stations are-addressed 14 by the program, that is correct.

15 MR. CARROLL: Okay. Because in reading Chapter 16 18, I found a place or two where they weren't included, but 17- there was sort of -- well, for example, a control room or

.18 TSC -- okay.

19 We had quite an. argument about that on ABWR, 20 System 80-?lus, with the staff, as to whether that should be 21 or should not be included in this program. I guess we 22 convinced them that it should be.

23- MR. KERCH: Once we have the centers and the 24 associated interfaces functionally defined, we proceed to 25 developing a set of design guideline documents, and these ANN RILEY & ASSOCIATES, LTD.

-\-

Court Reporters 1250 I Street, N.W., Suite 300

-Washington, D.C. 20005 (202) 842-0034

156 il - -guideline documents are developed to be

-( )

2 AP600-project-specific,-where we use' accepted industry

-3 standards and guidelines-but we: develop a set of

4 project-specific documents that each=of the-human-system 5- interface. designers can use while they're designing their 6 displays,_ while-they're designing their. alarm system,_soton,-

-7 . so that we have ar consistent design being applied,J you know, 8 across:all' designers, 1

9' You have your guideline = documents, you proceed 10 with your detailed design specifications,-these are your-11 detailed display designs, your detailed alarm system design, 12 your detailed controls, soft and dedicated controls, and so 13 on.

14 -Once you_have that, you can see we have a feedback

( 15 -loop in our process where we have a rapid prototype and 16 -concept testing phase that we use to -- as a feedback loop 17 in our design process.

LIB We use this to refine and improve our design based

-19 on the results of conceptz testing, where-we're using actual 20 ~ utility operators,-plant operators, as test subjects in a 21 concept test environment, 22- And finally, when we're done with-the concept test 23 phase,-we can catchLall our design documentation upito date-24L and' proceed with implementing the design, both hardware and 25' software, and you can see that we have elements eight and O ANN RILEY & ASSOCIATES, LTD.

\s Court Reporters 1250 I Street, N.W., Suite 300 Washington, D.C. 20005 (202) 842-0034

.__-.m_m__-----

157 1 nine, procedure development and training program f3 I

) 2 development. They're also inputs to the program and covered 3 by the program.

4 And then we have our element 10, which is the 5 verification and validation phase. Here, verification and 6 validation is broken up into five parts, one of which is an 7 integrated system validation that's performed on a plant 8 simulator, where you do a complete validation of the main 9 control room and the integrated human-system interface 10 environment.

11 MR. CARROLL: Now, this diagram and some other 12 things are all requirements of -- at least 0711 guidelines 13 for doing this. If that document didn't exist, would you 14 have approached the issue of human factors essentially the

} 15 way you did?

16 MR. KERCH: I think we would have, but certainly

17. we've got a program that is certainly more explicitly 18 defined now as a result of 0711.

19 We submitted our initial Chapter 19 of the SSAR in 20 June of '92, and at that time, I believe a draft of 0711 did 21 exist and was out for review and so on, but I do not believe 22 it was used extensively by, at that time, our design team 23 that put together Chapter 18, and there were similarities.

24 I mean we had a review of operating experience, 25 although we needed to document it better, which is what h'# ANN RILEY & ASSOCIATES, LTD.

Court Reporters 1250 I Street, N.W., Suite 300 Washington, D.C. 20005 g202) 842-0034

.158 11 we've,done as-a--result of 0711. We had task analysis

2e defined.

3 MR _ CARROLL: So, I guess the bottom line I'm 4 looking for is you hava-found 0711 to be a reasonable and 5 valid approach to doing all of this stuff.

6 MR. KERCH: Yes, . sir.

7 MR. CARROLL: Thank you.

8 MR. KERCH: So, what I want to do'is proceed to-9 each of the elements here and provide a summary'of each, As 10- I said, there are 10 elements,-and we just saw sort of a 11 block diagram of those 10 elements _and how-they're related 12 to one another.

13 Each of'the 10 elements were reviewed by the staff 14 at'a certain level of completion. Basically, this program-L15 defines a process, and that process,. certain parts of the 16 process, certain elements, are-either completed'or partially 17 completed.or were defined at a certain level.

18 So,- as I go through each-of the 10 elements, I'll 19 try to explicitly state at what level it's reviewed and what 20 . we do have existing and what we're asking be approved.

( 21 Element one is the HFE program management element.

22 Again, this was reviewed at a complete level, meaning that 2 3 _. this element is totally defined and described, and what it 24- presents is our human factors engineering program goals, ,

25 scope, assumptions, and constraints.

O ANN RILEY & ASSOCIATES, LTD.

U Court Reporters 1250 I Street, N.W., Suite 300 Washington, D.C. 20005 (202) 842-0034

159 1 We define our human-system interface design team j

) 2 and its organization, its qualifications, its composition, 3 its experience and background that those members must have.

4 So, this is where the design team comes in.

5 The plant design processes and procedures are part 6 of this element, where they're described and defined and o

7 presented.

8 We have a human factors engineering design issues 9 tracking system that is being used to track current 10 outstanding human factors-related design issues across the 11 plant, across the project.

12 And finally, we do have -- one of the significant 13 documents that presents the processes and procedures being 14 used is a WCAP document that provides the AP600 quality

) 15 assurance procedures supporting the staff's review of 16 element one and element seven, element seven being the 17 -actual human-system interface design element.

18 Element two is the operating experience review.

19 Here, using NUREG-0711 and one of the appendices as 20 guidance, we conducted a review of operating experience 21 where human factors-related issues were identified, and then 22 we addressed each of those issues and stated how the AP600 23 design tackles those issues and resolves them.

24 MR. CARROLL: What did you consider plant designs 25 similar to the AP600?

I\

\-

ANN RILEY & ASSOCIATES, LTD.

Court Reporters 1250 I Street, N.W., Suite 300 Washington, D.C. 20005 (202) 842-0034

160 1 MR. KERCH: We defined a plant -- basically, the

() .2 3-Westinghouse PWR experience we used as our reference. So, that included -- you know, in emergency response space, they 4 sort of break our. Westinghouse PWRs up into two groups, the 5 low-pressure plants and the others, high-pressure.

6 We took the collective experience of all of them.

7- We=did not want to set aside any experience that might be 8 relevant.

9 DR. MILLER: Did that include international 10 experience or just domestic? ,

11 MR. KERCH: That included Westinghouse 12 international experience.

13 DR. MILLER: So, it included the French plants?

14 MR. KERCH: Well, no. It included the Ringles, 15 the Be naus.

16 DR. MILLER: Okay.

17 MR. KERCH: No, it did not include the French 18 experience.

19 DR. UHRIG: What do you taean low-pressure, 20 high-pressure?

21 MR. KERCH: Westinghouse PWRs have a 22 safety-injection system.

23 DR._UHRIG:

Oh,_okay, it's a safety-injection 24 system.

25 MR.'KERCH: It's a safety-injection system.

() ANN RILEY & ASSOCIATES, LTD.

Court Reporters 1250 I Street, N.W., Suite 300 Washington, D.C. 20005 (202) 842-0034

1 161 1 DR. UHRIG: Okay.

()

2 MR. KERCH: So, the elements are written a little 3 bit differently depending on the plant, and we use both, 4 .because for the most part, they're very similac.

5 DR. UHRIG: I hadn't heard the term before.

6 MR. KERCH: For those issues that were not 7 addressed by our design yet or were outstanding issues, we 8 placed into our: tracking system, our human factors-

9. engineering-issues tracking system, so we don't lose track 10 of those and k'.w that they'll need to be addressed.

11 The document, then, that documents our-review 12 operating experience is this WCAP, which I have brought 13 along as my security blanket, along with the other 14 significant WCAPs.

r N.

(_) 15 MR. CARROLL: Pass it down, I'd like to take a

~

16 look at it.

17 MR. KERCH: Okay.

18 Element three, the' function requirements analysis 19 and allocation -- this is an element where we defined and 20 identified the AP600 critical safety functions. We then 4

21 identified the systems, components, and processes used to 22 achieve azid satisfy each of those critical safety functions.

23 We compared those with our -- again, our 24 reference, which is the collective AP600 -- or the 25 collective Westinghouse PWR experience; we j;1stified any ANN RILEY & ASSOCIATES, LTD.

Court Reporters 1250 I-Street, N.W., Suite 300 Washington, D.C. 20005 (202) 842-0034

t. . .. .

162 JL- differences-between the cwo.

() 2 We-then took those systems and processes that are l

3 used to achieve a given critical safety functim. and 4' identified whether the process was automatic or manual or a-5 combination of both, and again, we did tne same thing.with 6 our reference collection, and then we compared the-two-and

7. we provided a rationale on why any AP600 processes were 8- different from conventional plants, and we also described 9__ the methodology in this document on -- the methodology that 10 we used to arrive at this AP600 function allocation decision 11 -- that is, the level of automation, whether it's an 12 automatic process or whether it's a manual process, 13 And then we also, in the document, provide a 14 description of human factors activities that are conducted 13

(_,/ 15 as part of the HSI, the' interface design process, to verify 16 the' adequacy of our function allocation decision, 17 These are things like, you know, workload analysis 18- that is done in_part by our task analysis, workload analysis ,

19 that is done during concept testing, and of course, our 20 final V&V, and again, the document that documents this 21 function requirement analysis and allocation is WCAP-14644, 22 and this was a complete element level.of review conducted by

-23 the staff, just as the review of operating experience _was, 24 also.

25- So, the first three elements -- program O# ANN RILEY & ASSOCIATES, LTD.

Court Reporters 1250 I Street, N.W., Suite 300 Washington, D.C. 20005 (202) 842-0034

163 L:

V 1 management,-review of operating experience, and function 2 .-requirements analysis and allocation -- were reviewed atla 3 complete element level.

4 Element four, task analysis, was reviewed -- we 5 presented an implementation plan. So, we'are not complete 6 with our task analysis yet, but we provided a-plan to the 7 staff, and it's boen-reviewed, and this task analysis really 8 -- our task analysis is-subdivided into two activities, what 9 we call function-based task analysis and what we call'the 10- operational sequsace: analysis.

11 The operational sequence analysis is very similar 12 to the -- today's job and task analysis that it was 13 performed for today's-plants.

14 The function-based task analysis takes the' plant O

I q,/ - 15 and decomposes from two high-level functions and decomposes 16 those functions down into lower levels, two high-level 17 functions being protect the public from radiation and 18 generate electricity.

-19 We decompose from those two functions down to what 20 we call level four' functions. These level four functions 21! are things like control RCS pressure, control RCS 22 temperature, control steam generator inventory.

23 We then take each of those level four functions 24 - and do a task analysis on them.

25 .

First we identify the goals of the function, we

() ANN RILEY & ASSOCIATES, LTD.

Court Reporters

-1250 I Street, N.W., Suite 300 Washington,-D.C. 20005 (202) 842-0034

=- -

164 l

1 then identify the systems and processes used to accomplish

() 2 the goals of that function, and then we perform the

! 3 cognitive task analysis on it, wnere we apply Rasmussen's 4 decision-making model to each process, and there's a 5 standard set of questions that we develop out of Rasmussen's 6 human decision-making model, and basically, what we get out 7 of that cognitive task analysis is an identification of the 8 controls and indications needed by the operator to make 9 decisions about that process to accomplish the given goal 10 for that function.

11 We do that independent of looking at the actual 12 design. So, it's just a cognitive task analysis done using 13 -- just using the HSI design team members, 14 Once that identification of controls and

)- 15 indications is made using the cognitive approach, that list 16 is compared now with the actual design, and if there are any 17 discrepancies that exist -- say, you know, we say that you 18 need a pressure indication here in order for the operator to 19 determine and make this decision but there's not one in your 20 design -- we go back to the designers and we resolve -- that 21 discrepancy is resolved with the designers.

22 So, that's sort of a summary of our function-based 23 task analysis.

24 The operational sequence analysis is the more 25 traditional approach where we take a procedure or a

( ANN RILEY & ASSOCIATES, LTD.

\-

Court Reporters 1250 I Street, N.W., Suite 300 Washington, D.C. 20005 (202) 842-0034 1

165 g 1- procedural guideline and, _for-each step within that

() 2- procedure, again we identify those controls and-indications.

3 that the operator needs to execute or make the decision 4- about that step.

5 So, in a similar way but in a different way, we do L

6- the same sort of verification: Do those controls and-7 -indications really exist in the design once we've identified 8 them through our task analysis?

9 The implementation plan -- this implementation-10 plan I'm talking about is described in 18.5.2, and its 11 execution is the responsibility of the combined license 12 applicant, and what task analysis does, therefore, is to 13- provide input to the human syster interface design, to '

eg 14 staffing, the staffing analysis, toLprocedure development, h 15 and to--training.

16 MR. CARROLL: Now, the commitment to do this is '

17 going to be' contained in a-DAC?

18 MR. KERCH: It's an ITAAC. '

4 19 MR.' CARROLL: It's an ITAAC not a DAC, Okay.

20 MR. KERCH: The commitment to execute the plan and 21- provide the associated documentation to the NRC is an-ITAAC.

22' MR. CARROLL: And has that been prepared and -

23 submitted?

24 MR. KERCH: Yes, sir.

25 MR. CARROLL: And approved?

ANN R; LEY & ASSOCIATES, LTD.

Court Reporters 1250 I Street, N.W., Suite 300 Washington, D.C. 20005 (202) 842-0034

L 166 1~ MR. KERCH: Yes. As'farLas I know, there are no 2 open items'in that area. We have-a formal-ITAAC document revision that needs to go out'yet-to actually. formalize the

~

3 L4 very latest changes to our human f actors engineering program 5 ITAAC, but I believe wefre there.

6 MR. CARROLL: Thank you.

7 MR. KERCH: Thisofigure is taken out of-18 5 of 8 .the SSAR, and it basically just tries to show the 9 relationship between.our two task analyses, operational 10- sequence analysis and function-based task analysis, to our 11- human-system interface design, where both the OSA and.the 12 FBTA are used as= inputs -- was also used as inputs to our 13 human-system interface design -- our display design, our 14 alarm system design, etcotera. '

) 15 Again, I mentioned-completeness checks, display requirements,-display--design constraints, and so on. .

17 Once we've got a design,'we then perform a second 18 operational sequence analysis on it and we look at things-19 such as an estimate of operator workload based on=the

-20 operator working through this interface, performance time 21- estimates,. staffing issues, and so on.

222 Element five is staffing.

23 Here we say that staffing levels and the 24 qualifications of plant personnel are the responsibility of 12 5 - the combined license' applicant.

ANN RILEY & ASSOCIATES, LTD.

Court-Reporters 1250 I Street, N.W., Suite 300 Washington, D.C. 20005 (202) 842-0034 1

I

- 167 1 However, we do provide input from the designer

() 2- from us to the combined license applicant for the 3- determination of the staffing level of the main control-room 4 crew, and basically, in this document, we say that, look, we 5 have a design goal-for the main control room, and-our design 6 . goal is in -- I'll state it -- the design -- thank you very 7 much -- the design goal is to provide the appropriate -- to 8 design the plant -- to design the AP600 plant and the'  ;

9 interface in the main control of the main control room to 4 10 allow the plant to be monitored and controlled during all 11 conditions by one reactor operator and one senior reactor 12 operator. Again,'that's a design goal.

13 We then state in the same document that, hey, you 14 know, the-law, current law, 10 CFR 50.34 (f) , I believe, says

(,_,/ 15 that you have for a single unit two and two, two ROs'and two 16 SROs. We're saying that-the COL is responsible to address 17- that.

18 We have a design goal we're going in with. Our-19 design process will work towards achieving that goal, and l 20 then part of.that process.is the final V&V on a 21 plant-specific simulator, and that final validation will 22 show whether or not that. goal is accepted and whether or not 23 that goal is met, and whether or not it's met is irrelevant.

24 The COL is still responsible for addressinc -he staffing 25 levels of 10 CFR 50.34.

I ANN RILEY & ASSOCIATES, LTD.

/'

Court Reporters 1250 I Street, N.W., Suite 300 Washington, D.C. 20005 (202) 842-0034

168 i

1 DR. SEALE: Have you addressed anything related to

() 2 the issue that - perhaps you've come up with a design now 3 that one operator and one senior operator can handle --

4 MR. CARROLL: -In the control room.

5 DR. SEALE: --

in the control room. There's a 6 requirement for two and two. Have you picked the games that 7 these other two people are going to play to keep them awake 8 and so on? Parcheesi or whati 9 MR. KERCH: No , sir, we haven't. We have not.

10 DR. SEALE: I'm being facetious, but I'm not being 11 facetious. It seems to me that a person who has 12 responsibility has got to be challenged with that 13 responsibility, and if, in fact, there is a.fifth-wheel kind 14 of attitude that might develop, you're setting yourself up.

(O

,,/ 15 MR. KERCH: Conceptually, what we think the others 16 .will be doing -- we have the RO and we have the SRO. Okay.

17 The second RO, we. feel, will work within-the' switching and 18 tagging area of'the main control room, which in our design 19 is a little separate-area from the main control area-but 20 still part of the boundaries of the main control room, l 21 So, that second RO will work in that switching and 22 tagging area and be the primary interface between the 23 maintenance staff of the plant --

24 DR. SEALE: You haven't defined the magnitude of 25 that task, though, have you?

O ANN RILEY & ASSOCIATES, LTD.

Court Reporters.-

1250 I Street, N.W., Suite 300 Washington, D.C. 20005 (202) 842-0034

169 1 -

MR.'KERCH - We'have not.

() 2-3.

MR. CARROLL: Which is fine on day-shift, when most of the maintanance is being done. What does he do at 4 ' night? -Parcheesi.

5 DR.'SEALE: Certainly, the SRO might do something 6 -like maintenance management supervision or something. I 7- should say sequencing and control of access to plant and

8. that kind of thing.

9 CHAIRMAN BARTON: Which is what a lot of utilities 10 are doing today with a separate-shift. Yuu just won't need 11 a separate shift. These extra-people can provide that work.

12 control interface management between maintenance.

13 DR. SEALE: And that's fine for maybe 16 hours1.851852e-4 days <br />0.00444 hours <br />2.645503e-5 weeks <br />6.088e-6 months <br /> a 14 day, but what about the other eight? It seems, maybe, as we .;

I 15 look at human factors concerns, we may want to try to dig 16 deeper into these other issues.

17 MR. CARROLL: I notice George hanging over in the

-18 corner. Why aren't you at your place and_asking questions?

19 We looked all for you, since this is your chapter.

20' MR. KERCH: Element-six is the integration of 21 human-reliability analysis with human factors engineering.

22

Here-:again we presented an implementation plan to the staff.

23 The execution of that plant-le the responsibility of the COL 24 applicant.

25 Within this plan, we have defined critical human ANN RILEY & ASSOCIATES, LTD.

O- Court Reporters 1250 I Street, N.W., Suite 300 Washington, D.C; 20005 (202) 842-0034 .

1

170 1 actions, at-least what the criteria would be used:to

-(

! %)

i 2 identify _ critical human .ctions from the probabilistic risk 3 assessment, and risk-important tasks.

4 So,-in the WCAP that's referenced here --

5 actually, I don't reference it,_but it is WCAP-14651.

6 MR. CARROLL: It's up there.

7 MR. KERCH: Oh, I'm sorry, it is up there.

8_ WCAP-14651_ presents the plan and presents the 9 criteria used to select the critica. human actions and what 10 we call risk-important tasks.

11 What we found initially was that the way we define 12 _ critical human actions and one of the basic design premises 13 of the AP600 was that there are no critical human actions to

-14 respond to a design basis event._

15 So, to meet the intent of the 0711 criteEia, we 16 set up a-definition that we called risk-important tasks, 17 which accomplishes the same end.

18 It identifies tasks that are, relatively speaking, 19 more:important or more risk-critical than others, and with 20- those tasks, then, we input those-tasks into our task 21 analysis, and they're looked at and they're analyzed-

-22 specifically by our task analysis, by.our operational i

23 sequence analysis, and of course, the task analysis input to '

24 our HSI design, then, and to staffing and to procedure and }

25 training development.

O ANN RILEY & ASSOCIATES, LTD.

Court Reporters 1250 I Street, N.W., Suite 300 Washington, D.C. 20005 (202) 842-0034

171 1 The next figure comes out of ths WCAP and

() 2 basically tries to show the relationship between our human 3 reliability analysis and the HSI design cask analysis, the 4 HRA, which we define the cutoffs and definitions of critical 5 human actions, risk-important tasks.

6- We look at our PRA and we establish and identify 7 risk-important tasks. We use that and analyze those in our 8 task analysis.

9 The task analysis is input to procedure 10 development and the HSI or man-machine interface design, and 11- that feeds back into our human reliability analysis in the 12 sense of confirming and refining any assumptions that were 13 made in the human reliability analysis portion of the PRA.

14 We can use our RRA to establish test scenarion and

() 15 performance assumptions that will be needed in the final 1 16 validation and the defining test scenarios that would be 17 used on the simulator, and of course, we can validate those 18 assumptions that were made in the HRA through the final V&V.

19 DR. WALLIS: Can I ask a naive question? What's 20 the database for humca reliability in a nuclear control room 21 during a major accident *i 22 MR. KERCH: The database? You mean the source?

23 DR.--WALLIS: How do you make quantitative 24' predictions of how people are going to behave when the base 25 we have is TMI and a few situations like that?

O-ANN RILEY & ASSOCIATES, LTD.

Court Reporters 1250 I Street, N.W., Suice 300 Washington, D.C. 20005 (202) 842-0034

172 1 MR. KERCH: I can partially answer that. I'm not

() 2 3

a PRA expert, and we would need someone from PRA 'ot answer that in total.

4 I-know that there is a human reliability of the 5 section of the PRA. They do make assumptions based on our 6- concept ofrwhat we think our interface design is-and based 7 -- and I believe that there is a table of human error 8 probabilities that is used.

9 What the source of that table is, I'm not sure and 10 can't answer at this time, but we'll find out and get back 11 co you.

12 DR. UHRIG: Would simulator experience be one of 13 those quantities?

14 MR. KERCH: I don't know if a study has been done 15 to try to quantify human error based on simulator training f 16 or simulator experience.

17 MR. McINTYRE: This is Brian McIntyre.

18 There have been some studies done based on 19 simulator-type training, but I know also that they have 20 actions and they bias the times depending on how much stress 21 the operator is expected to be under, that if he's sitting 22 there and his Parcheesi game, as I think was the suggestion, 23 is interrupted, he'll do a really good job, but if there's a 24 do:en alarms going off, that he's probably not going to do 25 as good a job.

=

ANN RILEY & ASSOCIATES, LTD.

0% Courc Reporters 1250 I Street, N.W., Suite 300 Washington,-D.C. 20005 (202) 842-0034

173 1 So, that is taken,into account, and I believe that

() 2 those databases are based on not just the nuclear experience 3 but just the human experience in general.

4 So, it's not just a set set of numbers and error 5 -probabilities that are looked at, but there is a bias and a 6 way=to-try to--- to make it look like what we would expect =

7 to happen.

8 MR. CARROLL: George, what's the practice? It's 9 based on Swain, isn't it?

10 DR. APOSTOLAKIS: Well, that's why they are doing 11 ATHENA, you know, to try to LOvelop an approach to 12 developing those human error rate distributions, and 13 information on the simulators is being used, but it's not 14 the traditional data, because-you don't have enough.

() 15 Can you give us an example of how you did this, 16 .you know, you got the arrows that you show there-and some 17 important events and what you did with them?

18 MR. KERCH: Well, we haven't -- again, we've 19 submitted a plan to the staff.

20- .DR.-APOSTOLAKIS: Okay, 21 MR. KERCH: So, we haven't completed the plan, 22 althaugh in our WCAP, we provide a preliminary list of 2 3 -_ risk-important tasks that have been identified from the PRA, 24 based on PRA at that time.

25 DR. APOSTOLAKIS: And this is -- a human O ANN RILEY & ASSOCIATES, LTD.

Court Reporters 1250 I Street, N.W., Suite 300 Washington, D.C. 20005 (202) 842-0034

174 1 reliability analysis is coupled with possible failures of ,

2 the-I&C system?

3 MR. KERCH: Yes, I believe that is correct. You 4 know, certain scenarios are defined in PRA.

5 DR. APOSTOLAKIS: That's good.

6 DR._UHRIG: _One of the major things you do in the 7 initiation of a major problem is sit back and watch what's 8 happening, keep your hands off the system. Is that true 9 here? I assume it is, as it is in other plants.  ;

10 MR. CARROLL: More so.

11 MR. KERCH: Yes, I believe that is correct, that 12 is true here, and even more so.

13 DR APOSTOLAKIS: Is it 72 hours8.333333e-4 days <br />0.02 hours <br />1.190476e-4 weeks <br />2.7396e-5 months <br />? -What is the 14 time window?

k 15. MR. KERCH: Time window for?

16- DR. APOSTOLAKIS: The operators doing nothing, 17 they're not supposed to do anything.

18 MR. KERCH: Well, I'm going to be careful how I 19- say that,-because-we still have emergency _ response 20 guidelines, and-if you look at those guidelines for a 21 reactor trip or LOCA, you look at those guidelineE, it has 22 the operator doing things inside of 72 hours8.333333e-4 days <br />0.02 hours <br />1.190476e-4 weeks <br />2.7396e-5 months <br />, but those are 23 guidelines that are written to optimize the plant's 24- recovery, but you're right, the basic premise of the plant 25 design is that, from a safety point of view and looking at ANN RILEY & ASSOCIATES, LTD.

Court Reporters 1250 I Street, N.W., Suite 300 Washington, D.C. 20005 (202) 842-0034

175 1 the safety systems only, the operator can step back and not ,

l

() 2 3

do anything to respond to a design-basis event for, I believe, 72 hours8.333333e-4 days <br />0.02 hours <br />1.190476e-4 weeks <br />2.7396e-5 months <br />, that's correct.

4 DR. UHRIG: The main thing is to get the alarms 5 quieted down.

6 MR. KERCH: Yes. Or design the alarms so that 7 they're not a nuisance during an event like this. And that y 8 leads me right into the next element, the actual design of 9 the human-system interface.

10 In AP600, the human-system interface consists of 11 two parts, or we like to break it down into two parts, 12 consists of the operation and control centers -- and these 13 are a center from which someone can control or monitor the 14 operation of the plant,

,O

( ,/ 15 These centers include the main control room, 16 technical support center, remote shut-down room, the 17 emergency operations facility, local control stations, and 18 associated work-stations with those.

19 The human-system interface resources that are 20 available at these control centers are what we call the wall 21 panel information system. This-is a large screen display.

22 MR. CARROLL: In the control room.

23 MR. KERCH: In the control room, large screen 24 display, 25 So, all six of these resources are present in the

[/l

\ss ANN RILEY & ASSOCIATES, LTD.

Court Reporters 1250 I Street, N.W., Suite 300 Washington, D.C. 20005 (202) 842-0034

176 1 control room.

() 2 Two or more may be present as needed at the other 3 centers, at the other operation control centers, but the 4 wall panel information system large screen displays is in 5 the control room, the alarm system's in the control room, 6 plant information system -- the plant information system are

? your non-1E work-station displays, CRT displays, okay, if we 8 . ate to implement it today.

9 The computerized procedure system, the soft 10 controls and dedicated controls -- again, soft controls --

11 this is an MA station, manual / automatic station, that would 12 come up on a CRT screen, as compared to a dedicated control 13 that would be a physical switch on a control panel, but we 14 do have both in the AP600 control room, soft controls and

() 15 dedicated controls, and we have the qualified data 16 processing system.

17 The qualified data processing system are displays 18 -- these are your qualified 1E displays that include the 19 ODPS, or qualified displays, include the Class 1E 20 post-accident monitoring system variable.

21 DR. UHRIG: Those are all hard-wired, the 22 qualified data processing system.

23 MR. KERCH: No, sir.

24 DR UHRIG: They're not.

25 MR. KERCH: They are not. They are qualified so

()

ANN RILEY & ASSOCIATES, LTD.

Court Reporters 1250 I Street, N.W., Suite 300 Washington, D.C. 20005 (202) 842-0034 i

177 1 the software _that drives and generates the displays and the i f 2 QDPS displays come out of the protection and safety 3 monitoring system, the P&S system.

4 So, it is a fully-qualified system using qualified 5 software and hardware, as compared to the plant information 6 system displays which come out of the data -- the DDS 7 system, the data -- memory dump here -- the data display and 8 acquisition system, which is a non-qualified I&C system.

9 DR. UHRIG: What about the wall panel information 10 system? That's software?

11 MR. KERCH: That is also displays, but they're on 12 a large screen now, and that's a non-qualified system coming

13. out of-the DDS system.

14 The alarm system -- now, I'm saying all this --

() 15 what has been approved is, again, the design process for 16 this element. What we've submitted is an implementation 17 plan. We have not submitted a completed design.

18 We submitted an implementation plan to get us 19 there, and it was those boxes inside the dashed-box on the.

20 figure I had up earlier, which includes functional design, 21 design guidelines, design implementation, concept testing.

22 Okay?

23 DR. UHRIG: Now, the physical design of these will 24- follow -- what is, the NUREG-0711?

25 - MR. KERCH: The design process -- oh, I'm sorry.

O ANN RILEY & ASSOCIATES, LTD.

Court Reporters 1150 I Street, N.W., Suite 300 Washington, D.C. 20005 (202) 842-0034

178 1 The physical design will have to address the criteria that

() 2 are found in 0711, Revision 1, that's correct.

3 DR. UHRIG: Revision 1. That's the one that's 4 completed about a year ago.

5 MR. KERCH: Yes, sir.

6 DR. UHRIG: Which is for digital systems.

7 MR. KERCH: Yes. And I want to stress the point 8 that what has been approved is a design process here, and 9 the use of these resources in the control room, with the 10 exception of the computerized procedure system, and I'll 11 talk more about that when I get to the element on procedure 12 development, but the use of that resource has not been 13 reviewed and approved by the NRC nor precluded at this 14 point, and I'll talk more about that when I get to that 15 element.

16 What has been approved, again, are these resources 17 and their use and the fact that the main control area of the 18 main control room would have a reactor operator's console 19 that would consist of redundant reactor operator 20 work-stations and a senior reactor operator's work-station.

21 Each work-station would have multiple CRTs or l

22 visual display units from which the operator would access 23 his plant information system displays, he would access his 24 computerized procedures, he would access his soft controls.

25 The control room would also have a dedicated ANN RILEY & ASSOCIATES, LTD.

\ Court Reporters 1250 I Street, N.W., Suite 300 Washington, D.C. 20005 (202) 842-0034 l

l 179 l

I safety panel, and on those dedicated safety panels, that's  !

() 2 your Class 1E' panel, qualified panel that has and houses -

3 your QDPS, qualified display units, and your qualified

, 5 dedicated controls, and-this is your minimum set of 4  ;

5 dedicated fixed-position switchos.-

6 The operator can also interface from his

-7 work-stationLwith the wall panel large-screen displays.

i 8 Again, design details have not been reviewed-or j 9 approved, just the process and the use of these resources.

! 10 MR. CARROLL: For our newer members, we went 11 through this originally on ABWR, this notion that it would 12 be acceptable to. define a process at the time of '

13 certification, and the same was applied to the design of the 14 I&C systems, and the rationale for this is the fact that

() 15 - we're in a changing environment in terms of technology, and 16 it didn't make sense to lock people in to 1998 technology 17 when one of these plants may be ordered in 2005.

18 DR. UHRIG: If this process is'followed, then is

~19 there any subsequent NRC review of the details of the 3 20- system?

21- MR. CARROLL

Yes.

'22 MR. KERCH: There is an ITAAC that says that the 4 23 implementation plan as described within the SSAR will be

. 24 executed and the documentation'provided to the.NRC.

25 DR. UHRIG: For review and approval.

}

4 l

[ ANN RILEY & ASSOCIATES, LTD.

Court Reporters 1250 I Street, N.W., Suite 300 Washington, D.C. 20005 (202) 842-0034 a

- . - . , . . - , , . , , , . _ , . . . - 4 ,.,,,,_v,r.,,,',--n,,__,e ,..,.,,.~,,.,,,.,,,,,,n.- _ - ,

,,._,,--,-,,_-.n,-,n,--+,.~,w.nn,,--- , - - - - , ..,- = ,an

180 1 MR. KERCH: Yes, sir.

() 2 MR. CARROL: t The difference between this, 3 apparently, and ABWR is we called them DACs, in those days, 4 design acceptance criteria, and we also had ITAACs, but not 5 for this kind of thing. But now Brian tells me he's 6 changing the nomenclature on me. So, I'll wait till we get 7 to that and find out what he's up to.

8 DR. MILLER: So, if we look, say, internationally, 9 where the ABWR has been implemented -- of course, the NRC 10 doesn't have anything to do with their final implementation 11 -- would we expect the Japanese to follow the implementation 12 plan or not?

13 MR. CARROLL: The Japanese design was ahead of the 14 certification of the ABWR, but they did modify it to some

( 15 degree based on some of the things the NRC was doing. They 16 have a little different approach to human factors than we 17 do.

18 DR. MILLER: I understand that. That's why I'm 19 asking the question. If I went to the ABWR in Japan -- and 20 their control room is quite a bit different than, say, GE 21 design, but my question is would they have followed these 22 implementation guidelines?

23- MR. CARROLL: They didn't exist.

24 DR. MILLER: So, they didn't exist at that time.

25 MR. CARROLL: No. But Taiwan, I'm not sure of.

ANN RILEY & ASSOCIATES, LTD.

Court Reporters 1250 I Street, N.W., Suite 300 Washington, D.C. 20005 (202) 842-0034

181 1 DR. MILLER: Taiwan is using a different approach,

() 2- I understand. They're going to use more of the GE approach.

3 MR. CARROLL: The control rooms for JK-6 and 7 4 were designed by Hitachi and Toshiba.

5- DR. MILLER: _Right.

6 DR. POWERS: .You've made a substantial commitment here to-digital and software-based systems, and your 17 8 previous slide listed quite a few of them.

9 MR. KERCH: Uh-huh.

10 DR. POWERS: And the previous speaker said that, 11 elsewhere in the plant, in electrical systems, they were 12 'more interested in proven technologies. Why, then, is there 13 the discontinuity here? Why do we think these are proven 14 technologies or why do we have such commitment to them?

k 15 - MR. KERCH: We think that, as stand-alone

16. technologies, all these are proven, with the exception of 17 the computerized procedure system, and even that one may be, 18 perhaps, to a degree or an argument can be made.

19 For example, the wall panel information system 20 large-screen displays are used fairly extensively in the 21 telecommunications industry and also at some low dispatch --

22 I believe I'm correct in-saying some low dispatcher control 23 centers.

24 IHl. POWERS: The. reliability of those systems, at-25 .least in the telecommunication industry, is not so high as

=

ANN RILEY & ASSOCIATES, LTD.

~

Court Reporters 1250 I Street, N.W., Suite 300 Washington, D.C. 20005 (202) 842-0034

182 1 to be confidence-inspiring, is it?

2 MR. KERCH: Okay. We don't feel that the wall 3 panel information system is needed, and we don't want to 4 over-design that system to the point of -- and it was 5 . identified early in our process -- we don't want to 6 over-design it in the sense that the operator cannot operate 7 without it.

8 One.of the goals of the wall panel system and the 9 way -- the techno1cgy that we would use to implement that 10 today would be a rear-projection large-screen display unit, 11 and if the projection units were to fail and we lose those 12 large screens, the same displays would be available at the 13 worktstation CRT displays.

14 So, one, we don't want to design the system so we 15 can't operate without it, and two, we certainly want to 16 consider reliability and what we can do to enhance the 17' reliability of that system as we're designing it.

18 At this point, some of the objectives of that 19 system are to help provide a common frame of reference from 20 which the-operating crew in the control room can use to help 21 them maintain overall situational awareness of the plant.

22- You know, there's a concern that, if you're going-

.23 -to operate and monitor.the plant through a CRT, that you-24 could have what we-call the keyhole effect and lose the-big 25 picture.

ANN-RILEY & ASSOCIATESi LTD.

Court Reporters 1250 I Street, N.W., Suite 300 Washington, D.C. 20005 (202) 842-0034 7 ..

183 1 So, that's one purpose of that wall panel display, j

() 2 3

-to help maintain _the big picture and provide a common frame of' reference. [

4 The other means by-which we want to use that 5 large-screen display.is to help the operator locate 6 'information that he needs and locate it at the work-station.

7 So,-we want to use the large-screen display as a-navigation 8 aid, a navigational aid.

9. So, the operator may fly a cursor up-onto that --

20 some object on that wall panel display, salect it, and.he 11 gets a related, more detailed display at his work-station.

12 MR. CARROLL: I don't think you're answering the 13 question that Dana posed.

14 MR. KERCH: The answer to the question is that-( 15 those are the two objectives. We feel we can, through sort 16 of,a redundant means, provide the same objective at the 17 work-station by providing the same displays at the 18 work-station.

119 DR.' POWERS: The problem is that it's all hope.

20 MR. KERCH -. Pardon me?

21 DR. POWERS: It's all hope.

22 -MR. KERCH: Sure it is. That's correct. U1're

, 23 providing a plan'and not a completed design, 24 DR. POWERS: I mean you've made a commitment to 25 this. You've designed your entire -- that chart that you ANN RILEY & ASSOCIATES,-LTD.

Court Reporters 1250 I Street, N.W., Suite 300-Washington, D.C. 20005 (202) 842-0034

184 1 put up earlier is designed to produce this system, and the

() 2 3

question is, why would you do this when you have other ,

systems that at least you have a track record on?

4 MR. CARROLL: Other technologies, yes.

5 MR. KERCH: Utility input is one answer I can give 6 you.

7 The URD says -- I believe that a requirement in 8 the URD says thou shalt have overview displays, and when we 9 questioned our utility customers in our FOAKE program what 10 that meant and how they wanted that implemented, they said l 11 large-screen displays.

12- MR. CARROLL: Now, you're still talking about 13 large screens. I think --

14 Dana, aren't you talking about the whole notion of

) 15 using digital systems --

16 DR. POWERS: Digital softwata-based systems.

17 MR. KERCH: Oh, okay. We went with a compact 18 electronic work-station-type main control room for several 19 reasons, one of which is the URD. URD requires us to go 20 with electronic displays, electronic work-stations, and even 21 electronically presented procedures. Okay?

22 We feel, by going with this style of a control 23 room, we don't need as large of a physical space.

24 So, our main control area within the main control 25 room is physically smaller than I think what you'll find in ANN-RILEY & ASSOCIATES, LTD.

Court Reporters 1250 I Street, N.W., Suite 300 Washington, D.C. 20005 (202) 842-0034

185 1 most plants. The AP600 main control area is roughly 34 feet

'2 by 34 feet.

3- We do not nend a cable spreading room by going 4 with digital technologies.

5 DR. POWERS: .Okay. You've got a more cramped and-6 congested control room now. I mean there's another Fay to 7_ look at these things.

8 MR. KERCH: I don't think it is more cramped.

9 MR. CARROLL: But there isn't a cable spreading 10 room that's going to catch on fire.

11 DR.-POWERS: Well, there are other things that

1. will catch in fire.

13 What I'm trying to understand is why are these  !

14 things good? I mean the control room is a perfect example.

( 15 You say, well,.it's more compact and loan and mean, and I i

16 say, yes, congested, dark, dingy, claustrophobic -- all 17 those words come to mind, too.

18 MR, KERCH: I think, using the digital ,

19 technologies, you can bring information to the operator's  ;

?

. 20 _ fingertips that you can't do in today's conventional plant. ,

21 DR. POWERS: And we've concluded that operators  !

-22 .are information-starved, so they need more information. l I 23 'DR. MILLER: They need more information that is l 24 prioritized, I think, is the key issue, l l

25- I .hought all this came out of a number of 1

O . ANN RILEY.& ASSOCIATES, LTD.

Court Reporters 1250 I Street, N.W., Suite 300 l Washington, D.C. 20005 (202) 842-0034 i

186 1 studies, and of course, you referenced the studies, and I

() 2 know Woods, who did work for Westinghouse for some time, I 3 thought, came from that, and also what's been done in the 4 international arena -- say, the French, in four systems, has 5 -a wai'i panel display in it -- which concluded it is a more 6 effective human interface system than the old large wall 7 panels with dials and so forth, and the other-issue that the 8 French addressed was the qualification of those systems.

9 They introduced an intermediate _ qualification 10 -- called a -- they call it a 2E. They qualify all the human 11 interface systems at a different level.

12- DR. POWERS: Forgive me, but I ha"e not found the 13 French to be particularly on the cutting edge of digital and 14 software technologies. I mean the fact is ths Europe _is

) 15 'way behind most of this stuff that he's talkin3 about here.

16 How does it compare with your Sizewell experience?

17 Is that similar in nature?

11 8 MR.~KERCH: I talk a little bit about Sizewell.

19- Sizewell~has CRT displays. They have a system called safety 20 information display system, which.is our functional 21 equivalent of the QDPS.

22 I don't know much about their advanced alarm 23 system. We weren't involved with that, so I can't speak on 24 that, although I believe it's annunciateG windows.

25 They do not have soft controls, they have ANN RILEY & ASSOCIATES, LTD.

Court Reporters 1250 I Street, N.W., Suite 300 Washington, D.C. 20005 (202) 842-0034

187 1 dedicated controls, although I understand those dedicated 2 contro1r., are multiplex, at least some of them.

3 So, as far-as some of these resources, I believe 4 they're using some of these.

5 So, I would say that functionally equivalent plant 6 information system,- ODPS are two, at least, that are

_7 implemented.

8 DR. POWERS: Where they have not elected to go 9 with an identical system, conceptually -- this is still a 10 conceptual _ system -- where you have differed from Sizewell, 11 is the difference borne of inaccessibility at the time of

-12 Sizewell to the technologies that you want to apply, or is 13 there some educated election there?

14- MR. KERCH: I think, to some extent, it's 15 technology, other than that, I don't know. I don't know 16 why sizewell made the decisions they made.

17 DR. MILLER: I think, going back to the-French, I 18- think the N-4 reactor has everything this has got. It has a 19 wall display. It has soft controls. That system is 20 approached where the operator _never stands up to do 21 anythings he does everything from a sitting position.

22 I should have brought pictures of it. I have 23 pictures of their bystem. I think it has everything in 24 there. I'm not certain it will meet this guideline, 25 necessarily, one way or-the other.

ANN RILEY___& ASSOCIATES, LTD.

Court Reporters 1250 I Street, N.W., Suite 300 Washington, D.C. 20005 (202) 842-0034

188

1 DR. POWERS

I mean the question I'm really asking

() 2- is, if the operator can do everything sitting down, he never j 3 ~ has to stand up to do anything -- what I'm asking -- is that '

I j 4 a good thing? Why do we think that that's a good thing?

5 - DR. MILLERi I wasn't involved in the studies or s' 1 4

6 anything like that. I was over there, I asked the French

! 7 those questions, and they claimed the operators said that's j 8 what they wanted, and apparently there's a bunch of studies .

-t

). 9 done.

10 I looked at their list of references in there,-and 11 those references list - a lot of the studies that I'm aware

• 12 of have not studied that carefully -- that have concluded I 13 this is a more effective way to have human interface with 14 the plant, which is what we all have seen in the current 15 plants, which is dials and so forth.

16 MR. CARROLL: In fast-happening. events, it's nice i

17 for the operator to be in one place, with everything in l 18 front of him.

19 DR. POWERS: _But you-have to understand, he 20- - doesn't have everything in front-of him. He has to page 3

21 through to find these things.

] 22 MR. CARROLL: Yes.

23 DR. POWERS:-- Your hypothetical example, the t

24 overall display is gone, and he has to use a CRT, he's got 25- to page.

.A?iN RILEY & ASSOCIATES, LTD.

Court Reporters 1250 I Street, N.W., Suite 300 i

Washington, D.C. 20005 (202) 842-0034 r r_.ew,,-or,.,. m ,%~

,,,y,,_,,.,...,m.---- .

m.,wm.w-.,,,_... - . - . . .%, , - w,, _.,,-,-_,w.~e-. . ---w...ww,, r

189 1 MR. KERCH: He'd have to navigate to the 2 information.

3 DR. POWERS: He does not have everything in front I

4 of him.

5 MR. CARROLL: Yes, he does. He just has to page.

6- I've seen control boards that are just terrible from a human 7 factors point of view, where he'd have to go here and turn 8

this switch and run down to the other end and look at a dial 9 and come back and turn another switch.

10 DR. POWERS: What you're saying is that, when you 11 compare the worst of conventional boards to the best 12 hypothetical board, the best hypothetical board comes off

'13 well. I'm not surprised at that.

14- Compare instead for me the best of the

(~

15 conventional boards and the best of the hypothetical boards, 16 and why do you think that this hypothetical software system 17 is best?

18 DR. MILLER: Maybe we should ask you to compare --

19 the Westinghouse plant I've been in is kind of one of the 20 latest, is at Braidwood or Byron. How would this compare --

21 see if you can address Dr. Powers' question based on that.

22 Like procedure systems are not computerized. They 23 go back and they leaf through them with books and so forth.

12 4 How would_your computerized procedure system compare with 25 leafing through books?

ANN RILEY & ASSOCIATES, LTD.

\-- Court Reporters 1250 I Street, N.W., Suite 300 Washington, D.C. 20005 (202) 842-0034

l 190 1 MR. KERCH: We think we can reduce the operator's

() 2 workload through the use of computerized proceduros. An 3 example would be if you were to computerize the emergency 4 operating _ procedures.

5 One of the things that are done in emergency 6 operating procedures and following them is there's a set of 7 parallel -- we call it parallel information that the 8 operator has to keep track of in his mind or keep 9 remembering to go to as he's executing the steps of the 10 procedure.

11 For example, there's a fold-out page, for those 12 familiar with Westinghouse PWRs. There's a fold-out page of 13 items that he's got to remember to keep referring to 14 -periodically-and making sure that he doesn't have to take

() 15 some action for some item on the fold-out page.

16 The monitoring of the critical safety functions is 17 considered a parallel action that the operator or somebody, 18 a shift technical advisor, has to do while the steps of the 19 optimal recovery guideline are being executed.

20 DR. MILLER: Now,--in current plants, there is 2 21 critical function monitor which is computerized.

22 MR. KERCH: There are some plants that have --

23 DR. MILLER: Which are non-qualified.

24- MR. KERCH: That's correct.

25 DR. MILLER: They are used. Like at Watts Barr

-O ---

ANN RILEY & ASSOCIATES, LTD.

Court Reporters

^1250 I Street, N.W., Suite 300 Washington, D.C. 20005 (202) 842-0034

..,-.______-m_- - _ - - -

191 1 there was one.

() 2 MR. KERCH: Yes. And so, that's used as a tool, 3 really, to help reduce the operator's workload. He's~ busy 4 trying to -- you know, to get through and execute the 5 procedure.

6 So, we feel that we can have the computer monitor 7 those parallel activities and alert the operator when he 8 needs to take some action based on that activity.

9 So, in a sense, it reduces his workload.

10 DR. MILLER: Do che computerized systems give the 11 operator a more rapid indication of the entry levels to the 12 emergency operating procedure and that type of thing, entry 13 points to the emergency operating procedure.

14 MR. KERCH: We think we'll go a long way towards

) 15 addressing what some current plants see today as some time 16 constraints.

17 For exnmple, doing the presnurizer solic on an 18 inadvertent safety objection -- a lot of plants are having 19 .to address the fact that they.can't get to the SI 20- . termination criteria soon enough to prevent the pressurizer 21 from going solid, which is undesirable.

22 We feel we're going to go -- with computerized 23 procedures, we'll get to that point a lot sooner, and I'm 24 'not saying that the computer has control of everything.

25 This is an operator-controlled interface.

O- ANN RILEY & ASSOCIATES, LTD.

Court Reporters 1250 I Street, N.W., Suite 300 Washington, D.C. 20005 (202) 842-0034

~ m

192 1 The computer doesn't do anything except assess a

() 2 step and tell the operator and provide the supporting 3 information on the status of that step.

4 So, we feel we'll go a long ways toward addressing 5 those things, and another time-critical issue is the --

6 filling the steam generator and sending water down the steam 7 lines during a-tube rupture and getting to those appropriate 8 steps, cool-down and de-pressurization steps, soon enough to 9 prevent that from happening.

10 So, we feel that, you know, the computerized 11 procedure system goes a long way towards addressing that,-

12 and we don't think we're going to see -- we'll have that 13 problem.

14 Interfaceuis a way to reduce the workload-and get

() 15 it to the appropriate point.

16 MR. CARROLLi Back to Dana's thing about the 17 workstation.

18 You know, you just take a system that you worry 19 about, let's cay component cooling wnter. With digital 20 technology you can have the whole system up on the screen --

21 MR. KERCH: Yes, sir.

[ 22 MR. CARROLL: -- you can have all tho flows, what 23 pumps are running, what aren't. It's all there, and you can 24 do a lot more-than I'm just describing, whereas in the good 25 old days with the best of control board designs you had to ANN R1 LEY & ASSOCIATES, LTD.

Court-Reporters 1250 I Street, N.W., Suite 300 Washington, D.C. 20005 (202) 842-0034

193 1 visualize a lot of things.

()

2 You looked at a pressure gauge and you said, yes, ,

3 I guess that pump is running, and you looked at this or 4 that, and you can really manage the operation of systems a 5 lot better.

6 DR. POWERS: In fairness, you are still going to 7 have to look at a CRT and say, yes, I guess that pump's 8 running.

9 DR. MILLER: Well, I asked earlier whether you '

10 looked at international experience and I could have the same l

11 question that Dr. Powers has. When I looked at the N4 the 12 French have, where they have a back panel and display like 13 you have specified, and I_ looked at the one -- the ABWR in 14_ Japan, it was a totally =different approach.

()

15 In fact, the one in Japan looked somewhat like-the 16 old wall panels approach except it's more condensed, so I i

17 asked the question have you looked at what's been done '

18 internationally, say with those two plants and other plants.

19 DR. UHRIG: There is also-the experience in 20 Taiwan, the EPRI implementation of the KaChang reactor?

21 DR. MILLER: KaChang did the computerized 22 ' operation.

23 DR. UHRIG: Yes, emergency procedures, i

l

24. The studies that I saw on this basically said it L 25 didn't really help the best operators but it did a

() ' ANN RILEY & ASSOCIATES, LTD.

Court Reporters i 1250 I~ Street, N.W., Suite 300 t l Washington, D.C. 20005 (202) 842-0034

194 1 tremendous job of improving the performance of the newer and

() -2 the less experienced operators. At least that was a goal.

3 DR. MILLER: Yes, I would hope -- they have gone

4. way beyond with the KaChang now.

5 DR. UHRIG: This is five years ago?

6 DR. MILLER: Ten years ago, I think, and I am not 7 certain they ever implemented that program.

8 DR. POWERS: Would I be wrong if I was to 9 characterize everything that has been said in response to my 10 question with the description that says that I see a 11 tremendous amount of enthusiasm about how these new 12 ' technologies can solve the problems we have percoived to 13 exist in existing system, but relatively little attention 14 being given to what the new problems that these systems are 15 going to create.

16 MR. CARROLL: -I think, Dana, we do have some 17 experience with these new systems.

18 We have it in fossil power plants. We have it in 19 the aviation industry. I think we know what some of the

-20 tradeoffs are.

21 MR. KERCH: And we have looked at some of those 22 problems in our OER Report, the Operating Experisnce Review 23 Report, where we looked at -- there is a section in that 24 document that addresses interfaces that are not used in 25 current nuclear power plants or there is not a significant O ANN RILEY & ASSOCIATES, LTD.

Court Reporters 1250 I Street, N.W., Suite 300 Washington, D.C. 20005 (202) 842-0034-1

195 i

1 amount of operating experience in a nuclear power plant but

() 2 yet used in other industries such as soft control, so we i 3 have looked at that experience and identified human factor

! dL related issues from that experience and then stated how we l 5 -planned to address that in the AP600 design or put it in our i

6 issues tracking system.

d 7 DR. POWERS: Is this OER Report one that I should 8 have read?

9 MR. CARROLL: It's right here.

10 DR. POWERS: I mean the question is do I have it?

)

] 11. MR. KERCH: I think it will go a ways towards i

12 answering your question, yes, sir, especially looking at the 13_ use of soft controls,'our screen displays and advanced alarm 14- systems.

( 15 DR. MILLER: I know you did ask one, I think, l 16 critical question -- but what new things are we going to 17 have to deal with that we dca't anticipate?

18 I don't think.that question is very well answered.

H19 MR KERCH: Yes.

By the way, Dana, that is the subject

~

-20 DR. MILLER:

21 of a research program being proposed by the I&C Branch 22- here -- that is one of the new-research programs, to look at 23 interaction problems we don't know about when you use 24 computerized systems.

25 DR. POWERS: I think that if-I -- I'm glad that O ANN RILEY & ASSOCIATES, LTD.

Court Repo'iters 1250 I Street, N.W., Suite 300 Washington, D.C. 20005 (202) 842-0034

196 1 NRC has a research program proposed in that area.

() 2- I would be more comfortable if the designer had a 3 research program.

4 MR'. KERCH: We have I won't call it a research 5 program but part of our process, our design process for the 6 design of the HSI is this concept testing where we look at 7_ the adequacy of our concept and we use the results of 8 testing real-operators on the new interface and we feed 9 those results back into_the design of the interface.

10 DR. SEALE- It's always risky to take experience 11 in one area and try to graft it on unilaterally to a 12 different problem and in the case of the airline industry 13 versus nuclear power plant operations, of course the big 14 problem besides the high-end personal investment that the

) 15 individuals involved have, I mean life and death of the 16 airlir.c aicuation, is the timescale, 17 MR. CARROLL: That's correct.

18- DR. SEALE: And it's hard to say how much of the 19 positive attributes of certain things that are being done 20 are directly related to the timescale tyranny that is 21 imposed on decisions in the airline industry, but in the 22 case of the kind of -- you know, the barn's on fire and the 23 horses are gone thing like they had with the DC-10 that lost 24 the back end control surfaces and so on --

25 MR. CARROLL: Sioux City --

4 ANN RILEY & ASSOCIATES, LTD.

Court Reporters 1250 I Street, N.W., Suite 300 Washington, D.C. 20005 (202) 842-0034

197 1 DR. SEALE: Yes, the United process of integrated +

2 crew training and the fact that they had brought an 3 experienced person who was a deadhead on the flight and to 4 help provide input for the problem-solving that-had to be 5 done in order to get that airplane on the ground with some 6 survivors, it seems to me makes the point that multiple

-7 . assessments by individuals of problems and-the process of 8 picking and choosing that goes with that can have some very  !

9 positive consequences in coming to the right answer.

10 . Now what we have done here is we have said, oh, 11 boy, newer is better, digital is nicer, and all that. sort of 12 thing, and we're going to be able to reduce all this stuff 13 to the point to where one person can handle and the other 14 guy can play Paicheesi. ,

15- You know, it's all pretty good until you get to 16 that last statement. Somehow or another, it strikes me that 17 the ability to broaden the experience base that is involved 5 18 in serious accident assessment and-response is not a 19 capability we ought to throw away without thinking very, 20 very seriously about it.

21 That is what you are doing when you say you are <

22 going to put it all on one console with a bunch of call-ups 23 for different. files and so forth and let one person be the 24 navigator.

25 DR. MILLER

This is not proposed to have one '

ANN RILEY & ASSOCIATES, LTD.

4 Court Reporters 1250 I Street, N.W., Suite 300 Wcshington, D.C. 20005 (202) 842-0034 4-

-- _ . . - - . _ , . . , _ , , , . , _ _ . . .,._,-e..__,__...__ . _ _ , . _ ,___,._~.r,.._ _ . , , . . . . _ . . . _ . - , , , _ _ . . . _ , _ -

.. _ . ,_.,_,,_ 7

198 1 -person, is it?

2 MR. KERCH: - No, it's ' design, simply a design 3 goal and not only that, the werk stations will have multiple 4 CRTs.

5 Our starting point in design _right now is 6 CRTs 6 per workstation plus three flat panel displays at the 7 workstation, so that gives you six, 12, 18 windows that the .

8 operator at the reactor operator's console can look at.

9 Now granted he is only going to use-the six and 10 the three because 18 is overwhelming, but he'll probably 11 won't use the redundant workstation. He'll use the primary 12 workstation, if you will.

13 CHAIRMAN BARTON: Still, all your premises are you 14 only need one control room operator, one --

15 MR. KERCH- That's a design goal. Thnt's e,orrect.

16 DR. WALLIS: I would like to help you by 17 suggesting that if John Kemeny were alive today and-we could '

18 ask him if he would say what he said 20 years ago when he 19 chaired the President's Commission on TMI, that you should 20 use. digital as much as possible -- that's tae only way to 21 go, 22- DR. POWERS: Look, I just really am unpersuaded by 23- arguments by authority regardless of where that authority 24 comes from.

.25 (Laughter.)

ANN RILEY & ASSOCIATES, LTD.

Court Reporters 1250 I Street, N.W., Suite 300 Washington, D.C. 20005 (202) 842-0034' a

199

- - -7 1" MR. CARROLL: You should .sve Lecognized that

( ) 2 early-on, Graham, in your career.

3 DR.-WALLIS: It just takes longer to persuade some 4 people than others.

5 DR. POWERS: Well, yes, I guess authority will make me-among the unpersuadable. I'll define a limit n the 17: experience if all we have is-authority to argue-that, l

8- MR. CARROLL: But one of our own, Hal Lewis, was a 9 big exponent of the source of,the authority doesn't help.

10 DR. POWERS: Now if you can get my wife to endorse 11 it, I'll probably have to give in but --

12 (Laughter.)

13 DR. POWERS: -- but short of that authority --

14 DR. MILLER: Can I have her phone number?

() 15 DR. POWERS: That's going to look good on the R16 E. nrinacript,-you asking for my wife's phone number.

17z (Laughter.]

18 DR. KRESS: I'll give it to you.

19 [ Laughter.]

20 CHAIRMAN BARTON: Moving right along'--

21 MR KERCH: So that the human-system interface 22 design again processes is submitted as an implementation 23 plan as described in 18.8.1 and the execution of that plan 24- and the submittal of the documentation and review and 25- approval is.-- the submittal of that plan and the supporting I

ANN RILEY & ASSOCIATES, LTD.

Court Reporters 1250 I Street, N.W., Suite 300 Washington, D.C. 20005 (202) 842-0034

200 i documentation is finalized cc the applicant's

() 2 responsibility, and is aga!.n an ITAAC item, to execute and 3 document and submit to the NRC for review and approval.

4 I did mention that the one resource that there 5 isn't-sufficient operating experience in industry that we 6 would like to use is a computerized procedure system and I 7 placed this statement basically exists in Chapter 18 and 8 Chapter 13 that the computerized procedure system is not 9 part of the design scope that the NRC is being asked to 10 approve.

11 Its acceptability will be determined during the i

+

12 AP600 verification and validation on a simulator and b5 13 reviewed as part of an application for a combined license.

i f 14 DR. MILLER: Is Westinghouse in the process of 15 designing a computerized procedure system?

16 MR. KERCH: Yes, sir. We -- in fact, the 17 computerized procedure system is being installed at the 18 Beznau plant.

19 I believe I am correct in saying that the HSK has 20 recently said, and they put certain conditions on it, but 21 they I believe have given the approval with conditions that 22 Beznau can go ahead and install --

23 DR. MILLER: What plant is this?

24 MR. KERCH: Beznau in Switzerland.

25 DR. SEALE: And the HSK is their regulatory --

() ANN RILEY & ASSOCIATES, LTD.

Court Reporters 1250 I Street, N.W., Suite 300 Washington, D.C. 20005 (202) 842-0034

201 1 MR. KERCH: Yes, their regulatory body.

That's 2 correct.

3- Now they have given them certain conditions.

4 There are certain validation conditions and some more 5 simulator scenarios and training they want to see done, but-6 they can go ahead and install it now along with, in parallel 7 with their paper procedures and so on.

8 That design, it is a backfit design, is the i 9 starting point for our design. We feel we can make several 10 enhancements to that system that would improve it'.

11- DR. MILLER: Of course if there is some XYZ 12 Corporation came up with another one-that was comparable 13 that could be also used-in this plant, right?

14 MR. KERCH: Yes, sir.

) 15 Now procedure developmentHis Element 8 of the 16- program. Procedure development,- as stated when-I covered.

17 Chapter 13, is the responsibility of the-COL and there a 18 WCAP that provides to the COL the designer's input to that, 19 providing the information, as stated earlier.

20 Again, there is.a difference here. The 21 computerized procedure system is the interface or the medium 22 by which the operator would execute his procedures, okay? --

23 so procedure development is.still the COL's responsibility-24 whereas the HSI design, the computerized procedure system, 25 is simply an interface through which he would operate those ANN RILEY-& ASSOCIATES, LTD.

Court Reporters 1250 I Street, N.W., Suite 300 Washington, D.C. 20005 (202) 842-0034 Y,

l 202 1  : procedures.

2 DR.-MILLER: When you say "information" on AP600 3 emergency response guidelines, what exactly does that mean?

4- MR. KERCH: Actually, the ERGS, the AP600 ERGS I

were submitted to address Element 8 of Chapter 18 and the

~

S 6 DSER open items that were there, so-the ERGS are there, the-7 background documents are there, and this WCAP provides 8 information on how those ERGS n re developed and how you 9 would take those ERGS and go to procedures.and how you would 10 take-those emergency procedures and computerize them and 11- . implement them into our --

12 DR. MILLER: That would be the starting point, to 13 develop the emergency operating procedures?

14 MR. KERCH: Yes.

15 Training program development, covered in Chapter

-16 13, but also part'of the HFE-program, again COL applicant 17- responsibility, and we provide a designer's input.to that 18 -training program dev.elopment.

4 19- DR._WALLIS: I have a question about procedures, 20 as_to the kind of intelligent feedback frcm these computers.

21 I mean'there are procedures which are on the computer.

22- Presumably the operator looks at them arid does things.

23 Does the computer come back and say what you have 24 -just done_ violates this procedure? -- or are you sure you really want to do this? -- or does it give some kind of ANN RILEY & ASSOCIATES, LTD.

Court Reporters 1250 I Street, N.W., Suite 300 Washington, D.C. 20005 (202) 842-0034

_-- - _ - - _ _ - - - - - I

1 203 1 feedback to make sure that he is on track?

() 2 MR. KERCH: The system as conceptualized now would

-3 make a recommendation, would recommend let's say 4 transitioning -- say there is another procedure that the 5 operator should transition to.

6 The system would suggest transitioning to that 7 proce' dure for the following and provide supporting 8 information that it used to make that decision.

9 DR WALLIS: But --

10 MR. KERCH: It is up to the operator to make 11 that -- whether he agrees or not.

12 MR. CARROLL: But does the system have input from 13 an actual plant?

14 MR. KERCH: Yes.

(_) 15 MR. CARROLL: Okay.

16 MR. KERCH: Yes, it does. It receives --

17 DR. UHRIG: When you go through a mode change, for 18 instance, one mode to another, one. state to another, this-

19. would be reflected in the emergency procedures that were 20 presented to the operator at that point?

21 MR. KERCH: Well, you mean like transitioning from 22 mode 3 to mode 2 or mode 4 to mode 3, is that your question?

23 DR. UHRIG: Yes, or an accident progressed to 24 another stage, so to speak.

25 MR. KERCH: Okay. Yes,'if you are in the

/'N t\ ,) ANN RILEY & ASSOCIATES, LTD.

Court Reporters 1250 I Street, N.W., Suite 300 Washington, D.C. 20005 (202) 842-0034

204 1 computerized emergency procedures and you are in E-1

() 2 basically, I forget whether it is E-1 or E-0 but you meet-3: the~ccnditions to transition to SI termination, which is 4 another sub-procedure within the ERGS, the computerized

5. procedure system will tell you transition to -- and click j 6 here to do that, 7 DR.-UHRIG: Because it got information from the

-8 plant.

9 MR. KERCH: Yes, because it received information 10 from the plant and assessed that, but it will display to the 11 operator that information that it used to make that

'12 decision, and the operator is in control. He can agree or 13 not agree or he can go off and confirm that that information 14 is-correct.

15 -The operator is in control of that. The computer 16 doesn'tErun off and do it.

17 HDR. WALLIS: So is diagnosis part of this? As I

'18 remember with TMI, one of the problems was the procedures 19 seemed to say if'you have a guillotine break, this is what 20 happens, and if you have this -- there was much less 21= emphasis on getting'from the evidence to what must be 22- happening, the diagnosis of what is going on --

23 CHAIRMAN BARTON: We are now into a world of 24 system-based procedures.

25 DR. WALLIS: And your computer has the sort of x

ANN RILEY & ASSOCIATES, LTD.

Court Reporters 1250 I Street, N.W., Suite 300 Washington, D.C. 20005

=

(202) 842-0034 wG

205 1- intelligence to diagnose.what is going on?

() 2- MR. KERCH: The computer within the reactor trip 3 response guideline --Lyou eventually get,_after you have

-4 verified your immediate actions and all the safety _ systems 5 have worked the way they are supposed to, you actually_get 6 'to'what I would call diagnosis steps where the operator 7 reads --

8 DR. WALLIS: But-there isn't assort of plant 9 simulator capacity in the_ computer so it can anticipate what 10 might be the result'of taking some action?

11 MR. KERCH: No. '

12 DR. WALLIS: It can't do it real: time following 13 the accident.

14- DR. UHRIG: I know of no system that has_that kind 15 of thing.

16 MR'. KERCH: No, it's not - .it is not an AI.

17- system, an artificial intelligence system.

18 DR WALLIS: But are you working towards that? ,

i 19 MR. KERCH: We can, but I think there are a lot'of issues that would need to be overcome.

21 DR.~ WALLIS: I think you ought to look forward to 22 what's-better. -Don't be stuck in-just what regulations 23 restrict.you to. I am always bothered by following the 24 letter of some law which may be stupid-in view of what is 25 - possible next year.

O ANN RILEY & ASSOCIATES, LTD.

Court Reporters 1250 I Street, N.W., Suite 300 Washington, D.C. 20005 (202) 842-0034

206

'l MR. CARROLL: Then you are going to encounter

() 2 3-Dana, who is going to say --

DR. POWERS: Why? Why is that better?

4 MR. CARROLL: Our Luddite member --

5 DR. WALLIS: After 10 years of trying to implement 4 6- some of those systems, we.have zero implemented.

7 DR. MILLER: .Yes. As Bob alludes, there's been

-8 many who have tried exactly what you are including Bob, 9 including myself, and the issues are not easy.

10 Of course, what we have now is symptom oriented

, 11 procedures, which is far better -- and there is some 12 diagnostic information provided at the end.

13 MR. KERCH: Right, exactly.

1 14 DR. MILLER: But to have a fully developed

\ -15 computerized system which tells the operator when he is all

-16 done, okay, this is what happened, I think we are a long, 17 long way from that, unless Bob's new people down there have 18 developed more things than we have developed.

19- MR. KERCH: So that's right. There are some 20 diagnostic steps where the computer will assess those steps 21 and recommend -- transitioning to a loss of secondary 2 2 -. coolant or loss of primary coolant procedure, and-the ERGS-23 are kind of written that if wrong decisions are made, you 24 are at times reassessing and rediagnosing at various points 25 within the ERGS and making sure that you are at the right A

ANN RILEY & ASSOCIATES, LTD.

Court Reporters 1250 I Street, N.W., Suite 300 Washington, D.C. 20005 (202) 842-0034

207 1 place'and the computer will continue to do that and continue

() 2- tosmake the appropriate suggestions or recommendations, but 3 it is not an AI system right now,'that's correct.

4- DR. MILLER:

ButLthe ultimate goal would be to 5 have a computerized AI type system which in the end would 6 say, okay, here is the root cause of what caused --

7 initiated this event.

8 I think we are -- we have bitten off bits and 9 pieces of that over the years --

10 MR. KERCH: Right, that's correct.

11 DR. MILLER: -- but to do it for the entire 12 plant --

13 DR, UHRIG: The Taiwan experience from what I 14 understand was basically an expert system in its simple-form 15 and-it strikes.me that that -- well, effectively it was 16 reduced to a conventional program by the time.it was 17 finished -- with the AI aspects that were essentially.

18 eliminated,-but in-the design phase it played a big role ---

19 at least that's my understanding.

20 MR. KERCH: The system you mentioned earlier?

21 DR. UHRIG: Yes, the one in -- the one -- that EPRI submitted on the Taiwan plant.

~

22 23- CHAIRMAN BARTON: I think we need to move along.

24~ MR. KERCH: Element 10 is the human system 25 interface design test program and in AP600 we have really 1

b

\ /'

ANN RILEY & ASSOCIATES, LTD.

Court Reporters 1250 I Street, N.W., Suite 300 Washington, D.C. 20005 (202) 842-0034

208 1 two parts to our test, testing and-I have mentioned'one of 2 them, and that is the human-system interface design-concept 3' test.

4 This is an-actual.part of our HSI design process 5 and it is used to refine our functional and-detailed 6' interface design.

7 These concept tests are organized around 15 human 8- performance issues that are-identified by WCAP 14701 and 9 WCAP also presentsia description of how we identified those L10 15 human performance issues that are relevant to this

-11 advanced interface, okay? '

12 There are actually 17-issues identified there --

13 16 and 17 are your formal V&V testing issues, so the concept 14 tests that's part of the design process addresses the first-15 15 human performance issues.

16 Again, there-is a WCAP that presents our concept

17. test plan that is centered around-those 15 issues.

18 Finally, also part of the test program issthe 19 ' final verification and validation. We have another? I 20 document, WCAP.14401, that provides a programmatic level 21 description of the HFJ V&V so at a programmatic level the 22  ; NRC has been; asked to review this V&V, so an I'LLAC exists 23  ;

that says the detailed implementation plan will be developed 24' and executed and associated documentation submitted to the

-25 NRC for review and approval.

O' ANN RILEY & ASSOCIATES, LTD.

Court Reporters 1250 I Street, N.W., Suite 300 Washington, D.C. 20005 (202) 842-0034 a

209 1 This V&V really consists of five parts. I have 2 mentioned one of those parts was the integrated system 3 validation where you do a validation on a plant simulator, 4 validating the integrated control room and interface design.

5 There are other parts. There is a human factors 6 engineering design issues tracking verification where you go 7 and you verify that all the issues you had identified and 8 were tracking in your system have indeed been resolved and 9 documented.

10 DR. UHRIG: Are you using V&V in the same sense 11 that it is used in the computer programming, in that the 12 verification means that you are doing what you set out to 13 do, and validation means that it was the right thing to 14 do -- is that the concept that you are using?

k 15 MR. KERCH: Yes, sir. I believe that kind of 16 captures it pretty well.

17 There are, like I said, five parts to the formal 18 V&V. The validation that I mentioned on a simulator is 19 checking to see if indeed it was the right thing to do.

20 The other parts are verification phases where you 21 are going and you are checking have you done what you said 22 you were going to do, which includes the tracking system 23 resolution, HFE design verification, HSI task verification 24 are other parts of that five-part V&V program.

{

25 The last thing to talk about in Chapter 18 is G ANN RILEY & ASSOCIATES, LTD.

Court Reporters 1250 I Street, N.W., Suite 300 Washington, D.C. 20005 (202) 842-0034 I

-210

, 1- minimum. inventory.

() 2 Here-we-defined in Chapter-18 in response to-the 3 DSERicriteriatfive-criteria that we would use tolidentify 4 .the-AP600 minimum inventory of controls, displays and 5 alarms -

6- MR. CARROLL EWhat,is-the notion here? -Why are 7 you doing this?- What does minimum inventory mean?

8 'MR. KERCH: - -I think the background on this and the 9 rationale was that since we weren't submitting a-completed 10- design, the intent is that we should define a' minimum 111= inventory or a minimum set of controls,-displays and alarms )

12 that are needed to ensure that the plant can achieve and 13 maintain a safe shutdown --

14 '- MR, CARROLL: And that is something you should f 15 -know at this_ point? 1 16; MR. KERCH: Yes. So that1is what we set-out to-(17 _do.

18 DR MILLER: That minimum is listed here in the 19 FSAR someplace?-

'20 MR.-KERCH:- Yes, and I have it as an overhead in 21' my handout.

22 Five criteria that we negotiated with the Staff on 23 and agreed--to on how to pick and identify this minimum

-24 -inventory set was Reg Guide 1.97, types A, B and C, Category L25 l' instrumentation should be displayed as the minimum set.

O ANN RILEY & ASSOCIATES, LTD.

Court Reporters 1250 I Street, N.W., Suite 300 Washington, D.C. 20005 (202) 842-0034

_____m__..___-

211 1 You should have those dedicated controls for ll 2 manual, safety-related system actuation, such things as 3 manual containment isolation, manual feedwater isolation, 4 manual safety injection, manual reactor trip and so on.

5 MR. CARROLL: Now these are all hard-wired?

6 MR. KERCH: These are dedicated fixed positions, 7 switches which are hard-wired to a certain extent.

8 For example, ESF actuations I believe are 9 hard-wired to the ESF actuation cabinets and if Ken's coming 10 on next, he can address that further, the I&C chapter 11 author, but --

12 MR. CARROLL: Okay.

13 DR. WALLIS: Do you have a vessel Jevel?

14 MR. KERCH: Vessel level indication? We have --

15 DR. WALLIS: I forget --

16 MR. KERCH: We have a parameter that is used to --

17 DR. WALLIS: It does have a vessel level?

18 MR. KERCH: It's not a direct reactor vessel level 19 indication. ,

20 One of our paramecers here does that -- hot leg 21 water level.

22 DR. WALLIS: Hot leg.

23 MR. KERCH: Hot leg water level, reactor hot leg 24 water level is part of the minimum inventory --

25 DR. WALLIS: So if you lose that, you don't know ANN RILEY & ASSOCIATES, LTD.

Court Reporters 1250 I Street, N.W., Suite 300 Washington, D.C. 20005 (202) 842-0034

=212

-1 where it'is --

2_ -MR. KERCH: I'm sorry?

3 DR. WALLIS:' If the_ hot let is dry, you don't know 4 where the level is?-

5-- MR. CARROLL: When'it's dry you've got a problem.

6 DR. WALLIS: You knew you had a problem before.

7 [ Laughter.)

8 MR. KERCH: I can't answer that. I don't-know.

9 DR. WALLIS: The vessel level has been around for-10 decades. _It's a question. ~

Do you still not have a vessel

'11 level'--

12 MR KERCH: I know that has been an issue

-13' discussed with-the Staff and I am not the person who was the 14 lead on that.

_(,,/ 15 MR.-McINTYRE: Brian McIntyre, Westinghouse.

16 We have what we call hot leg level but it actually 17 measures from the bottom of the hot leg-to about the reactor 18 vessel flange level.

19 DR. WALLIS: So it does measure level?

20 MR. McINTYRE: Yes.

21 -DR. WALLIS: Okay, thank you.

22 MR. CARROLL: And the measurement principle is 23 .thermocouples.

24 MR. McINTYRE: No, it is BP.

25 MR. CARROLL: Oh, oh, oh.

i

() '

ANN RILEY & ASSOCIATES, LTD.

Court Reporters 1250 I Street, N.W., Suite 300 Washington, D.C. 20005 (202) 842-0034

213 1 DR. MILLER: It's not the same as you would in a

( ) 2 BWR. It's not inside the vessel.

3 MR. McIN7TRE: It measures the level from the 4 bottom of the hot leg to about the reactor vessel flange in 5 the system, which includes the vessel. In that region we 6 have an indication of level.

7 MR. CARROLL: And the flange is above --

8 MR. McINTYRE: The flange is -- yes, there's about 9 7 feet above the hot leg.

10 DR. WALLIS: Well, what would be the measurement 11 of uncovering the fuel if that happened? *fhat would be the 12 symptom or the measurement which would tell you that it 13 started to uncover?

14 MR. McINTYRE: You would know when the level fell

) 15 below the level of the hot leg from several instruments.

16 DR. WALLIS: That's not down to the fuel yet.

17 MR. McINTYRE: No, you still have some time to get 18 to the fuel. and after that it is the core active 19 thermocouples that tell you when you start to see --

20 DR. WALLIS: Do you have to use the thermocouples 21 to tell what is going on there?

22 MR. McINTYRE: That's right. I mean when you are 23 getting into a degraded cooling situation, that would be 24 your primary indicator.

25 DR. MILLER: You can also get information from

(/N- w) ANN RILEY & ASSOCIATES, LTD.

Court Reporters 1250 I Street, N.W., Suite 300 Nashington, D.C. 20005 (202) 842-0034

214 1-- your ex-core nuclear instrumentation.

() 2 3

DR, POWERS _ . That's hard to do, by the way --

that's hard to do, to use the ex-core nuclear 4 instrumentation to decide what the water level is.

5 DR. MILLER:_ But~you can tell -- you can't-tell

'6_ exactly what the water level is -- when the core becomes --

7- CHAIRMAN BARTON: You can tell it's pretty low.

8 DR. POWERS:

It's very low, yes.

9 DR. MILLER: If you go through the TMI scenario, 10- if somebody had understood the ex-core instrumentation, they 11 would have known the core was uncovered, 12 DR. POWERS: I think I went through that analysis.

13 I was there when that analysis was being done and it took 14 people literally-days to do-that.

() 15- It is'not a prompt indicator.

16- DR. SEALE; No.

zl7 DR. MILLER: If you understand what-it's telling

18 -you, it is.

19 I:think operators now are partly trained to look 20- at the ex-core.

21 DR. SEALE: Moving right'along --

22 MR. KERCH: Other criteria we used to pick this 23- -minimum list -- those controls, displays and alarms required ,

24, to perform critical manual actions as identified from the 25 PRA.

O' ANN RILEY & ASSOCIATES, LTD.

Court Reporters 1250 I Street, N.W., Suite 300 Washington, D.C. 20005 (202) 842-0034

215 1 This is coming out of Element 6, which is_the

() 2 3

integration of human reliability analysis-with the HFE program where we identify critical human or manual actions 4 and/or risk important tasks.

5 Here if there is anything that we are saying meets 6- the critical definition, then it is those controls, displays 7 and alarms needed to perform those actions are part of the 8 minimum inventory.

9_ Also, those alarms used in performing safety 10- functions to respond to design basis events for which there 11 is no automatically actuated safety function, any of these 12 alarms would be part f

'he minimum inventory.

13 Finally, sort of a catch-all criteria, those 14 controls, displays and alarms necessary to maintain the-() 15 critical safety functions-and safe shutdown conditions.

16- DR. WALLIS: As an event proceeds, the priority 17 -with which the operator needs information changes. Certain 18 things are not important at the beginning and become very.

19 important later on.

=20 Is there some sort of prioritization which 21- emphasizes things which are important at the time more-than

-22 all the other-information?

.23 MR. KERCH: We hope to use our operational  !

24 sequence analysis to identify that and then use the results 25 of that in identifying those important things at certain O ANN RILEY & ASSOCIATES, LTD.

Court Reporters 1250 I Street, N.W., Suite 300 Washington, D.C. 20005 (202) 842-0034 l

216 1 times to help us in our display set design and how you

-( ) 2 navigate to the display you need at the moment.

3 DR. UHRIG: Would that also be incorporated in the 4 emergency operating procedures, that information?

5 MR. KERCH: Yes. Sure. It's there, I mean as you 6 progress down through the optimal recovery guidelines within 7 the ERGS and therefore the emergency procedures tells you 8 what you should be looking at and to assess that step or 9 execute that step.

10 DR. WALLIS: It tells you what you should be 11 'looking at but I was wondering if it shouldn't sort of 12- emphasize them, either they are bigger on the screen or 13 they're something -- that they are in front of you and not 14 to the side or whatever.

(O,1 15 You_have to yourself prioritize if they are all 16 still coming at you.

17 MR. KERCH: We want to set up our navigation 18 scheme so that what he needs at that moment is right in 19 front of him where he can access it within a one-step 20 selection and navigation action.

21 MR. CARROLL: Would you say that is an improvement 22 over old-style control rooms?

23 MR. KERCH: I think you can bring it all up right 24 in front of you.

25 You can bring up the display that presents the

() ANN RILEY & ASSOCIATES, LTD.

Court Reporters 1250 I Street, N.W., Suite 300 Washington, D.C. 20005 (202) 842-0034

217 1- information to the operator and you can bring up right

() 2. adjacent to that the soft controls that go along with 3 executing, taking any controls that are associated with that 4 display you just brought up, whether it is a sequence of 5 steps and associated controls or whether it is a physical 6 display of a given system, and the soft controls that go 7 along with that system.

8 DR. POWERS: With the numerous options that an 9 operator has to display things, various things, and look at 10 them in ever greater detail, have we looked at incidents 11 that I know for sure have occurred in Department of Energy 12 reactors and I believe occurred recently at Zion, where 13  : operators get so enmeshed in all this thicket of paging 14 controls.and menus and what-not that they do neglect the big O)

(m 15~ picture?

16 MR. KERCH: That is an identified issue that came 17 out of the operating experience review and that is what we

-18 hope to use our wall panel displays to help address.

.19 DR. POWERS: Yes, but how do'you get him to look.

20. up from the CRT with all these devices created by Mr. Gates 21 to look at your wall panel?

22 I mean he didn't at Zion so why should do it here?

23 DR. MILLER: But he.didn't have-it at Zion --

24 ~ DR. POWERS: If you look at the inspection report 25 on the recent fuel rod motion issue at Zion you'll find ANN RILEY & A4 JCIATES, LTD.

Court Reporters 1250 I Street, N.W., Suite 300 Washington, D.C. 20005 (202) 842-0034

218 1- that ---

() 2 DR. MILLER: You are talking.aboutLthe Zion 3 insertion --

4' MR. KERCH: Yes, he is.

-5 DR. MILLER: Well, we canigo to that one. =I think 6- there's another reason for that mistake.

7- DR. POWERS:- It is an incident,that was taking, 8- place there, and I know we have had the problem at 9- Department of> Energy _ reactors which do have lots of CRTs=and

10. what-not in~the control rooms -- that people get enmeshed 11 in -- ah -- go deeper and deeper and deeper into one issue 12 while the world'is-going to hell on another.

13- MR. CARROLL: That's the shift supervisor.

-DR. POWERS: He is playing Parcheesi as I recall. ,

15 MR. KERCH: No. He should be:doing his job in- a 16 supervising role since the RO changes reactivity.

17- MR. CARROLL: Looking at the big picture ---

18 MR. KERCH: What he should be doing.

19 .MR. CARROLL: And that's where this display comes 2 0 :- -in-.

~ 21 - DR. SEALE: I know. I shook him up.

22- MR. CARROLL: I don't know.

23 DR. POWERS: But nevertheless it is an identified 24 issue --

25- MR. KERCH: Yes.

ANN RILEY & ASSOCIATES, LTD.

Court Reporters 1250 I Street, N.W., Suite 300 Washington, D.C. 20005 (202) 842-0034

f 219 l' DR. POWERS: '-- that you are going to try to 2 handle?

3 MR. KERCH: Yes.

4 DR. POWERS: But I don't think it's handled by 5; just puttint up the display,

6. RMR . KEPCH: No,--you're right. You've got-.to draw 7 his attention to that large screen display. That's there to 8 help maintain the big picture.

9- DR. SEALE: You might need one of the screens with 10 the fish hooks on it.

11 MR. CARROLL: Well, if I remember the KK6 and 7 12 design,.they.had thinga that did bring your attention to 13 off-normal conditions -- I don't know if it was flashing

/"

14 lights or what it was-at that particular-point --

(,),/' 15 MR. KERCH:- The results of applying these five 16 criteria are presented in SSAR Table 18.12,'2-1, and I have 17 that_as a handout and put them up there and will briefly 18 speak about how we apply these to or how we conceptualize 19 applying these to the main' control room and then also to the 20 remote shutdown workstation.

21 These are the controls, displays and alarms that 22 -we are saying are needed to achieve and maintain safe 23 -shutdown, so they therefore should be qualified and

'24 presented on qualified hardware and software, so these, as

25- far as applying these to the main control room, the displays ANN RILEY & ASSOCIATES, LTD.

Court Reporters 1250 I Street, N.W., Suite 300 Washington, D.C. 20005 (202) 842-0034

220 1 or the parameters that are listed here would be included in llh 2 the qualified data processing system displays, which are 3 your qualified Class 1-E displays coming out of the PSM or 4 Protection and Safety Monitoring System and found on the 5 dedicated safety panel, which is the one Class 1-E panel.

6 The dedicated fixed position controls are also on 7 the dedicated safety panel. Now again these are manual 8 backups to the automatic Protection end Safety Monitoring 9 System signals.

10 DR. POWERS: Why is it that containment hydrogen 11 is on display but not on alarm?

12 MR. KERCH: Is on --

13 DR. POWERS: Display but not on alarm.

14 MR. CARROLL: Previous slide.

15 MR. KERCH: Why is it displayed and not alarmed?

16 DR. POWERS: Yes.

17 MR. KERCH: The way we went through and selected 18 what should be alarme' was looked at parameters that are 19 needed to be assessed by the operator when he is evaluating 20 the status of the critical safety functions, so the one that 21 there is a critical safety function, if my memory serves me 22 right, on containment, containment environment, and I would 23 have to confirm this because I don't have a set of ERGS in 24 front of me, but I would assume that the reason that it's 25 not there in alarm is that it is not called out in that CSF, ll ANN RILEY & ASSOCIATES, LTD.

Court Reporters 1250 I Street, N.W., Suite 300 Washington, D.C. 20005 (202) 842-0034

221 p 1 although-I am not 100 percent sure of that. -

I have to go s

() ' 2_ .back and double-check.

3 'DR. POWERS: For design basis accidents, don't you 4 .use hydrogen igniters-to get rid of hydrogen-in the 5 containment?-

6 MR. McIN7TRE - No. For the design basis we have 7 theEPARS._

8 DR. POWERS: You are-using PARS for the DBA?

9 MR. McINTTRE: PARS forJDBA and igniters for 10 severe accidents.

11 DR. POWERS: Nevertheless, I think my point'still 12- stands.

13 Do you really want to turn igniters on with-

~14 hydrogen exceeding combustion limits?

15 MR. VIJUK: I don't know the answer but -- and I 16- am-a little surprised myself that there is not an. alarm on 17 that, but -- and I don't know what the rules were behind-i 18 it -- but we do have-the igniters and they are in the 19  : procedures.

20 I guess -- they may be in the next, may be in the-21 severe accident guidelines,-which we have an interface here-22 between the emergency response guidelines takes the operator

-23 as=far as it can take him until he gets an indication that 24 he is getting a degraded core.

-25 There is nothing he can do about containment

( ) ANN RILEY & ASSOCIATES, LTD.

Court Reporters 1250 I Street, N.W., Suite 300 Washington, D.C. 20005 (202) 842-0034

222 y 1 -

hydrogen or ncthing he needs to do until he is in a severe

() 2. accident, so I am guessing that the rationale for not-having 3 an alarm on that .s i that he doesn't have anything to do 4 about containment hydrogen until he knows he has got core 5 -exit temperatures above 1200 degrees, and he may be seeing 6 hydrogen in the containment because they would be picked up 7 in the severe accident guidelines, which he would phase into 8 from the emergency guidelines.

9 DR. POWERS: I think we have found an effective =

10 way to terminate a severe accident. Turn the igniters on 11 with 15 percent hydrogen in that containment.

12 MR. CARROLL: They I think they would stop being 13 igniters and become detonators.

14 DR. POWERS: So one nice thing about it is we '

15- could analyze the consequence easily ---puff release.

16- CHAIRMAN BARTON: Go ahead, Steve.

17 MR. DEUTSCH: Steve, this is Ken Deutsch.

18 Is it true that the containment hydrogen 19 concentration doesntt meet that criteria that you 20 established earlier? We have that on our diverse actuation 21 system displays along with the igniter controls.

22 MR. KERCH: You mean the indication of the 23 concentration itself?

24 MR. DEUTSCH: Yes.

25 MR. KERCH: Well, all we're saying -- that it ANN RILEY & ASSOCIATES, LTD.

Court Reporters 1250 I Street, N.W., Suite 300 Washington, D.C. 20005 (202) 842-0034

223 1- should-be displayed,-the hydrogen concentration should be-O

([ 2 displayed as part-of-the minimum inventory set, that's 3- correct.

4 MR. DEUTSCH:_ -Yes, but.that might be the reason we 5 don't show an alarm for that.

6 MR. KERCH: Because we are implementing it on the 7 DAS?

8 I am not sure. I would have to go back and look.

9 .I don't remember on that-particular on.i. I'll find out.

10 Okay, that's it.

11 MR. KERCH: :The only difference is that the remote 12 shutdown wor:: station we have the same minimum inventory set, 13 only_now the workstation itself is a non-1E workstation, 14 whereas in the main control-room this is at a qualified i

s) 15- dedicated 1-E safety panel.

1 16 At the-remote shutdown, it's a non-1E workstation 17 and the controls are not fixed position dedicated switches.

18 They would be soft controls.

j

-19 MR. CARROLL: They're what?

20 MR. KERCH: They;would be soft controls.

-21 MR. CARROLL: Okay. Now how do-you deal with the 22 old issue of activating the remote shutdown station and 23 whether-somebody could go to it and do things in a normal-24 ope;ation that shouldn; be done and all that kind of stuf f ?

25. MR. KERCH: Well, there would be again a transfer

() ANN RILEY & ASSOCIATES, LTD.

Court Reporters 1250 I Street, N.W., Suite 300 Washington, D.C. 20005 (202) 842-0034

224 1 of control from the main control room workstations to the

() 2 remote shutdown workstation and appropriate switches to do 3 that.

4 I forget where they are located at. Ken, do you 5- recall where the transfer switches would be located at to 6 control -- to transfer control from the main control room to 7 the remote shutdown?

8 MR. DEUTSCH: This is Ken Deutsch again.

9 I have a slide that shows how we implement that 10 transfer in the digital I&C. I think the physical location 11 of the switches is in a hallway, very close to the remote 12 shutdown workstation but not physically at the remote 13 shutdown workstation.

14 MR. KERCH: And if someone were to actuate those

/"

(,,N) 15 switches, first of al?, the control and access codes would 16 be administratively controlled and if someone were to do it, 17 you would get an appropriate alarm in the main control room 18 telling you that this has occurred and then you can take 19 appropriate action from there.

20 MR. CARROLL: But they've taken over your reactor.

21 MR. KERCH: Right. I think that concludes the 22 presentation.

23 CHAIRMAN BARTON: Any other questions or comments.

24. committee members or Staff?

25 (No response.]

Ih ANN RILEY & ASSOCIATES, LTD.

5l Court Reporters 1250 I Street, N.W., Suite 300 Washington, D.C. 20005 (202) 842-0034

_ _ _ _ _ _ _ _ _ b

! 225 l

1 CHAIRMAN BARTON: That concludes the review of I( ) 2 Chapter 18. We will now recess until cne o' clock. '

3 (Whereupon, at 11:59 a.m., the meeting was 4 recessed, to reconvene at 1:00 p.m., this same day.]

5 6

7 8

9 10 11 12 13 14 15 16 17

-18 19 20 21 22 23 24 25

.('

ANN RILEY & ASSOCIATES, LTD.

Court Reporters 1250 I Street, N.W., Suite 300 Washington, D.C. 20005 (202) 842-0034

. . . __ _ __-- J

l 226 l- 1 AFTERNOON SESSION (x 2 (1:33 p.m.)

3 CHAIRMAN BARTON: Come back to order.

4 We're at agenda item 8 on the schedule, 5 Westinghouse presentation of chapter 7, Instrument and

-6 Control. Before we hear from Westinghouse on the-instrument

7. and control system, Don Miller would like to make some 8 comments.

9 DR. MILLER: .Yes, just briefly, I prepared some 10 what I call questions and comments that each Oae of you 11 should have gotten the last few days. Westinghouse also has 12 a copy of those. Many of those are for my own clarification 13 or maybe for yours, if you're interested.

A 14 I think my concerns focus primarily on the nuclear

(,) 15 instrumentation, partly clarification, partly lack of l'6 understanding of what they're trying to do, and partly of-17 innate belief that the use of fission chambers is always 18 superior to the use of boron-type instruments, but that's an '

19 inbred thing that I have based on the fundamentals of Mother 20 Nature.

21 The other comment I'd like to make is yesterday we 22 heard an overview of the I&C system, high-level, of course, 23_ and the comment was made that a major change in a system is 24 use of digital technology instead of analog. I believe of 25 equal significance is the wide and broad use of fiber optics b)'

\' ANN RILEY & ASSOCIATES, LTD.

Court Reporters 1250 I Street, N.W., Suite 300 Washington, D.C. 20005 (202) 842-0034 I

227 1 in an optical-based technology throughout the I&C system. I

() 2 think that technology is of equal significance in the 3 advancement of_this particular I&C system. I think that was 4 also used in the CE to sone -- C2 system maybe to a lesser 5 extent in ABWR. But I think thLy both have -- it has a lot 6 of advantages, and I whipped out a little bit of a handout, 7 so to speak, to give you my views on why I think it's 8 important.

9 With that I think I'll just turn it back to John 10 and let'him start the meeting. But I thought I'd make those 11 comments before wo get started.

12 - CRAIRMAN BARTON: Thank-you, Don.

13 Brian, a,e you ready for chapter ??

14 MR. McINTYRE: Yes. The presentation will be made

/ 15 by Ken Deutsch from our I&C group.

16 MR, DEUTSCH: Can you hear me all right?

17' I'm going to discuss the material that we've 18 presented in the SSAR chapter 7, the I&C information for 19 AP600.

20 For hP600 the I&C systems are based on the use of 21 digital microprocessor-based technology, and by doing this 22- we can split the system up both functionally and physically 23 and distribute it throughout the plant, und this gives us a 24 lot of advantages over the traditional analog systems.

25' When I talk about --

O ANN RILEY & ASSOCIATES, LTD.

Court Peporters 1250 I Street, N.W., Suite 30G Washington, D.C. 20005 (202) 842-0034

l l

228 1 MR. CARROLL: What disadvantages?

That was for

() 2 you, Lieutenant Lud.

3 [ Laughter.)

4 MR. DEUTSCH: I honee'ly can't think of any '

-5 particular disadvantages.

6 DR. POWERS: I can think of several right off the 7 bat. They're a little more sensitive to acidic smokes and j 8 particulate. EMI/RFI a little bit. i l 9 MR. DEUTSCH: I actually think digital systems are

• l 10 probably less susceptible to EMI/RFI, because an analog ,

11 system is more susceptible to noise-in general. A digital l 12 system.you can have algorithms to filter those things out.-

13 Generally these digital systems operate on a much higher t 24 frequency than the noise, EMI/RFI sources, and so those are

-15 issues that have to be-addressed in the design of the l 16 digital nystem, but I don't-think that that's a particular -

. 17 disadvantage.

~18 MR. CARROLL: How about smoke?

19 MR. DEUTSCH

Smoke?- I am not aware of any 20 particular susceptibility to smoke. Certainly when you get

. 21- into an area of a digital system and ycu're making things smaller and you're putting more computing power onto a. piece l 23 of hardware and your tray sizes get smaller, then the leads

'24_ get closer together,-any kind of particulate matter could 25 have a greater effect. So you have to consider those

, ANN RILEY & ASSOCIATES, LTD.

Court Reporters-1250 I Street, N.W., Suite 300 Washington, D.C. 20005 '

(202) 842-0034 7

--.-._-.-- - - ..- - . . ~ - - . -- , - - -.

1 229  ;

i '

1 things, and you use conformal coating on'.x rds, qualify the

() 2 equipment that needs to work in a harsh onvironment, and-3 those things can be addressed.  !

4 D;R . POWERS: What is the qualification test for [

5 smoke-for digita2 systems?

• 6 MR. DEUTSCH: I'm not aware of any particular 7 standard. We qualify our I&C systenis, the safety-related 8 I&C systems, for a harsh environment. I think that the I&C 9 community is aware of issues relating to environment. I'm 10 not eware of any particular issues relating to smoke. But

• 11 those are factors that can be, you know, particulate 12 certainly. But by conformal coating the boards and 13 environmentally placing the devices, that's a problem that .

14 can be managed.

()

[

15 DR. PONERS: So what you're telling me is that the 16 design approach does not include an exaniination of those 17 things that might be downsides to these digital

~

[ 18 microprocessor-based technologies.

19 .MR. DEUTSCH: Well, I'm not sure I would agree h 20 with that. I mean, one of the bis issues that we addressed 21 on AP600 is the effects due to fire, and generally when we t

22 had fire areas we considered the most undesirable effect  ;

4 23. that could happen. So we basically were forced during that

24. kind of evaluation to assume our digital systems would fail, 25 and --

ANN RILEY & ASSOCIATES, LTD.

Court Reporters 1250 I Street, N.W ,. Suite 300 Washington, D.C. 20005 (202) 842-0034

- - . - _ -- . . . - = - ._. . _ - . _ _ . . _ . . . - . _ - , . . - . - - - - -.-

230 1 MR. CARROLL: Fail, or produce unwanted actions.

() 2. MR. DEUTSCH: It depends on the particular

-3 situation.

4 MR. CARROLL: But you did recognize that --

5 MR. DEUTSCH: Yes.

6 MR. CARROLL: Unwanted actions --

7 MR. DEUTSCH: Yes.

8 MR. CARROLL: Could result from --

9 MR. DEUTSCH: Yes. Generally if it was a simple 10 link we considered an undesirable action was potentially 11 possible. If it was a data link interface between 12 something, we recognized that there were no simple failure 13 modes that would cause this system-to communicate data that 14 would cause first actuation. So we did consider those

() 15 . things in our fire evaluation, and came up with a story to 16 handle all that, due to the location of the equipment, our l 17 ability to deenergize portions of the system, and things 18 like that. So that was considered.

19 MR. CARROLL: There's a classic example of smoke 20 interacting with digital I&c. I don't remember all the 21 details, but it occurred in a large machine shop that used 22 computer-controlled equipment, and I guess there's been some 23 in telephone-awitching centers too, and it's very subtle.

24- I mean, you pu-the fire out, you say well, I.

25 guess everything's okay, and you atart using.the equipment, f ANN RILEY & ASSOCIATES, LTD.

Court Reporters 1250 I Street, N.W., Suite 300 Washington, D.C. 20005 (202) 842-0034

231 1 and it gets squirrelier and squirrelier as time _goes on 2 because particulate matter is bridging over close 3 tolerances. So I'm a little surprised -- ACRS has been 4 yelling about smoke for --

5 DR. MILLER: Oak Ridge has completed a study along 6- with -- well, Sandia and Oak-Ridge together.

7 MR.-CARROLL: I hope it's better than the one thay 8 presented originally.

9 DR. MILLER: It's in a NUREG that you might want 10 to take a look at.

11 DR. UHRIG: Is it a draft NUREG or is it final?

12 DR. MILLER: I think it's final. The one I have 13 is -- it does address the issues of bridging-and so on. And 14 I don't know if the staff when they reviewed AP600 brought 15 that issue.up or not.

16 MR. CARROLL: Did they?

17 MR. DEUTSCH: They brought up the general 18 consideration of fire and qualification. They brought up.

19- EMI/RFI. I can't think that we had any extensive discussion 20 of-smoke.

21 MR. CARROLL: Does the staff have anything to add 22 _to that?.

23 MR. CHIRAMA: This is Matt Chirama or NRR, HICB.

24 No, we did not address smoke particularly because we had

25 ongoing study-going on at that time-in'research which we ANN RILEY & ASSOCIATES, LTD.

Court Reporters 1250 I Street, N.W., Suite 300 Washington, D.C. 20005 (202) 842-0034

232  !

I were looking that when the reg guide was -- i

()

~

2 DR. UHRIO: Will it be revisited? Will this issue 3 be revisited?

4 MR. CHIRAMA: Yes, when the reg guide comes out 5 we'll consider that.

6 MR. CARROLL: Well, wait a minute now. If you 1 cert $ fy 6.%: design, you're not going to be able to revisit

• 8 it for this particular --

9 MR. CHIRAMA: Well, from what I gather, the reg j 10 guide is not going to address smoke because it doesn't turn 11 out to be a big contributor to failures. As Ken mentioned, 12 the coating provided by the microprocessor chips and things 13 like that provide sufficient protection against smoke.

) 14 That's the outcome of the study.

15 DR. SEALE: And what we don't know won't hurt us?

16 DR. POWERS: When you say it provides sufficient 17 protection against smoke, I mean clearly I can define -- I

, 18' can design. smokes that nothing can protect against. So-

_ 19 there must be-come standard --

H2 0 . MR. CHIRAMA: They looked at the possible 21 . contributors from fires that occur in nuclear powerplants.

22 That's what the study focused on.

23 DR. POWERS: And you concluded that --

?-

24 MR. CHIRAMA: I don't know the final outcome of 25 the reg guide as-yet, so obviously research still is working ANN RILEY & ASSOCIATES,-LTD.

Court Reporters 1250 I Street, N.W.,-Suite 300 Washington, D.C. 20005 (202) 842-0034 4

wu..w,..-. ,y ,,,,~_v.--_._. .,.--my.,

,,y.e-. ~.,,y., _,...,_...,y__- ,, ,.._.,me--,_,-,__,,e- . . . , - . , . ,

233 1 on it.

() 2- DR. MILLER: The conclusion of that report was 3 that -- I think Matt was - my recollection was consistent 4 with what he just said, that smoke was somewhat less of a 5 stressor than others, including EMI/RFI and things like 6- lightning and so forth, 7 MR. CARROLL: So.

8 DR. SEALE: Yes, but Don, that's like saying that 9 diphtheria won't kill you unless you' catch it. I mean, if 10 you have a fire, your stressor is smoke.

11 MR. CARROLL: -Well, it can also be heat.

12 DR. SEALE: True. But it's not lightning, not 13 EMI, it's smoke. And so you really need to address that 1

14 issue against the possible exposure that you'd have in a 15 control-room fire or whatever.

16 DR. MILLER: I guess I'm not going to particularly 17 defend the report, I'm just reporting the results of the 18 report. We may want to look at it, but it was based on 19 their analysis that we put it all in the PRA and so forth, 20 it's not going to be an issue compared to some of the other 21 issues. It also can be --

22 DR. POWERS: They tell me they put these things in 23 the PRA. I turn around and say well, how did you put it in

-24 the PRA? I mean, I haven't got a clue how you do failure 25 probabilities on smoke. I just don't know how it would be

() ANN RILEY & ASSOCIATES, LTD.

Court Reporters 1250 I Street, N.W., Suite 300 Washington, D.C. 20005 (202) 842-0034

- .- _ .m. _ _ ___ _ . . _ _ . . _ _ . _ . _ . _ . _ _ _ _ _ _ _ _ _ _ _ _ . . . _ . _ . .

l 234 >

1 done. I don't know how'you handle -- the more difficult I

() 2 problem with smoke is that I have a fire here, and 3 frequently I have a fire here in an area that has no safety [

4 systems in it. That's where my fuel loads tend to be. It

, 5 produces smoke. Smoke goes up. -It's liable to go up into

.6- an area that has digital systems in it, and I may-not see 7 the results for a month after the fire is over. How do-I 8 handle that in a PRA?- It's a technology I don't know how to 9 do.

~

10 DR. MILLER: I don't know if -- Matt, can you help '

11 us out on those questions, or is anybody from research here 12 who maybe --

13 MR. CHIRAMA: Well, I guess physical separation 14 does buy you a lot in fire protection and -- fire and smoke

() 15 protection. One channel may be affected, but the other 16 threeLchannels will be okay in general.

'17 MR.-DEUTSCH: I think the biggest danger of smoke 11 8 for an I&C system in a smoke environment is not considering 19' the possible effects, and like I said, although I am not-20 -very aware of -- from the' hardware end what particular tests -

21 have been done on the equipment due to smoke, 22 I know that from a design end that we have 23 considered fires in the I&C rooms, fires in the control 24 room,.and we've looked at the I&C design and the overall 25 plant design and approach of our architecture with respect

) ANN RILEY & ASSOCIATES, LTD.

Court Reporters 1250 I Street, N.W., Suite 300 Washington, D.C. 20005-(202) 842-0034

_ _ - , _ . _ . - , , _ . . _ . . ~ . - , _ . _ . - , - _ _.-.. _ . _ _ _ . _ . _ _ _ . _ _ . _ _ _ - - . __. ._ -

l L 235 1 to these particular things, and things like by physically

() 2 distributing the system throughout the plant you do get some 3 benefit than if you had everything in the same room and you 4 had a smoke environment. So we do get some benefits.

5 We have considered adverse effects due to fires in 6 the I&C system. But I think they also use digital 7 technology in fire protection systems. I think you have to 8 design your equipment for the environment and you'need to 9 consider these environments, and if they're likely, you have 10 to consider those.

11 DR. APOSTOLAKIS: Isn' t it true, though, that at 12 this time we really don't know how handle the possibility 1 31 that the I&C system will give wrong information to the 14 opet rs?

) 15 MR. DEUTSCH: Well, from a safety-related

' perspective I think we would argue that we have the benefit

'17 of separation, so if you postulate a fire of smoke 18 environment, that that may affect, will.atfect a particular 19- division of our I&C systems, but --

'20 CRAIRMAN BARTON: Except if the fire's in the 21 . control room, where they.all come together.

22. MR DEUTECH: In the control room then we would

.23 evacuate the control room and go to the remote shutdown 24 roum.

25 DR. POWERS: Any fire causes you to evacuate the O ANN RILEY & ASSOCIATES,- LTD.

Court Reporters 1250 I Street, N.W., Suite 300 2 84 I b3

l 236 1 control room? Wastepaper fire causes you to --

() 2 MR. DEUTSCH: No. So you're talking about a small 3 fire that would not cause an evacuation of the control room 4 but would cause some data to appear on a CRT that would be 5 correlated with other data and would -- and might cause the 6 operator to respond incorrectly.

7 That would have to be looked at, but I think it is 8 a relatively low probability. I think that you could have a 9 fire in the main control room and your CRT goes out, your 10 CRT screen goes -- your data qualities go bad on a 11 particular screen, and you clearly know you have an issue, a 12 problem in the I&C system.

13 But I think we'are really concerned with problems 14 caused by the smoke that go unnoticed by the operator,.and I

() 15 think those are probably pretty small probability. Because 16' even though the fire is in the main control room, the CRT is 17 in the control room, the signals coming in to the computer 18 system, the computational portion of the computer system is 19 not in the main control' room. Effectively, all the control l

-20 room then is getting information from outside the control 21 room and processing it and putting it on a display. So --

22 DR. MILLER: You use fiber optics -- early on, I 23 said you'used fiber optics widely. You use fiber optics to 24 do the data transmission from those systems into the control 25 room, right?

-=

ANN RILEY &' ASSOCIATES, LTD.

Court Reporters 1250 I Street, N.W., Suite 300 Washington,.D.C. 20005 (202) 842-0034 w

237 1 MR. DEUTSCH: I can't say that we solely use fiber

() 2 optics. We rely on fiber optics a lot. When we need to 3 transmit a lot of data over a long distance, we use fiber 4 optics to provide us electric separation --

5 DR. MILLER Right.

6 MR. DEUTSCH: ---in our 1E equipment. But-in some 7

cases, we-can do the job _just as well, and less expensively, 8 using a copper connection, and then we would do that.

9 DR. MILLER: Well, one of the questions I had, was 30 there any qualification program for the use of fiber optics?

11 Which could include smoke and fire.

12 MR. DEUTSCH: There's -- there are none that I am 13 aware of. I know there's been issues of -- issues raised 14 relating to a long-tenn degradation of fiber optics,

() 15 radiation induced problems in fiber optics.

16 DR. MILLER: Well, you were careful to, at least 17 what I read, not put the fiber optics anyplace where you 18 would expect radiation.

19 MR. DEUTSCH = Yes.

20 DR. MILLER: Which I think is, certainly with 21 today's knowledge, is prudent.

22 MR DEUTSCH: Uh-huh.

l 23 DR. MILLER: But the question is other stressors 24 such as -- I don't know about smoke and fiber optics.

25 Certainly fire could be --

=

( ANN RILEY & ASSCCIATCS, LTD.

Court Reporters 1250 I Street, N.W., Suice 300 Washington, D.C. 20005 (202) 842-0034

l 1

238 1 MR. DEUTSCH: Yeah, I would think smoke would be a

() 2 stressor in fiber optics, because they are generally 3 jacketed, and the light is confined inside the fiber.

4 At some point, heat of the same kind of magnitude 5 that would cause insulation to file on a fiber conductor 6 would probably cause the fiber to fail, and you would lose

-7 your ability to transmit' data over that particular data v

8 link. But it sort of goes away -- it would be very 9 difficult-to postulate a spurious actuation caused by that 4

10 kind of failure.

11 DR. MILLER: I agree.

12 MR. DEUTSCHt And part of the advantage of the 13 -advanced I&C is we can distribute these functions physically 14 and we can also divide the functions functionally so that 15 particular subsystems are performing particular calculations 16 and we are not cramming everything into one space, into one 17 big mainframe computer.

18 MR. CARROLL: What's the down side of that?

19 MR. DEUTSCH: The down-side of?

20 MF. CARROLL:_ Distributing the system all over the 21 place.

22 'MR. DEUTSCH: I guess from a maintenance 23 standpoint, you have to go to more locations or something 24 like that. But-for these particular -- for this particular 25' application, the advantages are felt to far outweigh the

() ANN RILEY

• ASSOCIATES, LTD.

Court Reporters 1250 I Street, N.W., Suite 300 Washington, D.C. 20005 (202) 842-0034

239 1 disadvantages. This is the way the industry is going.

() 2 DR. APOSTOLAKIS: Now, all this analysis will be 3 found someplace? I mean the PRA looks at these things and 4 concludes that=the physical separation of the channels makes 5 failures of very low probability events?

-6 MR. DEUTSCH: These --

7 DR. APOSTOLAKIS: Or are you improvising right 8 now?

9 MR. DSUTSCH: No. The PRA considered the -- the 10 attributes of the I&C system in considering the thin; J it 11 considered.

12 Tom, can you help me here, what we did with the 13 fire?-

14 DR. APOSTOLAKIS: No, no. The failures of I&C.

O

\s ,/ 15 Not just fire.

16 MR. DEUTSCH: Oh, failures of I&C.

17 DR. APOSTOLAKIS: Not just fire-related. Fire 18 could be one of the --

19 MR.-DEUTSCH: Well, our I&C design-is built to the 20 regulations, the list of standards that we have said. So, 21 we are -- in Chapter 7 we present an architecture and an 22= approach to an I&C design. And even though we have 23 implemented many of these elements in plants, we are not 24 coming _here with a product and say this is -- this is the 25 product that.we will use.

O ANN RILEY & ASSOCIATES, LTD.

Court _ Reporters 1250 I-Street, N.W., Suite 300 Washington, D.C. 20005 (202) ~ 842-0034 l.

240 1 DR. APOSTOLAKIS: No, I understand that you comply

() 2 with the regulations. But you gave some answers to the 3 questions that were raised, and you said, you know, this can 4 not happen because of separation. In the control room, you 5 know, if there is a big fire, they will move out. If it is 6 a-smaller fire, then computations are somewhere else. All 7 these argumento are in the PRA?

8 MR. DEUTSCH: For --

9 MR. VIJUK - I can help you out here maybe, Ken.

10 MR. DEUTSCH: Yeah.

11 MR. VIJUK: Ron vijuk, Westinghouse. They are in 12 there, I believe, in the sense that, as we talked earlier, 13 we have four divisions of I&C and they are in separate fire 14 areas. And so when they look at a probability of failure

() 15 and they assume one division fails, because the other three 16 are independent, and you only need two to-accomplish a 17 safety function, now that is success in PRA space. So, to 18 that extent at least, these things are considered in PRA.

19 MR. CARROLL: Now, when you talk about four 20 separate systems from the fire point of view, how about from 21 a smoke point of view?

22 MR. VIJUK: I don't know explicitly how they 23 factored that. I would be surprised if they do anything 24 explicit with smoke. I think it is -- if you have a fire

.25 that can propagate from one fire area to another, because it ANN RILEY- ASSOCIATES, LTD.

Court Reporters 1250 ILStreet, N.W., Suite 300 Washington, D.C. 20005 (202) 842-0034

-241 1 is -- because there is nothing to prevent smoke from going

() 2 there, we would assume anything in that area fails as well.

3 The -- when I say they are in separate fire areas, 4 there's fire dampers that close and smoke doesn't propagate 5- in rooms and things like that.

6 DR. POWERS: You assume --

7 MR. VIJUK: And the same way, a fire in a 8 non-safety area can't affect the safety area.

9 DR. POWERS: You assume that --

10 MR. VIJUK: Or has a low probability.

11- DR. POWERS: -- failure probabilities of fire 12 penetration barriers are all zero?

13 MR. VIJUK: No , I am sure they have nome 14 probability on that, and depending on particular barriers.

() 15 We have=three hour barriers, we have one hour barriers. I 16 don't -- I can't swear that they factor all that into the 17 PRA, but I know they could if they -- if they get that 18 rigorous.

19 DR. POWERS: We will -- we hear about that in a 20- different chapter, right?

21 MR VIJUK: I-guess you will cover the PRA at some 22 point.

23 MR. CARROLL: Yeah,.but when we talk about smoke 24 in that future chapter, let's understand that one hour and 25 three hour fire doors are not smoke barriers, necessarily.

() ANN RILEY & ASSOCIATES, LTD.

Court Reporters 1250 I Street, N.W., Suite 300 Washington, D.C. 20005 (202) 842-0034

__ 9

4 242 l 1 MR. VIJUK: Okay. Yeah, I don't know how they 2 cover that in the PRA.

3 DR. APOSTOLAY.IS: Let's come back to the I&C. Is 4 it true that the principal defense against undesirable 5 events here is the physical separation?

6 MR. DEUTSCH: Yes, that would be -- that is a big

] 7_ defense that we use, is the physical. separation of the 8 equipment, yes.

]

9 DR. APOSTOLAKIS: Is there any functional 10 diversity anywhere? Any -- you would use different ways of

{

11 doing things?

. 12 MR. DEUTSCH: . Yes, we use functional diversity in 13 the reactor trip system in coming -- in using different-14 parameters to respond to particular initiating events. And 15 we also use a form of functional diversity in distributing 16 the functions to particular elements of the I&C system.

17 DR. APOSTOLAKIS: But --

18 MR. DEUTSCH: I will be showing some of that as we 19 go through.

20 DR. APOSTOLAKIS: But you did not, do though, or 21- you did not attempt to apply any of the methods that, 22 admittedly, are still evolving to - to see -- to do a more 23 formal failure analysis? You just trust that by taking 24 these measures, this is a.high reliability system?

25 MR. DEUTSCH: Right. We rely on the physical ANN RILEY & ASSOCIATES, LTD.

-Court Reporters 1250 I Street, N.W., Suite 300 Washington, D.C. 20005 (202) 84' -0034

243 1- separation and, typically, that kind of evaluation is done 2 as part of the PRA.

3 DR. APOSTOLAKIS: Now, again, this may be a dumb 4 question. Does a system like that exist now, or is it just 5~ design? Do you actually have --

6 MR. 0EUTSCH: The system that I am showing?

7 DR. APOSTOLAKIS: Yeah.

8 MR. DEUTSCH: The system that I am showing is 9 primarily based an implementation from Sizewell. It is not 10 necessarily the system that we would --

11 DR APOSTOLAKIS It is not the same.

12 MR. DEUTSCH: -- we would do in five years from 13 now.

14 DR. APOSTOLAKIS: How, the staff has prepared a 15 regulatory guide, which I am sure was issued after you guys 16 had done most of the work, that places the emphasis un

-17 controlling the process that produces the-software. You 18 have all sorts of requirements and so on.

19 MR. DEUTSCH: Yes.

20 DR. APOSTOLAKIS: If i took that now and compared 21- it against what you did, would I see that you came close to 22 that?

23 MR. DEUTSCH: I believe so. The staff and

'24- Westinghouse had a -- had a lot of interaction and exchange, 2d in the software design and verification process, and I j ANN RILEY(& ASSOCIATES, LTD.

Court Reporters 1250 I Street, N.W., Suite 300 Washington, D.C. 20005 (202) 842-0034

l 244 1 -think, yes, we definitely came.

() 2 DR. APOSTOLAKIS: Now, you mentioned Sizewell,-

3 there'the primary method they use to verify that this is a 4 very reliable system was tests, testing?

.5 MR. DEUTSCH: I think-that was one of the methods 15 _that was used. I think that the Sizewell software design 7 followed a very structured approach from the very beginning, 8 and I think the testing program that was done as-part of 9 Sizewell is just almost additional proof that, see, what we 10 did here worked.

11; Yeah, they -- I am not as cognizant of the whole 12 Sizewell test program, but I know it was quite extensive and 13 done by many different organizations. But that is not the 14 basis of -- that Sizewell software was verified. It was

) 15 done in a structured process from the very beginning.

16 Because Westinghouse believes that that is how you have to 17 design safety-related software, you have to start at the

.18 - beginning. That you can't have software and grandfather it 19 ih very well.

20 MR. CARROLL: But, George, you have also got to 21- recognize that, because of the policy c'he staff established 22 back in ABWR' days, that even if somathing goes squirrelly 23 with the software, we have got the diverse actuation system 241 as a backup to it.

25 DR. APOSTOLAKIS: I think that the -- in my

~

ANN RILE ~& ASSOCIA@ES, LTD.

O- Court Reporters 1250 I Street, N.W., Suite 300 Washington, D.C. 20005 (202) 842-0034

r 245 }

1 opinion, in my mind, a more scary situation.is for tha-  :

() 2 operators receive either confusing or wrong information, and

{

3 they start taking action. When things fail, I -- I don't  ;

i 4 know. It is not, of course, a desirable thing, but -- i 5 MR. CARROLL: At least you know it is -

6 DR. APOSTOLAKIS: At least you know it is not 7 working.

[

8 MR. CARROLL: Yeah.  ;

9 DR. APOSTOLAKIS: But if you get the wrong 10 information, then we really don't know what they are going >

11

~

'to do. And perhaps it is unfair to ask you to -- to do an 12 analysis of -- a detailed analysis. Some analysis is f 13 required, but I-don't know where you would even start.

14 DR. UHRIG: Is it any different than analog?

() 15 DR. APOSTOLAKIS: Pardon? i 16 DR. UHRIG: Is it any different.than analog that 17 goes bad?

18- DR APOSTOLAKIS: Yeah. -Because the failures here 19 .are discontinues.

20- DR, UHRIG: Okay. Well, an analog can be off.

21 DR. APOSTOLAKIS: ' Yeah, .And you can get crazy 22 stuff. You can -- yeah.

23 MR. CARROLL: But I think the notion is that if we 24 really start losing confidence that the digital system-is 25 functioning right, you -- you go over to the diverse ANN RILEY & ASSOCIATES, LTDi Court Reporters 1250 I Street, N.W., Suite 300 Washington,.D.C. 20005 (202) 842-0034

-_ _ - _ . . _ _ _ _ _ _ _ . . . ~ . _ _ _ . ._ _ ._ _ _ . _ . _ - . _ . . ._

246 l 1 - actuation system and go to high shutdown.

( 2 DR. APOSTOLAKIS: Yeah, Yeah. So it is really 3- . diversity -- diversity and redundancy.

4 DR. UHRIO: There are methods of detecting sensors 5 going bad.

6 MR. CARROLL: Oh, sure.

7 DR. UHRIG: And, in fact, there's even methods of 8 replacing a signal with a c rrogate signal.

9 MR. CARROLL: Yeah, I dcn't think it is sensors 10 George is so worried about as --

11 DR. UHRIG: Well, the whole channel, it's the 12 whole channel you are really testing, it isn't just the 13 sensor.

14 DR. APOSTOLAKIS: Yeah, but in all fairness,

) 15 again, this is really at the cutting edge of technology. I 16- mean people are beginning now to think about these things.

17 I mean all you see out there is some incidents mainly f om 18 the airline industry that are really strange. I don't know 19_ that anybody could have predicted or could have developed a

-20 methodology that would have predicted what actually happened 21 in some cases. Some weird things, I mean it is just that we 22 are learning now how to use this technology.

-23 So I guess using diversity in physical separation 24= is the best you can do right now, isn't it?' Although the 25 -best actually would be to try to answer these questions as ANN RILEY & ASSOCIATES, LTD.

~

Court Reporters 1250 I Street, N.W., Suite 300 Washington, D.C. 20005 (202) 842-0034

247 1 w 1. I mean at least to some extent.

() 2 DR. POWERS: I mean you re-raised the fundamental 3 question. Why not go with a proven technology rather than 4 these --

5 DR. APOSTOLAKIS: Yeah.

6- DR. POWERS:- -- things susceptible to weird, as 7 you describe it, events?

8 DR. APOSTOLAKIS: But I think they are forced to, 9 aren't they? I mean analog systems are out. Aren't you 10 forced to go to digital? It is going to be very --

11 MR. DEUTSCH: It is difficult to buy. analog 12 systems, you know. And the analog systems are proven, but 13 that is because we have come to grips with some of the 14 parameters and we know we have to deal with a drift in

-15 analog system and that is something that inherently is -- 's .

16 less of a problem in a digital system.

17 Does that make the digital system better? I mean, 18 in that particular area, yes. But we have.come to grips 19 with it.

20 DR. APOSTOLAKIS: Of course, we need to ask these 21 questions so that people can do the fire analysis.

22 MR. DEUTSCH: Yes.

23 DR. APOSTOLAKIS: And ask them to include these,

-24 whether they-included the values.

2L So, I don't know, I think we exhausted the

() -ANN RILEY & ASSOCIATES, LTD.

Court Reporters 1250 I Street, N.W., Suite 300 Washington, D.C. 20005 (202) 842-0034

248 1 subject, Mr. Chairman.

() 2_ MR. DEUTSCH: Okay. This is a functional 3 representation of the I&C elements that are discussed in 4 chapter 7. The operations and control center system is 5 primarily the things that Steve Xerch talked about earlier.

6 They are shown here because they have such a tight 7 connection with the I&C system, which really becomes an 8 #.nformation system to the operator.

-9 The data display and processing system is the 10 system that provides the hardware and the computational 11 tools to provide that information to the operator. It 12 consists of a distributed plant computer system, alarm 13 systems, procedures. The operator displays, those are the 14_ displays on the operator consoles. The large wall panel

() 15 display that -- the hardware and the computational functions '

16- is provided by the data displaying processing system.

17 The data displaying processing system also 18- prevides_what we're calling the monitor bus, which is the 19 backbone data highway that ties everything together. I'll 20 be talking about that in a little more de". ail in just a 21 minute.

22- Down below are the systems which prim e > .ly 23 interface with the plant, with the components out in the 24 plant. To the left, of course, we have the protection 25 nafety monitoring system that has three primary functions --

ANN RILEY & ASSOCIATES, LTD.

Court Reporters 1250 I Street, N.W., Suite 300 Washington, D.C. 20005 (202) 842-0034

249  ;

l i i the reactor trip,-the engineered safety features actuation, 2 and it also provides the qualified displays which the 3 operator uses. They're qualified to the harsh environments, #

4 the accident monitoring, and the minimum inventory that l 5 Steve was talking about earlier are shown on those qualified 6 displays. J 4  ;

. 7. Plant control system is where we have segmented I

8 the control of the non-safety related equipment. Where i 9 protection safety monitoring system relates to the control i l 10 of safety related components, the plant control system, 11 which obviously is much broader, relates to the control of i

12 the non-safety related equipment.

13 In-core instrumentation system we'll talk about a

]

14 little bit more. That consisted of the in-core 15 thermocouples and the fixed in-core detectors that we use 16 for flex mapping.

) 17 Special monitoring system'is -- right now consists j 18 only of the loose parts monitoring. system necessary to meet ~

{. 19 Reg-Guide 1.133, and we talked about the diverse actuation 20 system-as being a' separate system and relied upon to get-to 21 hot shutdown if we have a failure of our integrated I&C 22 . system, and-I will be talking about that in a little more 1

23 detail.  ;

24 MR. CARROLL: . Where does the instrumentation i

25. associated with mid-loop operation fit in?

4 i.

ANN RILEY-&' ASSOCIATES, LTD.

I -

Court Reporters J250.I Street, N.W., Suite 300

' Washington, D.C. 20005 (202) 842-0034

250 1 MR. DEUTSCH: That instrumentation consists of --'  !

I

() 2 MR. CARROLL: Level?

2

! 3 MR. DEUTSCH: ~ Level.

{

4 MR. VIJUK: You mean what is it classified? It  !

5 would be in PMP.

6 MR. DEUTSCH: Yes. If-the parameter that's used 7- .to monitor them, the hot leg level, for instance, is a 8 parameter-used by the protection safety monitoring system, 9 it comes into the IE system. We do have a way of

-10 communicating-information from the 1E system to the non-1E 11 system.

12 DR. UHRIG: That is broadcast only?

13 MR. DEUTSCH: Yes.

14 We talked about the monitor bus as the hackbone of

)

^

15 our highway, and it uses what's called a fiber distributed 16 data interface network, and that is a standard industry 17 physical protocol for a network, and in our design, this 18 network supports both periodic and aperiodic data transfer.

19 When I mean periodic, that's guaranteed information which is 20 transferred on a regular basis at least once a second, and 21 our design includes that periodic factor, and by doing that, 22 we can make sure that our data highway is not overloaded.

23 Our data highway can. handle over 200,000 points, 24 and those are not broadcast by exception, those are 25 broadcast every second. So by doing that, we have a design J

O ANN RILEY & ASSOCIATES, LTD.

Court Reporter" 1250 I Street, N.W.,. Suite 300 Washington, D.C. 20005 L (202) 842-0034

251 1 that works just as well normally as it does -- it works just

() 2 as.well under overload conditions during a plant transient 3 as it does normally. Aperiodic data transfer is more 4 associated with transfer of blocks of information like files 5 from one computer system to another.

6 DR. UHRIG: Is there a priority on the data 7 transmission? Do you have a hierarchial priority as to 8- whether the periodic comes before the aperiodic?

9 MR. DEUTSCH: They share bandwidth, but the 10 periodic is assigned a fixed portion of the bandwidth, so 11 nothing can interfere with that.

12 DR. UHRIG: So you don't get an overload 13 condition.

14 MR. DEUTSCH: Exactly. We can have all our points

) 15 that_are -C up to 200,000 points be changing, and that does 16 not represent an overload condition.

17 DR. UHRIG: That represents a certain fraction.

18 MR. DEUTSCH: Yes.

19 DR UHRIG: Ten percent, some --

20 MR DEUTSCH: -Yes.

21 -

DR. UHRIG: Do you know what that fraction is?

22 MR, DEUTSCH: I'm not certain. It's on the order 23' of 30 to 40 percent. On that order.

24' DR. UHRIG: So-the other is available for things 25 associated with a transient that are not normally ANN RILEY & ASSOCIATES,-LTD.

Court Reporters 1250 I. Street, N.W., Suite 300 Washington, D.C. 20005 (202) 842-0034

I 252 1 transmitted?'

2 MR. DEUTSCH: Yes.

3 DR. UHRIG: Or do you --

4 MR. DEUTSCH: That would be like associated with I 5 if the operator had requested some information that may=have 6 been archived in a historical storage-media, and he wants to .

-7~ bring that up. Do --

8 DR. UHRIG: So the transient data is the periodic 9 data. The data giving you information on the --  :

10- MR. DEUTSCH: The periodic data -- the periodic 11 data that we have listed there, that's associated with plant 12 process parameters.

13 DR. UHRIG: Many of which are important in a 14 ~ trannie,.t.

O V 15 MR. DEUTSCH: Yes.-

i 16 - DR. UHRIG: Do you have additional data that are 17 transmitted during a transient?

18 MR. DEUTSCH: No.

[ 19 DR. UHRIG: No. Okay, 20 MR. DEUTSCH: No.

21 DR. UHRIG: Then you're all right.

l 22 MR. DEUTSCH: Right.

23 Again, the topography of the monitor bus is a ring 24 topography which has the advantage that if you severe the 25 ring by cutting completely through it, the node ANN RILEY & ASSOCIATES, LTD.

Court Reporters 1250 I Street, N.W., Suite 300 Washington, D.C. 20005

, (202) 842-0034

i 253 1 concentrators can sort of seal themselves and you can still

() 2 communicate to all elements on the ring just with the -- if 3 you have that single failure through the ring.

4 Again, that follows the final point that even 5 though the monitor bus is not safety related, it's designed 6 so that no single failure can prevent-the transmission of 7- data to all the highway drops.

8- DR. POWERS: I have been puzzled over that last 9 sentence for a-minute or two here.

10 MR. DEUTSCH: Uh-huh.-

11 DR. POWERS: And I'm puzzled over it relative to 12 your previous slide, which is all these protection systems 13 going into the monitor.

14 MR. DEUTSCH: Well, I'm going to show you that in

() 15 a little more detail, and maybe that w.,ll clear that up.

16 I'm going to talk about the protection system, and then I'll 17 show you a more detailed architectural diagram of the 18 protection system.

-19 First of all, attributes of the protection system,

-20 we have four redundant channel sets, and the AP600 plan has

-21 four trains of-engineered safety-features.

22 The reactor trip and engineered safety features 23 actuation is based on two out of four voting logic. What 24 -this provides is that_at least two failures must occur to 25 give us any 6purious. actuation, and no single failure -- at

() '

_ ANN RILEY=& ASSOCIATES, LTD.

Court Reporters 1250 I Street, N.W., Suite 300 Washington, D.C. 20005 i

'(202) 842-0034 2

1 254 1 least two failures most occur to disable a safety function,

() 2 and no single failure results in a spurious actuation. It 3 would require two failures to give un a spurious actuation.

4 So it gives us a good defense and gives us a lot 5 of good attributes for testing and keeping the plant on 6 line, and I'll talk about that a little bit more in this I 7 next slide, but the reactor trip breaker configuration, and 8 I will show you what that configuration looks like, also 9 eliminates the need for separate bypass breakers, which are 10 typically on Westinghouse plants.

11 DR. APOSTOLAKIS: I don't understand your 12 statement at least two failures must occur to disaole the 13 safety function.

14 MR. DEUTSCH: Well, because we use two out of four

) 15 voting logic, and we have -- we have four channels and we 16 use two out of four voting logic --

17 DR. APOSTOLAKIS: So what kind of failures are 18 these? I mean, it's dangerous that there is a high value of 19 a parameter someplace and that particular channel says that 20 everything is normal? That kind of failure? Yes, Must be.

21 MR. DEUTSCH: There are no single failures that 22 can --

23 DR. APOSTOLAKIS: No , single I understand.

24 MR. DEUTSCH: And --

25 DR. APOSTOLAKIS: But at least two -- at least two

[#l

\--

AJM RILEY & ASSOCIATES, LTD.

Court Reporters 1250 I Street, N.W., Suite 300 Washington, D.C. 20005 (202) 842-0034

255 1 includes-two.:

2- At least two includes two.

(f MR. DEUTSCH:

3- DR.=APOSTOLAKIS -. Yes. With two failures, you're 4 not shut down.- You still have two-other-channels, 5" MR. DEUTSCH: Yes. Actually,-we --

6- DR.-APOSTOLAKIS: So-at least three.

7 MR. DEUTSCH: Yes. If we have nothing in bypass, 8 it would take: three f ailures to --

9 DR. APOSTOLAKIS: Yes.

10 NR. DEUTSCH: -- to disable a safety function, 11 yes. If we start with four channels -- I was a little 12 confusing there.

=13- DR. APOSTOLAKIS: So it's at least three.

147 MR DEUTSCH: Yes. Yes. It takes at least-two O

T/

s 15 failures to cause any spurious actuation, but if we have a 16 system that hasn't been -- that has no failures, you're 17 right, it would take.at least three fsilures before that 18' system'is disabled.

19 DR. APOSTOLAKIS: And again, the twofout of four 20 logic must be used several times, not just once.

~21 MR, DEUTSCH- Yes.

22 DR.'APOSTOLAKIS: Yes.

23 MR. DEUTSCH: Yes.

24 I wanted to show this slide on the bypass logic 25 for our reattor trip system and the engineered safety

, g ANN RILEY & ASSOCIATES, LTD.

Court Reporters 1250 I Street, N.W., Suite 300 Washington, D.C. 20005 (102) 842-0024 1

1 features actuation and show how tM i two out of four logic 2 provides us with the enhanced ca se ..lity of tr..intenance, If '

3 we don't have any inputs bypassed, our resulting trip logic 4 is two out of four, but if we have a failure in one channel 5 or we're doing maintenance on one channel, we can bypass 6 that channel and tha trip logic then reverts to two out of 7 three of the remaining channels.

8 So even in this condition, we still meet the 9 single failure criteria, and still no single failure can 10 cause a spurious actuation.

11 This architecture even allows us to bypass a 12 second input. In this case, the logic would revert to one 13 out of two. We assume it's a single failure criteria, but 14 we now are in a situation where, after having bypassed two 15 channels, a single failure could cause a spurious actuation.

16 DR. APOSTOLAKIS: What is the fraction of the year 17 where you expect to be doing this? That must be very low, la right?

19 MR. DEUTSCH: Yes. These are covered by the plant 20 tech specs -- 4 21 DR. APOSTOLAKIS: Yes.

22 MR. DEUTSCH: --

the length of time that we're 23 allowed to be in any of these situations.

24 MR. CARROLL: But typically it would be either 25 zero or one, not the others.

ANN RILEY & ASSOCIATES, LTD.

Court Reporters 1250 I Street, N.W., Suite 300 Washington, D.C. 20005 (202) 842-0034

257 1 MR. DEUTSCH - Yes. Certainly.

-( 2-- MR. CARROLL: Yes.-

3 DR UHRIG: The reason for the fourth channel 4 'there is if you have a-spurious failure, some momentary

5- failure of1some sort, a transient wire, one bypass, i.t still 6- doesn't take you down.

7 DR. APOSTOLAKISi- Right. But if-you have two 8 bypasses, it does.

9 DR.-UHRIG: If you get two bypasses,Lthe spurious 10- ---the signals will take you down, 11 DR. APOSTOLAKIS: They don't expect to be in that 12 state for too long.

13 MR. DEUTSCH: Ever.

14 DR. APOSTOLAKIS: Or ever. Yes. So why is it 15 there?

-16 CHAIRMAN BARTON: Just to show-you --

.17. DR APOSTOLAKIS: Oh, so that weLunderstand it.

18 DR. .UHRIG: If_you're in the process of repairing 19 an instrument and then another one fails on you, you can 20 still -- or if you're in the process of= testing an 21 instrument and a failure occurs, there is still protection.

22 DR. APOSTOLAKIS: Yes. Then they would not~ switch

-23 to'one out of two. If they're worxing-on one, they have two 24 out of three, and then this other failure occurs, they're 25 still okay.

ANN RILEY & ASSOCIATES, LTD.

Court Reporters 1250 I Street, N.W., Suite 300 Washington, D.C. 20005 (202) 842-0034

f l

t 258 1 DR. UHRIG: They're still okay, but it goes to a I 2 one out of two. All it takes is one out of the remaining 3 two --

4 CRAIRMAN BARTON: One spurious signal will take 5 you down, then.

6 DR. UHRIG: If you have two bypasses, one spurious 7 signal will take you down.

8 MR. VIJUK: What if you have two bypass?

9 DR. APOSTOLAKIS: Two bypass does not mean that 10 one has failed, does it?

11 MR. DEUTSCH: The tech specs do permit us to 12 bypass two, but bypass is sort of -- gets confusing. Even 13 though we bypass that channel, it automatically configures 14 the logic to one out of two.

O

(_,/ 15 DR, APOSTOLAKIS: Because one is placed on.

16 MR. DEUTSCH: Effectively, the --

17 DR. APOSTOLAKIS: One is placed on.

18 MR. DEUTSCH: -- functional logic does that, 19 right.

20 DR. APOSTOLAKIS: Yes. But '.> hat you were saying, 21 you were describing the situation where one is bypassed, 22 right, and the other fails.

23 DR. UHRIG: Yes.

24 DR. APOSTOLAKIS: Yes. So the logic is two out of 25 three. You can still tolerate one failure.

()

ANN RILEY & ASSOCIATES, LTD.

Court Reporters 1250 I Street, N.W., Suite 300 Washington, D.C. 20005 (202) 842-0034

259 l- 1 DR. UHRIG: That was the argument why we went to

() 2 four channels in St. Lucie One compared to three in our 3 Turkey Point plants. We wanted to be able to survive a

-4 transient of some sort when you were working on those two.

5 MR. DEUTSCH: Yes.

6 DR. APOSTOLAKIS: Why is two out of four better 7 than two out'of three?

8- DR. *! ILLER: -You have the reason up there, 9 DR.-APOSTOLAKIS: Well, two out of five is even 10 better.

11 DR. UHRIG: Two out of four is less likely to 12 happen than two out of three.

13 DR. APOSTOLAKIS: Which one is less likely to 14 happen?

() 15 DR. UHRIG: If you've got four channels, you can 16- stand two failures.

17 DR. APOSTOLAKIS: Yes, j i

18 DR. UHRIG: - If you've only got three channels, you 19 can literally only stand one failure.

20 MR. CARROLL: And you will have one channel down 21 'if you're testing.

4 22 DR. UHRIG: Yes. Any time you're testing, you --

-23 DR. APOSTOLAKIS: But the current reactors,

-24 though, I think the logic that's applied is two out of 12 5 three.

ANN RILEY & ASSOCIATES, LTD.

'+'

Court Reporters 1250 I Street, N.W., Suite 300 Washington, D.C. 20005 (202) 842-0034

l 260

1 DR. MILLER

Th).

.2 DR._UHRIG: No.

31 DR.-_ MILLER: Almost all of-them are two --

4 DR. UHRIG: All the modern reactors are two out of 5 four.

6 DR. APOSTOLAKIS: Two out of three is in a lot of

7. places.

8 DR. UHRIG: Well, --

-9 DR. MILLER: The older' plants.

10 DR. UHRIG:' The older plants only had three 11 channsla.

12 _DR. MILLER: Yes. Not very many are two out of 13 three.

14 DR. APOSTOLAKIS: Okay, b

- \s ) 15 MILLER:

DR GE BWRs are one out of two taken 16 twice,.which isTa:little bit different. It's still two out 17 of four.

18 CHAIRMAN BARTON: Are we okay on this? Can we 19 move on?

20 DR. MILLER: Yes, 21 CHAIRMAN BARTON: Okay. Ken?

22 MR. DEUTSCH: Move on? Okay. This is the 23 protection and safety monitoring system architecture that I

> 24 -was talking about. First of all,-I want you to notice that 25 we do send information from the protection safety monitoring ANN RILEY & ASSOCIATES, LTD.

Court Reporters

-1250 I Street, N.W., Suite 300 Washington, D.C. 20005 (202) 842-0034

261 1 system to the monitor bus. All that information is

() 2 isolated, and it is just information that's used by other 3 Systems. It's not -- the monitor bus is not critical to the 4 functioning of the protection and safety monitoring system 5 at all.

6 DR. MILLER: You have optical solators in each 7 one of those areas, right?

8 MR. DEUTSCH: Yes 9 DR. UHRIG: There i no way you could get a signal 10 back.

11 AR. DEUTSCH: That's right.

12 DR. APCSTOLAKIS: And wa are all supposed to know 13 uhat an optical isolator is, I suppose.

14 MR. CARROLL: Doesn't everyone?

15 (Laughter.]

16 DR. MILLER: Maybe Ken can give us a one-liner on 17 that.

18 DR. DEUTSCH: Well, that could actually represent.

19 a fiber optic data link or it could be a particular 20 component. It would have maybe a small portion of fiber 21- optic data link, or there could be other technologies used.

22 But we anticipate that this would be a fiber optic data link 23 to communicate from the protection safety monitoring system 24 to the monitor bus.

25 MR. CARROLL. That isolates it --

(\~ ANN RILEY & ASSOCIATES, LTD.

Court Reporters 1250 I Street, N.W., Suite 300 Washington, D.C. 20005 (202) 842-0034 l

m

t 262 1 DR. DEUTSCH:

. Supposed to electrically isolate it.

2 DR. MILLER: And it also means data transmits more

-3 than one way.

4 DR. DEUTSCH: -Yes_ It's not a duplex data link.

5 This is only a transmitter; that's only a receiver.

6 DR. MILLER: 'Yes. When you draw that nrrow, that 7- rw a y_means that arrow.

8 DR.-DEUTSCH: Yes.

-9 This is sort of the beginning of the protection

-10 safety monitoring system, the integrated protection 11 cabinets. The information from the field sensors comes in 12 -through the integrated protection cabinets. We really 13 actually have four sets of integrated protection cabinets 14 which are in separate rooms,. se'jarate -I&C rooms, each 15 associated with a separate protection division.

16 The integrated protection cabinets share-17 information with each other through fiber. optic data-links 18 within-themselves, and each integrated protection cabinet 19- -takes the'information from the sensors it has and-

'20 'information from the sensors received from the other 21 integrated protection cabinets and performs logic to 22 j determine if the reactor trip breakers associated with that ,

23'l :particular division should open.

24-l Again, the-logic performed on a channel basis is 25 two out of four, and the logic performed at the reactor trip O

Q ANN RILEY & ASSOCIATES, LTD.

Court Reporters 1250 I Street, N.W., Suite 300 Washington, D.C. 20005 (202) 842-0034

263 1- breaker level is also two out of four.

2 DR. MILLER: That system, though, and'I'm going to 3 -get a tutorial on it later, is substantially more 4 ' complicated.

5 DR. DEUTSCH: Yes. That's because you have read a 6 lot of information of how it was implemented in Sizewell and 7 you've seen the hardware issues associated with it.

8 DR. MILLER: The fur.damental is two out of four, 9 two out of four.

10 DR. DEUTSCH: Yes.

11 Dk. UHRIG: But it is not taking place in-the 12 sense that GE did.

13 DR. MILLER: No. No. Uses a very unique little 14 four coil in there to do a dynamic two out of four. I'm 115 going to-learn about that later. I' don't. fully understand 16- it yet.

17 IMt. DEUTSCH: Okay. The information from-the 18 integrated protection cabinet,-the information that it 19 receives is also transmitted a couple places. Some 20 -information is transmitted to the qualified data-processing

21. cabinet, and that is augmented by some information -- other 12 2 information that's received by the qualified data processing ,

23 cabinets.

24 For instance, the in-core thermocouples are not 25 used by the protection system and are not used by the -- to j,

4

\ '

ANN RILEY & ASSOCIATES, LTD.

Court Reporters 1250 I Street, N.W., Suite 300 Washington, D.C. 20005 (202) 842-0034

264 1 -. actuate any engineered safety features, so the in-core

( ) 2: thermocou;1es don't come in this path, but they come in 1 3 through thisfpath.

4 The qualified data processing system is a 5 two-train. system, it's not four train, it's two trains. It 6: uses separation divisions B and C, and so the integrated 7 protection cabinets can provide information to a qualified 8 data processing cabinets along with cata received from its 9- own sensors, and this two-train system provides information-t 10 to the operators on the qualified displays. E 11 h DR. MILLER: They're not 1E? '

12 -DR. DEUTSCH: This is 1E. These are the 1E 13 displays in the main control room, yes. Those are the 1E 1 Ldisplays.

15 As I mentioned earlier, some information-is 16 provided to the plant control system by. dedicated data 17 links, and that's-this path here, and some information --

18 and the integrated protection cabinet also does the 19 bi-stable function associated with the engineered safety 20 features actuation.

21 Engineered safety features actuation cabinets, in 22- AP600, there's four-sets of these cabinets, each associated 23 with its own engineered safety features actuation division.

24 'Each cabinet receives information from the-four. separate 25~ integrated protection cabinets, so the engineered safety O ANN RILEY & ASSOCIATES, LTD.

Court Reporters 1250 I Street, N.W., Suite 300 Washington, D.C. 20005 (202) 842-0034

.-____---a

~.

l 265 1 features actuation-voting -- two_out of four voting is done.

) 2 zat this_ level, and the system 1cvel logic for engineered 3 safeguards actuation is done here, and once it-has been 4- determined that a particular safeguards actuation is 5 necacsary, that'information is transmitted on the logic bus -

6 to sets of protection logic cabinets, which is the interface

7. 'with the plant components.

8 DR. POWERS: You persuaded me that I ought not 9 worry too much'about_the monitor bus, but for the ec.me 10 reasons,.now this logic bus seems to show up as something I 11 'ought to worry about.

12 DR. DEUTSCH: Yes. And this - the-monitor bus is 13 not safety related. This logic bus is safety _related.

14. There'are actually four separate logic buses, one associated

( 15 with each safety division. So whe; I said you didn't-have 16 to worryLabout-the monitor bus, that was-because it was not:

17- safety related and it is not key to the function of the 18 safety related protection' system. But-this --

19 DR. MILLER: There's only one bus, right, monitor

'20 bus?

21 DR. DEUTSCH: Yes.

22' DR. MILLER: What the diagram doesn't show is it 23 doesn't show four logic buses.

24 DR. DEUTSCH: Yes,-it doesn't show four_ logic >

25- buses.

ANN RILEY & ASSOCIATES, LTD.

Court Reporters 1250 I Street, N.W., Suite 300 Washington, D.C. 20005 (202) 842-0034

26CI 1 DR.-UHRIG: Is that whole system there on the 2' right duplicated three additional times?

-3 DR. DEUTSCH: Yes. Actually, think of it is as 4 coming out this way, where this is -- where this is Train A, 5 we would have a Train B, C and D.

6 DR. POWERS: So then the operator _has four 7 displays and four controls?

8 DR. DEUTSCH: Think of it right to here. These 1 9 were added later, these are -- the operator displays and-10 controls are not. safety related. Th!.s is a fiber optic data 11 link.

12__ DR. UHRIG: The multiplexers are not 1E;either.

13- DR. DEUTSCH: The multiplexers are 1E.

14 DR. UHRIG: They are.

15 DR. DEUTSCH: These particular multiplexers are

1. 6 associated only with a protection safety monitoring system.

17- Je do have multiplexers that are associated with the plant 18 control system.- This represents CRTs and operator soft 19 -control inputs, and they can~c7-- --

20- MR. CARROLL: Why don't we_have an isolator?

21- DR. DEUTSCH: This is fiber optic.

22 MR. CARROLL: Isolated?

23 DR. DEUTSCH: Right. You're saying why -- that's 24 because this was added to the figure at a later time.

25 MR. CARROLL: But if you're going to use that

\ ANN RILEY & ASSOCIATES, LTD.

Court Reporters 1250 I Street, N.W., Suite 300 Washington, D.C. 20005 (202) 842-0034

267 1- symbolism,'-it belongs over there, too?

( 2 DR. DEUTSCH: Yes.

3 MR.1 CARROLL:- Okay.

4 DR. APOSTOLAKIS: Can you tell me Icallquick 5- _ what's a multiplexer?

6. DR. POWERS: Before'he answers that question, 7 there can't'be an isolator between the operator displays 8 controls and multiplexer, not a one-way isolator the way 9' there.is up above.

10 DR. UHRIG: The signal:goes both ways.

11 IMt. POWERS: Yes, it goes both ways. We wouldn't 12 have control. Yes, it can't be a one-way isolator.

13 DR. DEUTSCH: That's right. It's not a one-way 14- . isolator. It's just an electrical' isolator.

15 -This morning, Steve was asked --

16 CHAIRMAN BARTON: Ken, before you go on, what's a

-17' -multiplexer.

18 DR. APOSTOLAKIS: Multiplexer, 19 DR. DEUTSCH: What was that?

20- DR. APOSTOLAKIS: What's a multiplexer?

21 DR. DEUTSCH: Oh, what's a multiplexer.

22 DR. APOSTOLAKIS: Oh, good. I thought it was 2s obvious.

24 -[ Laughter.]

25 DR. POWERS: He's only struggling with where to ANN RILEY & ASSOCIATES, LTD.

Court Reporters 1250 I Street, N.W., Suite 300 Washington, D.C. 20005 (202) 842-0034

'268 2? begin.

(j . :2 (Laughter.)

DR.-UHRIG:

How to make this simpler.

4 -(Laughter.]

1 5 DR. DEUTSCH: Whenever you're typing on your 6 keyboard on your computer, you have all those keys, but if

7. you would cut-the wire between there, you wouldn't see a 8_ separate wire for each of those keys.

9 MR. CARROLL: No , you would see maybe nine.

10 DR. DEUTSCH: Yes. And that's because you have a 11 multiplexer in your keyboard, and it takes those signals of 12 'all those keys and puts it on a few sets of wires. That's 13 called multiplexing, and then when it gets to your 14 processor,. it is demultiplexed. So it's a way of combining.

O (/ 15 data:on a few wires. Now, it's typically _a time-shared 16 -multiplexer. It used to be a frequency multiplexing.

.17- DR. MILLER: If you take the original ones way 18 back when, you would sample a piece of data, hold it there-19 while the other channels are being read out and keep-going

-20 around. Now it.is time-shared. But-the first multiplexers, 21 where the name came from, you really were only processing 22 ._one data bit at a time. Now they do time-shares. So the 23 word multiplexer is derived 1from the old way, not the new 24 way.

. 25 MR. CARROLL: It's a Greek word, George.  ;

() ANN RILEY & ASSOCIATES, LTD.

Court Reporters 1250 I Street, N.W., Suite 300 Washington, D.C. 20005 (202) 842-0034

269

,1 [ Laughter.)

2 CHAIRMAN BARTON: Go ahead, Ken, before you lose

-3 .us.

-4 DR. APOSTOLAKIS: The concept is probably Greek, 5 but the word is not.

6 (Laughter.)

7 DR, APOSTOLAKIS: The engineered safety features

-8 actuation cabinets, that's where the logic, the two out of 9 four thing is?

10 DR. DEUTSCH: For engineered safety features, yes.

11 DR. APOSTOLAKIS: And where is that physically?

12 -Where are they, these cabinets?

13 DR. DEUTSCH: Physically: the cabinets associated

~14 with a particular safety division is located in a common O

5 s,/ -15 room on AP600. So we have a room for Train A, room for 16 Train B, C and D. Sc physically, thet Tre located very 17- close to each other in AP600. This arch.cecture supports 18 _ putting.these out in the field close'to the actuated 19= equipment, but in AP600,-we have such a few nunter of-20 interconnected equipment that we have located this in the 21- same room.

22 Just like this architecture, when we talk about-23 .multiplexers, we're primarily interfacing with the CRTs in 24 the soft control workstations, but if we were to apply this 25- architecture with a conventional control room, with a long O ANN RILEY & ASSOCIATES, LTD.

Court Reporters 1250 I Street, N.W., Suite 300 Washington, D.C. 20005 (202) 842-0034

270 1- _ board:and switches,'we would still multiplex, we would still s

() 2 bring all those_ switches to the multiplexer and just 3 communic. ate with just a pair of wires or a datalink to the 4 rest of the protection safety monitoring system.

5 This provides a convenient way to do this main 6 control room remote shutdown room transfer that there was a 7 question on this morning, that these_ transfer switches, this i 8 main -- this_ transfer panel is located outside the fire zone ^

9 of the main-control room near the remote shutdown 10 workstation, and we don't have to transfer all these 11 electrical cables and all these lines.

12 Effectively, all we're doing is telling the system 13 now to don't listen to information coming from the main 14- control room, start listening to information-coming from the 13

\_) 15 remote shutdown room.

16 MR. CARROLL: It's or -- you can't get into a 17 situation where they're both telling it different things.

18 DR. DEUTSCH: That's right. Yes. We show it very ,

19 simply here,'but there's also some logic that resides down 20 here that-it knows which set of inputs to use.

21 And again, ev2n if we have a problem,-you isolate 22 the problem to one safety division, you know. _If we have a

- 2 3_ failure of this transfer, that only relates to one safety 24 divicion.

25 DR. UHRIG: And the reason that's located near the

() ANN RILEY & ASSOCIATES, LTD.

Court Reporters 1250 I Street, N.W., Suite 300 Washington, D.C. 20005 (202) 842-0034

271 1- remote shutdown workstation is that if you have-to evacuate

_( ) 2 the control room rapidly, you go over: and -- you don' t have L

i

_3 to hang around long enough to make that transfer, right?

4 DR. DEUTSCH: That's right, We want to keep this 5 in a separate. fire zone ---

6 DR UHRIG: Yes.

7 DR. DEUTSCH: -- from the main control room. and 8 then it just makes it -- you may as-well put it on the way 9 to the remote-shutdown room.

11 0 DR UHRIG: Yes, j 11 DR. DEUTSCH: We talked about the reactor trip E12 switch gear arrangement. These breakers are located in the 13' power line between the control rod drive motor generator 14 sets and the rod control cabinets. The power flowing --

15 that comes through this is non-Class-1E power, but the

~16 reactor trip breakers are controlled by the four protection 17 safety monitoring system division, and this configuration 18- provides a two cut of four logic for the four divisions.

19 DR. APOSTOLAKIS: Putting twice two out of four.

20 DR. DEUTSCH: Yes. Each breaker is tripped,'has 21- two trip mechanisms, an under-voltage attachment which 22 actuates the breaker based on loss of a signal and a shunt 23 trip attachment which actuates the breaker based on the

24 presence of a signal. So we actually have some additional 25 levels of defense and redundancy at this level.

ANN RILEY & ASSOCIATES, LTD.

Court Reporters 1250 I Street, N.W., Suite 300 Washington, D.C. 20005 (202) 842-0034

272 1 MR. CARROLL: These are your tried and true DB50.

O

g. 2 breakers?

.3 :DR. DEUTSCH: Yes.

-4 Mk. CARROLL: Yes. I don't think those are true.

5 DR. POWERS: I was-going to save it. I'm waiting 6 for the punchline here.

7 [ Laughter , ) '

8 MR. CARROLL: I'm always getting in trouble with 9 those.

10 DR. DEUTSCH: We talked earlier about our; defense 11- against complete loss of the protection safety monitoring 12 system being the diverse actuation system, and this is a 13 -blockEdiagram of the diverse actuation system for AP600.

i 14 I'm showing this to show you really how simple it'is and--how I 15 it comes down to the_-final actuated component separately 16 from the protection and safety monitoring system.

17- This represents an-air-operated valve, and maybe 18 to put that valve in its safety position, we would vent the 19 air. Normally the protection and safety monitoring system 20 would control an' air solenoid through a -- associated with

.21 1E power, and would vent the air to put that valve in a' safe 22 position.

23 The diverse actuation system can also put this 24- valve in a safe position separately from the protection and 25 ~ safety monitoring system.

ANN RILEY & ASSOCIATES, L TD .

f Court Reporters 1250 I Street, N.W., Suite 300 Washington, D.C. 20005 (202) 842-0034 u

273 1 In this case, we have the capability-for

() 2 automatically positioning that valve and manually 3 -positioning.that valve, and one:of the key elements I'm 4 trying to show here is the manual diverse actuation system

'S control does not_use any of the automatic portion of the 6' diverse actuation system. We actually wire the switches---

7 MR. CARROLL: Could_you point on the screen, because you're-8 . - - -

9= DR. DEUTSCH- I'm sorry..

i

- MR. CARROLL: -- making it difficult 1for those of 11- us on this side to see.

12 DR. DEUTSCH: We actually wire the owitches 13- directly to the~ actuated component in a two out of two 14 logic.

154- CHAIRMAN BARTON: Ruptured diaphragm, this valve 16- fails?

17 DR. DEUTSCH: Pardon me? ,

18' oCRAIRMAN BARTON: How does this-valve fail on a  !

-19 ruptured _ diaphragm? This is an air-operated valve?- l 2CF DR. DEUTSCH: Yes. This is an example of an air-operated valve. So its failure state is loss of air 22 puts it into its fail safe position. And this represents 23- two trains of diverse actuation system automatic processing.

24'

~

We have dedicated diverse actuated -- diverse actuation J25 system sensors that feed the automatic logic, and these 1 ANN RILEY & ASSOCIATES, LTD.

Court Reporters 1250 I Street, N.W., Suite 300 Washington, D.C. 20005 (202) 842-0034

274 1 sensors also feed dedicated diverse actuation system ll 2 indicators.

3 So the diverse actuation system provides the 4 capability for reactor trip, selected engineered safety 5 features actuation, and also provides a diverse indication 6 of selected plant parameters. Those are very -- those are 7 the more conventional meters, discrete --

8 DR. POWERS: You said diverse. Did you just mean M-9 redundant in this case?

10 DR. DEUTSCH: These are diverse from the 11 protection safety monitoring system, jl[ 12 DR. POWERS: Oh, okay.

() 13 MR. CARROLL: Now, is it permissible to use 14 digital components in this diverse manual actuation system?

15 DR. DEUTSCH: No. We have taken the approach that 16 our manual actuation is --

17 MR. CARROLL: Hard wired.

18 DR. DEUTSCH: Yes, wiring out there.

19 MR. CARROLL: Okay.

20 DR. UHRIG: Well, do you mean digital or do you 21 mean systems of software?

22 MR. CARROLL: I mean microprocessor.

-23 DR. UHRIG: Microprocessor where you would have to 24 program it as opposed to hardwiring.

25 MR. CARROLL: No. The reason I raised che 9 ANN RILEY & ASSOCIATES, LTD.

Court Reporters 1250 I Street, N.W., Suite 300 Washington, E.C. 20005 (202) 842-0034

275 1 question was ACRS fought a battle with the staff a long time 2

h ago on the principle that you should be able -- I mean, if 3 you want to hardwire it, fine, but you should be able to use 4 microprocessor-based technology, too.

5 DR. DEUTSCH: Well, we do use microprocessor-based 6 technology to perform the automatic actuation.

7 MR. CARROLL: Yes. Okay.

8 DR. DEUTSCH: Now, some functions don't even have E 9 an automatic actuation --

10 MR. CARROLL: Right.

11 DR. DEUTSCH: --

like cur automatic 12 depressurization system.

13 MR. CARROLL: Yes.

14 DR. DEUTSCH: We just nave the capability to do 15 that manually through the diverse actuation system.

16 MR. CARROLL: Right.

17 DR. DEUTSCH: I knew there were some questions 18 about the nuclear instrumentation on AP600. These are the 19 ex-core nuclear instrumentation. We have four separate 20 overlapping ranges. The start-up range, we use the BF3

?" proportional counters because Westinghouse has used that 22 forever. We have a large experience based on thcce, and 23 we've had great success with those.

24 We have four proportional counters located around 25 the core and provides about six decades of counting range.

ll ANN RILEY & ASSOCIATES, LTD.

Court Reporters 1250 I Street, N.W., Suite 300 Washington, D.C. 20005 (202) 842-0034

276 1 The intermediate range uses pulse-vision chambers, and that h( ) 2 covers about eight' decades of range.

3 Early on, we were planning to use compensated ion 4 chambers for the intermediate range, and that works fine 5 during normal operations; however, when we got into looking 6 at-our needs to meet Reg Guide 197 and considering a LOCA,-

7 the compensated ion chambers are very sensitive to the gamma-8- that would be present in a post-accident environment, and so 9 we went'and changed to pulse vision chambers so we wouldn't -

10 be susceptible to this gamma.

11 Reg Guide 197 requires a range of ten to the minus 12 six to 100 percent of full power, so the intermediate range

-13 by themselves meets the requirements for Reg Guide 197, and 14 the power range uses two segment of uncompensated-ion 15 -chambers.

16 DR. MILLER: I've got a number of questions on 17 this one.

18 DR. DEUTSCH: Sure.

19 DR. MILLER: Go ahead.

20' MR. CARROLL: So chat's two' locations for the 21 power range?

22 DR. DEUTSCH: No. There are four locations, and 23 each location has an upper and lower detector segment.

24 MR. CARROLL: Okay.

25 DR. MILLER: That gives you -- they can get --

ANN RILEY & ASSOCIATES, LTD.

Court Reporters 1250 I Street, N.W., Suite 300 Washington, D.C. 20005 (202) 842-0034

. .._.___m... - _ .. _ ___.__. _ _ _ ..._ .. _ .. _ _._._ _. _ _ .._.. _ _ . _ _ ... _ _ _

277 DR. DEUTSCH: Axial power distribution-2 information.

3- DR,LMILLER: Let me start. First of all, I have a-

-4 broad question -- why even specify what kind of detectors 5 you're using, Why not leave it --

just provide the-6 performance specs and say you can leave it open in lieu of 7 putting it into the SSAR.

8 DR. DEUTSCH: Yes, That --

! 9 DR. MILLER:

That's a broad question.

10 DR. DEUTSCH: Yes. That was just a call. It was 11 based on our experience'. I think either way would have been 1:E okay, but we sometimes get pressure to, you know, put in as

- 13 much detail-as we can.

14 DR. MILLER: Well,fI could ask the staff what they 15; -accepted. The reason:I ask is you indicate that the-16 intermediate CICs which you've used historically also are-

?

[ 17 susceptible to gamma rays, which I agree with, would never

. 18 meet Reg Guide-197, 19 -By-thstway, your WCAP, then, that you provided is l

20 out of date. That's not --

l

.1 2 DR. DEUTSCH: It certainly is. 'It.was written 22 about'1992.

l -- 23- DR. MILLER: Okay. All right. Because that was ,

, 24 my next question. You're not going to use the WCAP which

. 25- did specify the CIC.

,7%r-

ANN RILEY & ASSOCIATES, LTD.

-) ' Court Reporters 1250 I. Street, N.W., Suite 300 Washington, D.C. 20005

~

(202) 842-0034

278 1 DR. DEUTSCH: Yes.

h 2 DR. MILLER: Proportional counters are also BF3 or 3 susceptible to high gamma rays. I know you indicate you 4 have not had that experience, but many PWRs replace them 5 with their wide range vision chambers, either both --

6 DR. DEUTSCH: But we're not relying ora the 7 start-up range for post-accident environment.

8 DR. MILLER: That's true.

9 My next question is, what is a pulse vision 10 chamber? Can you explain what you mean by that? Is that a 11 chamber that operates in cambelling mode and --

12 DR. DEUTSCH: Yes. Actually, intermediate range 13 operates in both -- in the early part of the range, it 14 operates in a counting mode, and then as it gets to the 15 upper ranges and the pulses are coming too fast to count, 16 then it switches over to cambelling.

17 DR. MILLER: To cambelling. Okay.

18 DR. DEUTSCH- Yes.

19 DR. MILLER: Now, why not use the lower part of 20 that tor your start-up range, then, which is happening in 21 many PWRs, including Westinghouse, both domestically and 22 internationally. They've replaced their -- they've thrown 23 away those boro: --

24 DR. DEUTSCH: We have a -- I might not be using 25 the right term -- a low -- we want to keep as many neutrons f ANN RILEY & ASSOCIATES, LTD.

Court Reporters 1250 I Street, N.W., Suite 300 Washington, D.C. 20005 (202) 842-0034

279 1 in the reactor vessel as we can, and so there would be some f 2 question whether this low range of pulse vision chambers 3 would see enough neutrons with fresh fuel to give us the 4 required count rate. The proportional counter has roughly 5 two decades of measurement below this, so we do get enhanced 6 sensitivity using the proportional counters.

7 MR. CARROLL: But that sensitivity isn't needed 8 once you get the fuel irradiated.

9 DR. DEUTSCH: Probably, yes, 10 DR. SEALE: That's why they call it the start-up Il range.

12 DR. MILLER: But you're saying that that pulse 13 vision chamber has a lower loading of uranium than would be 14 the wide range channels like marketed by Gammametrics?

h 15 DR. DEUTSCH: I'm not familiar enough with --

16 DR. MILLER: Because the Gammametrics system does 17 go down where the boron trifluoride does, but it has a few 18 more milligrams of uranium in it than maybe that one does.

19 See, I don't know what the spec is on that one to go down to 20 ten to the minus six.

21 DR. DEUTSCH: I know that we -- there was some 22 concern, there was some concern in being able to see enough 23 neutrons --

24 MR. CARROLL: They put the reflector on it.

25 DR. DEUTSCH: -- you know, because we do ll ANN RILEY & ASSOCIATES, LTD.

Court Reporters 1250 I Street, N.W., Suite 300 Washington, D.C. 20005 (202) 842-0034

280 11 everything we can to keep _the neutrons inside the core, and

() 2

~

since we 1ocate these outside the core, you know --

3 DR. MILLER: Right.

4- DR, DEUTSCH: You know, what you're saying is 5 certainly feasible,--and I don't know if -- I'm not the_right 6- -person to answer thoseDparticular details. But what I 7 believe is that because of the sensitivity that we need, we 8 needed_to go with this start-up range.

9- DR MILLER: I'm questioning that. I guess my 10 recommendation overall is not -- unAess the staff has an 1 11 objection, do.'t specify the type of protections at alli 12 just specify what they have to do, performance, and down th3 13 line, let the owner or somebody else decide what detection 14 they need.

15 MR. . CARROLL: Westinghouse has a. proprietary 16 interest in specifying them and they want to provide them.

17 DR. MILLER: Well, I understand that.

18 DR. POWERS: See, Bill Gates isn't the only one 19 that knows aoout-bundling.

20. [ Laughter.)

21' DR. UHRIG: These are-all ex-vessel?

22 DR.'DEUTSCH: Yes. These particular ones. I'm

'23 going _to talk about our in-vessel ones in a minute.

24 DR. MILLER: -I think that those who have replaced 25_ their proportional counters have done it for a reason and 1

O' ANN RILEY & ASSOCIATES, LTD.

Court Reporters 1250 I Street,-N.W., Suite 300 Washington, D.C. 20005 (202) 842-0034

281 1- found it advantageous to have one' channel which does both

() 2 start-up and intermediate; if nothing else, simplicity. I 3 don't-know. I've never done the neutronics'to see whether 4: they're absorbing neutrons. I.would.be surprised-if-I'came 5 up with that.

6 DR.-DEUTSCH: Yes. Well,-I mean, I can think of 7- the perfect person to ask this question, and I don't know 8 where he would come. He might say yes, you're right, or he-9 may have a reason why there's a particular issue that you 10 and I aren't aware _of.

11 CHAIRMAN BARTON: Is this a safety question or is 12 this a choice of different detectors here?

13 DR. DEUTSCH: Yes, I think this is just a choice.

14 CHAIRMAN BARTON: Then proceed.

15 DR. DEUTSCH: _Okay.

16 In-core. instrumentation. This plant uses fixed

-17 in-core' detectors instead of movable flux detectors to do 18 the' flux mapping.

19 MR. CARROLL: Now, I forget, have fixed in-cores 20 -been put in later PWRs? l 21 DR. DEUTSCH: There's --

22 MR. CARROLL: What's Sizewell got, for example?

,23 DR. DEUTSCH: I don't know what Sizewell has.

24 Westinghouse has been getting some experience with the

25. vanadium fixed in-core detectors at St. Lucie, although that

() ANN RILEY & ASSOCIATES, LTD.

Court Reporters 1250 I Street, N.W., Suite 300 Washington, D.C. 20005 (202) 842-0034

r.

t-282 1 is not their system that they use to do their flux mapping.

() 2 MR. CARROLL: Okay.  ;

-3 DR. MILLER: Well, the Babcock and Wilcox und CE t

54 plants all have fixed.in-core.

5 MR. CARROLL: Oh,-I_know that.

6 DR. MILLER: I thought you said PWRs.

~7 MR. CARROLL: No, no, I meant Westinghouse.

8 DR. DEUTSCH: So our fixed in-core detectors, we 9 have 38 core _ locations and we make six axial measurements

10. per each location. We're using a vanadium detector.

11 At one time, we were looking at Rodium, but 12 recently Westinghouse has gotten much more experience-on the 13 Vanadium and it has much more desirable properties in that 14 it's a longer life. material, and you don't have to replace 15 it as often.

16 DR. MILLER: Less sensitive, though, right?

17 DR. DEUTSCH: It's --

18 MR. CARROLL: Rodium is what BMW has been using?

19 DR. MILLER: And CE also.

20 MR. CARROLL: And CE also.

21 DR. DEUTSCH: It's not -- it's not -- it doesn't 22 have-the prompt sensitivity that the platinum has.

23 DR. MTLLER: Okay, It's less sensitive than is 24 Rodium, 25 DR. DEUTSCH: Yes. It's not quite as sensitive.

ANN RILEY & ASSOCIATES, LTD.

Court Reporcers 1250 I Street, N.W., Suite 300 Wachington, D.C. 20005 (202) 842-0034 l

i

\

283 1 DR. MILLER: But it is faster, though. '

( 2 DR. DEUTSCH: Un-huh.  ;

3 DR. MILLER: But you're not using it for safety l 4 anyhow.

5 DR. DEUTSCH: That's right. This is --

6 DR. UHRIG: Wha *. is long in long life? Three 7 years? Ten years?

8- DR. DEUTSCH: That's'long life. relative to the 9 Rodium, and it's on the order of five years or so. Ron, is-10 that what-you recall?

11 MR. VIJUK: Yes. I think it's like five to ten 12 years they're expecting for Vanadium.

la MR. CARROLL: It still beats the hell out of the 14 movable in-core system.

() 15 _DR. MILLER:

Anything does that.

15 DR. UHRIG: There's always a problem.

17 DR MILLER: Anything does that, yes.

18 So again, your WCAP where_you-specified Rodium is 19 cut of date also. ,

20 DR. DEUTSCH: Exactly.

21 The in-core-instrumentation system also provides 22 the core thermocouples, Type K chromel-alumel thermocouples 23 that we use to +0nitor the core. Those are split up-into

~

24, two groups. Those are safety related and split up into two 25 separation divisions, although we do.have'four instruments i

O .

ANN RILEY & ASSOCIATES, LTD.

Court Reporteru

-1250 I Street, N.W., Suite 300 h4 -b3 2

u __ ~ - _ _ _ _

284

,_ 1- which are dedicated to the diverse actuation system.

() 2 DR. MILLER: Now, the SSAR did not specify what 3 type of fixed in-cores you're using. Are you going to 4 change the SSAR to do that "o# or am I just out of date?

5 DR. DEUTSCH: I thir.k that's one-place where we-6- chose to say that -- talk about a functional requi .ent 7 instead of a specific implementation.

8 DR. MILLER: I agree with you on that. Vanadium

_9 is okay, but it may not be the only way to do it, e 10 DR. DEUTSCH: Right.

11 DR. MILLER: Okay. So you're going to keep the 12 SSAR the way it is and not specify any --

13 DR. DEUTSCH: Yes.

14 DR. MILLER: Okay. So you're just giving this as b) g, 15 an example, the way it could be done.

16 DR. DEUTSCH: Yes. If we were building one.

17 tomorrow, this is the way we would go.

18 DR. POWERS: If your-digital equipment is 19 sensitive to radio frequency interference, isn't there an 20 attraction to using thermocouples that don't have a 21 magnitude leg?

22 DR. DEUTSCH: What was the question again? If our 23 --

24 DR. POWERSt Digital equipment is sensitive to 25 radio frequency interference, isn't_there an attraction to

- =

r

( ANN RILEY & ASSOCIATES, LTD.

Court Reporters 1250 I Street, N.W., Suite 300 Washington, D.C. 20005 (202) 842-0034

285 1 avoiding thermocouples that have one -- similkr metals as

.: magnetic?

3 DR. DEUTSCH: I can't answer that specific 4 question, but the thermocouples are such a slow speed that 5 any kind _of EMI-RFI interference can be easily filtered out.-

6 MR. CARROLL: They'rt ?own in the reactor.

7 DR. POW 3RS. They're the best antennas I know of 8 for radio frequency-interference because they have one leg 9 that's magnetic. One of the pairs is magnetic.

10 MR. CARROLL: Yes, that's right.

11 DR. POWERS: Okay. And at least when we use them 12 around induction-fields, it just kills us because we get 13- that -- we pick up everybody. We pick up pilots overhead 14 and everything else in there. So I just wondered why use a 15 magnetic lead thermocouple, 16 MR. CARROLL: I don't know, I'm not aware of any 17- ovoblems with --

18 DR. POWERS: Nor am I.

19 MR. CARROLL: -- alumel and PWR, 20 DR. POWRRS: Well, what you usually have to_do to 21 get rid of radio frequency interference on the 22 thermocouples, if you're trying to measure them, you have to 23 ground them about every 30 centimeters. I mean, you have to 24 shield them and ground them out all over the place if you've 25 got radio frequency interference, because-you've got -- I ANN RILEY & ASSOCIATES, LTD.

Court Reporters 1250 I Street, N.W., Suite 300 Washington, D.C 20005 (202) 842-0034 U

( ] 286 1 mean, you've got a great big magnetic leg in there.

() 2 MR. CARROLL: Sure. Sure.

3 DR. POWERS: And typically, the transmission lines 4 are also chromel-alumel as well. I don't know whether they 5 are in this case or not.

6 MR. CARROLL: Not always.

7 DF, DEUTSCH: Anything else on the in-core 8 instrumentation?

9 (No response.]

10 MR. DEUTSCH: Just to wrap things up, I wanted to 11 show the architecture associated with the plant control 12 system. What 3 am showing here is the t'unctional 13 segmentation of the control system to various distributed 14 controllers, the protection system, signals coming L:

15 through -- signal selection algorithm which is a combination 16 of algorithms which determine which of the redundant signals 17 should be used. It might be an average, it might be a 18 weighted average. It is might be the middle signal, depends 19 on the particular signals coming in and where it is used in 20- the system.

21 The plant control system shares a bus. We are 22 showing this -- we are calling this the process bus here, 23 but it -- it will probably be physically the same monitor 24 bus that is the integrating data highway for the plant. And 25 it is 1E controls and, again, it is non-1E components that ANN RILEY & ASSOCIATES, LTD.

Court Reporters 1250 I Street, N.W., Suite 300 Washington, D.C. 20005 (202) 842-0034

J 287 1 this controls and we do have non-1E multiplexers which are

() 2 associated with the plant control system that would 5 3 interface with the controls in the main control room and the 1 4 remote-shutdown room.

5 DR. MILLER: So the isolation in this case is only 6 in that -- over there under which you called signal

I '  :

selection?

8 MR. DEUTSCH: Yes. Those -- that -- in that case, i

-that is signals only flowing into the plant control system 9 '

4 ,

q 10 and so it is that transmit only from the protection system 11 to the plant control system.

12 DR. MILLER: So the two-way arrow means the signal- ,

13 is going both wayc.

14 - 11R. DEUTSCH

Exactly.

15_ DR. MILLER: The one way means it is really 16~ . electrically isolated?.

17 MR, DEUTSCH: Yes. Yes, that is trying to show -

18? the communication flow.

19 Okay.- l 20' -CHAIRMAN BARTON: Any other questions for the 21 Chapter 7? I t

22 (No response.]

t

! 23 CHAIRMAN BARTON: Thank you, Ken.

24 MR. DEUTSCH: Thank you.

25' lCHAIRMAN BAR7DN: We will now take a-15 minute

() ANN RILEY & ASSOCIATES, LTD.

Court Reporters 1250 I Street, N.W., Suite 300 Washington, D.C. 20005 (202) 842-0034

1 288 1 recess and come back at 20 after 3:00.

() 2 [ Recess.]

3 CHAIRMAN BARTON: We reconvene-the meeting. We 4 are going to--Item No. 10, Test and Analysis Program. Brian.-

5 MR.!McINTYRE: My name is Brian McIntyre, and I 6 got the short straw to come up and do testing. Mainly 7 because all the people who could really talk about this and

=8 really do a crackerjack- job and answer all the questions 9 that you are going to have are backing working on the 10 questions that you have. So I came_down, I will do a rather 11 high level presentation on where we are on the testing 12 program and the things that we are doing with the i 13 Thermohydraulic Subcommittee and with the staff.

14 The objectives of the testing program, which we (I 15 have pumped about $60 million into at this point, and we 16 started back in 1989, was we have two types of tests. The 17 first type was to look at verification of some components, 18 because there are some components in the AP600, let's face 19 it, that-are different or that we are using in different 20 ways, and we need to do some verification in that area.

21 The other primary reason is that we needed to get 22 _ data to go back and do the validation of the computer codes, 23 because the AP600 is different. It uses -- low pressure is 24 a lot more important in this plant than it is in current 25 -plants and we knew that we had the test data in that area.

=

0- 7JU4 RILEY & ASSOCIATES, LTD.

Court. Reporters 1250 I Street,-N.W., Suite 300 Washington, D.C. 20005 (202) 842-0034

d t

289 '

1 We also did some -- these are more the individual

()

2 code-type, single parameter effects tests, and we also did 3 some larger more systems-effects tests. And finally, we had 4' to run_some of these tests mainly because -- to keep the NRC 5 happy. These were run -- we would have probably run fewer

• 6 tests. The NRC had some other ideas. They reviewed our 7' testing programs and we did add tests for their -- at their ,

8 request.

9 DR. POWERS: Presumably, that was more than just a

10 to keep them happy, that was to satisfy particular concerns 11 they had about safety issues or potential safety. issues?

12 MR. McINTYRE: That is keeping them happy, yes.

13 [ Laughter.)

14 DR. KRESS: That is what you meant, right?

) _15 MR. McINTYRE: That is exactly what I meant.

16 The procesc that we went through, and_I am not 17 going to talk about all of these but wo identified what I 18 phenomena we needed to look at. We went_through scaling, i 19 ~ and scaling has been.probably the biggest bone of

-20 contention, that's probably a fair statement, that we have 21 -had with the Subcommittee.

22 I think, and this is clearly going to be my view, 23 -but in scaling, back when we started this, again, a lot of

. 24 .these facilities were started back in 1988, 1989, and 25 scaling has come a-long way in that time frame. You know, O ANN RILEY & ASSOCIATES, LTD.

1250 I Street, N.W., Suite 300 Court Reporters Washington, D.C. 20005 (202) 842-0034

l 290 1- quite honestly, the large scale tests that we have on the

() 2 containment was scaled by the largest tank we could get 3 under a railroad bridge on a flatcar and get to Pittsburgh.

4- That was -- that was the type of logic that we went at, and -

~

5 now we are having to go back and show that, yea, verily, 6 that is the right size and it scales up in a PIRT test, or 7 PIRT reports and things'like that. And we are doing that, 8 and that's what we are doing with the Committee.

9 MR. CARROLL: Why didn't you do that as a first 1

. 10 step?

11 MR. McINTYRE: In that time frame, that wasn't 12 quite that important as it is today. .This is -- I'm sorry 4

13 Dr. Seale isn't here, but this is to some extent where we 14 get into the raising the bar comment.  !

() 15 DR. POWERS: Well, I think that -- maybe a fair l 16 characterization is it wasn't quite important in your mind.

17 I think the. issues of scaling have been around since 18 Reynolds' day.

19- MR. McINTYRE: Well, it is how you scale and to 20- what level.you scale. I mean it wasn't just -- I am being a 21 little' flip when I say it was the largest tank we could get.

22- There were other ways that things were scaled. There was  :

23 dimensional scaling then inside based on that. But what we 24 are being asked to do now is more than -- I think, clearly,

- 25 - more than anyone envisioned at the time.

O IdRJ RILEY- & ASSOCIATES, LTD.

Court Reporters 1250'I Street, N.W., Suite 300 Washington, D.C. 20005 (202) 842-0034 em.,y- - =--, --vw- w -r - e--e v --w-h- * -"v'""-*'F'"""*"'*"#D~'* ' ' ' ' * ' ' " ' " * ' ~ " "

291 i We scaled those facilities the way we scaled other 2 tests that we had run, like flexi-set, and G-loop and J-loop 3 and all the ot.her tests that Westinghouse has performed over 4 the years.

5 We were done with the testing program, then we got 6 into the verification and validation of the computer codes 7 and the model development and, finally, the safety analyses.

8 And we, at this instant, primarily in -- in these regions in 9 discussions right now with the staff and with the ACRS, and 10 wts have done some analyses. We are also re-doing some 11 l analyses, particularly in the containment area, right now.

12 One of the ways that we looked at what tests we 13 needed to run, and this is just -- it says " example" at the

14. top. There was a book that -- a report that we put together 15 where we looked at the different types of phenomena. This 16 happens to be for small break LOCA -- this thing goes on for 17 several pages -- where you looked at what the process was, 18 whether or not there was something unique or different about

-19 the AP600 compared to current plants, whether or not there 20 was some type of information that we had that was out there, 21_ either in the literature or Westinghouse had, tihat was 22 proprietary, and then, particularly, what we would need to

-23 do -- whether or not, even if we didn't have that data, 24 whether or not we thought it was important enough to go back 25 and test. And, finally, there's just some general comments.

ANN RILEY & ASSOCIATES, LTD.

Court Reporters 1250 I Street, N.W., Suite 300 Washington, D.C. 20005 (202) 842-0034

292 1 And this was the type of process that we went

-2 through to determine if we needed to run and get more 3 testing data. So this was sort of the first-cut-at what we 4 decided we needed to do.

5 The tests have fallen into three general-( categories, the component and verification, and then we had 7 a series of passive containment cooling tests, and then the 8 series of passive core cooling tests.

9 For the component verification, these are just 10 some examples, we ran the reactor coolant pump. We did --

11 because this is a different design, remember it is the 12 -canned motor pump, upside down joined to the steam 13 generator, which is different, and not only-that there's 14 one, but there's two, and that is a different design for us.

15 So we ran'some air and water flow tests in models of this 16 till we knew what was going on, and what sort of flow 17 patterns we would expect.

18 The question had come up earlier, I think Dr.

19-- Powers had asked the question on the reactor coolant pump.

20 One of the biggest problems we had there is we had to build-21 the world's largest, some sort of thrust-bearing -- and, 22 again, I am obviously not the mechanical engineer, but the 23 bearing guys really thought-this was great. The RCP 24 flywheel that we built is depleted uranium with a stainless 25 steel shell welded around it. And we ran the test for two ANN RILEY & ASSOCIATES, LTD.

Court Reporters 1250 I Street, N.W., Suite 300=

Washington, D.C. 20005 (202) 842-0034

~_ .

l 293 1 reasons, one was to figure out what.the coast-down losses

() 2 were, but also to confirm that the bearing that we had-3 designed would indeed support this thing, because it has got 4 this very, very heavy mass resting on this bearing, you are 5 going to expect.it to run for quite a period of time.

6 -And what we found, quite to our surprise, is that 7 the original design of the flywheel wasn't'as good as we 8 thought it was from the standpoint of drag-losses, and we 9 found a better way'to machine it and make it a little bit 10 smoother and improve the coast-down characteristics.

11 For the reactor internals, we did some air flow 12 testing, again, to look and make suru that things were going 13 to work the way that we thought they should.

14 And, finally, this was a very-favorite, when we

() 15 first started, that we are expecting -- we don't have 5,000 16 horsepower pumps to drive _the check valves open, and we did 17 a series cf tests to determine exactly how a check valve

. 18 would behave if it was only being operated with, you know, 19- 15 or 20 feet-of water with-the thing, you know, open all-20 the way, would it open half-way. The original design of the 21 plant, we --

22 MR. CARROLL: Would it open at all?

23 MR. McINTYRE: Would it open? Well, we -- that

24. was a question -- a very. favorite question, and we designed 2 5.- that.out by putting.the squib valves in, because-to-run O ANN RILEY & ASSOCIATES, LTD.

Court-Reporters 1250 I Street, N.W., Suite 300 Washington, D.C. 20005-(202) 842-0034

I 294 1 that, a test like that, you would have to somehow run a 2 statirtically meaningful sample of tests, of check valves 3 that had been held shut for two years in an atmosphere that 4 probably wasn't the best. And a statistical meaningful ,

5 sample of that would have taken a long time, and it was

• 6 better just to design it out.

7 DR. UHRIG You mean the squib valve?

8 MR McINTYRE: An explosive valve. So, basically,-

9 -that is what is -- the check valve is seeing no pressure on, 10 and.the squib valve is seeing the 2250 psi pressure on the 11 lother side. So it is not held shut and it is not' going to 12 get welded shut, corroded shut in the event, because the 13 squib valve is basically serving as the pressure boundary.

14 MR. CARROLL: Now, I guess I have never seen or 15 heard a presentation on the results of this test program.

16 MR. McINTYRE: Which ones? ,

17. MR. CARROLL: The check valve program. Is it just 18 because I have been away for a while, or have you ever made- -

19 one?

20 MR. McINTYRE: I don't know if we have made a 21 presentation. There was a report on it, j 22 - MR. CARROLL: I don't know. Am I the only one 23 .that'would like to hear about this?

24 CHAIRMAN BARTON: I would like to hear about it.

~

25 MR. CARROLL: Okay. Why don't we put that on a ANN RILEY & ASSOCIATES, LTD.

Court Reporters '

1250'I Street, N.W., Suite 300  ;

Washington, D.C. 20005 (202) 842-0034

295 1 future agenda.

2 MR. McINTYRE: Ron, want to write that down?

3 MR. VIJUK: Okay. Which test are you tal. king 4- about, Brian? The one - -

5 MR. McINTYRE: The check valve.

6 _MR. VIJUKi- The ones we ran at tile plant?

7 MR. McINTYRE: The ones at the plar;ts. There are-8 also the ones that Larry ran.

9 MR. VIJUK: _There were some other ones.

10 MR. McINTYRE: Yeah.

11 MR. CARROLLi Basically, the question is, tell us 12 why you have confidence that they will work.

13 MR. McINTYRE: For the phenomena that we looked at 14 for the containment cooling, there's quite an extensive 15 series of tests tnat. Ort mn. In this case there was a 16 flat plate test that we ran at the Science and Technology 17 ' Center at Westinghouse. It was about three feet wide and 18 about six or eight feet high where we got very, very basic 19 data. We then also ran a larger- scale test, three feet in 20 diameter, 24 feet tall, that had steam on the inside and 21 water flow down the outside, which is referred to as a small 22 scale test.

23 There was the large scale test, which was the 24 one-eighth volumetric scale of the containment _that had the 25 steam on the inside, the operating-deck --

= . - --

ANN RILEY & ASSOCIATES, LTD.

Court Reporters 1250 I Street, N.W., Suite 300 Wainington, D.C. 20005 (202) 842-0034

. _- - - - - l

296 1 DR. WALLIS: Wait a minute. Wait a minute. The

() 2 size of the facility is one thing. What was the atmosphere?

3 MR. McINTYRE: On the inside it was steam.

4 DR. WALLISi There's no air in the steam?

5 -MR. McINTYRE: There was air in the steam.

6 DR. WALLIS: Was there stratification in the 7 steam? What was the thermodynamic condit. ion of the stuff 8 that is being condensed?

9 MR. McINTYRE- What was the --

10 DR. WALLIS: What was the thermodynamic condition 11 of the stuff that was being condensed?

12 MR, McINTYRE: There was air in there and there 13 was steam inLthere because we-had basically a steam 14 generator hooked up where they were quasi-steady state.-

() 15 DR. KRESS: We heard a presentation on this and 16 they actually varied the steam / air concentration but they 17 het a limited range set up in the test so they couldn't 18 cover a real -- there.are reports available.

19 DR. WALLIS: _Okay, so then I should -- I am going 20 to get a stack of reports, I guess about all this.

21 MR, McINTYRE: Paul will give you more paper 22 than you probably want to see.

23 DR. WALLIS: Than the size of the test facility.

24- MR. McINTYRE: It's close.

25 MR. CARROLL: You have got to understand, Graham, O- ANN RILEY & ASSOCIATES, LTD.

Court Reporters 1250 I Street, N.W., Suite 300 Washington, D.C. 20005 (202) 842-0034

297 1- that is a Westinghouse strategy there. When you ask a

() 2 ,q uestion, they give you a ton of paper --

3 DR. WALLIS: Maybe someone could point out which ,

4 parts of the paper I ought to read.  !

5 MR. McINTYRE: Yes. I think you should -- Novak 6 had that question and we gave him a little-roadmap at the '

7 last meeting and I think.I can ask Paul -- I hate asking the 8 staff to do thir.go.but Paul will put that at the very top.

9 -We ran water. coverage tests because that was 10 another question. Remember, we were putting water in the 11 top of the. containment and expecting it to flow down and be  :

12- distributed and redistributed.

13 We built in that case a_ full-scale segmant of the 14 -containment dome at Walt: Mill and poured water down the 15 outside and it was built to prototypical types of 16 construction tolerances. It had bumps. It had weld marks.

17 We measured the water,-the water flow distribution, down at 118 the' bottom and:we factored that into the way we do the 19 analyses.-

20 DR. POWERS: Do you worry about that water 21- distribution around the top and things like that? Do-you 22 worry about-things like corrosion, dust collection, that 23 sort of thing? Grease?

24- MR. McINTYRE: The staff does a lot. We have an 25 in-service testing-program. We're discussing right now with

() ANN RILEY & ASSOCIATES, LTD.

Court Reporters

'1250 I Street, N.W., Suite 300 Washington, D.C. 20005 (202) 842-0034

298 1 them is the paint safety related, is the paint not safety

() 2 related, and what sort of -- we have had to look for 3 bounding type ass.2mptions we can make in the calculation.

4 It is inspected. It's not'something you just put 5 on and expect to stay there for 60 years.

6 DR. WALLIS: But wetability is-a very iffy, 7 whimsical phenomenon and a little bit -- a deposit of 8 something, somewhere will make the water go somewhere else.

9 MR. McINTYRE: It's tested.

10 DR. WALLIS: But it is very. difficult to know what 11 to test because things change in a whimsical way. If you 12 touch something with your fingers, that changes its 13 wetability -- I_am not sure what I would test.

14 MR. McINTYREt Every 10 years, and I realize that 15 is a long time for something that is whimsical, but at the 16 start of the test they are actually going to put'a guy --

17 remember, this thing is 130 feet in diameter -- there is 18 going to be a guy, a thin guy --

19 [ Laughter,)

20 DR. WALLIS: Just-like you or I?

21 MR. McINTYRE ' Yes, we could -- our next career 22 could be doing this --

23 CHAIRMAN BARTON: An after-retirement job?

24. MR CARROLL: Actually he is going to get a 25 sitcom --

() ANN RILEY & ASSOCIATES, LTD.

Court Reporters 1250 I Street, N.W., Suite 300 Washington, D.C. 20005 (202) 842-0034

299 1 MR. Mc!NTYRE: They are going to put the water

() 2 down the outside and the guy is with a ruler (nr a yardstick 3- or a tape measure of some sort. He's going to measure 4 around the outside of this thing -- it will probably be more 5 than one person doing this -- to show that yea, verily we 6 are getting the right coverage _down the side of the thing, 7 DR. POWERS: Is it that sensitive? I mean suppose 8 you got 50 percent of it covered. Can that do the same --

9 do adequate cooling for you?

1 10_ MR. McINTYRE: It will depend on where the 50 11 percent -- if it is-all on one side --

12 DR. POWERS: That's random.

13 DR. KRESS: The trouble, Dana, is the coverage was 14 linearly related to the transfer because they didn't take

( 15 any account for conduction in either the water or the steel, 16 so it's whatever area -- if the water is not on the area 17 it's not conducting any heat.

18 DR. POWERS: Well, what they took credit for is 19 one thing. What the reality is is quite a different thing, 20 and I guess I am asking a reality question, 21 DR. KRESS: In reality I don't think it would be 22 quite that sensitive.

23 MR. McINTYRE: It's not at sensitive in reality 24 but if we are thinking it is 60 percent and in reality it's 25 10 percent, that is a problem.

ANN RILEY & A SUCIATES, LTD.

Court-Reporters 1250 I Street, N.W., Suite 300 Washington, D.C. 20005 (202) 842-0034

______a _ - - . _ . _--

300 1 DR. POWERS: Yes, but if it is -- if you are

() 2 taking 100 percent and in reality it is 90 percent, I don't 3 think that's a p.roblem.

4 MR. McINTYRE: Yes. Circulation and 5 stratification. 'There was some looks at that in the large 6 scale test facility. We are also looking at other 7 information that is available out there. You know, Ivan is 8 constantly sending us the tests that he knows about that we 9 can look at.

10 For environmental effects where it says " surface 11 aging" we haven't kept something out there for 10 years but 12 we have run an awful lot of tests on the surfaces that we 13 have had and we have only found that it has gotten better.

14 This is special paint. It is not Sears' best.

()

It 15 is a -- I have totally forgotten the name of it, but it is a

-16 paint that is designed to work. We are testing the paint --

17 CarboZinc is the name.of the paint -- we actually tested the 18 paint that we expect to be used, that's required to be used 19 on the containment surface.

20 DR. POWERS: That is the external surface that you 21 are talking about?

22 MR McINTYRE: And internal, yes.

23 DR. POWERS: You will also paint the internal?

24 MR McINTYRE: Yes.

25 DR. POWERS: And could you say the name of the O ANN RILEY & ASSOCIATES, LTD.

Court Reporters 1250 I Street, N.W., Suite-300 Washington, D.C. 20005 (202) 842-0034

t 301 1 paint again?

() -2 MR McINTYRE: CarboZinc.

3 DR. POWERS: CarboZinc?

4 MR. McINTYRE: Yes.

51 DR. POWERS: Do you ask for credit for iodine 6 retention on that' paint?

i 7 MR. McINTYRE: No, I don't believe so.

8 MR. CARROLL: You should.  !

9 MR. McINTYRE: That paint is a very good absorber 10 of iodine.

11 DR. KRESS . They don't need it. They have got 12 sprays that will r ke that out --

13 [ Laughter.)- I 14 DR. POWERS: Oh, sprays. I forgot about that --

(,,/

15 and it's a nicely engineered spray _too -- safety grade, 16 right?

17 MR McINTYRE: I want to thank you all for that, 18 yes.

19 DR. POWERS: Maybe with coating they don't need 20 the sprays. Have you thought about taking those out?

21- MR. McINTYRE: We probably couldn't get credit for 22 the coating. I have got that feeling.

23 DR. WALLIS: This test database for circulation 24 and stratification -- I know nothing about it. It's just my 25- impression that you'd need a pretty big database to really ANN RILEY & ASSOCIATES, LTD.

Court Reporters 1250 I Street, N.W., Suite.300 Washington, D.C. 20005 (202) 842-0034

l l

302 1 understand what goes on in terms of circulation and

() 2 stratification in this containment when you have got various 3 initial conditions for the steam and so on and various times 4 of release and all kinds of things happening.

5 Is it really comprehensive?

6 MR. McINTYREt And that is the question, probably_

'7 the biggest question that we have been hashing out --

8 DR. WALLIS: Struggling --

9 MR. McINTYRE: Struggling is another --

10 DR. KRESS: Well, it's probably the major hangup 11 with the subcommittee right now.

12 DR. WALLIS: So I should just wait and see what I 13 get for information?

14 MR. McINTYRE: When we get to what we are doing

() 15 on -- we are doing some work in that area, and that would be 16 your best-tact at this point.-

17 For the passive core cooling tests, again these 18 have all been completed, we ran the core makeup testing or 19 core makeup tank testing at Westinghouse in Waltz Mill, 20 where we looked at the gravity drain behavior of the core 21 makeup tank under different types of postulated 22 circumstances.

23 We did testing in a mockup of a PRHR heat 24 exchanger. It wasn't the design that we have in the plant 25- now. T* caanged from the time we started until the time l

O AW RILEY &-ASSOCIATES, LTD.

Court Reporters 1250 I Street, N.W., Suite 300 Washington, D.C. 20005 (202) 842 0034

303

( 1 that we built the plant. The Staff came back and asked us

() 2 questions on that. We provided some information. I think' 3 that that now has been a resolved issue.

4 In Italy we ran full-scale testing of the 5 automatic depressurization test including the lines from --

6 DR- WALLIS: Wait -- the CMT. I am trying to 7 think about this. The draining of a tank is the simplest 8 . thing in the world.

9 DR. KRESS: Yes, but the issue was that there's 10 steam on top of it.

11 DR. WALLIS: How about the refill by condensation?

12 Were you addressing something like that?

13 .MR. McINTYRE: Refill of?

14 DR. WALLIS: Like getting a vacuum in the CMT. Is

) 15 that also reverse-draining -- is that-tested too?

16 MR, McINTYRE: In the original design.that 17- probably could have happened when we put steam a the top, 18 but what we get into --

19 DR. WALLIS: A-vacuum breaker or something?

'20 MR McINTYRE: No, we took the steam line out.

21 .IMt. WALLISt- Oh, you took the steam line out.

22 MR. McINTYRE: That was a better solution and we 23 now take it off the hot leg so it heats up.

24 DR. WALLIS: So it boils inside or it flashes

'25 inside or --

ANN RILEY & ASSOCIATES, LTD.

O _ _ _. Court Reporters 1250 I Street, N.W., Suite 300 Washington, D.C. 20005 (202) 842-0034

1 l

304 1 MR. McINTYRE: No, it's -- the top of the tank is

() 2 hooked to the hot leg.

3 DR. WALLIS: Yes, that's right.

4 MR. McINTYRE: So --

5 DR. WALLIS: So that is the way it drains, right?

6 MR. McINTYRE: When it.is actuated, you start to 7 bring water in. Those are the same types of questions that 8 the-Staff had.

9 DR WALLIS: All right, and it's all been

-10 resolved.

11 MR McINTYRE: All been resolved. I think that 12 particular aspect is not an issue at this point. Alan?

13 MR LEVIN: This is Alan Levin from the Staff.

14 We don't have any open items as far as I am aware 15 on the CMT.

16 MR McINTYRE: But there will be information on 17 that included in the pile that Paul is going to ship you.

18 DR WALLIS: A lot of stuff.

19 As mentioned, the' ADS testing was one.to confirm 20 that the sparger worked as expected, that the valves worked 21 as expected, and to get some information so that we could 22 determine the pressure drop through the lines into various 23 two phase flow conditions.

24 We ran two integral systems tests. There was the 25 tall, skinny facility, the full height, full pressure ANN RILEY & ASSOCIATES, LTD.

Court Reporters 1250 I Street, N.W., Suite 300 Washington, D.C. 20005 (202) 842-0034

305 1 facility in Italy. That was one-395th scale and if you 2 looked at it, it was really tall and really skinny, but that 3 was a model of the AP600.

4 We ran a-larger scale test with low pressure at 5 Oregon State University. It was a quarter dimensionally 6 scaled -- that facility is also referred to as APEX and it-7 is now being run by the NRC. They are doing their 8 confirmatory testing there, and all that has been put 9 together into the passive core cooling testing program. '

10 The codes that we're using, this was different for 11 us, the AP600, because we've modeled all containments up 12 -till now as one single 3-million-cubic-foot thermal I 13- equilibrium homogeneous node, because it dicin't really-14 matter, you had the sprays -- we have sprays, too, but we 15 don't take credit for them -- for pressure reduction, and 16_ with that the modeling was really quite simple, 17 With the passive systems without sprays then it-18, gets a little more complex. We needed to have multi nodes 19 and we bought _ the-GOTHIC program, then we made some 2 0 -- modifications to it,-hence the W, still our Westinghouse in

'21- front of GOTHIC, And we have been discussing-that, hashing

[22 out, struggling, whatever word you want to use, with the 23- staff now for three or four years, and with the Thermal 24 Hydraulic Subcommittee.

'25 We think we've made a lot-of progress in that O- ANN RILEY & ASSOCIATES, LTD.

Court Reporters 1250 I Street, N.W., Suite 300 Washington, D.C. 20005 (202) 842 0034

306

1 area. We have some things I'll talk about in a minute that f

() 2 _ we're going to be turning in to the staff that will resolve 3 .the questions that you guys had in December. .

4 For large-break LOCA and long-term cooling we're 5 -using the Westinghouse COBRA / TRAC code, which was verified '

6 against the testing program that we've run. We have 7 reso?ved I think all of the staff; questions and concerns on 8 using WCOBRA/ TRAC.

t 9 For NOTRUMP, NOTRUMP was our post-THI small-break 10 LOCA code, but for current plants, small-break LOCA is a f

11 very high-pressure phenomenon for the AP600 because of the 12 ADS system. A small-break LOCA is a low-pressure 13 phenomenon. We've had to do an awful lot of work in that-44 area. We're still doing an awful lot of work in that area

() 15 to verify and validate NOTRUMP at the lower pressures and to 16 show that we a, 1culate the phenomenon approximately, and

-17 I'll talk about that in just a couple --

18 DR. WALLIS: .You've done, I take it, the same sort 19 of things that the staff has done with their codes in i

20 comparing.with data from these facilities like SPES and 21 so --

22- MR. McINTYRE: Yes.

23- DP.. WALLIS: And I've never seen any of that 24 either.

25 MR. McINTYRE: It's in the file, ,

L ANN RILEY & ASSOCIATES, LTD.

Court Reporters 1250 I Street, N.W., Suite 300 Washington, D.C. 20005 (202) 842-0034 l-

u. _ -_ ,, _._ .. - __ . _ . _ - -- -- -- - ---

307 1- DR. WALLIS: Okay. And your codes agree with the

() 2 data and with the NRC codes?

3 MR. McINTYRE: Well, I don't know if they agree 4 with-the NRC codes. We think that they agree with the data 5 adequately.

6 DR. WALLIS: That doesn't mean anything.

7 DR. KRESS: Well, they have a criterion for what 8 they mean by adequate.

9 DR. WALLIS: Yes, I didn't like that either.-

10 MR. McINTYRE: Well, it gets down-to how good is 11 good enough, and --

12 DR. WALLIS: Exactly.

13 MR. McINTYRE:

And if something is coming down and 14 you're off by 500 seconds in when something is going to

() 15 happen, is that important, is that not important? So if 16 you're going to look at it and expect the measured versus 17 predicted and everything's going to be 1 for the entire time 18 period, no. Is it adequate? We think so.

19 DR. WALLIS: Well, yes, I don't know how the rest 20' of the committee feels, but I don't get a good feeling when 21 people feel something is good or it looks adequate, and 22 there must be some better way of making a decision. Maybe 23 I'm under some illunion here.

24 DR KRESS: Well, I think they actually prepared

25 the --1 calculated curves for the various test sequences with ANN RILEE & ASSOCIATES, LTD.

Court Reporters 1250 I Street, N.W., Suite 300 Washington, D.C. 20005 (202) 842-0034

308 1 the test data and set up criteria on how well the I( ) 2 calculacions fit within the uncertainty of the data and set 3 several criteria related to the uncertainties in the data 4 and the uncertainties in the calculation. And you can s 5 debate whether or not they fixed the right criteria, but in 6 principle that's the right way to approach without ass.ing

7. whether or not a code is sufficiently valid as to deal with 8 its uncertainties and the beta uncertninties and to rectify 9 the two in terms of how good they mesh.

10 DR WALLIS: I guess my problem was not with the il philosophy but with this sort of the quantification of-what 12 we're doing, so that the match was.actually done in a way 13 which was as rigorous, let's say, as possible, rather than 14 being based on how people felt about it.

() 15 MR. McINTYRE: No, we've tried to, you know, 16 present the data, and then you look at it and you say oh, 17 why does it look like that, and is that adequate, good 18' enough, horrible? .You know, there's -- I think we had 19 horrible as one of our criteria, but actually it was back at 20 Westinghouse before we fixed the codes -- oh, that's 21 horrible, and we would then-look and see what we needed to 22 change or improve.

23 The good news is for the-LOFTRAN computer code, 24 which are the non-LOCA transients, the AP600 is not 25 significantly different, and that's what we used for the O ANN RILEY & ASSOCIATES, LTD.

Court Reporters-

'1250 I Street, N.W., Suite 300 Washington,- D.C. 20005 (202) 842-0034

309 1 chapter 15 analysis, and another version of-it for the tube 2 rupture, and those have been completed and turned in. The 3 staff has completed their SER, and I don't think that there 4 are any open issues in that area.

5 DR. KRESS: Our subcom-ittee hasn't looked'at G LOFTRAN, as far-as I know.

l7 MR. McINTYRE: I don't believe so.

8 DR. KRESS: Should we?

9 MR. McINTYRE: Should you? You'7.e asking me.

10 [ Laughter.)

11 MR. McINTYRE: Given that we are not using it 12 outside of its ranges, and it's been, you know, what we've 13 been using for operating plants since before I came to 14- Westinghouse, it's gone through obviously some iterations,

-15 you know, since then, but I don't think that it's something 16 that the --

17 DR. KRESS: Does it call into play the passive 18 ECCS systems?

19 MR. McINTYRE:.- I think no.

20 DR.-KRESS: You don't?

21- MR. McINTYRE: Because it'a not an ECCS, I think

'22 the only thing it models is the PRHR heht exchanger.

23 DR. KRESS: Um-hum.

24 MR. McINTYRE: That's the only thing that would be 25_ different. And we did do some comparisons with LOFTRAN to ANN RILEY & ASSOCIATES, LTD.

Court Reporters 1250 I Street, N.W., Suite 300 Washington, D.C. 20005 (202) 842-0034

310 1 the experiments, yes, but also I have the CMT would be

() 2 something that would be actaared. So if you look at the 3 primary codes that we're 1.:aking at and what test, that 4 gives you-an' idea of the types of feedback from the testing  ;

5 program'actually into the codes.-  !

6 DR. WALLIS: I can understand single-effects .

.7 tests, PRHR,-CMT, AD. How do-you do a single-effects 8 containment test?

9- MR. McINTYRE:-

A single-effects test for that?

10 DR. WALLIS: How do you do a single-effects 11 containment --

12 MR. McINTYRE: With a six-foot-tall, four-foot 13 wide heated plate.

14 DR. WALLIS: But inherently it's a 15 multiple-effects --

16 MR. Mc1NTYRE: Yes, but you start by getting some 17- single-effects data to look at and then build from that. ,

18 DR. WALLIS: Which would be to me when you're 19 saying what's happening in the boundary layer on the wall or 20 what's happening in this or --

21' MR. McINTYRE: Yes, that's more -- this was 22 more --

f ,

l 23 DR. WALLIS: It's a composite of many effects,.

24. really.

25- MR. McINTYRE: Those were the original ones where-i O ANN RILEY & ASSOCIATES, LTD.

Court Reporters 1250 I Street, N.W., Suite 300 Washington, D.C. 20005 (202) 842-0034 >

m.

311 1- the heat-transfer-tests in the webbing -- things that are in

() 2 'hore are things like this-water-distribution test. That's a 3 single-paran.eter ef f ects test. That was when you set the 4 boundary --

5 DR. WALLIS: I think that if you look at say 6 boundary layer on the wall you probably find you're at a 7 really number never before tested by man or beast.

8 MR. McINTYRE: I can't comment on-that.

9 The open issues. The good stuff. Stratification.

10 We have two open-items in this area. We've-turned in L13 responses to the staff. They're in the process of reviewing 12 it. We.are in the process right now of revising the scaling 13 report to account for the comments from the staff and also

-14 from the ACRS concerns from-the December meeting, and we

() 15 expect to have that in to the staff and the subcommittee 16 this month.

17 MR. CARROLL: Isn't the first bullet an I'2S 18 concern?

19 CIIAIRMAN BARTON: Yes, it is.

20 MR. CARROLL: Okay.

21 DR. KRESS: And we have additional questions. 7 22 don't know if you'd call them open issues. I'm sure the 1

23 staff _has raised some open issues.

24 MR. CARROLL: Has the TH Subccmmittee reviewed or 25 looked at the document that is referenced up there in the L

O ANN RILEY & ASSOCIATES, LTD.

Court Reporters 1250 I Street, N.W., Suite 300 Washington, D.C. 20005 (202) 842-0034

- l.

312 1 first bullet?

h 2 Mk. McINTYRE: To the responses?

3 MR. CARROLL: Yes.

4 MR. McINTYRE: I don't know if you have -- well, 5 we obviously don't know what you -- the royal "you" -- have J 6 received.

7 DR. KRESS: I don't recall reviewing anything --

8 reply to RAI for the first bullet. I'm sure I haven't seen 9 it.

10 DR. WALLIS: What is the approach here? Again, 11 I'm new to thic committee. Do we wait for the staff to do 12 something before it comes to us? It may be that the staff 13 is not competent to make -- to review this issue.

14 DR. KRESS: They're competent to do that, but I 15 don't --

sometimes it happens we wait, and sometimes we get 16 it before, and sometimes at the same time. I don't know if 17 there's any fixed --

18 MR. McINTYRE: I think Paul would like to say 19 something here.

20 MR. BOEHNER: Well, I was going to ask a question.

21 These responses -- are they something you just submitted 22 recently, because I'm not familiar with those two, with 23 these two open items that, you know --

24 MR. HUFFMAN: This is Bill Huffman of the NRC 25 project staff. According to the technical reviewer, those lf ANN RILEY & ASSOCIATES, LTD.

Court Reporters 1250 I Street, N.W., Suite 300 Washington, D.C. 20005 (202) 842-0034

313 1 came in around the 23rd of January, c

h 2 DR. WALLIS: What I have seen is some very mild .

3 criticism by some ACRS consultants. I don't see any 4 response to that here. Do you respond in some way to that?

S MR. McINTYRE: Those I believe are in these 6 response --

7 DR. WALLIS: Is that hot air or something?

8 MR. McINTYRE: Is which hot air?

9 DR. F.ELIS: Is the consu3 tant's comments hot air, 10 or do you respond to that directly?

11 MR. McINTYRE: Directly to the consultant? No.

12 DR. WALIIS : No? Directly to the ACRS?

13 MR. McINTYRE: No , the staff looks at the comments 14 by the ACRb and decides which of those are ones that they 15 would like to turn in t questions, send them to us, and we 16 send in responses.

17 DR. WALLI I'm not concerned about the loops 18 here. I mean, suppose Dr. Zuber makes a comment and '

19 suggests that you shouldn't use this equation, you should 20 use that one. Do you immediately run off and use it?

21 MR. McINTYRE: No.

22 DR. WALLIS: All right. Do you say we're not 23 going to use it, or -- I'd like to know before we come 24 around the loop again a month later and you've used 25 something which a present ACRS member may not believe is ll ANN RILEY & ASSOCIATES, LTD.

Court Reporters 1250 I Street, N.W., Suite 300 Washington, D.C. 20005 (202) 842-0034

I 314 1 appropriate.

G i] 2 MR. McINTYRE: We make -- if you'll -- if you 3 haven't been to the meetings, they are spirited 4 discussions --

5 DR. KRESS - But cordial, right?

6 MR. McINTYRE: I don't think that I would go that 7 far.

8 And then -- and there are a diversity of opinions 9 that you hear. You know, you can hear this is total baloney 10 and you can't possibly do it that way and you're wrong, 11 Westinghouse, and don't ever come back in here again.

12 There's some of thst. There's also some of-gce, that's 13 really a pretty good job, and sometimes you're talking about 14 the same issue, f,_,.)

(_) 15 IH< . WALLIS : Well, I don't want you in any way to 16 feel defensive about this, and I --

17 MR. McINTYRE: Oh, not at all.

18 DR. WALLIS: And I just want to feel.that we are 19 trying to resolve the issue, and I just want the mechanics 20 not to be so tortuous that it takes too long to do it, 21 because if you get -- and so on, trying-to ask questions 22 about how do we facilitate the communication so that you 23 don't go away and do some work and then come back and it 24 goes through the staff. and by the time it gets up here I 25 don't go and say I don't believe that that was the right way

\ ') ANN RILEY & ASSOCIATES, LTD.

Court Reporters 1250 I Street, N.W., Suite 300 Washington, D.C. 20005 (202) 842-0034 s

315 11 to'do it.

()

2 ':I just want'to avoid a-long,_ tortuous route to resolving something. I'm groping for how do's, . void people 4 doing work-_which-is-not directly. effective, that's all. I'm 51 notL-- maybe the rest of the committee could help me. -I 6 suspect there may be another very_ good procedure for various 7 -_ people-getting involved in stirring up this --

8 CHAIRMAN BARTON: -Let me try to help here. .I 9 . think through the subcommittee meetings with Westinghouse 10 issues have been raised by this committee, and they've been 11 documented, and Westinghouse is_ going to respond toLthose-12 issues. I think until this committee - -till the members of 13 this committee are satisfied with the responsc, un will  ;

14 . carry it:as an_open item.- -I.said yesterday, as we go

) ~15 - through the-review of these chapters, we will issue interim

-16 reports that have ACRS committee open issues identified in 17 the-interim report. We won't sign off on AP600.until those 18 issues are resolved to the satisfaction of1this committee.

19 Now, how long it takes staff or Westinghouse to 20 get back with those answers, hopefully, it is on the same 21 track that=we are on, to try to get'the certification-done 22 .whenever we said we were going to do-it, sometime this 23 summer, is the current target. And if it doesn't happen, we 24 just will carry that as an open item, or it will be in our 25 -certification letter.

ANN RILEY & ASSOCIATES, LTD.

Court Reporters 1250 I Street, N.W., Suite 300 Washington, D.C. 20005

_ (202) 842-0034

-w

316 1 DR. KRESS: And, in general, when we raise issues ll 1 or questions, the staff has been responsive making decisions 3 as to whether they have already had an answer to that, or 4 whether it is something they still need information on, and 5 including that in their request for additional information.

6 And, which sometimes carries a lot more weight than our just 7 asking the question.

8 DR. WALLIS: Well, I guess -- again, maybe you can 9 -- again, I am not familiar with how you operate, but I 10 would like to see a situation where if Dr. Catton has some 11 criticism or some comment of the type I have seen here, that 12 an answer could be furnished quickly instead of several 13 months later, I mean someone -- someone who is really on 14 top of it at Westinghouse maybe could reply right away, Or 15 is that something that is impossible?

16 MR. McINTYRE: We find it is very easy to ask 17 questions. And some of the questions that get asked do not 18 have five minute answers. A matter of fact, some of the 19 questions that Dr. Catton has, which are -- like your whole 20 methodology stinks -- that is not a five minute answer.

t 21 DR. WALLIS: Well, that is not a fair -- that is 22 not a fair question.

23 DR. POWERS: That is not a question.

24 MR. McINTYRE: Well, there have -- that is not a 25 question.

h ANN RILEY & ASSOCIATES, LTD.

Court Reporters 1250 I Street, N.W., Suite 300 Washington, D.C. 20005 (202) 842-0034

317 1 DR. APOSTOLAKIS: But what is not clear to me is-

)( ) '2 when we raise questions, they go to Westinghouse through the i

'3 NRC staff or directly?

4 DR. KRESS: Both.

5 DR. APOSTOLAKIS: Noel.

6 MR. DUDLEY: This is Noel Dudley from the staff.

7 The Committee responsibility is to report and advise the

-8 Commission on technical issues before it. Now, the 9 technical issue before it in this case is the AP600 design.

10 We are getting information both from Westinghouse directly 11 and from the staff concerning the staff's review of the 12 Westinghouse documentation.

1 13 The formal documentation for presentation of what 14 the Committee issues are is a letter either to the EDO-or-to O

k_,/ 15 .the Chairman, and it is advice to the Chairman and to the 16 -EDO on the consensus that is built- by the Commission,- cnr by 17 the Committee. Therefore, a single consultant to the ACRS 18 is not carried forward immediately, it has.to go through a 19 full Committee ACRS meeting, discussed with the ACRS, 20 consensus reached, and then a letter prepared.

21 DR. APOSTOLAKIS: Okay. So what you are saying is 22 that if we have any issues, they will be listed in the 23 letter to the EDO, example?

24 MR. DUDLEY: Yes.

4

-25 .

DR. APOSTOLAKIS: But we will not direct them to ANN RILEY & ASSOCIATES, LTD.

Court Reporters 1250 I Street, N.W., Suite 300 Washington, D.C. 20005 (202) 842-0034 1

318 1- .Westinghouse?

f) 2 -MR._DUDLEY: That is correct.  ;

3- DR. APOSTOLAKIS: EDO will decide what to do with 4- them.

-5 MR. DUDLEY: That is correct.

6 vA.;APOSTOLAKIS: Okay.

7 DR. WALLIS: Well, I am -- I am concerned ~with-8-- 'something else. How do you get the thermodynamics right?.

9 .And I don't really -- I am-trying to answer that question.

10 I don't_really' care about the politics. If Dr. Zuber has a 11 certain thermodynamic equation and-you have a certain '

-12 thermodynamic equation, and theyLare different, then maybe 13' you are both wrong, but you can't both be right.. And that 14L -is not a political question, it.has to be something which l 15 can be resolved, and someone has really worked it out, could 16 be' resolved in day- ,- And it seems rather preposterous to 17 spend months of going around--some bureaucratic system in

'18 order-to resolve' thermodynamics, which'could be resolved-in 19- a day,

20 DR. POWERS: But maybe that -- this is an issue 21_ you need -- we need to address internally. I am not Brian-L 22 is in a position to address this.

23 - MR. McINTYRE: I can't --

24 DR. WALLIS: No, no. This is.something that can 25 not wait. I mean it seems to me, it-just -- it will be O ANN RILEY & ASSOCIATES, LTD.

Court Reporters 1250 I Street, N.W., Suite 300 Wachington, D.C. 20005

,_02) 842-0034

319

-1 silly for you.to use some thermodynamics, have Dr. Zuber,

( 2 say, impose some thermodynamics on you, which then turns out 3; 'three months later-to-be inappropriate. There has got to be 4 a way to do it better than that.

5 .MR. McINTYRE: It has -- I would agree with that.

6 I can give you -- I think this will help some of your -- it 7 -is not going to make you feel any better, but I can, at 8 least I will explain where we are with these.  !

9 One of the issues.has bee" ATWS for the AP600. We 10 completed a report and got it in to the staff last month, 11 and they are in the process of reviewing that.

12- From-the December Thermohydraulic Subcommittee 13 -meeting there'were a number of issues that came up that were 14 testing related. And what the staff did is they turned 15 those into open items to us, they sent them to us, and we 16 are_ going-through, providing_ answers to them. And, I-agree,-

17 there probably could be better ways if we had a more direct l 18 interchange. We need to go back. Some of these are very 1 19 .hard questions. Lik.e I said, it is easy to ask the 20 question, it is very, very dif ficult to get the answer, 21 particularly when we are not directly interchanging with you 22 except every other month when we come into this room and 23 have our cordial discussions.

24 We are expecting to-have these answers to the 25 staff and they will then forward them. We send copies to ANN RILEY & ASSOCIATES, LTD.

Court Reporters 1250 I Street, ii.W., Suite 300 Washington, D.C. 20005 (202) 842-0034

320 1 Paull. Paul will: forward them to you. That -- we.,think we

() '2 3-are addressing.what the questions and comments were on the

.PIRT scaling report.

4 LFor ADS testing,;we are going'back and I -- I

.5 asked about this one, because this was Virgil Schrock's 6 comment,Ron whether or not that is right. We think it is 7 and_.this is a place where we are going to have, I believe, a

-8 technical difference of opinion. But this -- I understand,

.9 talking to Bill Brown, this is the approach, and Mike Young, 10 :this is going to be our proposal, so we will be-back in and

-11 this is going to be our answer. You are going to get this answer in to the staff in February, and then we will send it 13 to you.

-14 DR. KRESS: There was another aspect of-that, and 15 that had-to do with the fact that the flow-process.to the 16 ADS valvec is a Fanno process and there was some question 17 about how you reduced your test data.

18- - M R.,' M c I N T Y R E : .Right.

19 DE. KRESS: And actually_ determined what the 20 stagnation conditions were.

21 MR. McINTYRE: Yes. Right.

22 DR. KRESS: Is that part of your, what that --

.23 MR. McINTYRE: That's it. That's what that means.

24 DR. KRESS: Okay. Well, that's a little different 25 than the-momentum flux thing. Those are two separate O ANN-RILEY & ASSOCIATES, LTD.

Court Reporters 1250 I Street, N.W., Suite 300 Washington, D.C. 20005 (202) 842-0034 I

m )

321 1

1 issues. You could have --

( 2 MR. McINTYRE: Of whether or not he was using the-3 stagnation pressure.

4 4 D:R._KRESS: Right. Whether --

-5 MR. McINTYRE: Right.

6 DR. KK .d : You know, he could be using the 7 correct Fanno equation and may rot have reduced the. data

-8 correctly. It's two different questions actually.

9 MR. McINTYRE: Yes,:we are going-back and making 10- significant revisions to that report.

11 For COBRA / TRAC, this in not an AP600 issue, but it 12- was one that we were taken to task for, in December, as to l

13 whether or not we had the documentation _ completed for the 14- operating plants. It is my understanding that Westinghouse,

(^")\

(_ -15 another part of Westinghouse, is committed that that will'be l 16 in the staff's-hands by the end of March.

17 In the NOTRUMP, which is, I meutioned, the other.

18 area where you have probably the most discussion with the E19 - Subcommittee,-we are preparing answers for the documentation

-20 of the differential equations, the changes. We expect the 21 report will be in, along with the FSER changes -- the FSER 22- response to the questions the staff has asked in February, t

23 and that:will also address the flooding-and the pressurizer 24 -surge line, and ADS and how the IRWST penalty is calculated.

25 'DR. WALLIS: Is this,-the flooding and pressurizer

~

ANN RILEY & ASSOCIATES, LTD.

Court Reporters 1250 I Street, N.W., Suite 300 Washington, D.C. 20005 (202) 842-0034 1

3'42 1- . surgi.-line,'is that based upon experimental data?

2- MR; McINTYRE: Some of:it will-be.

. DR. WALLIS: Some of it will be.-

4 MR. McINTYRE: Our surge 11ine is,-if you-look at'

-5 ;the plant,-it is_not just a straight line, it-is - -no, it

-6 ~;is not even that_ good. It-is like a corkscrew. hat's the

-:7 way it --

8- DR. SEARLE: A dysfunctio1al corkscrew.

9- MR. McINTYRE: That's-the way --

10 CHAIRMAN BARTON: 'Does'it work?

11 M R'.. M c I N T Y R E : Yes. We believe that'it does. Our C 12- recalculating,-flooding there appropriately when the ADS is

-13 open, they are looking at that.

- 14 DR. KRESS: There was a question-of whether-you-15:- -were'using the correct critical flow model also.

MR1 McINTYRE: -Right.

17 DR. KRESS: -Is that still-in the' report?

-18. MR. McINTYRE: That -- I don't know. I-would'have 19- to go back and look. And in LOFTRAN there are no open=

120 items, no'known open items in the LOFTRAN. So'in ---

'21 We're done with.our testing-program. We think 22 that it's adequ' ate to support the - -'the design

~ certification application. We're in the-process of; 24 responding to-the--last open items and -- and'the ACRS 25 concerns, and we expect those will be done in February. I O ANN RILEY & ASSOCIATES, LTD.

Court Reporters 1250 I Street, N.W., Suite 300 Washington, D.C. 20005 (202) 842-0034

323

.1 believe there is one report that's done like the first week

() 2 in March. I'm trying -- just trying to mentally remember 3 all the dates, and that'will be in to the subcommittee and 4 'm sure we'll probably meet again to talk about it.

5 DR. WALLIS: Now, these ACRS concerns are the ones 6 you already know about?

7 MR. McINTYRE: Yes.

8 DR. KRESS: There are certainly more questions and 9 concerns than were on this list.

I think I would count 10 maybe a dozen each for the RCS and the containment, but not 11- all of them are serious. They're more -- a lot of them are

- 12 questions, and-some of-them could even-be characterized as 13 - being whining on the part of the subcommittee but -- but --

14- and I don't consider them to really be pathological with O( / 15 respect to this second bullet, but wnere they might be 16 pathological is -- is with respect to the schedule, and so 17 what I intend to do is put together what I think is a 18 complete list of the-ACRS concerns -- the subcommittee's 19 concerns and discuss them with the full committee during 20 this meeting and-perhaps be-sure they're -- so be sure-you 21 have the same list-that we do and -- and some indication of 22 which ones I think-or the committee thinks is more important 23 than the others.

24 CHAIRMAN BARTON: Because the staff -- do you have 25 the letter from the staff showing you which ones they O ANN RILEY & ASSOCla.ES, LTD.

Court Reporters 1250 I Street, N.W., Suite 300 Washington, D.C. 20005 (202) 842-0034

324 1 thought that'they passed'along to-us?

2 DR. KRESS: -I have that list.

-3I CHAIRMAN BARTON: Okay.

4 DR. KRESS: And then -that will Ime considered in my I 5 --

6 CHAIRMAN BARTON:: Okay.

~

DR. KRESS: -- development;of --nof the_ discussion 8- for the' full ACRS --

9 -MR. CARROLL: Hopefully you will include those 10 itema that the committee agrees on-in our interim report?

11- DR. KRESS: - Your interim 7;eport .

.12 MR. CARROLL: On review of --  !

13 DR. KRESS: So, just -- just -- -

14 MR. CARROLL: -- of this material.

15' DR. WALLIS': So we have to. agree on-this --'this c16 in the same month in which you are going to supply all_the 17 answers?-

L18. MR.-McINTYRE: No.

19 DR. WALLIS: .That's-a remarkable turn-around.

20'- MR. McINTYRE:- I doubt if you will do it this-21 month.

22 :DR. WALLIS: We will or you will?

23' -MR. McINTYRE: We will.

24 DR. KRESS: That's why I'm concerned about the -l 25: schedule.

-( ) ANN RILEY & ASSOCIATES, LTD.

Court Reporters 1250 I Street, N.W., Suite 300 Washington, D.C. 20005 (202) 842-0034 u

s 3251 '

1 DR. WALLIS: ~ That 's - February, 1998?

) 2- MR. McINTYRE: Yes, it is.

3 MR. CARROLL:- Mr. Graham, I think one point 4 nobody's made is that.atsubcommittee of ACRS does not really 5 recommend anything or do anything. They-are fact finding to 6- Lpresent their views to report to the full-committee. -So, 7 it's only after full committee deliberation does any 8 official position of ACRS'come out. Sometimes subcommittees 9~ tend to forget that and -- and tell people do this, do that, 10 and they're not supposed to do that.

11. Your second bullet. We believe the AP600 testing 12: program is adequate. Shouldn't that read will be shown to 13 be adequate?

14 MR. McINTYRE: _ No. Is is the right word.

15 MR. CARROLL: Because you believe it?

16 MR. McINTYRE: 'We believe it. We've even -- the-17 staff has looked at it for their own purposes and -- and

.18 said that it's - that it's adequate. ILjust-got a paper by

'19 an NRC consultant saying-hey,-it's a -- it's a good testing 20 ' program. He wants to know if we can release that. We said-21 yes, certainly that one can be released.

22 MR. CARROLL: Okay.

23 MR. McINTYRE: We think that it is an adequate 24 program.

25 DR. KRESS: I'd say it is a pretty extensive ANN RILEY & ASSOCIATES, LTD.

Court Reporters 1250 I Street, N.W., Suite 300 Washington, D.C. 20005 (202) 842-0034 A

326 p 1 program.

. 2.. 'MR, CARROLL: Oh,.I understand.that. I just -- I 3- think because there are:s'till some open issues --

4; MR. McINTYRE: Well, I was talking about the 5 testing program' .

6 DR . -- KRESS : ' Yeah. I think the open issues are-7J more in dealing with-the way they do --

8 MR. McINTYRE: 'It's what you're doing -- what 9; we're doing with the code --

'10 DR. KRESS: Well, there is some scaling questions, il but I don't think they're. pathological. I think they can be-12 revised.

13 MR._ CARROLL: Okay.

14: CHAIRMAN BARTON: Are there'any further questions s,) 15 of Brian on the-test' program at this time?

~16- [:No response.)

17= CRAIRMAN BARTON: Does the staff have.any comments *

18. .on the testing _ program?

19 --MR. McINTYRE: _ Yeah, before you_go to the testing,

'20 'I,just -- when are we doing Chapter.15 on your schedule?

21 CHAIRMAN BARTON: Which, 15?

22 MR'. McINTYRE: . Yeah, and 6.

f 23: MR. DUDLEY: Right now that's May.

24 MR. McINTYRE: _Okay.

'25' MR. DUDLEY: That's still negotiable, though.

A ANN RILEY & ASSOCIATES, LTD.

Court Reporters  ;

1250 I-Street, N.W., Suite 300 Washington, D.C. 20005 (202) 842-0034

327 L '

1 DR. KRESS: That's where a lot of this will show

() -2 up . .

3 MR. McINTYRE: -Right.

4 CHAIRMAN BARTON: We have -- we have six in April 5- it looks.like.

6; DR. KRESS: Yeah. Six in April, 15 in May.

7 CHAIRMAN BARTON: Six in April, 15-in May.

8 MR,-McINTYRE: Okay, so that's the time that most 9 of the accident analysis really needs to be done and agreed i

10 to, which means that we need to get the stuff to you, you.

11 need to look at it and we need to come down and d.iscuss it. j 12 DR. WALLIS: Can I ask something? -I'm just:trying

-13 to figure out what's. going on here. You have done this 14 -AP600 testing program. I understand that NRC is still doing 15 tests?

16 MR McINTYRE: Yes.

17- DR..WALLIS:- So,. where does that fit in? I mean,

-18 -suppose they-come up with some new-concern as a result of 19- tests-now in progress?

20 MR. McINTYRE: The tests that we ran were -- the  !

21.  : . tests that they are running are the-beyond the design basis 22 l -- what happens if this doesn't work and this doesn't work 23 I-and=this doesn't work and this doesn't work, the ones that 1

24 l we were looking at were the design basis type testing.

So, 25  ! any tests that we've had to rely on to do design I

i O

i l

ANN RILEY & ASSOCIATES, LTD.

Court Reporters 1250 I Street, N.W., Suite 300 Washington, D.C. 20005 (202) 842-0034 l

328 1 certification they told us to go run, and that -- that's j 2 .what was important to have them on' board to run them, 3 :DR. WALLIS: I guess the design basis, again, is 4 this a philosophy-if you so specify that_only thece things 5 are-likely_to happen or only_these are the things that we 6 have to officially worry about happening, Is that what-7 design. basis is?

P MR. LEVIN: If I can jump in here for a second --

9- this'is Alan Levin, _ Westinghouse is required to assess-10- plant behavior beyond the design basis in the context of the 11 PRA, and they are required to do analyses -- and a lot of 12 the confirmatory testing has been done, which is done by '

13 research at NRR's request -- is to confirm that A, 1 14 Westinghouse hasn't missed anything significant and -- and:

I 15 in the original justificat.on for performing a confirmatory

~16 test program, we committed to the Commission and informed 17 Westinghouse that if we did identify anything4 that we felt 18- significant'that Westinghouse had missed, we would-require

-19 ~ -

them to do additional-testing within the design 20 -

certification test program to resolve those issues, 21 ' The confirmatory testing'for beyond the design 22 basis is-to confirm that their analyses with codes like MAAP 23 also are basically -- accurate's a little too strong a word, 24 but at least they're, again, they're not missing.anything.

25 There are no cliffs to fall off right-beyond the design O ANN RILEY & ASSOCIATES, LTD.

Court Reporters 1250-I Street, N.W., Suite 300 Washington,-D.C. 20005 (202) 842-0034 l

l-329 1 -basis,-and that's sort of been the guiding philosophy with a

()

2 lot of the confirmatory test program. So, it -- ic'stnot 3- quitezso cut'and dried as, you-know, design basis beyond 4 design basis. There's kind of a gray area d

., between there.

5 DR. WALLIS:_ I-guess I'm-getting at the question-6' which is a cosmic question'rather than-a_ question of your 7 immediate concern. If I wanted to test a sports car, I 8 would have maybe the type of NRC tests where you drive into 9 a brick wall or'you drive-it around the curve at a certain

'10 speed with certain specifieu grade and'a certain ---and then 11- _you put salt -- put: soap on the road or something and you do 12 1all these things which somebody has said you should-do, but 13 really, if I wanted to buy this car I'd probably want to put 114' some person in it and have them drive around and see if they 15 -could make'it have a -- an accident which had not yet been 16 -anticipated, and that doesn't seem to be ever done. These-17 seem to-be all-very specified things.

-18 MR. McINTYRE: Well, I think those are closer to

=19 the tests that,-you-know --

20 DR WALLISi But more likely to happen =to a 'real-

-21 reactor is someone driving it --

22 MR. McINTYRE: I guess --

-23 DR. WALLIS: -- or:something that --

24 MR McINTYRE: 'Well, let's just pursue your --

25 DR. WALLIS: ---wasn't in the design basis.

ANN RILEY & ASSOCIATES, LTD.

Court Reporters 1250 I Street, N.W., Suite 300 Washington, D.C. 20005 (202) 842-0034

330 1 MR. McINTYRE: -- your analogy because I like Q

fy' 2 this. This is fun. I would say that we ran the -- the 3 . drive it into tne brick wall.

4 DR. WALLIS: Yeah.

5 MR. McINTYRE: And Alan ran the let's drive it 6 around real fast around a corner and have a beer or two and 7 see if -- what really crazy things could happen to it, but 8 did Alan --did we find, you know, each and every possible 9 perhaps thing that could happen? No, we didn't, but I -- it 10 was approached from a reasonable standpoint. Is that fair, 11 Alan?

12 DR. WALLIS: But then you drive it around and you 13 find that these radial tires that sort of changed their 14 geometry in some way and you didn't know that unti.1 you

) ~

experienced the wobble in the steering.

16 MR. LEVIN: I understand Dr. Wallis' concern, and 17- my comment is that.-- that to a certain extent, this is --

18 this is sort of a philosophical argument I think, more than 19 it is a technical argument. To a certain extent, you're 20 constrained in your testing by having to be able to 21 characterize things reasonably well so you can analyze them.

22 You -- you can't just, you know, punch a hole in the system 23 somewhere and see what happens because if you want to model 24 the thing, you'd like to have a well defined brake. You 25 know what the brake size is, you know where it is. You know

() ANN RILEY & ASSOC 7.ATES, LTD.

Court Reporters 1250 I Street, N.W., Suite 300 Washington, D.C. 20005 (202) 8 2-0034

331 1; how to modmiswhat you've got, and so there's -- the point of

() 2 this -- this is not -- these things are not AP600

'3 simulators.- It's to develop computer-code models to predict 4 the way?the plant's going to behave, and to do that, you've .

5 _got to haveLreasonably well characterized tests. I think if 6 you look at the broad spectrum of testing, including what 7 Westinghouse has done and what the staff has done in the 8- reactor systems over a range of_ scales over a range of brake 9 sizes, locations, thermohydraulic conditions and so forth, 10 there's a pretty reasonable coverage'of the map. You might 11 be able to find some other. places to -- to kick the tires 12 and-slam the doors, but on the whole, I think it's a

-13 reasonable coverage.

14 DR. WALLIS: I'd like to have sort of the test 15 '. pilot, if you were to take this thing and see if he could 16- make it get into a situation where you'd have a serious =

17 accident. That's what you would do with an airplane.

18 DR.=KRESS: The trouble is, you have no airplane E19 to fly.

20 DR. WALLIS: You have--to have a really good 21 simulator.

22 DR. KRESS: Well, the simulators don't do it,

'23 either. What you're touching on is the -- the historical 24' problem this committee has had with design basis space 25 l versus reality-for years and years, and the philosophy there l

1 l

ANN RILEY & ASSOCIATES, LTD.

Court Reporters 1250 I Street, N.W., Suite 300 Washington, D.C. 20005 (202) 842-0034

332-1 is9-- is you develop these design basis accidAnts eus -- as

() _2' reasonable enveloping surrogates actually for -- for all 3 conditions andLyou do this not because-there's some proof 4 that it works. The only proof that it works is -- is-the 5 experience.- What we do is we've done this over and over 6 again with the presentireactor,-operating-reactors,-and the 7 proof that it works is now you go_back and you do PRA for 8 each of these. The PRA is supposed-to be reclity where-you 9 look at all these accidents in as much detail as you can-and 110 you end up with a: risk level that is acceptable, hopefully.

-11 That's the philosophy, and there's no other real way to do 12 -it other than just rely.on the PRA in the-first place, which 13 is -- is your driving the test pilot because you don't have-14 a reactor to play around with. You have to do -- you_have-A

\% ,)- 15 to use a simulator, and the simulator in this case is the 16- PRA. and --

17 DR. WALLIS: The_ simulator is the PRA?'

18 DR. KRESS: That's the -- if you want reality.

19 The simulator -- the simulators we're talking about in the

'20 training for operators doesn't get you into the risk area 21 very far. So -- so if you're really dealing with risk, 22 Twhich ie the issue we're dealing with, safety and risk, you 23 --1the only real simulator we have is the PRA, and it does 24 supposedly touch on all of these issues you're talking 25 about, as much as we're able to visualize in them and put

ANN RILEY & ASSOCIATES, LTD.

Court Reporters 1250 I Street, N.W., Suite 300 Washington, D.C. 20005 (202) 842-0034

4 333 1 them into a PRA. So, that's the phiiosophy, and I don't

() 2 -think we have much alternative other than to keep doing it

. 3 that way.-

4 DR. WALLIS: I don't want to pursue this anymore '

5 at the present time. Thank you.

6- MR. HUFFMAN; This is Bill Huffman, one of'the 7 P600 PM's. I have some slides that I could present. They 8 reiterate the open issues on GOTHIC, NOTRUMP and the testing 9 program that Brian McIntyre discussed with you. We have no 10 quarrels with the identified open issues. I guess my only 11- point would be that --

that some of these are -- are not 12 obviously resolvable, especially in the area of GOTHIC, and 13 the staff is looking at them very hard, taking in advisement 14 and in consideration what discussions have taken place at'-

/3

. - (_) ' 15 the ACRS thermohydraulic subcommittee meetings..

16 The issues in the reactor systems codes and 17 testing, we looked through the transcripts and considered 18 all the comments in great detail'and issued specific open

19 items, and in fact, kept those items open in the draft SER 20 the. staff has prepared, and until those items are 21 satisfactorily resolved, will -- will not be able to reach 22 -closure. So, I'm stating the obvious, but that -- that's 22- really all that we had to say at this time. We're still 24- working for closure of these things.

25- DR. KRESS: Do you guys want to --

s

, ANN RILEY & ASSOCIATES, LTD.

Court Reporters 1250 I Street, N.W., Suite 300 Washington, D.C. 20005 (202) 842-0034

l-I 334-1

-1: CHAIRMAN BARTON: Are~they any different than the

() .2 ones that we saw? The staff's issues?

3 _DR. KRESS: I think_they would probably expand on 4 -it, btit dif ferent words.

5 CHAIRMAN BARTON: Well, if the subcommittee would 6- -like to have a presentation'--

7 MR. HUFFMAN: Let me get up -- ,

8 CHAIRMAN BARTON: -- we can do that. At least we 9 need the copies of the --

10 MR. HUFFMAN: Yes. I can hand out the slides, and 11 -I will get up and then go over them.

12 CHAIRMAN BARTON: Okay, fine.

13 MR. HUFFMAN: They're not much detail.

14 CHAIRMAN BARTON: Not much detail. ,

15 MR. HUFFMAN: Again, I'm Bill Huffman. The . issues 16 in NOTRUMP -- pardon, the slide must have -- that says code 17: numerics. The issues in NOTRUMP, I believe, are the'same 18: _ones that were reiterated -- were reiterating what was in 19 Mr. McIntyre's slides, benchmark deficiencies in the 20 momentum flux by calculations using actual two-phase. flow 21 equations, including the effects-of compressibility and 22 constant entropy.

23 DR. WALLIS: There is disagreement about what the-24 basic two-phase flow equations might be.

25 MR. HUFFMAN: We have a member of the technical

()

ANN RILEY & ASSOCIATES,.LTD.

Court Reporters 125C I Street, N.W., Suite 300 Washington, D.C. 20005 (202) 842-0034

335

1- staff here, Mr.-Ralph Landry. If-you'want to ---I am not

() 2

-3 able to address specific questions you may have on that.

DR. WALLIS:

It;;seems to me there's got to be t.

4 reference--to data. Whenever dealing with almost any 5 , two-phase flow situation, there's got to be a basis of data 6 because there are no fundamental equations of two-phase flow

[ 7 which are always valid.

8 DR.-KRESS: Well, they used a drift flux model.

9 It's homogeneous. They applied it for both vertical and 10 hori=ontal, and there was~some question as to whether you 11 can actually-do that for a horizontal. i 12 MR. HUFFMAN: Yeah, I saw a comment on that.

13 DR. KRESS: 'They did -- you included the momentum, 14' .which I think'they could deal with because they're only A

(s ) 15 important during-periods -- phases of the accident in-16 certain lines, and-they can deal with that probably.

17 DR. WALLIS: This is probably a worthwhile thing 18- they're saying. I just -- I just feel that if-you really.

19 want to be sure, you need some evidence based on data, or 20 you need to take a philosophy of maybe bounding.something-by 21= estimating how big it could be or small it could be, that 22 approach.

23 MR.-HUFFMAN: -I'm not sure what Westinghouse's --

24 DR. WALLIS: I don't know-what you.can do with 25 benchmark deficiencies.

ANN RILEY & ASSOCIATES,.LTD.

Court Reporters 1250 I Street, N.W., Suite 300 Washington, D.C. 20005 (202) 842-0034

336-1 MR . c HUFFMAN - -- approach is, but'I -- I'd asked

() 2 Mr.fLandry to -- to address-what he's expecting --

3 DR.-WALLIS:- Oh, so you-asked -- you asked them to 4 do it?'

5 MR.:HUFFMAN:' Yes, yes.

6 DR. WALLIS: And you_see what they do?

7 MR.- HUFFMAN: Yes,:yes.

4 8 MR. LANDRY: Ralph Landry from the staff. We 9 haven't talked to Westinghouse or seen what Westinghouse's 110 approach is going to be yet. We're still waiting to see 11 that. We're aware of the. concerns over;what equations are 12 used.

13 DR. WALLIS: Couldn't you save a lot of time by

.14 agreeing up front what the approach should-be and then n

L(,j) :15 sayingfif you follow this approach, we'll accept _it?

16 MR. CARROLL: That's.not the way of the NRC.

17= DR WALLIS - We went off.

,18 MR. LANDRY: The applicant has to propose it to us 19 and then we willoreact to it. Nothing has been presented to 20 us yet'.

21 DR. WALLIS: If you: agreed up front that you will 22 accept certain types of equations or certain types of 23 formulations of two-phase flow, then they could use them -

24 instead of having to submit something and go around and,

-25 'again, this tortuous loop and then finding out that some

'(/)

ANN RILEY & ASSOCIATES, LTD.

Court Reporters 1250 I Street, N.W., Suite 300 Washington, D.C. 20005 (202) 842-0034

337 1 consultant-has some otherzidea. -

2- MR.:LANDRY: That's not --

that's not the position 3- of,the staff'to define what equation will-be used.

4 DR. KRESS:t Sometimes that's done in reg guides.

5- MR. LANDRY: Yeah, but when it is done in the --

l- -6 in-the. Code-of Federal-Regulations and in guides,-there's

? specific purpose, but in this purpose, we have not defined 8 because this is a benchmarking case. We have not defined 9- what equation will be used. We want to see-what

'10 Westinghouse will propose.

11 MR. LEVIN: If I can also add -- this is Alan 12 Levin again. Dr. Wallace's comment with regard to data --

13 that,--part of this is -- is the -- the issue ofi 14' benchmarking the code against-the experimental-data. -There 15 were also questions about how the data were analyzed and - ^

16 and what assumptions were made in that regard- , DSo, there's 17 -- there's sort of a dual issue here. But don't -- don't 18 take this aus meaning that. there's an absence of -- of- test 19 data involved here because that's not the case.

MR. HUFFMAN: Okay, the next --

21- EDR.-WALLIS: Well, the effects of constant entropy 22 is never constant, so what's that mean?

Entropy always 23 increases and you've got a lot of evidence for that in the 24 way that all of us here operate. So what -- what do -- what-25 do'you mean by constant entropy? If you can make that ANN RILEY & ASSOCIATES, LTD.

Court Reporters 1250 I Street, N.W., Suite 300 Washington, D.C. 20005 (202) 842-0034 i

338 1 happen, you're something less.than mortal -- more than 2 mortal.

3 DR. POWERS:- Actually, he's simply; operating at.

4 equilibrium.

5 DR. KRESS:- Now, that's -- no, that -- that was 6' .the-question that --

7. DR. POWERS: -Reversibly constant.

8 DR. KRESS: -- that whether they used the proper 9 fanno relationships which, of course, are not constant.-

10 DR WALLIS:_ So what do you mean, you're asking 11 them to do something involving constant entropy? Is that 12 true?

13 DR.-KRESS: No, they did constant entropy. Now, 14' they.'re asking them --

-15 DR. __ WALL 7.S : For non-constant entropy? _

'16 DR. KRESS: Yeah, what would you got if you didn't

'17 .use constant Eentropy?-

18 DR. WALLIS: So you -- does the word missing not 19 constant entropy? Is that -- so do we have non-constant 20 entropy or increasing entropy?. Should.that be the right 21 word?

22 MR. LANDRY: I think this was typed wrong. I

23 think it should be constant enthalpy. This is stagnation.

24 DR. WALLIS: Oh, it's enthalpy. Okay.

25 MR. CARROLL: Thank you. Very good. I was going ANN RILEY & ASSOCIATES, LTD.

Court Reporters 1250 I Street, N.W., Suite 300 Washington, D.C. 20005 (202) 842-0034 l

.. . .- 0

_ . . -_ __ __ , _ , . . _ _ . _ _ _ . _ _ _ _ _ _ _ . m _ _ . _ . _ _ _ _ _ _ _ . . _ . _ _ _

339 f

-1 to guess that, actually, but -- t 2 MR._HUFFMAN: Than!t you, _ Ralph.

3' DR. WALLIS: Well, how do you get constant '

4 enthalpy and have momentum flux? I don't want to quibble. j 5 I just think you already get the words right if you're 6 asking them to do something, not send them off on some wild 7 goose chase.

8 MR. HUFFMAN: I believe that this is a paraphrase 9 of the specific questior.. If the words are incorrect in the 10 slide, it's -- it's a possibility that obviously I erred in

-11 preparing the slide. However, the staff technical -- the o 12 technical-review staff did prepara the specific questions.

13 I believe the ACRS has seen those questions and -- and no

, 14 one has raised any issues in the way the question was 15- phrased.

16 DR. SEALE: I -- I think the issue arose because '

17 they used a constant enthalpy -- they used a constant

a. 18 enthalpy calculation, and then the question was raised well, 19- then you have neglected momentum flux. What effect does 20' that-have? And so that's what's addressed in here. It's a

• 21 .little-bit cryptic-in it's statement,_but that's -- that's.

22 what's involved.

23' DR. KRESS: These four items are on my list, but 24 they're not exactly complete with respect to my list. I 25 have'a-couple of other items-that we discussed and I was --

ANN RILEY & ASSOCIATES, LTD.

Court Reporters 1250 I Street, N.W., Suite 300 '

Washington, D.C. 20005 (202) 842-0034

~.. _ _ _ _ . .. _ _ _ _ _ _ _, _ . _ _ . _ _ . . _ _

340 1 thought I'd bounce them to see if they're just missing or

() 2 you think they're not important. One of them had to do.with 3 the treatment of the entrainment at the T for the ADS 4, 4 entrainment-of liquid, and the subsequent effect on the 5 pressure loss equation through that line. And in fact, it i 6 -- it went back to the missing of the pressure or the flow 7 rate through both the ADS 4 line and the DDI line and -- and '

l 8 in preparing the code with the -- with the teatr. So, 9 that's sort of -- those two are kind of related in my .aiqd.

10 I don't see that on there, and it's -- it was a questien ,

11 that was --I thought was still an issue with -- at leadr.

i

, 12 with the subcommittee.

13 MR. LANDRY: Yes, we had that in the list. That's 14 in the SER. That was, we decided when we were helping Bill (f 15 put these together, would be better handled as a test 16 response rather than as a code response but jes, that is in 17 the -- that is in the list. All the commitments that were 18 .made at the last-subcommittee meeting were taken out of the '

[ 19 transcript and were transmitted to Westinghouse from the 20 staff, not only as commitments they made to the 21 subcommittee, but as requests for additional information 22 from the staff.

23 DR. KRESS: . Wonderful.- Thank you.

24 DR WALLIS: So, is it the staff or Westinghouse 25 g or a combination of them that's going to figure out how you ANN RILEY & ASSOCIATES LTD.

. Court Reporters 1250 I Street, N.W., Suite 300 Washington, D.C. 20005 (202) 842-0034

341 1 -- how you predict what comes out of a pipe into a separate

() 2 pipe or the T when you don't knov what the flow regime is in 3 the. big pipe? Is that -- who's responsible for getting the 4- right answer?

5 MR. HUFFMAN - Westinghouse is-responsible for 6 convincing the staff that they have correct answer.

7 DR. WALLIS: How -- okay.

8 MR. HUFFMAN: Anyway --

9 DR. WALLIS: That wasn't quite the question I 10- asked, but I guess it's all right.

11 MR. CARROLL: The right answer.

12 DR WALLIS: The right answer is defined by staff 13 acceptance. What is right is defined b*/ staff acceptance.

14 MR. HUFFMAN: Okay.

96 15 DR. WALLIS: Is that right?

16 MR. HUFFMAN: I agree. The staff has to reach an 17 acceptable safety position.

18 DR. WALLIS: It's all right. I agree. Somebody 19 has to make a decision.

20 MR. HUFFMAN: Yes, The other bullets, you can 21' read theta. I think one of the surprises out of the last 22 ACRS thermohydraulic subcommittee meetint was that there had 23 been some changes to the code numerics that the staff was 24 unaware of, and -- and this ---this, I think, is one of the

-25 items that's taken Westinghouse a long time to address and-O ANN RILEY & ASSOCIATES, LTD.

Court Reporters 1250 I Street, aL W . , Suite 300 Washington, D.C. 20005 (202) 842-0034

i 342 1 -- and I think we'll see this an,wer around the 15th of 2 February. '

3 DR. KRESS: There were two parts of that. One was t

4 the finite difference equations that were used and the other 5 had.to do with noding between the tests versus the AP600 t

6 model. So, it's --

4- 7 MR. CARROLL: That's the third bullet.

1 8 MR. HUFFMAN: That's the third bullet -- -!

4 9 DR. KRESS: Oh, I'm sorry.

10 MR.iHUTFMAN: --

that discusses that on the 11 NOTRUMP.

12 DR. WALLIS: -I think my -- my colleagues are going i

-13 to hate.me for keeping them from dinner or whatever as we're

, 14 going --

15 CHAIRMAN BARTON: No problem.

16 RDR. WALLIS: -- but this -- this code business, in 17 spite of the fact that there's documentation in the form of '

38 saying Sne used- these equations and then we worked them into .

zl9 finite difference form and so on, there's a question of-  !

20: what's in the actual-computer code. It's many, many lines,  :

1

21 and I know in the case of Trap, the all kinds of people 4 22 stuck in various things here and there to make it-work at

23 various times, Does stameone go through every line of this

[ 24 code and say what's this line doing and why is it there or 25 is it sort of a -- is that infeasible?

ANN RILEY & ASSOCIATES, LTD.

Court Reporters 1250 I Street, N.W., Suite 300 Washington, D.C. 20005 (202) 842-0034

~

1trvy , y,i -g y r W 9 'F-* FTwT *' '- ' -k-#w -9w wN' "--*--ww'-e- y- 1y ni mw w -r .ey e e we---,v,www - w w-ry w q -,,,wwyy-,,et---- , -w e --

--a,-i-w .-,,,-g-9yy--g --ww-tm--iim-t - w

l 343 1 DR KRESS: What the staff has done there is, for

() 2 example, compare the calculated results of a given code like 3 NOTRUMP to their version, which would be RELAP, the GOTHIC 4 to t: Jir version, which would be CONTAIN, and by virtue of 5_ looking at the -- how --

6 DR. WALLIS: So, it's a macroscopic evaluation?

7 DR. KRESS: It's -- yeah, it's a -- it's that kind 8 of an angle.

9 DR. WALLIS: Do you ever go into the code and --

10 do you a tually get the code or is it so proprietary you 11 cannot get-it?

12 MR. HUFFMAN: We have access to the code, although 13 --

14 DR. WALLIS: Does anyone ever look at it?

15 MR.-HUFFMAN: Mr. Landry?

16- DR WALLIS: Does anyone ever look at it and say 17 what's going on in this part of the ccde, or -- there might 38 be something in there which, in spite of the fact that it 19 fits the data, violates some-law of thermodynamics or 20 something else or -- again, this is a kind of large 21 question. When you've got these humongous-codes that you're 22 relying on and there could be all sorts of details, 23 weirdness about what's going on there.

24- MR. LANDRY: Ralph-Landry again from the staff.

25 No, we do not take the codt ,

the source code, and go through ANN RILEY & ASSOCIATES, LTD.

Court Reporters 1250 I Street, N.W., Suite 300 Washington,_D.C. 20005

-(202) 842-0034

344 1 line by line and check the code. The %ay we rely on the

() 2 assessment program is'through -- the way any code is 3 assessed through small scale tests, separate effects tests, i 4 - integral system tests, to determine that the code is a

5 performing in an acceptable manner for a particular l 6 phenomenon and particular events that we would expect to see 7 in a -- in a test program, that's the only way we have-to  ;

. 8 verify and to validate the code. We don't go line by line

-9 and check the code, _ Whether it's an applicant's code, a 10

vendor's code or an NRC code, we've never done that with the

, 11 NRC codes, either.

1 >

12 DR. WALLIS: I guess there could be some bizarre j 13 things where this sort of statement, you said if the scale 14 is of the scale of the test run, make this assumptio:.. If s 15 - the scale is AP600, make a different assumption ~. I'm not 16 _ saying, suggesting anyone would do that, but it's quite f

. 17 possible to have a code that operates that way and works

, 18 fine on the tests, but it's not extrapolable to --

19 MR. LANDRY: We do - we do have the option --

- 20 DR. WALLIS: But this has something about scaling:

4 21 in it.

I 22 MR. LANDRY: We have the option at any time we i 23. want to get the source code. There's no code _that's so proprietary we cannot have access to the source code, go in i

24 25 au examine the coding'itself if that is what'we choose to i

() ANN RILEY & ASSOCIATES, LTD.

Court Reporters 1250 I Street, N.W., Suite 300-Washington, D.C. 20005 (202) 842-0034 t- e im e- y -=ag-,vrre- v r e c e v e - *- rm- en- -r-e=e- ---++-r----e -- - ----,w w e ---evv--v-ew-w-- -ws-, , &re - w.i=eeen,--- -eM---=-e w- w

345 l

1 do. '

() 2 DR. WALLIS: I would think that if you took a --

3 an 18-year old and gave this 18-year old who's very smart 4 with computers a hunting license to find out what's wrong in 5 this code, you might find something, and that would be much 6 better than having a -- a large group of older folks look at 7 some-macroscopic results.

8 MR. LEVIN: Can 1 just in here?

9 DR. WALLIS: Yeah.

10 MR. LEVIN: This is Alan Levin. The -- there is 11 more to it than this. We don't do a line by line review,

-12 'but we do do audite. We have done quality assurance audits.

13 We do look at calculation of notebooks. We look at the way 14 the codes,are developed. We look at the basic equations and

() 15 how they.are implemented in the code. It's not a line by 16 line review, but we get a sense for how well that the 17 quality assurance and the documentation is.

18. Also, as you've heard a 1.ot of -discussion here 19 previously, an awful lot of work has been done on scaling, 20 and the ultimate aim of -- of having Westinghouse do these 21 -- these very elaborate, as Mr.-McIntyre has explained, 22 scaling analyses and -- and a careful-review of them, is to 23 have some assurance when we get to the end that, in fact, 24 the results from the test facility do scale to the AP600 and 25 _that the proper phenomena as identified in the high

-

0- ANN RILEY & ASSOCIATES, LTD.

Court Reporters 1250 I Street,.N.W., Suite :300 Washingcon, D.C. 20005 (202) 842-0034

346 1 importance phenomena have been covered and are properly

() 2 scaled and it's -- there's the top down and the bottom up 3 and look at_the phenomena and the systems interactions as 4- well.

t 5 So -- so, by the time we get to the end of this, 6 the -- the criterium we work through here is reasonable 7 assurance, and -- and that's what we look for, is to have 8 reasonable assurance that the codes are working the way 9 they're supposed to and are capable of doing the 10 calculations that they're supposed to be able to do.

11 DR. KRESS: Well, I think his question went to 12 potential scaling aspects of the code itself, not the test, 13 and one -- one example of that is you use the same noding, 14 for example. Although the size of these nodes are clearly

() 15 different in the tests than they are in a big one, and there 16 are certain input parameters that could be called scaling 17 parameters, certain knobs you twist, and I think you do look 18 at the input parameters closely and -- and you do kind of 19 require certain fidelity of the noding itself.

20 MR. McINTYRE: I --

I think one thing, your --

21 your comment that maybe there's something in there that says 22 well, if it's a' full size plant, get this answer and if it's 23 a -- you realize you're being facetious, but remember that 24 the staff not only does audit calculations of every test 25 facility we've had. but they also do' audit calculations.of

, ANN RILEY & ASSOCIATES, LTD.

Cout't Reporters 1250 I Street, N.W., Suite 300 Washington, D.C. 20005 (202) 842-0034

347 1 the plant. So, we're -- we're looking at a. set of staff

() 2 answers, . This is why they do the audit calculations.  !

3 DR. WALLIS:

Does_the staff run your codes?

4- MR. McINTYRE: No.

+

5 DR. WALLIS: Why not?-

6 MR. McINTYRE: I don't know.

7 DR. KRESS: Well, it's probably more -- more 8 instructive for them to run a different code --

9- MR. McINTYRE: -Yeah.

10 DR. KRESS: -- to compare the results I think-11 because they're familiar with the codes they're runnlag, and 12 then supposedly it's -- it's a bit of diversity added into 13 the system.

14 DR.- WALLIS: That seems to me very strange.

4

() 15 MR. LANDRY: The staff -- the staff does not-run ,

16 the cades of the vendor.because running any of these codes 17 is as much art as. science, and we have not been trained on >

18_ the -- their particular code. and: we don' t want to take the 19 time to l' earn their code -- '

20 DR.'WALLIS: Well, that's what concerns me.

21 - MR. LANDRY: -- and become that proficient, but --

22 DR. WALLIS: Is this the code which is a work of art so that it gives the answer desired'by the runner?

~

2S 24 MR.'LANDRY:' But one of the things that we do --

25 DR. WALLIS: It's much better to have someone who l

t-ANN RILEY & ASSOCIATES, LTD.

. Court Reporters i 1250 I Street, N.W., Suite 300 Washington, D.C. 20005 (202) 842-0034

%--.,-mq re- .e -.-m w,v. -

4- --g- - e- m -ey 9ug ..w p-ww.-. = - - - - - yg y .r w g y g,, p , y- -3ge -.e e www 9g ym- p w-->--r y----

348 1 did not design it run it. If you take an ACRS fellow and

() 2 say take the Westinghouse code, take the code used by the 3 staff and you have a hunting license to tan these things and 4 see where something might not be quite the way it would be 5 best for nuclear safety.

6 MR. LANDRY: We believe that is done in the way in 7- which the code has been assessed by the vendor and tne way 8 in which we have run our own codes against the test 9 facilities and in predicting the operation of the plant, 10 that by comparing the runs made with our codes and runs H11 which were made with the vendor's code, we're not_seeing 12 dramatic differences that would say once code is completely 13 out to lunch.

14 DR. WALLIS: Well, let's say what you've got then 15 is the -- the results of the vendor's code that the vendor 16 shows you. You don't get the results of the vendor's code 17 that some independent person might be able to get.

18 MR. LANDRY: Well, we --

19- DR. WALLIS: If they got a result they didn't want 20 to show you, they just didn't have to show it to you.

21 They'd run the code until they got something they wanted to 22 show you.

23 MR. LANDRY: We have the option of QA-ing, doing a 24 quality assurance review, which we did in November. We 25- spent a week at Westinghouse going through their bocks.

.. ==_. __

_ =

0- ANN RILEY & ASSOCIATES, LTD.

Court Reporters

'250

. I Street, N.W., Suite 300 Washington, D.C. 20005 (202) 842-0034 i

'J49 l

1 They are required, under the Code of Federal Regulations, to

() 2 document everything they do. We checked their calculation 3 notebooks, checked their code inputs, checked their results, 4 and determine that yes, what we're seeing is factual.

5- How, if we're getting to the point that, if I take G what you're saying another step, that they're lying to us --

7 DP. WALLIS: No, no, no. I was --

8 MR. LANDRY: Now we're getting into a legal 9 matter. If they -- they don't have the option to show us 10 only what they want to show us. Wo have the option to go up 11 -- and we did -- and QA their work and --

12 DR. WALLIS: No, I'm --

13 MR. LANDRY: -- found out that yes, they are 14 showing us-the right material.

15 DR. WALLIS: My assumption is that all parties are 16 behaving in a professional manner. No one is trying to 17 conceal anything or distorting, and I'm just concerned that 18 --

that if I wanted to be sure that some code I wrote was 19 really working, I'd give it to somebody else to see how

-20 robust it was in comobody else's-hands 'ecausea I know how 21- difficult it is to find errors in your own work and avoid 22 only looking at what you want to look at. So, I'm trying to 23 think through how one could do this better, and maybe this=

24 isn't the right forum for that.

25 MR. CARROLL: I don't think so.

=

ANN RILEY & ASSOCIATES, LTD.

Court Reporters 1250 I Street, N.W., Suite 300 Washington, D.C. 20005

-(202) 842-0034 ,

5

350 1 DR. POWERS: When you think about theae L.41.ngs --

() 2 MR. CARROLL: -I think that's a thermohydraulic 3 subcommittee.

4 DR.. POWERS: Graham, when you think about these 5 things, understand that the objective here'-is not have a 6 good running code or -- or any kind of code at all. It is 7 really to have a safe cility.

8 DR. SEALE: Well, I think there's an opportunity 9 sometime that we'll have to create to discuss some of these 10 things.

11 DR. KRESS: I think these are. provocative comments 12 that --

13 DR. SEALE: That's right, but I think we ought to 14 go ahead.

} 15 DR. KRESS: -- we ought to consider in.the 16- thermohydraulic subcommittee on how to dorthis better in the 17- future, perhaps.

18 MR. CARROLL: Except it doesn't'look like there's 19- . going to be any fucure for the foreseeable future.

20 CHAIRMAN BARTON: Go ahead, Bill. 'Inat negative 21 comment will' continue.

22' MR. HUFFMAN: Right, okay. The next area is 23 GOTHIC review, and on this slide, I did take the liberty of 24 b.s.sically -- this is-intermeshed with the scaling issues and 25 the test-data for containment, buc I lump them all together

.+ - = -

- =  %

ANN RILEY &' ASSOCIATES, LTD.

Court: Reporters 1250 I Street, N.W., Suite 300 Washington, D.C. 20005 (202)-842-0034

1 351 1 under the basic GOTHIC verification and validation, and the  !

() 2 four issues itemized here are the major ones, I believe, 3 constitute the issues in that area. I think we talked --

4 Westinghouse talked about these in their presentation. The 5 only one that wasn't, I guess, specifically addressed was 6 the location of the main steam line break. I don't believe 7 I've heard anything on that.

8 MR. McINTYRE: Bill, when you were talking, you 9 thought the GOTHIC issues were not resolvable?

, 10 MR. HUFFMAN: I did not say that --

11 MR. McIN7TRE: Okay.

12 MR. HUFFMAN: -- and I don't believe that's on the

13 record.

14 MR. McINTYRE: Okay.

15 MR. HUFFMAN: As I said, you characterized them

, 16 as, or there seemed to imply that these were all on a clear 1

17 pass the resolution, and I'm not certain that that's -- that 18 is the case in GOTHIC.

19 MR. McINTYRE: Okay.

20 MR. HUFFMAN: I just wanted to make is clear that 21 the optimism that Westinghouse has may not necessarily be 22 reflected in the staff.

23 DR. KRESS: Once again, these are on my list, but 24 I'm not sure all of my things are included. For example, 25 there was the issue of the -- of representation of the C sub

() ANN RILEY & ASSOCIATES, LTD.

Court Reporters 1250 I Street, N.W., Suite 300 Washington, D.C. 20005 (202) 842-0034

352 1 P and C sub V terms with partial derivatives in cancelling

() 2 out e.f the equation, and thereby having the wrong scaling 3 equation for GOTHIC. I'm not sure that's on here or not.

4 The other issue was raised was -- was -- I don't 5 know if it's under the test proaram or where, but it was the 6 fact that the LST tests were all steady state and we're 7 really dealing with transits, and -- and you need to justify 8 that, although it may not be a big issue.

9 MR. HUFFMAN: In my slide, again, I took a lot of 10 liberty-here --

11 DR, KRESS: Yeah, 12 MR. HUFPMAN: -- because most of these items have 13 not -- they've not progressed to a significant state from 14 the thermohydraulic rubcommittee.

/~%

k) s- 15 DR. KRESS: Right.

16 MR. HUFFMAN: I tried to summarize them.

17 DR. KRESS: Okay, 18 MR. HUFFMAN: And under this documentation, I 19 guess you get that the revised scaling report document is 20 scheduled to be submitted and --

21 DR. KRESS: You can -- you can put a lot under 22 document.

23 MR HUFFMAN: Yes, and essentially that's -- I 24 believe that all is captured there.

25 DR. KRESS: That -- that would include the'doing

() ANN RILEY & ASSOCIATES, LTD.

Court Reporters 1250 I Street, N.W., Suite 300 Washington, D.C. 20005 (202) 842-0034

353 1 --

2 MR. HUFFMAN: There are a let of --

3 DR. KRESS: -- of some of the pi groups and 4 re-looking at them and --

5 MR. HUFFMAN: Yes 6 DR. KRESS: Okay.

7 MR. HUFFMAN: I believe there's a lot of detail, 8 and I did not come prepared to address the -- the large 9 amount of detail, specifically because this is still a 10 generic open item.

11 DR. KRESS: There was -- there was an issue raised 12 about the -- the scaling also of the containment stuff, and 13 that had to do with -- with the normalization using the peak

,, 14 containment pressure, which was raised. I think it ought to

\ 15 be addressed. I personally am not very concerned about that 16 because I think it's -- it's a very appropriate and 17 reasonable scaling parameter to use. The consultants didn't 18 think so. I think perhaps there ought to be some addressing 19 of things like that, even though I'm not sure our:

20 subcommittee will endorse it as a -- as a problem, but 21 they're on the list. The full committee hasn't discussed 22 them yet. I don't know how the full committee will come 4 23 down on issues like that, but -- but, you kr. these are 24 things on my list that I'm not sure whether it shows up here 25- 'or not.

_ = _

() IMN RILEY & ASSOCIATES, LTD.

Court Reporters 1250 I Street, N.W., Suit: 300 Washington, D.C. 20005 (202) 842-0034

.. _ _ - - - _ J

l L

354

! 1 MR. HUFFMAN: My slides are not intended to

[-

\

2 reflect --

3 DR. KRESS: Okay.

4 MR. HUFFMAN: -- your list, and I r+ly on the 5 staff to -- to take the -- the concerns of the subcommittee 6 and express them as the specific issues with -- with-the 7 vendor and -- and I believe that the -- the technical review 8 staff here feels that they -- that they have or are 9 addressing the issues that -- that you've raise 10 appropriately. As far as whether you think the staff is or 11 is not, I cannot say for sure that they've captured every 12 single items raised by your consultants.

13 DR. KRESS: Well, and generally they have, and you 14 know, I've just -- it's probably -- some of them are x_,J 15 probably folded inte these things.

16 DR. SEALE: Yeah, that was that thing about the 17 size of the leak affecting whether or not you got 18 stratification, too.

19 DR. KRESS: That certainly was, and rather than 20 just the design basis accidents --

21 DR. SEALE: Yeah, right.

22 DR. KRESS: -- do you have to deal with 23 stratification issues for the whole spectrum --

24 DR. SEALE: That's right.

25 DR. KRESS: -- including the small break loca.

_ =_ .

() ANN RILEY & ASSOCIATES, LTD.

Court Reporters 1250 I Street, N.W., Suite 300 Washington, D.C. 20005 (202) 842-0034 a

355 1 So, that was an issue that --

( ) 2 DR. SEALE: Yeah.

3 DR. KRESS: -- that's on my list. You know, it 4 may be folded under this question of stratification they 5- have on here.

6 DR. SEALE: Yeah.

7 DR. WALLIS: Can I ask how much of this gets in 8 the public domain? Suppose that there is use of a lump 9 parameter model for this containment which is proposed by a 10 vendor and approved by the staff anG the ACRS places it.

11 And does it then get out in the public domain so that some 12 wise professor somewhere could say this is not the way to do 13 it, and --

14 DR. KRESS: Yes, b'

(_j/ 15 MR. HUFFMAN: The items that a.a nonproprietary do 16 get out into the nonpublic -- to the public domain.

17 DR. WALLIS: So you don't just have to satisfy an 18 in group. You have to satisfy the world?

19 MR. HUFFMAN: All nonproprietary documentation 20 gets sent to the public domain.

21 DR. WALLIS: And these issues then get into the 22 public domain?

23 MR. HUFFMAN: Th. questions all are in the public 24 domain, yes.

25 DR. WALLIS: So we are answerab'e to the outside ANN RILEY & ASSOCIATES, LTD.

Court Reporters 1250 I Street, N.W., Suite 300 Washington, D.C. 20005 (202) 842-0034

356 1 world?

l 2 MR. HUFFMAN: Yes, we are.

3 DR. KRESS: Yes, right, but still -- it's still 4 HRC's responsibility to pass judgement on the health and 5 safety issues.

6 DR. WALLIS: Well,-I think the --

7 DR. KRESS: The intervenors could possibly come in 8 and do something, but it'c still NRC's responsibility.

9 DR. APOSTOLAKIS: It seems to me that somebody 10 would have to spend considerable amount of time and effort 11 just to find out what the issues are.

12 MR. CARROLL: Yeah. There's probably easier 13 issues.

14 DR. APOSTOLAKIS: These things are not published

) 15 in technical journals.

16 DR. KRESS: Yeah.

17 DR. APOSTOLAKIS: You really have to know what 18 you're after. Even though they're in the public domain, 19 it's not so ease to find out what is going on, unless 20 somebody tells you,_look for this.

21 CRAIRMAN BARTON: Go ahead.

22 MR. HUFFMAN: Okay. The last slide deals with the 23 reactor system testing area, three different-areas. One is 24- the ADS test program, and Westinghouse is -- needs to 25 ' provide additional ADS test data evaluation to the staff,

\> ANN RILEY & ASSOCIATES, LTD.

Court Reporters 1250 I Street, N.W., Suite 300 Washington, D.C. 20005 (202) 842-0034 l

l

357 1 demonstrate the testing, cover the appropriate

() 2 thermohydraulic conditions. I think this is -- this 2' somewhat captures your first concern that you wanted to 4 . express about NOTRUMP.

5 In addition, key aspects of the ADS valve 6 qualification the staff is-expecting to be-added to the SSAR 7 and to the ITAAC to insure that the -- that the valve 8 characteristics are achieved as expected.

9 Scaling and PIRT closure, the demonstrate OSU is 10 appropriately scaled between the end of ADS blowdown and the 11 establishment of the IRWST injection, and this is, I 12 believe, the second hashmark under scale and PIRT closure is a -- is a technical staff issue.

~ 4 13 I don't believe this was 14 specifically raised by the thermohydraulic subcommittee, but k 15 it certainly is an area that you -- you might look to 16 provide additional documentation on the model's assumptions 17 and determination of pi groups, including comparison of the 18 hand calculated pi groups versus the database values.

19 DR. WALLIS: List the origin of this pi group 20 study, Where did -- what initiated that?

21 MR. HUFFMAN: I'll have Alan.

22 MR. LEVIN: This is Alan Levin. This is part of 23 the Oregon State University facility scaling analysis. It 24 was -- the background is -- is to a large extent, at least 25 the original scaling analysis background to a large extent,

_ _ __ =_ __ _

ANN RILEY & ASSOCIATES, LTD, Court Reporters '

1250 I Street, N.W., Suite 300 Washington, D.C. 20005 (202) 842-0034

358 8 1 was drawn from Zuber's H2TS methodology, and Oregon State 2 produced the scaling report derived the parameters.

3 DR. WALLIS: A long time ago.

4 MR. LEVIN: '95 Was the finsi, and -- and then 5 Westinghouse was requested to do a more involved multi-loop 6 analysis, more of a top down system effects analysis based 7 on that Wolfgang Wulff has done. So, this is all factored 8 in there together, and the -- the comparison between the 9 Oregon State University facility and the AP600 is based on 10 what Westinghouse calls hand calculated values for certain 11 parameters, flows and so forth, since there's no 12 experimental data for the plant, and then there are also 13 database values derived from the OSU data themselves, and 14 that -- that= bullet goes to a little bit more explanation as 15 to where there are differences, why those differences exist.

16 DR. WALLIS: So, there's nothing in the law that 17 says thou shalt calculate pi groups?

18 MR. LEVIN: No. There -- there's -- this has been 19 a negotiation between the staff in Westinghcuse, if you 20 will, as to what comprises an adequate scaling valuation.

-21 It's p'rt of -- I mean, there -- there's a -- if you look at 22 -- it ,cu look at the design certification rules, Part 52, 23 -and I'll give you the citation if you want -- 52.4? -- it 24 says that there must be an adequate database, and that you 25 must be able to demonstrate that the data are applicable to ANN RILEY & ASSOCIATES, LTD.

Court Reporters 1250 I Street, N.W., Suite 300 Washington, D.C. 20005 (202) 842-0034

359 1 the plant. The only way to do that is a scaling evaluation 2 and what coraprises an adequate scaling evaluation is a -- a 3 matter of techni- ,,udgment, I guess, engd.neering judgment.

4 DR. WALL.1: So, this is somethint which was 5 invented for this problem, this ADS' This APG00 problem?

6 DR. KRESS: It was invented for -- to deal with 1 7 scaling from --

8 MR. LEVIN: -

From a genetic standpoint.

9 DR. KRESS: -- test facilities to -- that are not 10 full scale, to the full scale.

11 DR. WALLIS: Well, I know that, but the -- these 12 pi groups that appear in this analysis are not pi groups 13 which are accepted by a broader technical comrr(nity as far 14 as I know. They have been generated for AP600.

15 MR. LEVIN: That's --

16 DR. WALLIS: They don't have a genealogy or 17 something. They don't -- you don't have names which are 18 recognizable and - ,

19 MR. LEVIN: Well, some of them do. I mean, some 20 of them are the conventional dimensionalist parameters that 21 you deal with like Reynolds numbers and Richardson numbers 22 and things like that, and a lot of them are not. They are

,23 combinations of the dimensionalist parameters. At -- I'm 24 not at a position to say if you did a comparable scaling ,

25 analysis on-a different reactor design, you might come up ANN-RILEY &' ASSOCIATES, LTD. '

Court Reporters 1250 I Street, N.W., Suite 300 Washington, D.C. 20005 (202) 842-0034

360 1 with a lot of the same dimensionalist parameters whether

-( ) 2 they've got names or r.ct.

3 DR. WALLIS: 1e11 all I'm trying to do is get a 4 sort of, again, a broader cosmic view of things, of what's 5 going on here rather than trying to just question what 6 you're doing. It appears that this -- these pi groups do 7 occur with the AP600, but -- and you're going beyond what 8 one would say sort of the broader technical base in a 9 community of experts in engineering.

10 MR. LEVIN: Well, yeah. The -- the objective here 11 is to demonstrate that the -- that the test facilities 12 represent the thermohydraulic behavior of the plant, and 13 that includes all those unique characteristics of the AP600 14 like CMT's and ADS's and things like that that you might not 15 have in a generic plant design, let's say, and so there are 16 going to be a lot of dimensionalist groups that are 17 essentially specific to the AP600 and its behavior in the 18 transient.

19 DR. WALLIS: This is in the -- this is in the 20 public record?

21 MR. LEVIN: Yeah.

22 DR. WALLIS: So, someone could design a chemical 23 plant next year using your scaling approach and then find 24 out that it didn't work, and this would then be.new evidence 25 about the viability of these pi groups, which would then be

-( ) ANN RILEY & ASSOCIATES, LTD.

Court Reporters 1250 I Street, N.W., Suite 300 Washington, D.C. 20005 (202) 842-0034

1 1

361

( 1 considered by this broader technical community. And this --

() 2. I'm just looking at the broader picture. Maybe it's not-3 appropriate, but again, I'm just worried about the sort of 4 arcane nuclear safety community or procedures that does 5 certain things that eventually has to become part of a much

-6 -broader technical base for our whole society.-

7 MR. LEVIN: Well, I think that --

8 DR. WALLIS: And cannot-be independent of it.

9 DR. KRESS: I think that the -- that the overall 10 methodology, the H2TS methodology which, by the way, was 11 developed initially, as I understand it, for more of an 12 accident perspective, is a -- is, if you want to call it 13 that --

I'm not sure that-I'm using exactly the right word, 14 but of a generic app)icability. It's a process for deriving 15 the dimensionalist parameters that you'd need to evaluate 16 the --the specialized specific scaling for any given 17- specific design, 18 Now, the AP600 scalihg parameters as derived 19 through that methodology might not be applicable to anything 20 aside from the AP600. If the technical community were to 21 make a determination in its wisdom that what Westinghouse

22. has done is incorrect and that, in fact, even though 2.1 specific scaling parameters are not applicable to the AP600, 24 1 guess we'd have to go back and take a look at it, but I 25 haven't seen anything to suggest that that's not the -- that O ANN RILEY & ASSOCIATES, LTD.

Court Reporters 1250 I Street,.N.W., Suite 300 Washington, D.C. 20005 (202) 842-0034

362 1 in fact that would be the case.

() 2 MR. CARROLL: Up to the time you certify the 3 design.

4 MR. LEVIN: That's correct. If new information 5- becomes available, it would tend to call the staff safety 6 evaluation into question and obviously we'd have to go back 7 and take a second look.

8 DR. WALLIS: Well, I guess I -- I'm going to be 9 quiet. I've said too much, but I just feel, trying to 10 figure out what -- what I'm doing here and what this 11 committee does and how all these things happen, and that the 12 way in which these decisions are reached and the way in 13 which the technical evaluations are made, all sort of become 14 habitual to the agencies involved eventually have to be O)

\_ 15 transparent to a much broader community, and that may not 16 always -- inevitably may not always confirm decisions made.

17 That's inevitable, and that -- that's just the nature of the 18 world. I just -- I just would hope that -- that this 19 doesn't-become too much of a closed community doing dances 20 according to some kind of -- yeah, some dances which are 21 evolved in almost a tribal ritual rather than as part of a 22 broader sort of technical awareness and then you toss aside.

23 CRAIRMAN BARTON: Thank you, Bill. Let me just 24 try to summarize.

25 MR. McINTYRE: Mr. Barton?

ANN RILEY & ASSOCIATES, LTD.

Court Reporters 1.40 I Street, N.W., Suite 300 Washington, D.C. 20005 (202) 842-0034 1

363 1 CHAIRMAN BARTON: Yes?

() 2 MR. McINTYRE: I think we had chased down some of 3 the answers to the questions. Can we do those now?

4 CHAIRMAN BARTON: Yes, if you've got them, that 5 would be fine.

6 MR. McINTYRE: Okay, Ron. Ron, I think quickly.

7 MR. VIJUK: These are from yesterday's 8 discussions. There was cne relative to the fuel and a 9 question about Zerlo and whether it was -- how it compared 10 to Zirc Niobium, and it is --

11 MR. CARROLL: It is a Niobium clad.

12 MR. VIJUK: Well, Zirc Niobium, it turns out, it 13 also has some tin and iron in it, which our folks believe 14 makes it better than a binary Zirc Niobium cladding, so that 15 -- that's the story I got on that one.

16 The -- another item was relative to higher 17 burn-ups, and I confirmed that we're using the 62,000 18- megawatt days per metric ton lead rod speed limit and -- and '

19 what our people tell me is the staff this November 20 reconfirmed that speed limit on burn-up and the methodology 21 for reactivity insertion accidents associated with that 22 burn-up, and that that's what we're using on AP600.

23 DR. POWERS: When you -- when we talked about this 24 yesterday, we talked about an average bundle limit. Now 25 you're saying it's a lead rod limit?

ANN RILEY & ASSOCIATES, LTD.

Court Reporters 1250 1 Street, N.W., Suite 300 Washington, D.C. 20005 (202) 842-0034

364 1 MR. V1JUK: Well, the lead rod limit was the thing

() 2 that was tied to the reactivity insertion events, okay.

3 There's some average bundle limits associated with this 4 incomplett rod insertion observations we've had, and those 5 are being dealt with also.

6 Again, in -- when we get into broad ejection 7 accidents in Chapter 15, we can talk about the methodology.

8 There was a question about the -- if we used leak 9 before break on the steam line, and it turns out we do use 10 it on the portion of the main steam lir) inside containment.

11 The number of tubes in the steam generators is --

12 MR. CARROLL: But you are not allowed to do it on 13 the feed water line, right?

14 MR VIJUK: That's correct. We do not use it on 15 the feed line.

16 MR. CARROLL: You wanted to, but they wouldn't let 17 you?

18 MR. VIJUK: That's right, i..tactly right.

19 MR. CARROLL: Bill.Schack's going to find out why.

20 DR. POWERS: He's going to recover from whatever 21 illnesc is afflicting him.

22 MR. CARROLL: Schack is sick, too?

23 CHAIRMAN BARTON: I don't think so. I think 24 Schack had a commitment. He had to work.

25 DR. SEALE: Then he lied because he said that he ANN RILEY & ASSOCI TES, LTD.

Court Reporters 1250 I Street, N.W., Suite 300 Washington, D.C. 20005 (202) 842-0034 4

365 l

1 was sick.

() 2 MR. CARROLL: Well, that's not mutually exclusive.

3 I'm-sorry.

4 MR. VIJUK: Just a couple more things. The number 5 of tubes in the steam generators-in 6307 in each generator.

6 The -- we have confirmed that the reactor coolant pumps do 7 not include an anti-rotation device, or an anti-reversing 8 device.

9 Let's see, I guess that was it. The other 10 question was relative to the --

11 MR. CARROLL: You've got to have more starting 12 than --

13 MR. VIJUK: -- type of hydrogen monitor in the rad 14 waste system, and we really haven't specified that at this 15 point.

16 MR. CARROLL: -- the other one and it's turning 17 backwards when you hit the brake.

18 CHAIRMAN BARTON: Thank you for the feedback. To 19 summarize the last two days, the subcommittee on advance 20 reactor designs has completed its review of the testing and 21 analysis program and -- well, don't get excited. Let me 22 finish -- in Report Chapters 1, 4, 5, 7, 8, 11, 13 and 18, 23- At this point in time, we still, since we do not have draft 24 FSER chapters, when they are available, the staff will make 25 presentations to the sub' committee on those SER's, and at

[ AJhi RILEY & ASSOCIATES, LTD.

Court Reporters 1250 I Street, N.W., Suite 300 Washington, D.C. 20005 (202) 842-0034

366 1_ this point, we will discuss results of the last two days

() 2 with the full -- full committee. We intend to, based on the 3 deliberations at the full committee, prepare an interim 4 report on what we heard the last two days, including any 5 open issues that the committee has at that time rogarding 6 the information and the presentations that we have heard.

7 If any member of the subcommittee has any open 8 items, I'd appreciate it if they got them -- any other 9 questions or open items, I'd appreciate it if they got them 10 to Noel today or this evening or tomorrow so we can make -

11 sure that we present all the issues that we've heard in 12 front of the full committee on tomorrow, 13 At the full committee tomorrow, I think we just --

14 from Westinghouse and the staff, just hear a summary of what 15 was presented today.

16 MR. CARROLL: Today and yesterday?

17 CHAIRMAN BARTON: And yesterday, yes. We don't 18 need to go through the whole two-days again tomorrow at the 19 full committee meeting.

20 DR. SEALE: How long of a time is allocated?

21 CHAIRMAN BARTON: We have -- right-in front of me 22 -- we have from 13:15 until noon tomorrow. It's an hour and 23 45 minutes.

34 Are there any other questions or comments at this 25_ time?

ANN RILEY & ASSOCIATES, LTD.

Court Reporters 1250 I Street, N.W., Suite 300 Washington, D.C. 20005 (202) 842-0034 9

367

( l' (No' response.)

() 2- The. subcommittee meeting is adjourned.

3 [Whereupon, at 5:04 p.m., the meeting was 4 concluded.)

5 6-J 7 8

9 10 11 12 13 14 15 16 17 18 19

'O.

21 22 23 24 25 i

ANN RILEY & ASSOCIATES, LTD.

Court Reporters 1250 I Street, N.W., Suite 300 Washington, D.C. 20005 (202) 942-0034 B

-_ _ -_------J

i REPORTER'S CERTIFICATE This is to certify that the attached proceedings r-t (N) before the United States Nuclear Regulatory Commission in the matter of:

NIME OF PROCEEDING: ADVANCED REACTOR DESIGNS SUBCOMMITTEE MEETING DOCKET NUMBER:

PLACE OF PROCEEDING: Rockville, MD were held as herein appears, and that this is the original

_ transcripc thereof for the file of the United States Nuclear (s / 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.

8%

on Hundley Official Reporter Ann Riley & Associates, Ltd.

O O

_ - - - - - J

gh.

I O

ASSESSMENT PROCESS O ,

Stephen D. Floyd Nuclear Energy Institute Meeting of Advisory Committee on Reactor Safeguards February 5.1998

• O d' !

O COMMENTS ON STAFF l APPROACH TO ASSESSMENT General Comments

- Process may result in some improvement in efficiency O - Improvements in effectiveness not evident

- Purpose and objectives of revised process still not clear

- Appears to be a " repackaging" of the current process O

'd'

O COMMENTS ON STAFF .

APPROACH TO ASSESSMENT Specific Comments

- Criteria for " white" " yellow" and " red" .

not defined

- Relationship of process to public g health and safety standards not evident

- Lise of industry average data still a problern

- Question whether economic data is really leading and therefore useful

- Management ef ectiveness g assessment not in rec ulatory sco ae h'

4 0

0 O

= .

c N .

E N -

w 9 ,

i.

Q 4

y ..

a O $[e a

O G N

I T

E E

M T

N E

M E

G A N N O A

M E E

R T O T I I N M E M S O C S t E 8 t H YD e T RR 9 OAU 9

1 r

r F S G a

O I VE ,

B E D F 5 S AA y O A B

S E R H O r

a J

d N T T u r O r a I O C b e

h c

T A

TAE F i M NR R R O O I F T N A I T N

E E H S T E R

O P T

S T

N E

M E

V O

R P

M I

O l lll

PROGRESS TO DATE

- Arthur Andersen recommendations to improve'the process and information base of the NRC senior management meeting (SMM)

- Significant process improvements have been implemented

- More' structured presentation of information

- Greater participation by HQ and regional managers

- Formal process for eliciting views of managers

- Performance Trend Plots have been developed and tested in the January 1997 SMM cycle

- Financial indicators identified and used during SMM screening meetings

- Supplemental information only

- Plant Performance Template developed and given to regions

- PIM data tc be coded in template subcategories

- Template has been applied in proposed Integrated Assessnent Process

~

O O O .

NEXT STEPS Pilot application of Template in July 1998 SMM cycle

- Public comment on Template, Trend Plots and financial indicators

- In conjunction with proposed Integrated Assessment Process

- Decision regarding assessment of management effectiveness

- Implementation i ;onjunction with Integrated Assessment Process

~~

O O- O .

QUESTIONS FROM OPERATIONS SUBCOMMITTEE

- ' Status of-WANO indicators

- Staff will have access to all indicators on proprietary basis

- One time, three year historical indicator data

- Raw data will also be available

- No access to INPO Performance Indicator Index i

- Evaluating whether we can meet SRM requests without Index

- Absolute vs. relative threshold for trend plot

- Staff evaluating both options

- Management Ef fectiveness

- Andersen and:GAO reports recommended assessment of management

- NRC practice is to infer management performance from PIM

- RES working on measurement methods

- Decision needed on assessing management as leading indicator

_