ML20135F178
| ML20135F178 | |
| Person / Time | |
|---|---|
| Issue date: | 03/07/1997 |
| From: | Advisory Committee on Reactor Safeguards |
| To: | |
| References | |
| ACRS-T-2095, NUDOCS 9703110259 | |
| Download: ML20135F178 (465) | |
Text
_ ___. .. .. . - - - - - - - _ _ -_ - - - - - - - - . - - - .
g Enwaw p Official Transcript of Proceedings lg g g I NUCLEAR REGULATORY COMMISSION E EE -
~
~
g KF"2,098 A;$
$ D uw ;.,
qqm W
=~ $9.-
- . 9 .%
- h. , i
- .,.Wc.
- h. - -
d
Title:
Advisory Committee on Reactor Safeguards :M y%p4::.
- 439th Meeting Eyg q
TRO4 (ACRS) y//jbo ri. w MM RETURN ORIGINAL g .,.
TO BJWHITE ' .cg,yIw N ,%g .
M/S T-2B26 M Docket Number: (not applicable) 415-7130 @4s!@
'.i: 8 55 h..K
- l. Location: Rockville, Maryland @4.acc MS
- $h@$
R f.'l$fW .5M Date: Friday, March 7,1997 Q1%
M VM i
, ~
z A,28 f, ,,ilCO CO["j - 302 T y }
u - r.,
o
3!I isc =m Q Work Order No.: NRC-1038 Pages 244-623 d"2; ': 2 r.__ - .; ;:
- a m-=3 o NEAL R. GROSS AND CO., INC.
Court Reporters and Transcribers ; ;
[./
1323 Rhode Island Avenue, N.W. j Washington, D.C. 20005 ,
-=
< .-mnn (202) 234-4433 ,V - - . , j 9703110259 PDR ACRS 9'0307 'pl' ,
E 1 i
T-2095 .kd,i f6 l PDR 3l 1 0
4 l
DISCLAIMER PUBLIC NOTICE BY THE UNITED STATES NUCLEAR REGULATORY COMMISSION'S ADVISORY COMMITTEE ON REACTOR SAFEGUARDS MARCH 7, 1997 i
The contents of this transcript of the proceedings of the United States Nuclear Regulatory j Commission's Advisory Committee on Reactor Safeguards on
}~ MARCH 7, 1997, as reported herein, is a record of the l r.
G discussions recorded at the meeting held on the above date.
This transcript has not been reviewed, corrected ,
e l and edited and it may contain inaccuracies. l 2
I l
v' NEAL R. GROSS COURT REPORTERS AND TRANSCRIBERS 1323 R1IODEISLAND AVENUE, NW
.(202)234-443 L WASillNGTON,DC. 20005 (202)234-4433 J
1 244 i 1 UNITED STATES OF AMERICA
,- 2 NUCLEAR REGULATORY COMMISSION
\
) 3 +++++
l 1 439th MEETING 5 ADVISORY COMMITTEE ON REACTOR SAFEGUARDS (ACRS) l 6 +++++ l 7 FRIDAY i
8 MARCH 7, 1997 9 +++++ l 10 ROCKVILLE, MARYLAND j 11 + ++++
l l
12 The Advisory Committee met at the Nuclear I 13 Regulatory Commission, Two White Flint North, Room T2B3,
(')
u ,/ 14 11545 Rockville Pake, at 8:30 a.m., Robert L. Seale, 1
15 Chairman, presiding.
l l
16 ,
l 17 COMMITTEE MEMBERS:
l l
1 18 ROBERT L. SEALE CHAIRMAN 19 DANA A. POWERS VICE CHAIRMAN l
20 GEORGE APOSTOLAKIS MEMBER 21 JOHN J. BARTON MEMBER ,
l 22 MARIO H. FONTANA MEMBER 23 THOMAS S. KRESS MEMBER l 1
I 24 DON W. MILLER MEMBER
, (~g
( ! 25 WILLIAM J. SHACK MEMBER l NEAL R. GROSS COURT REPORTERS AND TRANSCRIBERS 1323 RHODE ISLAND AVE., N.W.
(202) 234-4433 WASHINGTON, D.C. 20005 3701 (202) 234 4433
245 1 ACRS STAFF PRESENT:
2 JOHN T. LARKINS Exec. Director i,~.s)
\
3 ROXANNE SUMMERS Tech. Secretary 4 SAM DURAISWAMY 5 CAROL A. HARRIS 1 1
6 RICHARD P. SAVIO l
7 PAUL BOEHNERT l l
8 NOEL DUDLEY '
l 1
9 MEDHAT M. EL-ZEFLAWY l
. I 10 MICHAEL MARKLEY 11 AMARJIT SINGH l 12 )
13 ACRS CONSULTANTS PRESENT:
Or
(_ .L 4 IVAN CATTON 15 HAROLD W. LEWIS 16 17 ACRS FELLOW PRESENT: I 18 RICK SHERRY
]
19 1
20 ALSO PRESENT:
21 ED JORDAN l
22 JACK ROSENTHAL 1
i 23 DOUG CHAPIN 24 JOHN GANNON (7
ms
/ 25 LARRY DAMON l NEAL R. GROSS l COURT REPORTERS AND TRANSCRIBERS 1323 RHODE ISLAND AVE , N W.
(202) 234 4 433 WASHINGTON. D.C. 20005-3701 (202) 234-4433
246 1 ALSO PRESENT: (CONT.)
7s 2 CHRISTINE MITCHELL f )
3 CARMELO RODRIGUEZ 4 DAVID HILL j 5 JAMES CURTISS 6 JAMES WHITE 7 STEPHEN SOHINKI 8 JERRY ETHRIDGE 9 MAX CLAUSEN 10 GERALD SORENSEN 11 12
[
'% .)
( )
%J NEAL R. GROSS COURT REPORTERS AND TRANSCRIBERS 1323 RHODE ISLAND AVE., N W.
(202) 234-4433 WASHINGTON, D C. 20005-3701 (202) 234-4433
247 1 A-G-E-N-D-A gs 2 Acenda Item Pace s
3 Opening Remarks 248 l
4 Independent Safety Assessment of the 5 Maine Yankee Atomic Power Station 252 l 6 National Academy of Sciences / National 7 Research Council (NAS/NRC) 8 Phase 2 Study Report 294 9 Department of Energy Proposal for Tritium 10 Production 412 11 12 13 f'~N,
(_) 14 15 16 17 18 19 20 21 2-2 23 24 p.
(_) 25 NEAL R. GROSS COURT REPORTERS AND TRANSCRIBERS 1323 RHODE ISLAND AVE., N.W.
(202) 234-4433 WASHINGTON, D.C. 20005-3701 (202) 234-4433
248 1 P-R-O-C-E-E-D-I-N-G-S c's 2 (8:30 a.m.)
! \
~
3 CHAIRMAN SEALE: The meeting will now come to 4 order. This is the second day of the 439th meeting of the 5 Advisory Committee on Reactor Safeguards. During today's 6 meeting the committee will consider the following:
7 (1) Independent safety assessment of the 8 Maine Yankee Atomic Power Station; 9 (2) National Academy of Sciences / National 10 Research Council Phase 2 Study Report; 11 (3) Department of Energy Proposal for Tritium 12 Production; 13 (4) Future ACRS activities; I )
\/ 14 (5) Reconciliation of ACRS comments and 15 recommendations; and 16 (6) Proposed ACRS reports.
17 This meeting is being conducted in accordance 18 with the provisions of the Federal Advisory Committee Act.
19 Dr. Sam Duraiswamy is the designated federal 20 official for the initial portion of this meeting.
21 I would remind you that we have an interview l
22 with a potential candidate today at noon. And some of you 1
23 have been -- particularly those of you who don't know the 24 candidate -- have been scheduled to spend part of that
/^
( ,\) 25 time with him.
NEAL R. GROSS COURT REPORTERS AND TRANSCRIBERS l 1323 RHODE ISLAND AVE., N.W. l (202) 234-4433 WASHINGTON, D C. 20005-3701 (202) 2344433
249 1 We have received no written comments or
,-s 2 requests for time to make oral statements from members of
\) 3 the public regarding today's sessions.
4 A transcript of portions of the meeting is 5 being kept, and it is requested that the speakers use one 6 of the microphones, identify themselves and speak with 7 sufficient clarity and volume so that they can be readily 8 heard.
9 There are a couple af things. D9:.a and I have 10 been talking about the wall o'_ things that seem to be 11 piling up toward the end of this calendar year with the 12 AP600 review, the PRA standard review plan, after comments 13 review, and other things and the like. So that is some O) i
\_/ 14 advance planning that we are going to have to do.
15 We're trying to get a better handle on that 16 schedule and we've asked one of the Fellows to take a look ,
1 17 at the material currently in the AP600 Section -- or in 18 the AP600 area -- to identify those chapters that might j 19 require significant review. Things like the containment, 20 probably parts of the I&C, the PRA certainly, a few things 21 like that.
22 There's some value judgments that we're going 23 to have to make as well, there. For example, the 24 expectation is that the staff and the vendor will apply 1
,~3 (v) 25 the same general approach to things like ITAACs and so NEAL R. GROSS l COURT REPORTERS AND TRANSCRIBERS 1323 RHODE ISLAND AVE., N W.
(202) 234-4433 WASHINGTON. D C. 20005-3701 (202) 234 4433
250 1 forth.
,, 2 That is, those items whose treatment is l
/ \ ;
l
\)
3 outlined and for which standards are set in the 1
4 certification process but the responsibility for ultimate 1 5 delivery lies with the applicant. That is, the site --
l 6 well, the utility, basically. l l
7 And it's questionable as to whether or not we 1
1 8~ need to do anything in-depth on that beyond verifying that )
9 they're using basically the same methodology, the same 10 approach. There are questions like th.'t that we're going 1
1 11 to have to make some decisions on, and I would urge you to '
12 be thinking about them as we come cloler and closer to l i
13 having to implement that. l
(\ i
(._) 14 MEMBER FONTANA: Who is handling this?
15 CHAIRMAN SEALE: Gus is looking at it. In 16 that regard, if you can anticipate any requirements that 17 might be coming up in that same period -- namely, the last 18 quarter of this year, the first quarter of next fiscal 19 year, same thing -- we'd like to know about it because it 20 looks like we're already going to be pretty heavily 21 overtaxed.
22 We have a letter that we had looked at last 23 time and had not quite finished, and it's not on the 24 schedule that was published but we plan to discuss it this n
( ) 25 afternoon at 2:30; and that's the letter on the plant-ss NEAL R. GROSS COURT REPORTERS AND TRANSCRIBERS 1323 RHODE ISLAND AVE , N.W.
(202) 234-4433 WASHINGTON, D C. 20005-3701 (202) 234-4433
251 1 specific applications of the safety goals. About 2:30.
, 2 So those of you who might be interested in that --
3 Jocelyn, you might make sure they hear about it.
4 Okay. The first thing this morning is the 5 independent safety assessment of the Maine Yankee Station 6 and that's the Plant Operations Subcommittee. John 7 Barton.
8 MEMBER BARTON: Thank you, Mr. Chairman. The 9 purpose of this session is to hear a presentation by the 10 NRC Staff regarding the results of the independent safety 11 assessment for the Maine Yankee Atomic Power Station.
12 A little background. In late 1995, concerned 13 citizens forwarded anonymous allegations to the State of
/;
(_) 14 Maine which were subsequently forwarded the NRC. The 15 allegations were regarding inadequate technical analysis 16 to support an increase in the rated thermal power at which 17 Maine Yankee may operate.
18 NRC Office of the Inspection General completed 19 an inquiry into the allegation, established that Maine 20 Yankee had experienced problems with and made 21 modifications to, the RELAP/5YA computer code which was 22 used in the emergency core cooling analysis for a small 23 break loss of coolant accident.
24 OIG also found weaknesses in the NRC review
' p)
(v f 25 and follow-up activities which contributed to NRC's NEAL R. GROSS COURT REPORTERS AND TRANSCRIBERS 1323 RHODE ISLAND AVE., N.W.
(202) 234 4433 WASHINGTON D.C. 20005-3701 (202) 234-4433
252 1 failure to detect these deficiencies. Response to the ,
i 1
.-,s 2 findings as well as response to concerns by the Governor
)
~'
3 of Maine about the safety and effectiveness of the l l
4 regulatory oversight of Maine Yankee, Staff was directed 5 to perform an ISI of Maine Yankee.
6 The assessment was to be coordinated with the 7 State of Maine to facilitate participation by State 1
8 representatives consistent with the Commission's policy on l l
l 9 cooperation for states with commercial nuclear power 10 plants.
11 The briefing this morning is for information 12 only. There is no committee action required. And at this 13 time I'd like to turn the meeting over to Mr. Edward
,r^g k--) 14 Jordan who will lead the presentation by the Staff on the 15 independent safety assessment. Ed?
16 MR. JORDAN: Okay. Thank you very much. I 17 appreciate the opportunity to meet with the ACRS this 18 morning.
19 The team -- I was identified as the team 20 manager reporting to the Chairman. Bccause of the nature 21 of this independent safety assessment the object was to 22 provide independence within the staff. I was in AEOD at 23 the time and selected a team leader, Ellis Merschoff, who 24 was Division Director for Projects in Region II.
25 Jack Rosenthal is here as both a member of the (w.:)
NEAL R. GROSS COURT REPORTERS AND TRANSCRIBERS 1323 RHODE ISLAND AVE., N W.
(202) 2344433 WASHINGTON, D C. 20005-3701 (202) 234-4433
253 1 team -- he led the portion of the effort that reviewed the fs 2 code analysis that Yankee Atomic did for the Maine Yankee l t )
i'~' / \
3 plant, and so he'll provide some perspectives on that )
4 work. That was somewhat unique.
5 The team was set up as has been discussed, 6 based initially on allegations about the use of RELAP/5YA.
7 The objectives that I have in the discussion is to 8 describe the process -- and I'll try to be brief about the 9 process -- d tuss the findings and conclusions and then i
10 talk briefly about the regulatory lessons learned from 1 11 this effort.
12 We can skip 3; that's pretty well covered.
13 The next slide, 3a, sort of gives the present status and w/ 14 how we got where we are.
l 15 I think important along this is that in l 16 January of 1996, based on the IG's review and the staff's 17 follow-up to the review of the RELAP/5YA, power level at 1
18 the plant was restricted to 90 percent, which was their 1
19 original license power.
l 20 They had gone through a pcwer level upgrade to 21 2700 megawatts because of limitations in the analysis, and 22 so from that point on the plant operated at an authorized 23 power level of 2440.
24 Staff -- the bottom item -- the staff did
/'~T
(,,) 25 initiate NRR in the region, some lessons learned from the l
NEAL R. GROSS COURT REPORTERS AND TRANSCRIBERS 1323 RHODE ISLAND AVE.. N.W.
(202) 234-4433 WASHINGTON, D.C. 20005-3701 (202) 234-4433 l
254 l l
4 1 effort to that point, l
2 The independent safety assessment -- I should
/,s.h I
'- ') : mention -- there was an interaction on the part of the 1 4 Chairman with the S* ate of Maine, and this effort then had 5 a larger State involvement than previous investigations 6 and reviews that the NRC has done, and it was a very 1
7 pleasant and useful interaction.
8 Presently, the plant is shut down. They have ,
1 9 a number of restart issues related to design control, l 10 cable separation. They identified some power cables that 11 were non-safety grade and safety grade trays, and they've 12 had several problems that have caused them to remain shut 13 down.
(_) 14 They now have a number of fuel leakers and a 15 batch of fuel that they're concerned about, and so they're 16 changing part of this particular fuel loading. So they 17 are currently shut down.
18 The features of the independent safety 19 assessment -- first of all, there was an independence from 20 NRR and region 1; that is, the composition of the team was 21 made up from other than NRR and region 1 so that it would 22 avoid the criticism of the staff that were involved in 23 Maine Yankee reviewing itself.
24 It was a very well experieiced team; that is, O)
! 25 reporting to the Commission we had the pick of the Agency NEAL R. GROSS COURT REPORTERS AND TRANSCRIBERS 1323 RHODE ISLAND AVE., N.W.
(202) 234-4433 WASHINGTON, D C. 20005-3701 (202) 234-4433
l 255 1 from offices other than NRR and region 1. "So the other I
,_ 2 regions, Research and AEOD, all provided team members.
t \
l \ /
3 And there was participation by the State of Maine.
4 The process and the measures, the procedures, 5 were based on a modification of the diagnostic evaluation 6 program. And that is a formalized program, so we revised l l
7 the procedures in order to have a sound basis.
8 The mission of the ISA -- there were four l
9 mission statements. The first was to provide an l
10 independent assessment of the conformance to the -- Maine 11 Yankee to its design and licensing basis. And that was a 12 unique feature at the time; that is, for the NRC to look 13 in detail at the design and licensing basis. And this was
/")
_ _ ,) 14 derived from the problems that the RELAP/5YA code l
15 implementation had shown already.
16 The next was an independent assessment of 17 operational safety performance. That is not unique. The 18 risk perspectives -- we went a little further in using PRA 19 and risk considerations. We had a contractor on board who 20 was a PRA person that supported the team, and Jack 21 Rosenthal and others from the team had PRA backgrounds.
22 The third was to evaluate the effectiveness of 23 licensee self-assessments, corrective actions and 24 improvement plans. That is directly out of a diagnostic
/~N (s_- ) 25 evaluaticn.
NEAL R. GROSS COURT REPORTERS AND TRANSCRIBERS 1323 RHODE ISLAND AVE., N.W.
(202) 2344433 WASHINGTON, D C. 20005 3701 (202) 2344433
256 l l
1 And the last also, is a direct outgrowth of
-s 2 diagnostics; of providing a root cause analysis at the V' 3 end.
4 This was a large team with three State 5 representatives who were team members operating under an 6 existing procedure that the NRC had and specific 1
7 agreements with these individuals.
8 So the total team size was 23 people: three 1
1 9 members who were performing the operations and training i 10 review; four members who were doing the maintenance and 11 testing; a fairly large component of engineering design 12 and technical review -- the tech support area -- six 13 members; three members in management and organization; and
! )
x/ 14 three analytic code support.
15 And Jack Rosenthal was the leader for the 1
16 analytic code support. Most of that work was done at the l 17 Yankee Atomic offices; in fact, 99 percent of it was done 18 in those offices.
19 The level of State participation as I 20 mentioned, was extensive. Three members were actually on 21 the team. There was a process team that observed the team 22 activities at key milestones. Peter Wiley is a 23 representative of the Governor, and Dr. Remick you're all 24 quite familiar with. Both observed the process and met
()
25 with the team on occasion.
NEAL R. GROSS COURT REPORTERS AND TRANSCRIBERS 1323 RHODE ISLAND AVE., N W.
(202) 234-4433 WASHINGTON, D C. 20005-3701 (202) 234-4433
257 1 And we also briefed a citizen's group that the
,ss
, 2 Governor selected, periodically. There were three
( /
3 briefings of this group to keep them informed of the 4 progress. These were people from business and from 5 educational areas and former State employees. They were 6 not involved with the plant, nor were they State 7 officials. And we also briefed the Governor on two
, 8 occasions.
9 MEMBER BARTON: I have a question. On the 10 State representatives, were they active participants in 11 that? They did participate in interviews or were they 12 just observers in those areas?
13 MR. JORDAN: No, they were active p.
w- 14 participants. I guess I would say the limitation was that 15 we did not send the State team members out alone. And so 16 they were with an NRC person or an NRC person was with 17 them through the process.
18 But they were full-fledged members and 19 contributed findings and they participated in the 20 discussion of findings. And I would say, and yet the NRC 21 didn't rely solely on their information.
22 Important to point out is, when we send a team 23 of this size to a site -- and this has evolvr.d over time I 24 think, with the diagnostics and other team reviews --
/ \
(,) 25 there is a licensee -- I'll call it a shadow organization NEAL R. GROSS COURT REPORTERS AND TRANSCRIBERS 1323 RHODE ISLAND AVE., N W.
(202) 234-4433 WASHINGTON, D C. 20005-3701 (202) 234-4433
258 1 or support organization -- that is established at least I l
p- 2 equal to the size of the team.
( )
3 And so they're cognizant on a daily basis of l l
4 the findings of the team. They in fact, once we enter an 5 area of review, do their ovn parallel review, and so j 6 they're an amplifier to the effort. And it was a very, I l 7 think, useful connection. And we've had it appear in the l 1
8 past and it was especially conducive at this time.
9 They brought in a manager from -- a Maine 10 Yankee employee who had been on an assignment to INPO --
11 brought him back and made h m as the Utility Team Manager.
12 And so there were many fir. dings that were associated that 13 were in fact, derived from this effort.
x
\
Y. / 14 CHAIRMAN SEALE: I think that's called playing 15 a man-to-man defense.
16 MR. JORDAN: Yes, it is. But it was -- I 17 would not say it was a defense.
18 CHAIRMAN SEALE: Yes, I agree.
19 MR. JORDAN: And that was really the point I 20 wanted to make sure I made; that it was a complementary 21 aev beneficial complement. The schedule of this effort --
22 and this is following from the diagnostic procedure -- a 23 great deal of team preparation.
24 The team was brought together in mid-June,
~
/ 's
() 25 went through an internal training process, obtained NEAL R. GROSS COURT REPORTERS AND TRANSCRIBERS 1323 RHODE ISLAND AVE., N W.
(202) 234-4433 WASHINGTON, D.C. 20005-3701 (202) 234-4433
l 259 1 documentation from the site -- we do what we call a bagman
,. 2 trip to go to the site z.nd collect large quantities of
,' ( )
3 information -- studied the FSAR tech specs, previous 4 inspection reports, interviewed people in the NRC here 5 that did have or have had, responsibility for reviewing 6 performance of this plant to get a full understanding, and 7 develop then, an inspection plan or a review plan based on 8 previously knowledge but not dependent for their findings 9 on that previous knowledge.
10 MEMBER FONTANA: Pardon me. Over this period l
11 of time, what fraction of full-time do these team members l l
l 12 put into this? )
1 13 MR. JORDAN: About 120 percent.
f7 \
l l
-'\_ / 14 MEMBER FONTANA: Oh, okay. All right, i 15 MR. JORDAN: So these people were dedicated, l l
i 16 embedded --
17 MEMBER FONTANA: And worked very hard.
18 MR. JORDAN: -- during this entire timeframe.
19 We conducted a public entrance meeting. And I'11 20 interject along the way comments that we've received and 21 things that we would do differently.
22 We held a public entrance meeting on plant 23 property and you have to understand that the Maine public 24 is concerned about the plant. This had made the media en 25 quite frequently.
(v )
NEAL R. GROSS COURT REPORTERS AND TRANSCRIBERS 1323 RHODE ISLAND AVE., N.W.
(202) 234-4433 WASHINGTON, D.C. 20005-3701 (202) 234-4433
260 1 And having this public entrance meeting on
,_ 2 plant property -- although it was open to the public --
i )
3 was viewed as an inappropriate thing and we got feedback 4 subsequently that we should have had public meeting in a 5 truly public, off-site location. Next time we would.
6 The on-site evaluation -- there were two 2-7 week periods with two weeks back in Washington after the 8 first 2-week period, in order to assess what we had 9 learned and to decide on any directional changes. During 10 that timeframe, each day the team interchanged of course 11 with the Utility and had discussions with the Utility on 12 any substantive findings.
13 And any time during the period when a safety
,7 s._,/ 14 issue was identified, immediate health and safety issue or 15 operational question was identified, the team manager --
16 team leader -- Ellis Merschoff, met with the Utility i
17 management, explained specifically what the problem was, 18 called the regional office and NRR, and made sure that in i t
19 real time, those issues of operability, immediate safety, 20 were taken care of.
21 So this was not something that grew in the j 22 night and finally hatched on October 7th; it was an 23 interactive thing.
24 At the end of the first two weeks a summary r'N l
! ! 25 discussion was held with the Utility, and of course at the ,
x_/ .
NEAL R. GROSS COURT REPORTERS AND TRANSCRIBERS 1323 RHODE ISLAND AVE., N.W.
(202) 234-4433 WASHINGTON, D C. 20005-3701 (202) 234-4433
261 1 end of the second two weeks a summary discussion was held
-s 2 with the Utility. October loth, after the report was
'7
)
3 issued the 7th, we held a public exit meeting -- another 4 lesson we learned.
5 That's not sufficient time between issuing the 6 report and meeting with the public. Even with modern 7 technology, Internet and fast mail, the public in the 8 vicinity did not have time to have obtained the report, 9 read it, and be in a position to discuss it.
10 The exit meeting was held in two parts. The 11 first part was with the licensee with public observation.
12 We adjourned, then held a meeting with the public -- with 13 public participation. And this was an opportunity for the
!n i v.) 14 public to ask questions, make statements, and interact 15 with us.
16 So the next time we do this sort of an effort 17 I would expect to allow at least a week between issuing 18 the report and conducting the public meeting, and make 19 sure that we disseminate sufficient hard copies so that 20 there's not a problem of reproduction.
21 We sent a copy to the PDR. The PDR was 22 overwhelmed with copying requests. Everybody's not on 23 Internet yet.
24 CHAIRMAN SEALE: Did you hold that second
-m (J) m 25 meeting in a neutral site so that they --
NEAL R. GROSS COURT REPORTERS AND TRANSCRIBERS 1323 RHODE ISLAND AVE., N.W.
(202) 234-4433 WASHINGTON, D.C. 20005-3701 (202) 234-4433
262 1 MR. JORDAN: Yes sir, we did. In a Middle l 2 School in Wiscasset. The process, I think we pretty much
!n\
NJ' 3 talked about, but I will say something about the 4 horizontal assessment. We look at functional areas, that 5 is: engineering, maintenance, operations. Those are the 6 functional areas that we go across the entire plant 7 activities.
8 And then we take -- and you've heard these 9 discussions before -- a vertical slice of selected 10 systems. Both of those two features we've done in the 11 past and have a clear understanding of. The analytic code 12 review we have not, so this was a pioneering effort and I 13 hope you'll find interesting.
i i
\s' 14 Interviews -- with diagnostics and 15 investigations we do extensive interviews. And this was 16 the case here; and I've talked about the root cause 17 analysis.
18 The manner of conducting this effort is !
19 physically walking down systems, putting the Operations 20 personnel on shift for extensive control room 21 observations. During this four weeks on site the plant 1
22 was both operating, had a shutdown, and a restart, so the 23 evolutions of the plant were very evident. So it was 24 excellent opportunities for review.
A 25 The systems that were selected were service (J
NEAL R. GROSS COURT REPORTERS AND TRANSCRIBERS 1323 RHODE ISLAND AVE., N W.
(202) 234 4433 WASHINGTON, D C. 20005-3701 (202) 234-4433
263 1 water and HPSI and those were full, vertical slices. The
~s 2 aux feed water and emergency diesels were a limited
( )
3 vertical slice.
4 MEMBER POWERS: Was there a reason for picking 5 these particular systems?
6 MR. JORDAN: Yes. Certainly from the risk 7 perspective these were important systems and the extent of 8 them, and then the connection to the analytic code 9 reviews, which Jack will make evident.
10 MR. ROSENTHAL: On the code side, there had 11 been --
12 MR. JORDAN: You need to get to the mike.
13 MR. ROSENTHAL: Jack Rosenthal, AEOD.
- i
'v' 14 DR. CATTON: The code was RELAP?
15 MR. JORDAN: No, that's a point I should have 16 made early. RELAP had been done to death, so it had been 17 reviewed and examined and was on a separate track. So the 18 concern here was, were there failures to use and operate 19 the codes -- other codes -- based on the conditions and 20 limitations?
21 So we were looking across and we'll describe 22 in some detail, the other codes that we looked at. Go 23 ahead.
24 MR. ROSENTHAL: I was just going to offer the
/%
(j 25 point that, because the LOCA work had been looked at NEAL R. GROSS COURT REPORTERS AND TRANSCRIBERS 1323 RHODE ISLAND AVE., N.W.
(202) 234-4433 WASHINGTON. D.C. 20005-3701 (202) 234-4433
264 1 extensively, we chose the steam line break because it was
-s. 2 the next most challenging design basis event in terms of
(' >
' "/
3 thermal-hydraulic dynamics. And then we chose the dropped 4 rod -- I mean, it's a boring event thermal-hydraulically, 5 but because it uses so many of the codes in the code suite 6 to produce the setpoints of the plant it provided an 7 opportunity for a broad-scale look.
8 MR. JORDAN: So Jack, if you can both change 9 slides and talk, I'll let you handle this part of it. I i 1
10 didn't tell him I was going to that.
11 DR. CATTON: Jack's very capable.
12 MR. ROSENTHAL: Thank you. Okay, I think what 13 I just did is the introduction of why we chose to look at N_) 14 the steam line break in the drop rod. In fact, that there 15 were 13 codes that they used in the course of doing their 16 analysis. I 17 And we looked at all them for the purpose of 18 just simply saying, were there conditions in SERs that the l l
19 staff had written, and was Yankee Atomic, Maine Yankee, 20 following those conditions?
21 Yankee did not have a well documented path of 22 following those conditions, nor are they required to, but 23 it would have made life easier for them and us in 24 retrospect. And that was one of the issues that spills (Qj 25 over from the LOCA analysis.
NEAL R. GROSS COURT REPORTERS AND TRANSCRIBERS 1323 RHODE ISLAND AVE., N.W.
(202) 234 4 433 WASHINGTON. D.C. 20005-3701 (202) 234-4433
l 265 1 So what we had to do is say, well what is the
/
s 2 SER, what does it say, sometimes, what's the topical j
\ )
3 report? We read all the topical reports, sometimes back 4 code manuals, trying to figure out what were the real 5 conditions that were placed on the analysis.
6 In the case of COBRA we went back to the i
7 authors of COBRA and actually made phone calls to Pacific 8 Northwest to make sure thr.C we understood --
9 DR. CATTON: Don Rowe?
10 MR. ROSENTHAL: To his successor, to 11 understand what the conditions were. And then we looked 12 at their calculations and we concluded that, even though 13 it may not have been written down in a nice, fancy form, em
( )
\_/ 14 in fact, they had followed the conditions in the topicals 15 or the SERs, etc.
16 For example, if the staff had said for the r.
17 fuel work, okay, this is okay up to a burnup of 30- or 18 40,000 megawatts days per ton, in fact that's how they 19 were running the plant and doing their cales.
20 The overall conclusion was that the reload )
21 analysis -- I'll get to the codes in a minute -- was done '
22 very well, and that's something that they do all the time.
23 And you know, they are not co-developers -- as you'll see 24 in the list of codes in a minute -- but they know their r m.
(J) w 25 plant very well.
NEAL R. GROSS COURT REPCRTERS AND TRANSCRIBERS l 1323 RHODE ISLAND AVE., N W. 1 (202) 234-4433 WASHINGTON, D C. 20005-3701 (202) 234-4433 l
266 1 And it's my -- I worked for an NSSS at one 2 point -- that I think they know their plant better than a 7~
() 3 typical NSSS would know the plant in terms of valve 4 lineups, configurations, which RTDs are working slow, 5 etc., what the real pump curves are from real test data, 6 etc.
7 So that's a strength. And then -- I'm 8 reluctant to use the word weakness -- they are not code 9 developers writing code from scratch.
10 MEMBER FONTANA: Jack, going back a couple of 11 viewgraphs -- you don't have to do that -- but it says 12 here, main steam rupture RETRAN had errors but results 13 were not affected?
/^% '
/
(- -
14 MR. ROSENTHAL: Right.
15 MEMBER FONTANA: Were they aware of that?
16 That it had errors but it wasn't affecting the results?
17 Or did they just --
18 MR. ROSENTHAL: That was constructed -- I'll 19 get into that -- in the course of our very active month 20 with them.
21 MEMBER FONTANA: Okay.
22 CHAIRMAN SEALE: Jack, could I ask you a 23 question? You are very familiar with a wide variety of 24 codes and you've gotten your fingers dirty with them. Do
( ,) 25 you believe that it is fair to expect an organization that NEAL R. GROSS COURT REPORTERS AND TRANCRIBERS 1323 RHODE ISLAND AVE., N W.
(202) 234-4433 WASHINGTON, D C. 20005-3701 '202) 234-4433
~_-
267 ;
I has the kind of code interaction that these people to l l
3 2 have, to attract competent code developers? l
,' )
\~/
3 And if they had people that you would identify j 4 as code developers, do you believe that it would be l
5 reasonable to expect those people to be first-rate? l 6 My perception is that the answer to that is,
)'
7 they really don't have a challenging load for code i
8 developers and it would be -- I mean, a guy would have to j i
9 be a fisherman and a moose hunter to be happy at that site 10 if his analytic and professional skills were as a code l l
11 developer. I just don't expect you'd find a first-rate l 12 code developer. l 13 DR. CATTON: Bob, I think in this case Tom !
x ;
( )
i' N_/ 14 Fernandez was at Yankee, wasn't he?
i 15 MR. ROSENTHAL: Yes. j 16 DR. CATTON: And he is first-rate, but he had l
17 left.
18 CHAIRMAN SEALE: Well, yes --
19 DR. CATTON: He's on the LOCA site -- he was 20 there to sort of bring them into the world of using their 21 own codes and then he departed.
22 MEMBER FONTANA: Okay, but what I'm saying is, 23 you can't expect to hold people like that for --
24 DR. CATTON: It's the keeper that's tough.
,r g i
, ) 25 MR. ROSENTHAL: RES has spoken, I think about NEAL R. GROSS COURT REPORTERS AND TRANSCRIBERS 1323 RHODE ISLAND AVE., N.W.
(202) 234-4433 WASHINGTON, D C. 20005-3701 (202) 234-4433
l 268 j i
1 the code work needing three different kinds of people: j i
7 .. 2 the developer -- numerical type jock -- the analyst, and '
e )
3 the interpreter of the results.
4 And the analyst -- how do I model this real 5 plan, how do I do these real problems, and interpreting 6 the results -- they surely have strengths in. But the 7 pure code developer, they have lesser strength, yes.
8 Clearly.
l l
9 If you look at the codes, okay COSMO and 10 SIMULATE are commercial codes -- that's Sudzic & Company.
i 11 RETRAN is EPRI, and so we have to look at the code work by j J
l 12 EPRI. EPRI's got to make up for the strengths, Dr. !
l a
13 Seale --
r%
? i N/ 14 CHAIRMAN SEALE: Yes, sure, sure.
15 MR. ROSENTHAL: -- that Yankee or Duke Power 16 wouldn't have. STAR started out as a Ph.D. project at MIT i
i 17 and ultimately evolved into a 3-D space / time kinetics i 1
1 18 code. FPOSSTEY actually came from GAPCON which was an 19 NRC-developed code at PNL, and then -- it's interesting <
20 the way we approached it, I think.
]
21 There was FROSSTEY-1 which was received from ]
i 22 Yankee to the staff -- and then the staff had a lot, a lot i 23 of questions, and used PNL as the reviewer. Remember, PNL j s
I 24 was the developer of GAPCON. So they really got a (D
(_) 25 workover on FROSSTEY, to the point that they renamed it i
NEAL R. GROSS COURT REPORTERS AND TRANSCRIBERS 1323 RHODE ISLAND AVE., N.W.
(202) 234-4433 WASHINGTON, D C. 20003-3701 (202) 234-4433 !
l 269 1 FROSSTEY-2 by the time they were done, and it was PNL who
,s 2 was doing that review.
t i s"'/ 3 But when we saw that history of it, we were 4 confident that that had received a very good staff review.
5 The space / time kinetic stuff was looked at by Brookhaven 6 who had a lot of experience in that, and that also gave us 7 confidence.
8 COBRA was developed by the NRC so they're not 9 code developers. The CFH -- the heat transfer coefficient ;
10 -- actually, I thought that they did a very good job l 11 here -- and it's a contemporaneous, generic question.
l l
'2
_ Maine Yankee has Westinghouse, CE, and Siemens l l
13 fuel in the reactor, and they'll buy the next batch of l
[)
\_/
l 14 fuel from whoever is the low bidder, and good for --
15 that's proper. But they have to ensure the compatibility 16 of fuel. This cycle they're using, Westinghouse IFBA fuel 17 -- burnable poison is integral to the fuel and they have 18 some leakers and they're trying to find them.
19 So it's their obligation then, to ensure the 20 compatibility across the batches. They have the Columbia 21 data for the heat flux coefficient -- the CH correlation -
22 - and they're very good about their procurement specs such 23 that the Columbia test data is applicable.
24 And there are certain features for example, of p) v 25 Westinghouse fuel with additional mixers, which they have NEAL R. GROSS COURT REPORTERS AND TRANSCRIBERS 1323 RHODE ISLAND AVE., N W.
(202) 234 4433 WASHINGTON, D C. 200C5-3701 (202) 234-4433
270 I 1 not bought such that they keep the fuel consistent with I
,_ 2 test data for the heat transfer coefficient.
i \
~'
3 So they're very cognizant of mechanical l
l 4 properties and materials properties and chemical 5 interference type stuff, which I would expect everybody to 6 be. But then they are also very conscious of maintaining 7 the applicability of the heat transfer coefficient. And 8 we thought that was actually the strength.
9 MEM5ER POWERS: Do you know how far they're 10 taking their fuel as fal d '; burnup?
11 MR. ROSENTHAL: Yes, except I did this come i l
12 months ago and I don't remember the number.
13 MR. JORDAN: I don't remember either.
U)
(
14 MR. ROSENTHAL: It's in the report.
15 MR. JORDAN: We'11 come back to you.
16 MEMBER POWERS: Really, what I'm curious about i
17 is, are they getting to the points where the fuel 18 microstructure is changing around the rim and some of the 19 experiences that PNL's likely to have and PNL-developed 20 codes are just going to start breaking down badly?
21 MR. ROSENTHAL: Right. The answer is -- we're 22 talking about Maine Yankee?
23 MEMBER POWERS: Yes.
24 MR. ROSENTHAL: The answer is, no. They're f3 t, 25 right up against the approved edge of what the staff has NEAL R. GROSS COURT REPORTERS AND TRANSCRIBERS 1323 RHODE ISt.AND AVE., N.W.
(202) 234-4433 WASHINGTON, D.C. 20005-3701 (202) 234-4433
271 1 permitted and are participating -- for our discussions
,~ 2 with them they are aware of these emerging high burnup l i )
'~'
3 issues and have given papers at some of the meetings.
4 MEMBER POWERS: I understand that --
5 MR. ROSENTHAL: I think that they're right up 6 there. i l
7 MEMBER POWERS: I understand that the staff's 1
8 approved well into an extrapolation range and relative to 9 where we have experimental data to validate our codes. We 10 have a database that goes up to maybe 33,000 megawatt days 11 per ton, and staff's approving up into the 50s.
12 MR. JOPDAN: We'll gib you a en?ber back, 13 unless somebody's found it in the report. I would add g~g
'ss' 14 that the present problems that Maine Yankee has with 15 leakers is fresh fuel and it's mechanical. So it's not a 16 burnup-related failure of cladding.
17 And I would add that, in Jack's comments he 18 was not suggesting that the Utilities or even their 19 support, necessarily need to be code developers.
20 But I think what he conveyed was that the team 21 that we sent was made of people that were code developers 22 or had worked extensively with codes, and they gave the 23 Utility, Yankee Atomic, I think a very good workout.
24 This was a deeper look than we had gone in
, ~s i
25 previous work with codes And I think this particular (d
NEAL R. GROSS COURT REPORTERS AND TRANSCRIBERS 1323 RHODE ISLAND AVE., N W.
(202) 234-4433 WASHINGTON, D C. 20005-3701 (202) 234-4433
272 1 1
1 organization was in better shape than many we would come 1
7- 2 to in that fashion. l N ,Y
~ ,
3 CRAIRMAN SEALE: My question was a paraphrase l 4 of an observation that fellow that used to be around here 5 quite often would make and that is that we didn't expect 6 these people to be world-class in that area, but that we 7 just wanted to be sure they were competent.
1 8 MR. JORDAN: Yes.
9 CHAIRMAN SEALE: And we should note for the 10 record that Dr. Lewis just came in and his outer garment 11 is not a T-shirt.
12 (Laughter.)
13 MR'. ROSENTHhC: Going down the list, they do l n
I xs' 14 statistically combine their uncertainties and their 15 thermal margin setpoints. And we broached the subject
. 16 with them that one of the things that engineers most 17 misuse are statistics, and they assured us that they had 18 used some academician from Boston to help, to 19 independently review the statistics.
20 And then we checked on what the staff had done 21 and in fact, the work had been looked at by some of the 22 statisticians on the staff.
23 At any incident in time there will be an 24 analysis of record for that plant, and the steam link r~'s
( ,) 25 break goes back to cycle 11 -- I believe it's cycle 11 --
NEAL R. GROSS COURT REPORTERS AND TRANSCRIBERS 1323 RHODE ISLAND AVE., N.W.
(202) 234-4433 WASHINGTON, D C. 20005 3701 (202) 234-4433
273 1 but that was RETRAN 02 Mod 2. If they were to do it today
, ~\ 2 they would use RETRAN 02 Mod 5. They represented the core
\ )
'~'
3 as a single flow channel.
4 If they were to do it today they would break '
5 it up to better represent the effect of what's going on 6 inside of the vessel on the broken steam generator -- the 7 faulted steam generator -- and then the good ones. So 8 there's an evolution in the degree of sophistication the 9 code used.
10 Nevertheless, the analysis of record for them 11 was the RETRAN 02 Mod 2 analysis, so that should be 12 correct. And there was -- we did find some errors. That 13 was the point that we found the most differences.
p t
N/ ) 14 On the other hand, the way they were modeling 15 steam line break they were forcing an almost unphysical 16 cool-down of the pritrary system. And they don't get a 17 return of criticality, and without a return of criticality 18 it's no danger to the fuel. So that we could conclude 19 that the overall analyses was okay. Nevertheless we did 20 see -- we did have a lot of questions about how they were 21 doing the modeling.
22 Let me point out that CHIC-KIN is an old 23 Westinghouse betescode that was used for the injected rod 24 analysis. And GEMINI is a very, very, very simplistic
/
(,) 25 model. And so what you have to do is look at the codes NEAL R. GROSS COURT REPORTERS AND TRANSC91BERS 1323 RHODE ISLAND AVF , N.W.
(202) 234 4 433 WASHINGTON, D C. 20005-3701 (202) 234-4433
274 1 within their application,
,_. 2 We would never approve GEMINI, which is just a i
'J 3 new-node model, for a steam line break analysis or a LOCA 4 analysis, nor would they use it that way. And that for 5 things like the drop rod where hardly anything's going on, 6 you only have a few degrees change in temperature, it's 7 okay to use a simplistic model. So you always have to l I
8 look at the model versus the application.
1 9 That's all I want to say.
10 MR. JORDAN: Okay. I think I'm back up here.
11 In order to adequately express our findings we sort of ,
i l
12 took it in a 3-way approach. First of all, the 1
13 conformance to the regulations. l (y
- \
'\s / '4
. Second was an assessment in the manner of a 15 systematic assessment of licensee performance. We use 16 that language which got us in trouble subsequently, but 17 nevertheless we used it. And then finally we looked at 18 the risk with respect to the various findings to try to 19 express it in terms of risk.
20 The overall performance -- and this is I guess 21 you'd say the first place we got in trouble in terms of 22 the public's concerns -- the idea that overall performance
[ 23 is adequate for operation.
24 That's kind of a weak statement in a public r~s
() 25 meeting, and there is an expectation on the part of the NEAL R. GROSS COURT REPORTERS AND TRANSCRIBERS 1323 RHODE ISLAND AVE., N.W.
(202) 234-4433 WASHINGTON. D C. 20005 3701 (202) 234-4433
275 1 public; that overall performance shall be without flaw, 2 and shall be totally consistent with regulatory
/
l
') I 3 requirements. With the level of review that we did, we 4 didn't expect going in that it would be without flaw or I 5 without identification of non-conformance, non-compliance.
6 And so I guess in retrospect, in entering this 7 process, we would need to make that a clear statement to 8 the public; that this is going to be painful to everybody 9 and a team like this is likely to find problems including 10 violations of regulatory requirements.
11 So that didn't play very well and I think I 12 learned a lesson from it.
13 The design and licensing basis which is the i i
(_,/ 14 fundamental reason that we went in, was generally in 15 conformance. That is, there were not serious problems 16 found with respect to the licensing basis.
17 There have been spinoff problems in addition 18 to those that the team found, with regard to the licensing 19 basis that have now resulted in the plant being down for a 20 time until they're resolved. And so in that case it's to 21 the licensee's credit that they extended these reviews and 22 have pulled the string and found additional issues.
23 The operation side of it in terms of the 3
24 control room procedures, training, etc., for the operative
/N
( 25 staff was very good. These are experienced plant
\_s)
NEAL R. GROSS COURT REPORTERS AND TRANSCRIBERS 1323 RHODE ISLAND AVE., N.W.
(202) 234-4433 WASHINGTON. D.C. 20005-3701 (202) 234-4433
a 276 1 operators, they have a lot of pride in their work, and 2 they were doing a good job. Maintenance was generally
(, ,)
'~'
3 good.
4 Testing -- there were weaknesses there. We 5 called that acceptable in the SALP-type language.
6 Engineering was good. Self-assessment and corrective 1
7 actions were acceptable. A lot of weaknesses there. And 8 that will show up in the summary of root cause analysis.
9 CHAIRMAN SEALE: Ed?
10 MR. JORDAN: Yes sir. l l
11 CHAIRMAN SEALE: You characterized your I
12 overall performance of adequate for operation as kind of 13 almost damning with faint praise.
?~s
/ )
i, / 14 MR. JORDAN: Yes. l 1
15 CHAIRMAN SEALE: If I look at the six items l 16 that are there, I have a hard time coming up with a 17 weighting of each one of those items which would lead me 18 to that conclusion. I mean, you're very good in 19 operations and you're good in engineering, and you're good 20 in maintenance. And I have a hard time adding those 21 things up with a weighting factor and coming up with 22 something that is acceptable for operations.
23 MR. JORDAN: Okay. And it's the other three 24 things that -- in the design basis there were features at
(
(3 ,
j 25 the plant that in fact, did not -- would not perform as NEAL R. GROSS COURT REPORTERS AND TRANSCRIBERS 1323 RHODE ISLAND AVE., N W.
(202) 234-4433 WASHINGTON. D C. 20005-3701 (202) 2344433
l 277 1 they should.
,-- 2 And then I'll give an example. If you're in a i \
' '~
3 recirculation mode in an accident where you're pumping 4 from the sump, they had an NPSH problem for the pump and 5 they had a problem of submergence of electric equipment 6 instrumentation with the sump level that would be 7 attained.
8 Those had to be fixed, and once they were 9 fixed than that limitation on operation went away. But 10 they were in fact, indicative of the potential for other 11 problems of that nature. And that's where, when the 12 Utilities pulled the string they found other problems.
13 So that was the cide of what caused this i r x/ 14 adequate for operation. Once those deficiencies that we 15 identified were fixed, then the plant was back operable.
16 But there were still weaknesses.
17 In the testing, there were initial -- I'll say 18 surveillance tests that were not done. The FSAR required 19 systems to operate in a particular fashion, and it turned 20 out the testing program that was done periodically did not 21 demonstrate all the logic for those systems.
22 When the Utility was requested to do the 23 testing, to show us that the logic worked, it didn't work 24 in a couple of cases. And that was a very dramatic one.
I 25 One of the reasons it didn't work was a piece of wire was NEAL R. GROSS COURT REPORTERS AND TRANSCRIBERS 1323 RHODE ISLAND AVE., N W.
(202) 234-4433 WASHINGTON. D C. 20005-3701 (202) 234-4433
278 1 missing; it had been cut out somewhere during the life of 7w 2 the plant.
+
\
3 MEMBER BARTON: Ed, was that identified by the 4 team or was that self-identified by the Utility?
5 MR. JORDAN: That was identified by the team.
6 MEMBER BARTON: It was? All right.
7 MR. JORDAN: The team didn't do the testing --
8 MEMBER BARTON: You identified the --
9 MR. JORDAN: But we identified that here is --
10 MEMBER BARTON: The deficiencies in the 11 surveillance testing program?
I 12 MR. JORDAN: Right. So by look'.ng at the i 13 FSAR, looking at the --
(\ I
, )
\/ 14 MEMBER BARTON: Surveillance --
l 15 MR. JORDAN: -- wiring diagrams and then l
16 looking at the test procedure, the surveillance test 17 procedure that the Utility used, we said these parts of 18 logic aren't tested. Why not? And so that thing led to l 19 the testing.
20 And in any one of these kinds of efforts it 21 really takes a smoking gun for the team to obtain 22 credibility with the Utility, and that happened early l 23 enough that they got very strong credibility. And the 24 Utility was quite responsive: in following up on those
/ )
(,) 25 kinds of issues.
NEAL R. GROSS l
COURT REPORTERS AND TRANSCRIBERS 1323 RHODE ISLAND AVE., N.W.
(202) M 4433 WASHINGTON, D.C. 20005-3701 (202) 234-4433
279 1 And those are the kinds of things that caused 73 2 us to say, okay, they corrected that problem and they
( )
3 looked further to see other, similar problems, but that's 4 where we get to say, okay at that point, once fixed it's 5 adequate for operation, but it's not great.
6 MR. ROSENTHAL: Let me give an example -- and 7 this permeates the plant. This would be a HPSI flow --
8 high pressure safety injection flow. What L.t did is
)
1 9 they took the -- actually, before the plant is built you l l
10 do an analyses, so you take what you think is the design l
11 flows that you're going to have, you knock off ten l
12 percent, and that's what you use in your first round of l
13 LOCA analyses. And you demonstrate then, that that's l
("T t a l
\j 14 acceptable. l l
15 Those curves then became the values that they 16 were testing to. No margin, because they had -- it had 17 already been taken into account. So at that point -- and 18 it's not unique to high pressure safety injection -- they 19 have to test to show that they're within the assumed 20 curves used in the safety analysis but with no margin.
21 And in turn, every test becomes very, very 22 critical. In this case, the question of how you set up 23 the discharge -- you know, you flow trend -- the HPSI flow 24 to each injection point became critical, and small changes
,m
) 25 would have made the safety analysis no longer valid. And NEAL R. GROSS COURT REPORTERS AND TRANSCRIBERS 1323 RHODE ISLAND AVE., N W.
(202) 234-4433 WASHINGTON, D C. 20005-3701 (202) 234-4433
280 1 it permeates the plants.
7- .
2 MR. JORDAN: I think we ve sort of covered
( )
3 this -- and I would jump to the bottom line in the design 4 basis. We did find cther features in the analysis that 5 did not support 2700. In this process it identified a 6 number of other things, other than the RELAP/5YA analysis 7 that were a problem.
8 And so for the Utility to regain that upgrade 9 they will have to do further analysis, maybe physically 10 change some equipment.
11 The next one is simply sort of a brief listing 12 of some of the operabilities issues that were identified.
13 I talked about the logic circui.ry and the containment
- i
(/ 14 spray pump, PSH. Equipment qualification was a )
15 submergence.
16 The reactor water storage tank, that was a 17 level indication problem. That one was fairly subtle in 18 that the instrumentation that initiated logic -- there 19 were obvious redundant sets of instrumentation. One train 20 of the instrumentation had heat tracing on it. It was set 21 at a remarkably high temperature beyond the design of the 22 equipment.
23 And so the sensor itself, if that temperature 24 was aged beyond its useful life by some considerable
(% )
g 25 margin and there was a question of what was the precision x/
, NEAL R. GROSS COURT REPORTERS AND TRANSCR:BERS 1323 RHODE ISLAND AVE., N W.
(202) 2344433 WASHINGTON, D C. 20005-3701 (202) 234-4433
281 1 and capability of that -- so that was a subtle feature but I
s
\
2 it was I think -- the level of structure the team looked
\/
3 at and that there were problems of that nature.
4 Equipment ventilation -- I'm sorry, the 5 ventilation system -- there were a number of problems with 6 the ventilation system with respect to accident conditions 7 and those temporarily were resolved for the summer by 8 blocking intake air open. And of course for the winter 9 that's not a good solution; it has to be fixed. That's 10 enough on that.
11 We talked about the operation sufficiently.
12 We talked a little bit about maintenance, and this is 13 where PRA was interesting. There is a weakness
[ )
k/ 14 identified, inconsistent equipment reliability.
15 The team looked at plant data -- that is, went 16 back to contrc'. logs, maintenance logs, etc. -- and then 17 constructed the reliability and surveillance logs, the 1
18 reliability of the principle safety systems and trains of ;
1 19 system $, and identified through that effort that the steam 20 driven aux feed pump system had a remarkably poor 21 reliability.
22 And the Utility was surprised. And so 23 although they were using PRA to, at least an average or 24 better extent in the plant, they hadn't used it rigorously
/~'s (j 25 at all in looking at equipment reliability. And so the --
NEAL R. GROSS COURT REPORTERS AND TRANSCRIBERS 1323 RHODE ISLAND AVE., N.W.
(202) 234-4433 WASHINGTON, D C. 20005-3701 (202) 234-4433
282 1 you know, failure to start, failure to run data -- over a g3 2 period of time clearly established that that steam driven
\ l 3 aux feed pump which is very important in the loss of power 4 sequence, was not reliable.
5 MEMBER BARTON: That's interesting. At the 6 time of your inspection when the maintenance rule should 7 have been in full implementation at the site, because your 8 inspection period was after July, right?
9 MR. JORDAN: Yes. And there is n, I think an 10 inherent weakness at not looking rigorously at reliability 11 but looking at it in a subjective or deterministic way.
I 12 And that can be done.
p_
13 MEMBER POWERS: Did you look to see how their
! ) 1 k/ 14 quality assurance program was operating? With regard to l 15 these taings. l l
16 MR. JORDAN: Yes, and I think maybe that would 17 show up on the slide 19, sort of in the self-assessment 18 corrective actions. That was a fairly weak area. That 19 one, their self-identification is not very strong and two, 20 once identified, there was a weak implementation of 21 corrective actions if they knew there were deficiencies.
22 MEMBER POWERS: Did you try to identify a root 23 cause for why it would be weak, or --
24 MR. JORDAN: Yes, we -- well, what we did was
,m
! \
( ,/ 25 roll all of these strengths and weaknesses up to try to NEAL R. GROSS COURT REPORTERS AND TRANSCRIBERS 1323 RHODE ISLAND AVE., N.W.
(202) 234-4433 WASHINGTON. D C. 20005-3701 (202) 234 4433
283 1 then find the basic root causes from the middle root
~s 2 causes --
\'~] 3 MEMBER POWERS: All right.
4 MR. JORDAN: -- rather than root causes by 5 area. But it clearly addressed that --
6 MEMBER POWERS: I was looking for more focused 7 root cause, specifically in the QA area.
8 MR. JORDAN: I don't have direct recollection.
9 Does that -- do you have anything to add to that?
10 MR. ROSENTHAL: Let me say, on the code side 11 they are committed to the ANSI standard on QA. Don't look 12 to the QA Department to find errors or questions on the
,_ 13 modeling, the nodalization vf the thermal-hydraulics.
/ \
's /
' 14 MEMBER POWERS: No, no.
15 MR. ROSENTHAL: But what they did, what you'd 16 expect of a QA organization in terms of having procedures 17 in place, oversight, etc., we were satisfied with them at 18 Yankee Atomic, and I know that we looked at Maine Yankee.
19 MEMBER POWERS: What I'm wondering in the 20 hardware QA area, that maybe dilution of effort is the 21 root cause for this.
22 MR. JORDAN: Dilution of effort?
23 MEMBER POWERS: Yes. Asking too small of a 24 group of people to do too much.
/^N
( ,) 25 MR. JORDAN: Ah, Yes, resources were a NEAL R. GROSS COURT REPORTERS AND TRANSCRIBERS 1323 RHODE ISLAND AVE., N.W.
(202) 234-4433 WASHINGTON, D C. 20005-3701 (202) 234-4433
l l
284 ,
l 1 problem, and that does show up in the root cause. Let's
,_ 2 discuss it when we get back to there.
( )
3 MEMBER BARTON: You'll get to that.
4 MR. JORDAN: Yes. The next is the testing, I 5 and I've already indicated that there was a problem. They 1
6 did have of course, an extensive outage for steam i 1
7 generator tube crackina and had sleeved tubes and had done 1
8 an extensive program there. So there was a strength in 9 the work that they did in that area. And there were 1
10 weaknesses that we've talked about.
11 Safety assessment engineering. A small 12 engineering staff, they received a great deal of support 13 from Yankee Atomic. There was an interface problem e a
\_./ 14 between the engineering staff on site and Yankee Atomic.
15 And the example I would give is, while Jack 16 was in the Yankee Atomic offices and the team was on site, 17 Jack would phone Ellis Merschoff and advise him of his 18 findings over the past 24 hours2.777778e-4 days <br />0.00667 hours <br />3.968254e-5 weeks <br />9.132e-6 months <br />, and then we would meet 19 with the Utility engineering team who had gotten l l
20 information from the Yankee Atomic organization about his l 21 findings over the past 24 hours2.777778e-4 days <br />0.00667 hours <br />3.968254e-5 weeks <br />9.132e-6 months <br />, and in many cases they l 22 were hardly recognizable. That communication didn't work 23 very well.
1 24 And so that was a problem that Yankee Atomic l
) 25 was providing an overall product but the coordination l
NEAL R. GROSS COURT REPORTERS AND TRANSCRIBERS 1323 RHODE ISLAND AVE., N.W.
(202) 234-4433 WASHINGTON, D.C. 20005-3701 (202) 2344 433
285 1 between the two sets of engineering support were not 7s 2 sufficiently strong. And there was, at the plant, real N)
3 weaknesses in problem identification and resolution.
4 Self-assessment and corrective actions. That 5 was a weak area. There were significant problems with l 6 their ability to find and fix their own problems. And 7 then I'd like to get to the root causes.
8 And the root causes are linked very closely --
9 and we used too many words. We ought to put it on one 10 sheet somehow. Because that Utility -- after all, their 11 purpose is to provide electricity at a low cost -- but 12 their pressure on their economic side to be this low-cost 13 producer in Maine has caused the staff and the management
- \
\/x 14 to feel that they're under very tight constraints.
15 So they really did have economic pressures 16 that limited their resources. An example would be how 17 much engineering support they had. They had a huge 18 backlog of work. Their corrective action program -- big, 19 big backlogs. And the identification, because of these 20 backlogs and because of the limitations, people didn't 21 identify -- I'll say the lesser nuisance issues that 22 accumulate.
23 So the plant had a lot of nuisance problems --
24 these are what we call workarounds -- that they lived i,v) 25 with. Because if they put it on the list it would make NEAL R. GROSS COURT REPORTERS AND TRANSCRIBERS 1323 RHODE ISLAND AVE., N W.
(202) 234-4433 WASHINGTON, D C. 20005-3701 (202) 254 4433
286 1 the list longer but it wouldn't get fixed.
,- 2 DR. CATTON: How do you decide -- how did you
(' l
'~J 3 come to the conclusion that the available resources were 4 limited? Do you have some numbers in mind when you look 5 at how they're spending their money?
6 MR. JORDAN: Not at all. What we're looking 7 at is that they had a growing backlog of maintenance items 8 and the things out in the plant that didn't get on the 9 maintenance list --
10 DR. CATTON: So they weren't increasing the 11 allocation of resources to take care of that?
12 MR. JORDAN: They did not; that's right.
13 MR. ROSENTHAL: To the contrary, you know, n
\
x_) 14 when they were shut down to do this major steam generator 15 work -- and good for them for using advanced technology to 16 examine their tubes -- it would have been an ideal time to 17 work off your backlog. After all, you're shut down, you 18 can get in, you can work on the plant. I 19 But that's not what happened and in fact, they 20 took funding from maintenance and engineering and Yankee,
? .
21 etc. -- everybody had to contribute part of tneir budget i
I 22 to the steam generator tube inspection work. So that's an 23 example where you can clearly see their limited funding.
24 MR. JORDAN: They have an unusual structure in I r~~N
! ) 25 that they have a number of owners and they do not put NEAL RL. GRO'SS l COURT REPORTERS AND TRANSCRIBERS 1323 RHODE ISLAND AVE., N.W.
(202) 234-4433 WASHINGTON, D C. 20005-2701 (202) 234-4433 k I
287 1 money into a -- I'll say a fund -- for the purposes of
,y 2 future maintenance.. And so each time a crises is reached
)
3 they go back to the owners and obtain funds to get over 4 that crises.
5 DR. CATTON: Do you think if the money flow 6 were monitored, if you knew what it was. you could just 7 look at that and come to some of these conclusions?
8 MR. JORDAN: You asked the question early on, 9 did we have something in mind. The NRC tries very hard 10 not to have a value in mind. We're looking -- trying to 11 look at the performance side of it --
12 DR. CATTON: I know that; I know NRC does e
13 that. But I happen to think that if you were to watch the
\~ / 14 dollars you'd learn a hell of a lot. People put their 15 money where their heart is -- is a popular phrase these 16 days.
17 MR. JORDAN: Then I'll jump to -- we had a --
18 and in fact, you got a briefing yesterday I think, on the 19 Arthur Andersen study, and I was involved in that. What 20 we are looking for on a longer range is, is there evidence 21 that a Utility is under economic stress? And this would 22 be from commercially available facts regarding their 23 financial resources. Then look for the safety strains at 24 those plants that are under the greatest economic stress.
( ,/ 25 DR. CATTON: In other arenas, people who are NEAL R. GROSS COURT REPORTERS AND TRANSCRIBERS 1323 RHODE ISLAND AVE., N W.
(202) 234-4433 WASHINGTON, D C. 20005-3701 (202) 234-4423
288 1 very successful often use that as the quick look where the
,3 2 dollars are going.
i \
i
/
3 MR. JORDAN: Yes.
4 MEMBER BARTON: I think what complicates that 5 Ivan is that -- and we heard it yesterday and it's well 6 known in the industry -- that some of the better 7 performers are also low-cost performers.
8 MR. JORDAN: But I agree with you. It's where 9 --
10 DR. CATTON: But you can tell when you look at 11 the overall structure of the money flow.
12 MR. JORDAN: Since the next group is already 13 here I'd better move. The other root cause is the idea of
) 14 a lack of questioning attitude or culture, and this is the 15 inability of the Utility to self-identify. And the view 16 that the team came to that the management was complacent 17 at this facility with --
18 MEMBER KRESS: Ed, how do you measure the 19 extent of a lack of questioning culture -- the extent of a 20 questioning culture in the first place?
21 MR. JORDAN: Well, first of all, when someone 22 finds a ravel, if you don't pull it then you really have a 23 lack of a questiening attitude. And they had a lot of 24 ravels. They had indications of equipment deterioration
(_,/ 25 in limitations of workarounds for the operators and they NEAL R. GROSS COURT REPORTERS AND TRANSCRIBERS 1323 RHODE ISLAND AVE., N.W.
(202) 234-4433 WASHINGTON, D.C. 20005-3701 (202) 234-4433
. .. - . - . . .. . . . . - - . . - . . . . ~ . . . _ _ - . - . _ - . - - - . - . - . - , . . .
289 1 just weren't facing them. They were just' letting them 4
2 alone. i
$I
- 3. And that was the evidence of the questioning i.
-4. culture and.the idea of complacency on the part of the I
5' management. They were happy with.what they were'doing and
'6 where'they were and did not~ have a sense that -- I'll say'-
7 the world was passing tihem by in terms of ~ their level of ;
- 8 safety performance.
Seems like you're' measuring..
, 9 MEMBER KRESS:
p 10 attitudes and motivations -- kind of --
i 11- MR. JORDAN: I'd love to be able to do that.
4 I-i 12 We're inferring from poor p'erformance -- and so this is a
- 13- rollup and this'is -- these seem rather obtuse perhaps, 14
~
and high level, but the team of 23 people spent a great:
15' deal of time trying to say, okay, what's really behind 16' this? And from the interviews and from the actual-s 17 performance of the various functional areas, this is'how
~18.
they came out.
19 MEMBER BARTON: Tom, I think if you -- I'm
- 20. sorry, go ahead.
'21 MR. JORDAN: And I think the piece that I'm v
22 very pleased with is that in subsequent meeting with the 23 Utility after conveying the findings to'them, the Utility 24 itself said, yes, we had become complacent and we're not 25 keeping up with the times in terms of the questioning NEAL R. GROSS COURT REPORTERS AND TRANSCRIBERS 1323 RHODE ISLAND AVE., N.W.
(202) 234-4433 - WASHINGTON, D.C 20005 3701 (202) 234-4433
290 l 1 attitude.
,- 2 MEMBER BARTON: I think if you look at the f
)
first sentence of that second root cause, one way you can
~
3 4 measure this, how effective a Utility is, is you look at 5 the effectiveness of their deficiency reporting system and 6 how low the threshold is. It gives you a good indication 7 of, are people identifying issues and at what level, and 8 then how is the Utility responding to those in taking 9 corrective actions. And I think that you've shown that 10 that really was not evident.
11 MR. JORDAN: It was not evident.
l l
12 MR. ROSENTHAL: Let me say that, I can take 13 examples involving the single steam generator atmospheric i
\_) 14 dump valve that I showed, or an issue on testing of air-1 15 operated valves at the bottom of a mort-like structure and l 16 work it up to support these. So we have hardware examples 17 to support these conclusions. ;
l 18 MEMBER FONTANA: Okay, I think the question l 19 comes to mind that is probably not answerable at this 20 time, but under a more performance-based regulatory 21 structure, how would you go about finding these things?
22 MR. JORDAN: Well this -- I guess I'm pleased 23 to say that what we did was a performance-based review.
24 And so we derived these root causes based on performance.
,,3
!, ) 25 This was not a touchy-feely look. This was really, does NEAL R. GROSS CCURT REPORTERS AND TRANSCRIBERS 1323 RHODE ISLAND AVE., N.W.
(202) 234-4433 WASHINGTON, D.C. 20005-3701 (202) 234-4433
291 1
1 the equipment operate, what deficiencies are there, and 73 2 what kind of maintenance are they actually doing? So this 3 was performance-based.
4 MEMBER BARTON: This was a more performance 5 than compliance-based look? I 6 MR. JORDAN: Yes.
7 CHAIRMAN SEALE: And of the work list, the 8 work that has to be done -- the ma:.atenance list . Is it 9 changing?
10 MR. JORDAN: Right.
11 CHAIRMAN SEALE: Are things moving through it?
12 Are things sitting on it for three years?
13 MEMBER FONTANA: Okay, so this in your mind, I
k ) 14 was a pretty good test of that approach?
15 MR. JORDAN: It was, yes.
16 MEMBER FONTANA: Would it have been triggered 17 the same way? The whole approach, it was triggered 18 because of a miscalculation on --
19 MP. JORDAN: Yes, well that's -- let's go to 20 the last slide. I'm glad you asked that. Every time we 21 do something of this magnitude the NRC learns things about 22 itself. And this is a partial list.
23 The analytic code validation -- this turned 24 out to be, I think a positive with this Utility -- the p
(j 25 very thing that sent us in was not flawed, and that was NEAL R. GROSS COURT REPORTERS AND TRANSCRIBERS 1323 RHODE ISLAND AVE., N.W. j (202) 234 4433 WASHINGTON, D.C 20005-3701 (202) 234-4433 l
292 1 good news. The other things that we looked at while we
, s, 2 were there ascessing this licensee, there were problems.
i
)
3 And some of those are related to what the NRC does or 4 doesn't do.
5 Compliance with the safety evaluation reports.
6 , The staff does not use those, had not used those to the 7 extent that they should in inspection and licensing 8 reviews, and so that's a weakness. And so we have l 9 internal lessons learned to address that. l 10 And there was a question of licensing reviews 11 for power uprates, and so we have a lesson to re-examine 12 some of those other power uprates to verify that in fact 13 they're okay or that there are deficiencies.
-;, \
'/
14 The NPSH requirements. It's a generic issue 15 and it's being addressed. The inspection program and the 16 training program that supports it -- we should do a better 17 job to equip our people, give them the tools that they can 18 find things in a more effective fashion. Don't like 19 surprises and we got too many surprises out of this 20 effort.
21 That's all I have. Be glad to address any 22 other questions.
23 MEMBER BARTON: Are there any other questions 24 the ccmmittee has? Well if not, I want to thank Ed and
,m k ,,) 25 Jack for their presentations. Mr. Chairman, I think the NEAL R. GROSS COURT REPORTERS AND TRANSCRIBERS 1323 RHODE ISLAND AVE., N.W.
(202) 2344433 WASHINGTON, D C- 20005-3701 (202) 234-4433
293 1 Plant Operation Subcommittee will continue to follow the 2 progress at Maine Yankee.
I 3 In addition, I think our work with the staff, 4 as we discussed yesterday following the presentation we 5 had on the Arthur Andersen report -- the strength in the 6 SMM process -- we'll continue to follow that as well to 7 ensure in the future that plants that have a good 8 reputation like Maine Yankee did have, don't get surprised 9 and end up on the watch list.
10 Thank you.
11 MR. JORDAN: Thank you.
12 CHAIRMAN SEALE: Thank you. Before you leave,
,_ 13 Ed, we should have congratulated you -- or commiserated,
' ')
14 as the case may be -- on your recent promotion. You 15 certainly -- you put in your time I'm sure and we wish you 16 very well in that position. We look forward working with 17 you, very much.
18 MR.. JORDAN: Thanks.
19 CHAIRMAN SEALE: Okay. Well, it's time to go 20 on. The next item, which will go from now until 12:30, at 21 which time we will break -- or we'll go for lunch -- there 22 will be a break in there somewhere. I just want to remind 23 you of that.
24 The National Academy of Sciences National i
g\
\ ,/ 25 Research Council Phase 2 Study Report will occur now. Don NEAL R. GROSS COURT REPORTERS AND TRANSCRIBERS 1323 RHC.DE ISLAND AVE., N.W.
(202) 234-4433 WASHINGTON, D C. 20005-3701 (202) 2344433
l 294 l 1
1 1 Miller is the Chairman of the I&C Committee and I'll turn p-~ 2 it over to you, Don.
( l
'~'
3 MR. MILLER: Thanks, Bob. First of all I 4 think we're making arrangements for a majority of the 5 panel to sit in front -- and I assume Doug Chapin is here 6 somewhere, who'will sit in the middle. So I'll give them 7 time to kind of settle themselves in the room here.
8 Okay, I'd like to -- as people are now getting 9 settled I'd like to open this session with a few remarks.
10 But first of all I'd like to say I'm delighted that so 11 many of the members of this panel could attend this 12 meeting and certainly look forward to your presentation.
13 But before we get started we also have with us ip
'N_)
14 several other guests. First of all, I'd like to introduce 15 Dr. Harold Lewis. Hal, I'm glad to have you with us. I i
16 know you had some, I guess role in getting this all 17 started in the early '90s, and so I'm certain you're l
18 anxious here to see what this committee says. l 19 So we welcome you here all the way from 20 California. And you sat many years around this table so 21 I'm certain you'll have some questions also.
22 In addition, I would expect that we have 23 members of the NRC Research and Regulatory Staff somewhere 24 in the room; I certainly welcome them, And I would like
(\
( ,) 25 to also comment that many members of this committee, even NEAL R. GROSS COURT REPORTERS AND TRANSCRIBERS 1323 RHODE ISLAND AVE., N.W.
(202) 234-4433 WASHINGTON, D C. 20005-3701 (202) 234-4433
295 1 though we're busy on many issues, have made very
,s 2 interesting comments on the report, and I have all those i
L )\
l 3 in my file here. And there are some of those in our 4 notebook.
5 And finally, I' d like to welcon.e Doug Chapin l 6 who Chaired this committee, and I'm certain as goes 7 along he'll introduce all the members who are in 8 attendance here.
1 9 Just a bit of background. This effort began, 10 I suppose around 1990 or '91, when the ACRS became 11 concerned about digital I&C and how it was being 12 implemented into the nuclear power plants.
13 And a series of letters were written which
(_) 14 ended up recommending a study by the National Academy of 15 Sciences, and that all culminated in a meeting which was 16 held in January of 1995, in which the then Chairman of 17 this committee, Tom Kress, kind of kicked off the study.
18 And he posed a number of questions and issues of which 19 I'll not repeat, but of which of course I'm looking at to 20 see if we're answered those questions and issues.
21 And he ended up his comments by stating that 22 the ACRS believes -- this is in January of 1995 -- the 23 central focus of this study should be the important safety 24 and reliability issues in the use of digital computers
( ,) 25 within nuclear power plants.
NEAL R. GROSS COURT REPORTERS AND TRANSCRIBERS 1323 RHOCE ISLAND AVE., N.W.
(202) 234-4433 WASHINGTON, D C. 20005-3701 (202) 234-4433
296 1 The committee then was given the marching l l
,s 2 orders by Chairman Kress, and they ran off and very
- l
'~
3 quickly turned around and gave us a phase 1 study report l
4 which occurred in October of '95 and which we then wrote a 5 letter to the chairman of the commission, Chairman Shirley 1
- 6. Jackson. !
1 1
7 In the phase 1 study they identified six 8 technical and two strategic issues, certainly I'll repeat 1
9 here, as they moved on into phase 2. )
1 10 So the purpose of this meeting is to have the l 11 study panel report on the phase 2 study which we received l 12 in hardcopy form in late January 1997. So that bit of 13 background and direction, I'm going to turn the meeting 4
i
\/ 14 over to Doug Chapin who will then lead us through this I
15 report, and I'm certain I'll have appropriate input from i l
16 members of this panel.
17 Doug, good luck on the -- you've had an l
18 interesting time I'm certain, and the panel's been very l l
l 19 dynamic as my understanding goes, i l
l 20 MR. CHAPIN: Well, we're delighted to be here l 21 and I think at least part of our delight is that it 22 somehow signals the end of the tunnel. And we hope very 23 much there's not a train coming in the other direction.
l 24 We'll see in our briefings today.
l
(/'~-
v
)1 25 I want to harken back to Tom Kress' remarks l 1
NEAL R. GIUDSS I COURT REPORTERS AND TRANSCRIBERS l 1323 RHODE ISLAND AVE., N W.
(202) 234-4433 WASHINGTON, D C. 20005-3701 (202) 234-4433 l
297 1 when he met with us. He was very careful in that meeting fx 2 to say that the basis for his appearance there was that he
, )
3 was the Chairman of the ACRS I think, at that time, but 4 that he was not speaking for the ACRS; that he wanted to 5 make it very clear that he was not.
6 And I would like to turn that around for the ;
l 7 moment. The committee wants me to speak in 14 voices a 8 little bit at the outset, and so I have some remarks which l
9 we have gone over amongst ourselves, and I would like to l 10 give you an introduction.
I 11 I'm very much aware of your predilection to l 12 ask questions and make comments. We'll try to deal with 13 those as we go along, but our intent is to give you some
)
N' 14 introductory remarks which to us, lay out the high points, 15 some of the peaks and valleys of the work, and then the 16 members who are here and I, will be glad to address your 17 questions and comments.
18 There are a large number of specific 19 conclusions and recommendations -- too many for us to go 20 through one by one -- so we really will take our lead from 21 you as to what topics are of special interest to you.
22 Don has given you a brief history. Our 23 history really starts in December of 1994 when it was 24 recognized that the nuclear industry and the staff had
' ,) 25 worked for several years on how to best safely introduce NEAL R. GROSS COURT REPORTERS AND TRANSCRIBERS 1323 RHODE ISLAND AVE., N.W.
(202) 2344433 WASHINGTON, D.C. 20005-3701 (202) 234 4433
298 l l
1 digital I&C technology into nuclear power plants, but they 1 3 2 had a lot of difficulty in reaching a consensus.
)
3 And so that led the NRC to request the 4 National Research Council to do this study and the 5 committee was officially formed in December of '94 and the 6 first meeting .f the committee took place in January of 7 1995.
8 The members of the committee are as listed l 9 here. I'm delighted to tell you that there are eight of l 10 us present. On my left, Larry Damon, Jim White, Dave 11 Hill, Carmelo Rodriguez, Jim Curtiss, Christine Mitchell, l
12 and John Gannon at the far end of the table. And I think l l
13 they left this empty seat here for the slings and arrows i
!" . 1 r !
\/ 14 that we may hear later.
15 I want to point out to you that the committee 16 has a very broad makeup. The people are from industry, 17 from academia, from government, national laboratories.
18 They bring a wide variety of experience and a great depth 19 of experience to this endeavor.
20 It's important to note that the way the 21 Research Council operates, they deliberately structure the 22 committee to develop what's called a balanced point of 23 view. And specifically, we had sessions at the start of l 24 both phases of the study where the committee sits down and
- /N
,) 25 tells each other what our backgrounds are and what our l
NEAL R. GROSS COURT REPORTERS AND TRANSCRIBERS 1323 RHODE ISLAND AVE., N W.
(202) 234-4433 WASHINGTON, D C. 20005-3701 (202) 234-4433
f.
299 1 biases and special points of view are.
.s 2 And these sessions are led by a facilitator I
I 3 from the National Research Council and they're intended to -
4 make sure that the committee agrees and the Research 5 Council is also comfortable, with the fact th'at there is a 6 balance in tue committee and that all members understand i 7 the bases for the opinions and the like that are being 8 expressed.
9 Given all of that, I can assure you that these 10 discussions we had were -- let me say vigorous. There 11 were a lot of pointed exchanges. Members presented views; 12 they were attacked by some members of the committee and i 13 they were buttressed by others. l
/ \ i
- > \
\~/ 14 So we feel like we've had a long, arduous l 15 journey to get where we are, to develop the consensus, and 16 that's what we want to talk to you about a little bit this i 17 morning, is the results of that consensus as reflected in 18 the final report.
19 The report is complete; it has been delivered 20 to the staff and released to the public. There are a 21 couple of key members from the National Research Council 22 who are here. They're in the row behind you. Tracy 23 Wilson is on the left, Jim Zeketto, and then Dave Devmani 24 is here as well.
/~x
( ,) 25 Now, let me go back and make some specific NEAL R. GROSS COURT REPORTERS AND TRANSCRIBERS 1323 RHODE ISLAND AVE., N.W.
(202) 234-4433 WASHINGTON D.C. 20005-3701 (202) 234-4433
300 1 comments about what Don said. We really did do it in two 1
,.- 2 phases. The report that you have covers both phases, so i i
3 i i
3 l
you don't need to go back to the phase 1 report. !
4 In phase 1, the committee was charged to l
5 define the important safety and reliability issues, and l
6 those concerned hardware, software, and human machine !
7 interfaces that arise from the introduction of digital l l
8 instrumentation and control technology in nuclear power l I
9 plant operations. And of course those operations include 10 normal, transient, and accident conditions.
11 And in response to that enarge, the phase 1 l
12 charge, the committee identified eight issues associated 13 with the use of digital I&C. Now, we don't make any claim
(~'S l
's.) 14 that these are the only issues or the topics of debate or l l
15 concern, but the committee thinks that these are key j 16 issues, that they are important issues, and if we can l l
17 develop a consensus on these issues this will be major 18 step forward and will accelerate the use of digital I&C in 19 nuclear power plants.
20 From our point of view, the issues separate 21 into six technical issues and two strategic issues. The 22 six technical issues are listed here as well as the 23 strategic issues, and we'll come back to those issues and 24 talk about specific conclusions and recommendations later.
,m
( ) 25 Now, the phase 2 charge is a more detailed NEAL R. GROSS COURT REPORTERS AND TRANSCRIBERS 1323 RHODE ISLAND AVE., N.W.
(202) 2344433 WASHINGTON, D C. 20005-3701 (202) 234-4433
301 1 charge, and it's longer and so I've got it on two 7- 2 viewgraphs. Let me see if I can get this up here so j
! l 3 they're all there.
4 In phase 2 -- and I think it's important that 5 we gc chrough this; we make sure we are together on what 6 we were trying to do -- we were charged to identify 7 criteria for review and acceptance of digital 8 instrumentation and control technology in both retrofitted 9 and new reactors of advanced design.
10 We were to characterize and evaluate 11 alternatives approaches to che certification or licensing 12 of the technology.
13 And where there was scientific basis, we were
(.
\#- 14 to recommend guidelines on the basis of which the USNRC 15 can regulate and certify or license, digital information 16 and control technology, including means for identifying 17 and addressing new issues that may result from the future 18 development of the technology.
19 And then where insufficient scientific basis 20 exists to make such recommendations, the committee was to 21 suggest ways in which the USNRC could go forward to get 22 the required information.
23 That's a lot to say; ladies and gentlemen, 24 it's a lot more to do. It's a very large charge and it's r~N
( s)
~
25 simply not practical for us to do everything that's in 1
NEAL R. GROSS l COURT REPORTERS AND TRANSCRIBERS 1 1323 RHODE ISLAND AVE., N.W. I (202) 234-4433 WASHINGTON, D C. 20005-3701 (202) 234-4433 l l
l
302 l 1 this charge in the entire area of the digital I&C world. ,
, s 2 I think we talked about that when I was here last time;
{ \
t 3 stated this was an enormously broad charge.
4 So how do we do it? Well, in phase 2 the 5 committee first limited its work to those issues that we :
1 6 identified in phase 1. But despite the fact that we 1
7 limited it to those issues, we still found we had a very 8 difficult problem and it's hard to develop a consensus and 9 make recommendations and conclusions.
i 10 And this is true simply because the issues are l
11 difficult and controversial, and in fact, we picked them '
i 12 because they were challenging and they would require hard 13 effort on the part of an expert group to make some
! 8
\/ 14 progress.
15 Now, having done the work, if you look at what 16 we've done we identified two major themes associated with 17 the use of the digital instrumentation and control.
18 The two themes are, first dealing with the 19 technology itself as applied to nuclear power plants, and 20 the second one is dealing with the technology that is more 21 advanced if you will, than what is presently, widely used 22 in the industry. And importantly, the technology is 23 advancing at a rate and in a direction which is largely i
! 24 uncontrolled by the nuclear industry.
() 25 But at the same time, that technology has a NEAL R. GROSS COURT REPORTERS AND TRANSCRIBERS 1323 RHODE ISLAND AVE., N W.
(202) 234-4433 WASHINGTON, D.C. 20005-3701 (207) 234 4433
303 1 significant impact on the operation and regulation of the
-,xi 2 nuclear industry.
~
3 So the technical issues that we've dealt with 4 are primarily related to the digital technology and the 5 strategic issues are primarily related to adopting this 6 technology to the nuclear industry. And although that's 7 not a perfect way to sort things -- because the 8 interaction between the issues is so important --
9 nevertheless, it's a useful way to help us thinking about 10 it.
11 Now, in working on the study and seeking 12 consensus on the issues, we very rapidly discovered a 13 major impediment, and that's the communication barrier.
/ \
- 14 MEMBER APOSTOLAKIS: Excuse me, before you 15 take that off. The technology is more advanced than that 16 what you used in nuclear power plants. There is a very 17 large number of IEEE standards, and a smaller number of 18 IEC standards out there regarding digital I&C. Does this 19 sentence here mean that those standards would be advanced 20 enough for us to use? Says they are dealing with 21 technology that is more advanced than the nuclear power 22 plants.
23 MR. CHAPIN: We deal with the standards and 24 the issue of the standards at another point.
t
(_) 25 MEMBER APOSTOLAKIS: Okay, so we'll come to NEAL R. GROSS COURT REPORTERS AND TRANSCRIBERS 1323 RHODE ISLAND AVE., N W.
(202) 234-4433 WASHINGTON, D C. 20005-3701 (202) 234-4433
304 1 this?
p_. 2 MR. CHAPIN: I'd like to come back to that.
3 MEMBER APOSTOLAKIS: Okay, I'm sorry. We'll 4 wait.
5 MR. CHAPIN: The thrust of this is I think, 6 that the digital technology is a state-of-the-art
, 7 technology and it's moving very quickly. And as we 8 discussed I think, the last time that I was here, the 9 power plant business is one which typically has not moved 10 very rapidly and in fact, has good reasons to be very 11 conservative in its choices.
12 And so I think that what this sentence 13 addresses is simply the sort of timeline dichotomy, that
(_) 14 there is this technology which is moving to more advanced 15 technologies to more complex technologies, and that's 16 different than what's in the plants now.
17 The problem that we discovered very early on 18 was the communication barrier that exists among the key 19 technical communities and the individuals involved.
20 We thought we had a committee, most of whom 21 are active participants in one or more of the technical 22 areas associated with digital I&C. They have a wealth of 23 experience and expertise, and we could sit down and just 24 talk with each other in a very straightforward way.
r
( ,y) 25 Nevertheless, we discovered that it took us an NEAL R. GROSS COURT REPORTERS AND TRANSCRIBERS 1323 RHODE ISt.AND AVE., N W.
(202) 234-4433 WASHINGTON. D.C. 20005-3701 (202) 234-4433
305 i
1 extraordinary effort to first develop a common language
~, 2 that we could talk in, and then even more important, to 3 reach a common understanding among ourselves, and finally, l
4 to reach a consensus on what we were trying to do.
5 And further, it became very clear from our 6 interaction with the people who came and talked to the i
7 committee, and when we went back into our home communities 8 and talked to them, that this communication problem was .
l i
9 not unique to the committee. It's really embedded and i i
i 10 it's widespread. ;
11 The reason for the difficulty is pretty 12 apparent and that is, that there are these different i
13 technical communities, that there's a lot work which is -
14 s- simultaneously going on at a rapid rate in a lot of 15 different areas, and each one tends to use its own 16 technology, its own research agenda, its own focus.
17 But unfortunately, they use a lot of common i
18 terms which are of completely different meanings in the j 19 different communities. And let me give you three examples 20 that we struggled mightily with when we started talking.
21 One is the simple term, component. I thought
.. I 22 I knew what a component was when I sat down in the room, 23 but my definition of component was very different of that l 24 which many people in the software industry considered.
m
~_-
j 25 Diversity was another one. There were those of us who had NEAL R. GROSS COURT REPORTERS AND TRANSCRIBERS 1323 RHODE ISLAND AVE., N W.
(202) 234-4433 WASHINGTCN, D C. 20005-3701 (202) 234-4433
306 1 been operating in the nuclear community and we thought we l l
73 2 knew what diversity meant. Very different to software or l l ) '
3 the digital computer design people. i 4 Configuration management was another one. I )
i G thought I knew what configuration management was based on l
- ~' my own experience.
5 Again, the software, computer design 1
1 7 people have a different definition of configuration 8 management.
l l
9 So this problem is really significant for the l l
10 nuclear power industry and the regulators because they're ]
l 11 not dominant in this technology. They're not going to set l 12 the rules and they have to synthesize the information and i 13 experience that they get from a variety of sources and
.f~h i
N_/ 14 apply it in the plants in an area where safety hazards are ;
1 l
15 important, and it has to be done under rigorous public ,
i 16 scrutiny.
17 So to us it's worth talking about and l l
18 identifying actions which can be taken to address this l l
19 particular issue. So I want to go through those with you l l
20 and they're summarized on this slide.
21 We think there are five areas of need to 22 address the communication problem. The first and foremost l
23 is we need better, clearer, crisper statements of the 24 regulatory concern and the appropriate acceptance criteria
,a
!, j) 25 that are valid at any point in time.
NEAL R. GROSS COURT REPO 1TERS AND TRANSCRIBERS l
1323 RHODE ISLAND AVE., N W.
(202) 234-4433 WASHINGTON, D.C. 20005-3701 (202) 234-4433
307 1 We prefer -- the committee prefers focused, 73 2 succinct statements of regulatory problems, criterias, and t
'% ~ //
3 standards. And this is to be contrasted with what we i
4 consider the sort of current U.S. NRC approach which is 5 characterized by relatively complex statements of the l
6 requirements and often created by interconnected 7 endorsements of -- and caveats of a family of standards or i l
8 related documents.
9 Now we understand that the U.S. NRC's staff 10 particularly has taken this path because it's effective in 11 terms of time and effort. But we're concerned that the 12 gain and efficiency may be offset by a loss in clarity in 13 defining the acceptance criteria, p
i k 'i 14 VICE CHAIRMAN POWERS: Isn't there a legal 15 concern as well?
16 MR. CHAPIN: I suspect there is.
17 VICE CHAIRMAN POWERS: One is Public Law that 18 says that they should use consensus standards whenever
- they're applicable.
20 MR. CHAPIN: Yes, I think you can use 21 consensus standards. I think -- our opinion is that maybe 22 you can use the consensus standard, but you can boil 23 things down so that they're more succinct and easily 24 worked with.
(^N,
() 25 MEMBER APOSTOLAKIS: Is this the time where NEAL R. GROSS COURT REPORTERS AND TRANSCRIBERS 1323 RHODE ISLAND AVE., N.W.
(202) 234-4433 WASHINGTON, D.C. 20005-3701 (202) 2344433
308 1 we're discussing standards?
,- 2 MR. CHAPIN: I know you guys don't want me to, w/
3 but I would like to get through in about 20 minutes, and 4 then I'll turn you loose. Okay, so let me come back and -
5 - the standards are in here, and I'll be glad to discuss 6 those, but let me get through this.
7 MEMBER APOSTOLAKIS: So you want us to be 8 quiet until the end of the presentation?
9 MR. CHAPIN: I would like for you to be quiet.
10- VICE CHAIRMAN POWERS: We're not good at that.
11 MEMBER APOSTOLAKIS: That will be a major 12 achievement.,,.
,_ 13 MEMBER MILLER: I'll get my gavel out here.
I )
l 14 MR. CHAPIN: On the other hand, we understand 15 that it's r,.L r.coei).1.e for the criteria for the regulation 16 to always L_ r;nen -it cive and objective. And in today's 17 complex systems, that's clearly not feasible. And a good 18 example is human reliability.
l l
19 And we recognize and endorse the concept that l l
l 20 one must certainly do the best he can with a mix of 21 objectivity and expert judgment and a finite budget. So 22 we're not looking for everything to be written down and 23 say go do it in accordance with this number.
24 The second important area that we see is that
(~) 25 l
( ,) we need more proactive participation by the nuclear power NEAL R. GROSS COURT REPORTERS AND TRANSCRIBERS 1323 RHODE ISLAND AVE., N W.
(202) 234-4433 WASHINGTON, D C. 20005-3701 (202) 234-4433
309 1 industry and the U.S. NRC in the relevant technical f 2 communities. And that involvement is needed to assure (3)' ~ '
3 that the issues and constraints which are unique to the 4 nuclear power industry are recognized and addressed in 5 this larger community.
6 And active involvement also will help the 7 nuclear power community by providing increased access to 8 the closely related non. nuclear fields like software 9 engineering.
10 We think the third important point is to 11 strengthen the technical infrastructure in the nuclear 12 power industry and the U.S. NRC in digital systems. And 13 we think in that case, it's especially necessary to work o \
(. / 14 cooperatively and creatively because we have to husband 15 and multiply the available resources.
16 And we think that means working together, and 17 it means carefully selecting the topics to be pursued.
18 The committee recognizes that the regulator has to be 19 independent, but we see that it's possible to work in this 20 cooperative way and still maintain their independence.
21 The committee has seen some progress in this 22 regard even in the course of our study where the 23 regulators have been involved in reviewing and assessing 24 industry research programs and guidelines and development (3
i s ) 25 efforts and in new training programs for the staff. And NEAL R. GROSS COURT REPORTERS AND TRANSCRIBERS 1323 RHODE ISLAND AVE., N.W.
(202) 234-4433 WASHINGTON, D C. 200C5-3701 (202) 234-4433
310 l
1 we commend those efforts, and we urge that they be
'n \ 2 expanded to make best use of the limited funds.
3 We think that you need in your documentation 4 to formally address the communication problem. Now that 5 means make clear definitions of terms and context, and we 6 also suggest that increased informal communication between 7 the staff and the industry in those areas where there is 8 uncertainty or there's lack of clear regulatory guidance.
9 And finally, to address this major impediment 10 of communications, we think that one needs to modify the 11 regulatory mechanisms that are employed when an advanced 12 technology like digital I&C has temporarily outpaced the 13 regalations.
/ \
- }
'V 14 And such a mechanism that we see is something 15 like 10 CFR 50.59. And we believe that it's a l
l 16 fundamentally sound process and it should continue to be l
l 17 used. But as we discuss in the part of the report that l l
18 talks about case by case licensing, we think there are a 19 number of changes that could be made to the regulatory l 20 process, and that they would make -- help make the intent ,
1 l
21 on the basis of the regulatory decisions fully ,
I i
22 communicated.
23 Now consistent with all this background, we 24 took seriously the charge that we identify criteria for l m
(u ,) 25 review and acceptance of the digital I&C technology and NEAL R. GROSS COURT REPORTERS AND TRANSCRIBERS 1323 RHODE ISLAND AVE., N W.
(202) 234-4433 WASHINGTON, D.C. 20005-3701 (202) 2344433
) 311 1 recommend guidelines for regalation certification.
s 2 However, in carrying out the charge, we recognize there's I
3 some limitations.
4 We clearly have only a limited number of 5 issues that we could deal with. We worked for about a 6 year and a half, and we had to liuit the scope. Second, 7 we clearly recognize that general high level criteria were 8 not particularly useful. There's an abundance of 9 instructions like write good software and design I&C 10 systems carefully and make sure that they are V&V'd with
- 11 great caro.
12 We don't think that's particularly useful. We 13 also recognize that the final criteria are legally the
/~N 4 N/ 14 N.tC's responsibility. And important1 f, those licensing 15 criteria need to be hammered out in a detailed interaction l 16 between the regulators, the industry, and the public.
17 We had a wide range of expertise and l l
18 experience in this committee, but we're not a surrogate l l
19 for that interaction among the stakeholders. And so we 20 think the committee could best serve by focusing on l 21 delineating and defining the issues and then providing l
22 guidance to resolve those rather than working on specific !
23 licensing criteria. l 24 Further, based on extensive review and I f3 (s-) 25 interaction with all the other interested communities and l NEAL R. GROSS COURT REPORTERS AND TRANSCRIBERS 1323 RHODE ISLAND AVE., N W.
(202) 2344433 WASHINGTON D.C. 20005-3701 (202) 2344 433
312 1 a number cf agencies who are active in regulating this 3 2 thing, there are simply no silver bullets out there. This N.,]
~
3 is not something where if went on a broader search or we 4 go look, that we can pick, as some of the committee 5 members have said, low lying fruit and come back and say 6 aha, here is the way to solve this problem.
7 So as a result, we have not provided a cogent 8 set of principles or design guidelines or specific 9 criteria which can be readily used by the NRC to assess, 10 test, license, certify those kinds of activities. Rather, 11 what we've presented are recommendations and conclusions 12 that are related to each issue and are primarily addressed 13 to the NRC for their consideration, r\
14 The recommendations and conclusi$ns that we 15 have fall into four broad categories. And we emphasize j 16 that we think there's a lot of work by the U.S. NRC and 17 the nuclear industry to extend the work that we've done to 18 the stage of doing this detailed criteria development.
19 And what we have offered are four broad 20 categories of recommendations and conclusions. One is, 21 there's some places where we consider that the current 22 practice of the NRC and the industry is essentially 23 satisfactory and doesn't require any major work or maybe 24 some fine tuning.
() 25 We think there's some points of weakness in NEAL R. GROSS COURT REPORTERS AND TRANSCRIBERS 1323 RHODE ISLAND AVE., N.W.
(202) 234 4433 WASHINGTON, D C. 20005 3701 (202) 2344 433
313 1 the U.S. NRC approach, and we've tried to point those out.
rw 2 We think we've identified some issues in areas that merit f i V
3 further inquiry and research before satisfactory 4 regulation can be developed. And we've identified some 5 places where we think it is unreasonable to expect 6 criteria and guidelines in the near future.
7 Now turning to the high level issues more 8 specifically related to the digital technology, the 9 committee has some high level conclusions that we want to 10 emphasize. The first one is that we consider the 11 deterministic assessment methodologies including design 12 basis acciden;, hazard analysis, and the other formal
,s 13 analysis procedures all applicable to digital systems as D 14 long as the.y' re applied with care.
15 So we think the basic hazard analysis kind of 16 approach is appropriate. The second thing is, there is 17 considerable controversy within the software community as 18 to whether you can make an accurate failure probability 19 for software or even whether software fails randomly. We 20 had some extensive arguments, vigorous exchanges on that 21 topic.
22 However, the consensus of the committee agreed 23 that a software failure probability can be used for the 24 purpose of performing probablistic risk assessment in
/ \
( ,) 25 order to determine the relative influence of digital NEAL R. GROSS COURT REPORTERS AND TRANSCRIBERS 1323 RHODE ISLAND AVE., N W.
(202) 234-4433 WASHINGTON, D.C. 20005-3701 (202) 2344 433
314 1 system failure on the overall system.
.. 2 MEMBER KRESS: Are you going to expand on
\
s 't 3 that?
4 MR. CHAPIN: Yes, we'd be glad to come back 5 and talk about that.
6 We think explicitly including the software 7 failures in the PRA for a nuclear plant is preferable to 8 ignoring the software failures.
9 And let me come back to that, Tom.
10 I know this is one of the things that you 11 specifically talked to us about and we set out -- and I 12 know this is a topic of interest.
13 Another major high level issue is we think y.
\s.)) 14 it's important that you not address digital systems in !
l 15 terms of hardware or software; that we think hardware and {
16 software must be treated together as a system; and that if 17 you focus on one or the other, you should do so with great 18 caution.
19 And let me point out two examples in the 20 report. First, we have a chapter that's called Common 21 Mode Software Errors. But if you look at that chapter, it 22 leads far beyond the boundaries of the software itself; 23 and our belief is the successful resolution of that :
24 problem emphasizes the treatment of the systems as a
( 1
(
ss
) 25 whole.
NEAL R. GROSS COURT REPORTERS AND TRANSCRIBERS 1323 RHODE ISLAND AVE., N W.
(202) 234-4433 WASHINGTON, D C. 20005 3701 (202) 234-4433
315 1 And the second example is the t.reatment of 7 2 complexity. It's important to assure that system
)
'd' 3 complexity is addressed. And for example, digital system l
4 complexity issues are not resolved by greatly simplifying 5 the software while dramatically introducing more complex 6 hardware, and maybe the hardware will be less testable, l 7 and vice versa.
8 So we think you have to not push the 9 complexity from one place to another and hope that you've 10 solved it.
11 And finally, we came to this consensus that 12 most practical digital I&C systems cannot be exhaustively 13 tested. And therefore, we just can't show them to be free
/,
I ')
(/ 14 from any and all errors. However, the committee is 15 convinced that adequate approaches exist and can be 16 applied within practical resource constraints to support 17 the use of digital systems in safety critical applications 18 in nuclear power plants.
19 DR. LEWIS: I wonder -- Bob, may I ask a 20 question?
21 CHAIRMAN SEALE: Certainly.
22 DR. LEWIS: It may be premature. You may 23 cover it later, but you will tell me if that's true. But 24 your mention of mixing the hardware and software issues n
25 together, with which I agree, raises to me a question of (a)
NEAL R. GROSS COURT REPORTERS AND TRANSCRIBERS 1323 RHODE ISLAND AVE., N.W.
l (202) 234-4433 WASHINGTON, D.C. 20005-3701 (202) 2344433 l
l
316 l
1 time scales. In this whole business, there are many time I; l
,7. . 2 scales involved, and you started the whole discussion by l f )
\
3 saying digital technology is rapidly changing.
4 Everyone says that. And what most people mean 5 by it is that you have to buy a new personal computer 6 every three years because the old one is obsolete. The 7 reason for that is usually hardware. It's not software.
8 And it's the hardware that has been changing rapidly over 9 the last 20 or 30 years, and the -- most of us are still 10 writing programs in C.
11 C was invented 30 years ago or something like 12 that. The general developments of formal processes and 13 other means of verifying software relianility change at r^s ,
i
)
/ 14 what I would call an academic time scale which is rather 15 different from an industrial time scale. And there is a 16 time scale for the NRC to develop regulations, there's a 17 time scale for them to learn to use regulations.
18 I remember when the communists first took over 19 in the Soviet Union and promised that women would be given 20 equality. Somebody said yes, after the last generation of 21 males dies. You know, there are some things that have 22 that kind of time scale.
23 So mixing them all together is -- the ways in 24 which you do that in a practical matter is fairly
} (~T 25 important knowing what the time scales are. So when you -
( ,i NEAL R. GROSS COURT REPORTERS AND TRANSCRIBERS 1323 RHODE ISLAND AVE., N.W.
(202) 2344 433 WASHINGTON, D C. 20005-3701 (202) 2344433
317 1 - you trigger this saying something about hardware and
,- 2 software.
' ~ '
3 In a sense, the scale of NRC regulations makes 1
4 it possible to regulate things that are dependent on 5 software changes or that adapt to software changes, but 6 not to hardware changes because that will happen -- you 7 know, they'll be obsolete by the time they -- as the 8 hardware changes rapidly.
9 Are these issues of time scale integration 10 covered anywhere in your report? I didn't see them.
11 MR. CHAPIN: I don't think they're covered 12 explicitly, Dr. Lewis. I think we are very aware of the 13 time scale effects. For example, in the section on COTS,
)
x/ 14 the use of commercial off the shelf equipment, that is 15 largely driven by the reality of what's happening outside 16 the nuclear industry.
17 And when people have equipment that becomes 18 obsolete or the like, they have to go and what comes back 19 is something different than what they had. So I think 20 we're aware of that. I think what you said is very true.
21 We haven't covered that explicitly.
22 I think the point that we are emphasizing here 23 is that we think there is also a tendency in the industry 24 for some to say I have a problem in these systems and
( ,) 25 perhaps it's a software problem. And so let me deal with NEAL R. GROSS COURT REPORTERS AND TRANSCRIBERS 1323 RHODE ISLAND AVE., N.W.
(202) 234-4433 WASHINGTON, D.C. 20005-3701 (202) 234-4433
_ _ -. _ _ -. ._. .._._..._ . _ ._ _ ..- _ ...m
. . ... .. ..... _ ,._ ,m ..___m 318 p 1 the software problem by making'the software problem go' l
l ^2 away by working.in the hardware. t a i 3 And that doesn't make the problem go away, is .
f 4 what.our message is. It's that you-need-to make sure that
- 's if'you apply a software solution to a hardware problem or' 6 vice versa, that you face the fact that you still have t ,
7 this . issue.of complexity in really these practical real i
8 time type systems.
9 'DR. LEWIS: True. But in the past, there have 10 sometimes been software developments that' people have ,
11 ignored because there was no hardware that would support g 12 them; and yet, one could be reasonably confident that in l
13 three years there would be. So it is relevant if you're i O
14 trying to develop criteria for computing system 15 reliability.
1 16' I don't want to -- you don't have enough time 17 for this.
l 18 MEMBER MILLER: We have time later.
l 19 DR. LEWIS: I doubt it.
i 20 CHAIRMAN SEALE: Perhaps, !
21 MR. CHAPIN: Let me conclude by going through )
l
'22 some summary conclusions with you, and they're summarized 23 on this slide.
24 In summary, the committee notes that the (N
V 25 digital I&C is state of the art technology. It's widely NEAL R. GROSS COURT REPORTERS AND TRANSCRIBERS 1323 RHODE ISLAND AVE., N.W. l (202) 234-4433 WASHINGTON, D.C. 20005-3701 (202) 234-4433 )
319 1 used inside and outside of the nuclear industry. It
,x 2 offers powerful capabilities that can affect nuclear power
)
3 plant safety. And therefore, those systems should be 4 treated carefully, particularly in the safety critical 5 applications.
6 It appears to us that the U.S. NRC and the 7 nuclear power industry are moving forward with the 8 procedures, processes, and the technical infrastructure 9 needed to assure continued safe operation of the plants.
10 We have suggested several improvements.
11 And given this overall situation, the 12 committen considers that the use of digital I&C systems in 13 new nuclear power plants and in modifications and upgrades
'ws/ 14 of existing plants is appropriate and desirable. For 15 existing plants, we think it's particularly true where the 16 digital systems replace these older systems and equipment 17 for which vendor support is no longer readily available.
18 Now as I told you before, there were a lot of 19 conclusions and recommendations in the report. There were 20 a total of 44 conclusions and 39 recommendations by my 21 count. They're scattered over the eight issues. And 22 because of the large number, we don't intend to put each 23 one of them up and go through them like we did with the 24 issues.
() 25 And we'd like for you to question us about NEAL R. GROSS COURT REPORTERS AND TRANSCRIBERS 1323 RHODE ISLAND AVE., N.W.
(202) 234-4433 WASHINGTON, D C. 20005-3701 (202) 234 4433
320 1 those, and we'll try to respond. In order to help you do i
,s s
2 that,' let me explain very quickly how we did the study. I
\ '~' /
3 In phase one, the committee essentially acted as one 4 committee. We met together, we worked primarily as a l 5 group to develop and define the issues.
6 For phase two, once we had to work on the
)
7 detailed issues in order to be efficient, we broke into a l
8 number of parallel working groups, one for each of the )
l 9 issues identified. And the members of those working l
10 groups and the lead individual in that group took the 1
11 responsibility for identifying people that we should talk 12 to, identifying references in documents that we should l 13 review, and/or bring people in to particular in roundtable
,x
( I i/ 14 discussions.
15 They also took the lead in preparing the 16 initial draft for the papers that eventually became the 17 final report. But the whole committee spent considerable 18 time and effort reviewing and diccussing each issue and 19 agreeing on its associated chapter content.
20 so the document that you see before you is a 21 consensus document. It doesn't reflect the breadth of 22 opinion that may have existed on any given topic. We're 23 certainly grateful to everybody who interacted with us.
24 The staff came several times and met with us, and we very
() 25 much appreciate their input.
NEAL R. GROSS COURT REPORTERS AND TRANSCRIBERS 1323 RHODE ISLAND AVE., N.W.
(202) 234-4433 WASHINGTON, D C. 20005-3701 (202) 234-4433
. - . . - - - - . - . . - _ - . . _ . . _ . . . ~ - . - . _ . - . - - . . . . . . - _ . . _ _ ~ ..
I 321
- 1 I want to publicly thank these people who are
- -2 here and the members of'the committee who are not. 'They V 3; worked very hard. We had a difficult charge. We were
4 pushed?along on a very challenging schedule. And I think
~
- 5 they've.done a good job, and I would like to thank them i .6 publicly and commend them for a job well done.
! . a 7 I'd also like to particularly commend Tracy-i 8- Wilson who's just been a pillar in the National Research l
[ 9- Council staff who worked to support us. And I've already 3 t
- j. 10 introduced the folks who are here. And'at this point, I'm '
4 l 11 willing'to unleash the dogs and you guys can help out. ;
h 12 MEMBER MILLER: You don't know what you've
]- 13 said there.
W.
i 14 ' (Laughter . ) - "
i
- j. 15 CHAIRMAN SEALE: We have a junkyard dog.
}
q , 16 MEMBER MILLER: Yeah, we've actually debated i
17 the work unleashing'around here.
l MR. CHAPIN:
.18 I've been here before.
19 MEMBER MILLER: So I'm certain there is many i J
20 questions, and I will unleash the -- my fellow committee ,
21 members to begin asking questions. I know George to my 22 right is -- I'm trying to hold him back just for a few 23 more microseconds even -- to go ahead with questions, and 24 I know I have my questions. So I will take priority to {
25 the chair once in a while and try to get mine in edgewise.
NEAL R. GROSS COURT REPORTERS AND TRANSCRIBERS 1323 RHODE ISLAND AVE., N.W.
(202) 234-4433 WASHINGTON D.C. 20005-3701 (202) 234-4433 1
l 322 1 With that, I'll turn it to George, and I know l
i 73 2 he had a few questions and comments about certain areas of 3 the report.
4 MEMBER APOSTOLAKIS: Okay, thank you.
5 Yeah, there's a number of questions that I 6 would like to ask, but I will start with one that I think 7 is very important and very current. This committee had to 8 review the revision -- actually, it's an addition to the 9 Standard Review Plan to include digital I&C, right, 10 Chapter 7, or was it a revision?
11 MEMBER MILLER: Yes, I wanted to -- did want 12 to make a point of clarification. What part of the 13 Standard Review Plan did you have the opportunity to I l id 14 review?
15 MR. CHAPIN: We had access to the revision. I 16 would like to tell you that we did not consider that our 17 charter was to review the Standard Review Plan revision 18 and to generate a set of comments. So we were aware of, 19 we took it into account, but we thought we were asked to 20 do a study of the general topic rather than review the 21 thing.
22 But we did have access to the revised --
23 MEMBER MILLER: Did that include the branch 24 technical positions?
(x
(_) 25 MR. CHAPIN: Yes.
NEAL R. GROSS COURT REPORTERS AND TRANSCRIBERS 1323 RHODE ISLAND AVE., N W.
(202) 234-4433 WASHINGTON, D.C. 20005-3701 (202) 234-4433
323 1 MEMBER APOSTOLAKIS: Now one -- that document, 2 along with the branch technical position 14 which was more 7S
( !
3 specific, placed extreme emphasis on controlling the 4 process of production of the software with some testing --
5 maybe a lot of testing, And the major assumption there 6 was that if you control the process, you will get a very 7 reliable product.
8 And I'm having a lot of problems with that 9 idea that this is in fact true. Now, I searched a little 10 bit and I found that another industry -- in manufacturing, 11 for example -- they do place a lot of emphasis on 12 controlling the process and they do get a very reliable 13 product.
.,_ 't
\- 14 But at the same time, there are people there 15 who have spent careers identifying the failure modes of 16 whatever the process is producing so that the control of 17 the process specifically addresses chese failures modes.
18 And I'm not sure we have that in our case here.
19 Now, there is a sentence in your report that 20 says neither controlling the process nor testing the 21 product by themselves will guarantee high reliability. So 22 -- and the other thing that is related to the question of 23 the standards is that our documents cite many IEEE 24 standards, which themselves cite other IEEE standards, 10
(,,) 25 which themselves cite other IEEE standards.
NEAL R. GROSS COURT REPORTERS AND TRANSCRIBERS 1323 RHODE ISLAND AVE., N.W.
(202) 234-4433 WASHINGTON, D.C. 20COS-3701 (202) 234-4433
324 I
1 And in my opinion, the very last standard says
,_ 2 do a good job.
(3) 3 MEMBER MILLER: George, --
4 MEMBER APOSTOLAKIS: That's my opinion.
5 MEMBER MILLER: Yeah, I know.
6 MEMBER APOSTOLAKIS: I said my opinion.
7 So I would like to know -- well, first of all, 8 I don't think that your report, although I may be wrong, I l
9 placed -- made a lot of comments on this issue of i
10 controlling the process. But I would like to know the 11 opinion of the ccmmittee or the members that are present 12 as to whether this is a wise approach.
13 And I fully applaud your recommendation that
?a
'is
14 the U.S. NRC -- see, now we have to put the U.S. in front 15 so we don't have any confusion -- that the U.S. NRC 16 develop its own standards and its own -- define its own 17 terminology and so on so we will not use the IEEE standard 18 so much.
19 But I would like to know what the opinion of 20 this committee is on two issues. First, is it reasonable 21 to expect a highly reliable product if we control the 22 process of producing that product? And second, what is 23 your opinion of the IEEE standards?
24 MR. CHAPIN: Okay, let me give you a -- sort A
( ,) 25 of the Chairman's response, and then I think there are a NEAL R. GROSS COURT REPORTERS AND TRANSCRIBERS 1323 RHODE ISLAND AVE., N.W.
(202) 234-4433 WASHINGTON, D.C. 20005 3701 (202) 234 4433
325 1 couple of people here that we can call on who are more
- 2 expert in one area maybe with regard to the standards and
\
' 'l 3 in another area with regard to the software.
4 There is a chapter in the report on software 5 quality assurance. And I think what we are coming to 6 there is that you do -- you have to do a number of things 7 to do a good job on software. And that these various 8 approaches that people take of process control and 9 requirements review or testing all have roles.
10 And in particular, I will cite conclusion five 11 in that chapter where we say exposing software flaws, 12 demonstrating reliable behavior of the software, finding 13 unintended functionality, and flaws in requirements are pm
,,-) 14 all different concepts and they have to be addressed by a 15 combination of techniques.
16 And then we say systematic inspections and 17 planned testing with inputs from different parts can help 18 determine if there are flaws in the software. Functional 19 tests can be used to look for normal or boundary cases, 20 and you can look for test coverage with that kind of -- if 21 you do tests based on large numbers of random inputs, then 22 you can assess the likelihood that the software will fail 23 under specific operating conditions.
24 Requirements inspections can be effective if n
25 the people who are doing it have the kind of domain (v)
NEAL R. GROSS COURT REPORTERS AND TRANSCRIBERS 1323 RHODE ISLAND AVE., N.W.
(202) 234-4433 WASHINGTON, D C. 20005-3701 (202) 234-4433
326 1 experience that you're talking about. For example, we
,- . 2 don't think doing -- trying to do -- look for hazards by i )
~ ~
3 doing random testing is very effective. It's the needle 4 in the haystack approach, and you really need the domain 5 experience to say gee, these are areas that are important 6 and I need to look at.
7 And the system level hazard analysis is 8 important because it can tell us which states lead to 9 accidents and which environmental conditions are 10 important; and therefore, we ought to deal with those.
11 So in general, I think what we're saying is 12 you need to do both, and there's an appropriate 13 combination, and there is some intelligence and experience
\w 'i 14 which exists in the industry which can be used to guide 15 those selections.
16 Now let me take the software one particularly.
17 And John Gannon is here and is a representative with us --
18 has done a lot of work in the software.
19 And would you like to comment?
20 MR. CHAPIN: I'll try and say two things about 21 that. One about the software process. There are people 22 who would advocate a pure process approach. The Software 23 Engineering Institute at Carnegie Mellon University has 24 certainly advocated that approach.
,( ,) 25 And certain parts of IBM in the clean room NEAL R. GROSS COURT REPORTERS AND TRANSCRIBERS 1323 RHODE ISLAND AVE., N.W.
(202) 234-4433 WASHINGTON, D.C. 20005-3701 (202) 234-4433
327 1 approach have advocated pursuing a more process oriented s., 2 approach. There's very little empirical evidence that 3 supports that process only approach. We'll allow you to 4 state anything about the resulting product.
5 bo we wanted to be -- we wanted to make sure 6 that that caution was in the report. About the IEEE 7 standards, there was an article in communications to the 8 ACM last year which characterized them as ad hoc and 9 unintegrated and developed in a piecemeal fashion.
10 MEMBER APOSTOLAKIS: Can you give me the 11 citation?
12 MR. GANNON: Yeah, it is in the report.
13 MEMBER APOSTOLAKIS: Very good.
I ')
- - ' 14 MR. GANNON: Yes.
15 MEMBER APOSTOLAKIS: I appreciate it. Thank 16 you.
17 MEMBER MILLER: Yeah, I'd like to bring up the 18 point on that issue that --
19 MR. CHAPIN: Could I --
20 MEMBER MILLER: Okay, I'm sorry.
21 MR. CHAPIN: Larry, did you want to chime in 22 about the standards or not?
I 23 MR. DAMON: Well, it depends on how far we 24 want to go. There's no question of the fact that the IEEE
,.y
) 25 standards regarding control systems applied to nuclear NEAL R. GROSS i
COURT REPORTERS AND TRANSCRIBERS 1323 RHODE ISLAND AVE., N.W.
(202) 234-4433 WASHINGTON, D C. 20005-3701 (202) 234-4433
328 1 power grew with the industry. And our digital technology 2 standards go way back when you were using TTL logic.
- ')
3 And there are a number of things that perhaps 4 could have been addressed in those standards that were 5 not. The electromagnetic interfer nce, for example, has 6 been around since day one. You have your pulse driven 7 radiation monitoring equipment more vulnerable to EMI/RFI 8 than any of your digital state change kind of electronics.
9 Those have been working for a long time.
10 They've been dealing with EMI/RFI. There are standards 11 out there in the military that have been available to 12 them. The practice has been folded in. So the standards 13 issue goes really beyond clear statements of just the
/ i
% 14 IEEE.
15 It goes to what it takes to make the things 16 work. The problem that we're facing with the digital 17 technology is that the people in the industry who grew 18 with the standards understood the standards because they 19 participated in that development over years of time.
20 And nuclear licensing is really an art form.
21 There's no course you can take that's going to really make 22 sense of it. You know, you've got to grow with it. It's 23 process of some kind of an intellectual osmosis. And for 24 us who have been in the industry for a long time, it all s ) 25 makes sense NEAL R. GROSS COURT REPORTERS AND TRANSCRIBERS 1323 RHODE ISLAND AVE., N W.
(202) 234-4433 WASHINGTON, D C. 20005-3701 (202) 234-4433
329 1 But trying to explair. it to -- that's one of 2 the things we're having trouble with the committee.
( )
3 Trying to explain that nuclear licensing process to people 4 who have not been engaged in nuclear licensing which 5 communications difficulty. Take a long time to come to 6 that.
7 As standards grew with the industry, what has l
8 happened here is that the standards grow by the committee 9 process at a relatively lethargic pace. They don't change 1
10 fast. For good reasons too. It's good that they don't, !
11 because they'd destabilize the licensing environment if 12 they did. Conversely, the use of digital technology, 13 especially with the advent of the microprocessor, went
('~h
't.) 14 faster than the committee process was -- that we have in 15 the nuclear industry was able to follow on.
16 Now, they did a good job, as best they could, 17 trying to continuously upgrade this through the various 18 versions of 7.4.3.2 and the reg. guides that are coming 19 out dealing specifically with software. The difficult is 20 though that fundamental criteria which you'd call the 21 first principles in nuclear design like 279, and later 22 IPEEE 603, stayed dormant while digital technology 23 evolved.
24 So the idea of independence criteria put
(- .\
i ,) 25 together originally by 279 got hung on the horns of a NEAL R. GROSS COURT REPORTERS AND TRANSCRIBERS 1323 RHODE ISLAND AVE., N.W.
(202) 234-4433 WASHINGTON, D.C. 20005-3701 (202) 234-4433
. . - . - - - . . - . . . ~ - - . - . - - - - - ..- .-. . . - ~ . . - . - .. .. -
, 330 p
'l dilemmaofcommonmodesofbwarefailure. Now'what's 2 happened -- and we had the discussion yesterday on this --
3 is that-the NRC' staff and everybody recognized.that the J
4 idea of not:being able to have perfect software meant that 5 there.was a potential' failure' mechanism.
}~
6 And the rules of our industry say that failure
.]
i
- 7. mechanism has to be. treated as a clear and present danger, ,
8 and you have to accommodate for it from pipe ~ break on i
- 9. down. So when ccmmon mode software failure came up and
? ;
10 there was no showing, no matter.how much QA you put into 4..
11 software, whetherjit was module and simple, it's still a
) 12 possibility -- you had to treat it as that. l
- j. 13 And that meant that software common to 14 redundant channels was a failure mechanism that couldn't j: 15 really meet 279 basic principles. The NRC recognized this 16 and at the same time recognized the advantages of digital 17 technology. So they somewhat finessed the 279 18 requirements by allowing the use of equipment diverse'from 19' the basic equipment that's being changed in the perfection 20 system to be employed to show alternate means of shutdown.
21 Now in showing alternate means of shutdown, 22 they got into where they have to also, under rule 50.62, 23 show diversity. Diversity definition then became a big 24 problem. And they had to revise the Standard Review Plan
() 25 then to accommodate two things.
NEAL R. GROSS COURT REPORTERS AND TRANSCRIBERS 1323 RHODE ISLAND AVE., N.W.
(202) 234 4433 WASHINGTON, D.C. 20005-3701 (202) 234 4433
331 l 1 One was to raise all of the various 7- 2 requirements that have grown like topsy and try to get
( )
'~#
3 them in one place so that they had a pretty clear i 4 statement of what you were going to require to bring the 5 software quality as high and as understandable as you 6 could make it.
7 And the second thing is, they had to go in to 8 allow the use of ATWS equipment as a separate shutdown 9 mechanism to more clearly define what diversity really 10 meant. And to make -- to allow this concession, 11 essentially they depended upon things like NUREG-6303 and 12 other criteria that were developed for the advanced 13 reactors and pulled them back to be applicable to K/ 14 operating units.
15 Now, in a nutshell, that is how all of this 16 came together. Are the requirements now sufficient?
17 Reviewing the Standard Review Plan here in front of me and 18 having gone through 7.4.3.2 a number of times, my feeling 19 is yes, it is. But is it clear and concise and cogent and 20 readily understandable so that a reviewer can go into a 21 utility and sit down, go through the issues and say okay, 22 this is acceptable to me and the safety evaluation?
23 That's going to still take an awful lot of I
24 work. The point is, what's an acceptable standard? Is it
()
/ \
25 one that you can employ as a complete body of requirements FULAL R. GROSS l COURT REPORTERS AND TRANSCRIBERS 1323 RHODE ISLAND AVE., N.W. l (202) 234 4433 WASHINGTON, D.C. 20005 3701 (202) 234-4433
i 332 ;
l 1 and after a little bit of work come to a resolution yeah, l s 2 we can accept this system? or is it a standard that says I
> \
'J 3 crisply and clearly as we would like to see it you do j i
4 these things and you're okay to go ahead?
5 The biggest factor in this is that that
{
6 clearness isn't there, and it's going to cost utilities an 7 awful lot of work and time to go through the staff and l
8 work these issues out. The committee can't do that. It's .
l l
9 going to have to like get into the licensing environment.
10 And whether or not the utilities will go through that 11 effort to clean up the questions is an open question.
12 MEMBER MILLER: Can I follow up on that 13 comment there? Do you believe that branch technical I
x/ 14 positions have at least made a first step in clarifying 15 some of the issues you bring up, particularly branch 16 technical position 14?
1 17 MR. GANNON: Well, 14 is going to give you a 18 lot of assurance on software. I went through these again 19 last night, and I think that 14 is the one that is not 20 well copied.
21 MEMBER MILLER: I've got one that's marked up 22 if you want to look at it.
23 MR. GANNON: Well, my problem is, in the 24 review that I did of this last night, 14 came out like
(\
) 25 this. And I didn't bring my own copy with me. I got one NEAL R. GROSS COURT REPORTERS AND TRANSCRIBERS 1323 RHODE ISLAND AVE., N.W.
(202) 234-4433 WASHINGTON, D C. 20005-3701 (202) 2344433
~
333 1 from Tracy. .
i
,e' 2 MEMBER MILLER: See, I think 14 answers a
\ )
3 number of the questions that are raised by the report -- ,
4 addresses a number of them, and also addresses the 5 questions that George has brought up a number of times.
6 MR. GANNON: Well, 14 has to do with --
7 MEMBER MILLER: George may not agree with me.
8 MR. GANNON: -- assuring the highest level of 9 software quality that you can get. Now, I'm not a 10 software engineer. I think the person who answered fully 11 how far 14 goes to giving you the most reasonable result -
12 - cost effective result, if you will, will be somebody 13 like John over here or Nancy Leveson.
'N- 14 MEMBER APOSTOLAKIS: Well, I'm having big 15 problems with 14. And I'm glad that the chairman 16 mentioned the word specificity several times and the word 17 criteria and so on. Now, I'm not a software engineer l 18 either, so maybe this guidance means something different 19 to software engineers.
l 20 But it seems to me when I see guidance that 21 says -- and I may be oversimplifying now -- have a plan 22 through V&V --
23 MEMBER MILLER: 14 says have a plan that has 24 these certain acceptance criteria in it.
,m
( ,) 25 MEMBER APOSTOLAKIS: I'm not sure it has the NEAL R. GROSS COURT REPORTERS AND TRANSCRIBERS 1323 RHODE ISLAND AVE., N W.
(202) 234-4433 WASHINGTON. D.C. 20005-3701 (202) 234-4433
334 1 words acceptance criteria. Maybe it says of these 7 elements, but --
( )
~#
3 MR. DAMON: It's a process.
4 MEMBER APOSTOLAKIS: It's a process, but it 5 doesn't really give me any advice as to when the process 6 has been implemented satisfactorily. That's my 7 impression. The existence of the process seems to be the 8 goal of the process. And I'm having a hell of a problem 9 with that. I'm willing to be convinced otherwise; but 10 right now, I'm having a problem.
11 MR. CHAPIN: Yeah, I think in places where 12 there isn't a process and people are proceeding and 13 writing software in, you know, sort of a hit or miss
,7 m
's_ ! 14 unorganized way and you're writing software, it's hard to 15 not do that. There is an important requirement that says 16 you ought to have a plan, you ought to have a process, and 17 you ought to proceed through it in a systematic way.
18 I think that's a necessary condition.
19 I would agree with you, George; and I think 20 the ccmmittee would agree with you that that's not a 21 sufficient condition.
22 MEMBER APOSTOLAKIS: That's my problem.
23 MR. CHAPIN: But we -- I don't want to turn 24 this into the committee has reviewed branch technical (g) 25 position 14 or the Standard Review Plan and we have the NEAL R. GROSS COURT REPORTERS AND TRANSCRIBERS 1323 RHODE ISLAND AVE., N W.
(202) 234-4433 WASHINGTON, D C. 20005-3701 (202) 234-4433
335 1 following comments, because we didn't.
7 ~, 2 MEMBER APOSTOLAKIS: I understand.
t :
4
/
~'
3 MR. CHAPIN: But I think that -- you know, our 4 point is that you need to have that requirement. You need 5 to apply a thoughtful process. I think there's -- I'm 6 going to look in the report again, but I think there's 7 some words in the report, you know, to the effect that say 8 when you do the process, you make certain assumptions 9 about what the effect of controlling the process is on the 10 final result.
11 And it's important, if you do that, that you 12 understand what those assumptions are and that you have 13 some feel that you -- there is some connection, that you
(_ l 14 verify it in some way that if you are systematically 15 organizing the process, that you will in fact produce more 16 software.
17 But there's not any clear -- the other thing, I 18 in sort of defense of the staff, I think it would be l l
19 difficult even for John or the software guys to tell you !
20 okay, here's a piece of software, do the following three 21 things to it, check these boxes, and it's okay.
1 22 It's still going to come down to some soft 23 edges and some experience and some qualified people to 24 make those kind of judgements.
n
( ) 25 MEMBER APOSTOLAKIS: And I agree.
1_/
NEAL R. GROSS COURT REPORTERS AND TRANSCRIBERS 1323 RHODE ISLAND AVE., N.W.
(202) 234-4433 WASHINGTON, D.C. 20005-3701 (202) 234-4433
336 1 MEMBER MILLER: If the process includes
,y 2 developing a test plan to test for these things like l
,'~#
)
3 accuracy, completeness, and so forth, then the process I
4 requires you to report that test plan and the results. Is 1 5 that getting into the testing of the product? Because you 6 do that at each stage.
i 7 If it does that, is that what we're looking l 1
8 for?
9 MR. CHAPIN: Yeah, the FAA has a similar kind 10 of a document. And their document says there are certain l 11 attributes of the software, and then there are certain 12 pieces of paper that we look at to make sure that those l 13 attributes have in fact been achieved.
,m i I
's / 14 MEMBER MILLER: Been also -- where 15 appropriate, been tested and the results reported.
l 16 MR. CHAPIN: Yeah, and I think --
l 17 MEMBER MILLER: The results show that they 1 18 meet the requirements.
19 MR. CHAPIN: We could get John to talk again i l
20 if he'd like to comment again. But I guess what I'm 21 saying is I think those are all elements. But it's 22 important that we not get hung up on making sure the guy 23 has a plan and making sure that it's documented and then 24 concluding that the software is okay.
g
() 25 MEMBER APOSTOLAKIS: Good.
l NEAL R. GROSS COURT REPORTERS AND TRANSCRIBERS 1323 RHODE ISLAND AVE., N.W.
(202) 234-4433 WASHINGTON, D C. 20005-3701 (202) 234-4433
337 1 MEMBER MILLER: But if the plan includeo
,S 2 testing -- reporting test results, seems like that closes 8 )
v 3 the loop. )
4 MR. CHAPIN: Yeah, and then you have to look l
5 at the test and say what sort of tests have been run, what l 6 sort of coverage did the test provide. )
7 MEMBER MILLER: Then you can depend'on the j l
8 judgement of the -- of course, the licensee and the l 1
9 regulator to make those judgements on whether that's 10 appropriate.
11 DR. CATTON: We've got a comment over here.
12 MEMBER MILLER: Okay.
13 MR. DAMON: If you look at 14 as a proof kind o
I )
x/ 14 of document, you're going to have a problem. But if you 15 look at the 14 technical position as a quality assurance 16 approach, then you don't have a problem. You see, it's 17 how you view it is really important. Essentially, you can 18 liken 14 to the quality assurance program that you have 19 for main steam line piping.
20 You do everything in the world to make sure --
21 you can trace the middle back to the ore in the ground.
22 But then you automatically assume after you've done all l
l 23 that it ain't going to work. You're going to have a line 24 break. And then you go into that.
(_,/ 25 Well, 14, in the context of the whole review NEAL R. GROSS COURT REPORTERS AND TRANSCRIBERS i
1323 RHODE ISLAND AVE., N.W.
(202) 234-4433 WASHINGTON, D.C. 20005-3701 (202) 234-4433 l
l
'\
338 1 plan is do the very best you can. If you do it this way,
,e~ 2 you've got a pretty good probability of coming out at a
'\s' 3 high level. But, ATWS equipment is really the thing 4 that's going to take care of the common mode software 5 failure because you're only presuming it happens --
6 MEMBER MILLER: Oh , yes; 14 doesn't address 7 the common mode problem, no.
8 MEMBER APOSTOLAKIS: Again, being relatively 9 new to this committee, I had to understand the legal 10 aspects of the whole thing. So maybe BTP-14 was okay, I 11 think. So I asked the staff to give me similar documents 12 from other areas that involve thermal hydraulics or other 13 things.
t '
\
N '/ 14 And as I expected, I found much more specific 15 guidance there regarding temperatures and relative 16 humidity which I thought made perfect sense. What I see 17 in these documents is a very qualitative guidance to the 18 extent, you know, some -- you know, have a plan, do tests.
19 You know, there are tests and tests --
20 MEMBER MILLER: Say it's comparable to 21 software.
22 MR. CHAPIN: I have two people who would like 23 to comment if you guys want.
24 MEMBER APOSTOLAKIS: Yeah, I would like to be
,7 3
(,,,)
25 enlightened.
NEAL R. GROSS COURT REPORTERS AND TRANSCRIBERS 1323 RHODE ISLAND AVE., N.W.
(202) 234-443 WASHINGTON, D.C. 20005-3701 (202) 234-4433
339 1 MR. CHAPIN: Christine had her -- we had a
,~s 2 procedure when we had -- I think we all had name tags and
( /
3 we wanted to talk, you put your name tags up because 4 people got tired of holding their hands up.
5 So Christine, I think, was up; and Carmelo, 6 did you want to comment?
7 MS. MITCHELL: Yeah, I thir.a the -- to put 8 this in context, especially for people who aren't software i l
9 engineers, is there's a tremendous amount of software that i l
10 didn't have a plan. And by a plan, it wasn't -- we didn't 1 11 mean -- and I don't think the regulations mean that 12 somebody has a formulation in their head that they may or !
l 13 may not have communicated.
,r (s/ 14 Software engineering has gotten to the point l 15 where it wants a plan with particular elements that needs 16 to be written down. It needs to be followed. It needs to 17 be auditable in both a legal and technical sense. A:1d 38 that hasn't been present in many software engineering 19 applications.
20 They say in aviation, for exsmple, that no one 21 could explain to you how the flight management system for 22 a Boeing 757 runs. They just -- they couldn't explain it.
23 There was no formal plan that was followed -- written 24 down, followed, documented. I mean, that has certainly
((~
v
) 25 changed the Boeing 777. Software exists in a much NEAL R. GROSS COURT REPORTERS AND TRANSCRIBERS 1323 RHODE ISLAND AVE., N W.
(202) 234-4433 WASHINGTON, D C. 20005-3701 (202) 2344433
340 1 different environment, but that's how the software
,, 2 engineering community has matured.
( :
\/
3 So it's not just a fuzzy term, have a plan; 4 but very technically, it's written, it's verifiable. And 5 if something goes wrong five years later, you can find out 6 what was done, how it was done, what those data were, etc.
7 MEMBER APOSTOLAKIS: So what you're saying is 8 that merely asking for a plan is a good step forward 9 because --
10 MS. MITCHELL: Well, a plan with specific 11 elements and the results of the process that was guided by 1
12 those elements. So as I said, there's a technical meaning 13 to this as an engineer. There would be a legal meaning, I o
(s/ 14 would think. And there would be an audit trail at the l l
l 15 end.
16 MEMBER MILLER: What are the key elements you 17 would believe would be -- the plan for an industry such as 18 nuclear or the aerospace, which I think are somewhat 19 comparable?
20 MS. MITCHELL: And I'm going to defer because 21 I'm not a softwarc ungineer. I'm a software user and a 22 software writer. So I only know what happens when you 23 don't do it right. I know the data say that the 24 requirements specification is where the safety critical --
o
() 25 when you have errors in other industries, a very small NEAL R. GROSS COURT REPORTERS AND TRANSCRIBERS 1323 RHODE ISLAND AVE., N.W.
(202) 234 4433 WASHINGTON, D.C. 20005-3701 (202) 234-4433
341 1 percentage are attributable to the code.
~., 2 And a very large percentage -- and John may
/ i i/
3 want to talk about this just in terms of all these recent 4 reports. A very large percentage have to do with the 5 formulation of the requirements in the process much 1
6 earlier. And so I couldn't tell you what's good. I can 7 just tell you what happens when you don't have it.
8 MEMBER MILLER: Should I f'y in 757's now?
9 MS. MITCHELL: Good pilots.
10 (Laughter.)
11 MR. RODRIGUEZ: I'd like to address the issue 12 of specificity because I think that's the main ingredient 13 that is missing. I think in your example of the
,n, t
xs' 14 manufacturing industry, you know that if you're making 15 polyethylene and if you use ethylene concentration of 5%
16 and catalyst concentration of so many parts per million 17 and the temperature is 500 , then your polyethylene comes 18 out of a high quality.
19 Meaning that it comes out of the right density 20 and the right melt index and so forth. I think that the -
21 - in the nuclear industry, and particularly in the digital 22 instrumentation systems area, what is missing is that 23 specificity that will tell you that will address the 24 things that are different between the digital system and
,/ y
() 25 the analog systems.
NEAL R. GROSS COURT REPORTERS AND TRANSCRIBERE, 1323 RHODE ISLAND AVE., N.W.
('202) 234-4433 WASHINGTON, D C. 20005-3701 (202) 234-4433
342 1 And I think what is missing is the need to 7\ 2 address the basic differences which are basically the five
! )
'J 3 that digital systems operate in a sample data systems.
4 Data don't flow continuously. In a digital control 5 system, data are samples, so it uses the steam pressure 6 right now.
7 And you don't see it again for another few 8 milliseconds or seconds or whatever the case may be. And i 9 then the -- so data flow in a sample manner, and the 10 control actions to regulate processes or to take 11 protective actions also partaking in a sample manner.
12 They're taking one step at a time.
13 And furthermore, the other element is that I '
'w-14 these systems share resources. So that means that a 15 particular processor may be storing a pressure at a time.
16 But a few seconds later, these memory location in the 17 processor may be storing nuclear flux, to make an extreme 18 case.
19 In an analog system, that doesn't happen. In 20 an analog system, you have a capacitor in an amplifier 21 that is dedicated to storing a value that represents the 22 water level in a tank. Okay, and that's it. That 23 capacitor is not going to store nuclear flux.
24 Okay, so I think that what is missing in all t'x kj 25 these standards, the IEEE standards and so forth that we NEAL R. GROSS COURT REPORTERS AND TRANSCRIBERS 1323 RHODE ISLAND AVE., N.W.
(202) 234-4433 WASHINGTON, D C. 20005-3701 (202) 234-4433
343 1 see, is chat they don't have any teeth because they don't 7
2 get into the mode of saying okay, let's address these
( '~' )
3 differences and let's have a recipe that will allow you to 4 come up with an architecture of a digital control system 5 that will ma coh the common practice that has been common 6 in the nuclear industry.
7 In other words, don't have one computer that 8 is going to regulate pressures and flows and temperatures 9 and nuclear fluxes and so forth. Don't do that, because 10 that's bad. The NRC doesn't allow that in the fluid 11 systems and mechanical systems.
12 Have guidelines for partitioning these things 13 so that the feedwater system is controlled by a dedicated
/,x
! \
2 14 feedwater control system; the main steam system is 15 controlled by a dedicated control system; and the nuclear 16 flux is controlled by a dedicated system. And then you 17 have a supervisory system that is like an orchestra 18 director that doesn't play the instruments.
19 I think those are the standards that are 20 missing.
21 MEMBER APOSTOLAKIS: But isn't that the trend 22 though, to use these embedded systems?
23 MR. RODRIGUEZ: Yes, that is the trend.
24 That's exactly the trend. But in looking at the IEEE
,n,
(,) 25 standards and looking at the software standards, it's like NEAL R. GROSS COURT REPORTERS AND TRANSCRIBERS l 1323 RHODE ISLAND AVE., N W. l (202) 234-4433 WASHINGTON, D.C. 20005-3701 (202) 234-4433 I
344 1 you say; do things right. Don't do them wrong.
, ~3 2 MEMBER APOSTOLAKIS: I'm glad you agree. I am N] 3 so glad you agree.
4 MR. RODRIGUEZ: That's exactly -- in my 5 opinion, I think that that's what is missing.
6 MEMBER APOSTOLAKIS: Thank you very, very 7 much.
8 MEMBER MILLER: I'd recommend we -- we had 9 scheduled a break at 11:00. This might be a good time to 10 break and return at --
11 MEMBER APOSTOLAKIS: So that we can celebrate.
12 MEMBER MILLER: We'll return at ten minutes 13 after 11:00.
/N t i ss' 14 (Whereupon, the foregoing mat er went off the 15 record at 10:56 a.m. and went back on the l
16 record at 11:12 a.m.)
17 MEMBER MILLER: I have been advised to retain 18 control of this meeting. And I'm certain that Dr. Lewis 19 would have a very eloquent statement of how I cannot 20 retain control of this meeting. But be that as it may, I 21 would say we would recommend going on with our questions.
22 Did we bring closure to that last set of 23 discussions, or would you like to add more to it?
24 MR. CHAPIN: I'd like to comment with regard
,-\
) 25 to the specificity. And that is, I think the committee --
NEAL R. GROSS COURT REPORTERS AND TRANSCRIBERS 1323 RHODE ISLAND AVE., N.W.
(202) 234-4433 WASHINGTON, D C. 20005-3701 (202) 234 4433
345 1 this committee; I'm sitting on this side of the table.
1
,, 2 You know, the idea that the specifics are just sitting out I x
\ ')
3 there somewhere waiting to be gathered, we don't -- it's l
i 1
4 very desirable to have specificity, but it's not a matter l l
5 of that they've been overlooked or that they're going to l
6 be easily obtainable. I I
7 So I think -- you know, the more specificity l
l 8 is something that we desire. We have no complaint with !
l 9 that. But I don't think that in this area, the software, !
I 10 that those specifics are going to be easily obtainable.
11 And if one wants to go after those, then he should mount 12 an effort, and it will be an effort I think is what we're 13 saying.
.,^s 14 MEMBER FONTANA: Could the specifics be so l 15 application specific that you get a different set for 16 everywhere you look fc an application? In other words, 17 can you do -- we'd like to see more specificity, but is it 18 feasible to do much more without getting directly into the 19 application? l l
20 MR. CHAPIN: I don't -- I could give yc,u an l 1
l 21 opinion.
22 John?
23 MR. GANNON: I think there is some acod 24 examples of things that heve worked particularly in the (ws ) 25 nuclear industry and that we should try and follow the NEAL R. GROSS COURT REPORTERS AND TRANSCRIBERS 1323 RHODE ISLAND AVE,, N W.
(202) 234-4433 WASHINGTON. D C. 20005-3701 (202) 234-4433
346 1 good examples that we have.
,- s 2 MEMBER MILLER: I would like to -- I've been
?
\
s )
3 advised I've been kind of favoring my right side here.
4 I'd like to see if members of the committee over on the 5 left might have comments. I know Dana Powers, who has 6 written me a long statement on the report, I am confident 7 would have questions equally penetrating as my colleagues 8 to my right.
9 So Dana, would you like to jump in here?
t 10 VICE CHAIRMAN POWERS: In some cases, your 11 comments have been very interesting, but they do lack 12 specificity. Maybe you could give us some interpretation 13 or additional insights on how we should read the report to
,m
- )
N_' 14 understand better.
15 In the beginning, of course, with just this 16 issue of product versus process specificity, you say you 17 should regulate the process and evaluate the product as is it's appropriate, but you didn't give us much guidance on 19 what that "as appropriate" meant.
1 20 Another area is you have used what I call the 21 commensurate kinds of recommendations. You've said this 22 or that should be commensurate with the magnitude of the 23 change and whatnot. In particular, you suggest that the ,
l 24 reviews of the digital systems ought to be commensurate t ) 25 with their complexity.
%/
NEAL R. GROSS COURT REPORTERS AND TRANSCRIBERS 1323 RHODE ISLAND AVE., N.W.
(202) 234-4433 WASHINGTON, D C. 20005-3701 (202) 234 4 433
347 1 And then later in the report, you say ah, but x 2 there is no measure of complexity. And I was a little bit
! \
i /
'~'
3 puzzled how I could make something commensurate with 4 something that I couldn't measure. And similarly, there 5 are other commensurate things. I should make evaluation 6 of the human-machine interface commensurate with the 7 magnitude of the change.
8 And again, I don't know how to make -- I'm 9 left at a loss of understanding how I make things 10 commensurate which obviously have radically different 11 units.
12 MR. CHAPIN: Okay, maybe some of that we can 13 clarify by perhaps citing some examples. Take -- and I'll
! )
'w / 14 ask my friends to think about these as well. But for 15 example, in the product versus process environment, we 16 have a sense which I think is reflected in the report that 17 the NRC has a tendency to emphasize the coding end of the 18 problem, the software itself, the production of the 19 software problem.
20 And I think the experience in other industries 21 and the experience of the committee would suggest that 22 there are a bigger bang for the buck if you pay more l 23 attention to the earlier parts of the software cycle, the 24 definition of the functional requirements, the translation t
y,/ 25 of those requirements into requirements for the software.
! NEAL R. GROSS l COURT REPORTERS AND TRANSCRIBERS I 1323 RHODE ISLAND AVE., N W.
(202) 2344433 WASHINGTON, D.C. 20005-3701 (202) 234-4433
348 i
i So-in product versus process, we say there l 2 should.be a balanced point of view. I think we do a f-little better than that in, for example, in the software 3
4 area by saying, for example, don't, focus so much on 5 software itself, but go back and spend more effort -- got i 6 more focus on the earlier part of the cycle.
7 That would be one place that I would~say we do 8 some of that. You mentioned the human-machine interface. '
9 Well, let me take another one. Commensurate with 10- complexity -- the example I would-cite there, I think, is 11 the commercial off the shelf hardware and eoftware where 12 there is a wide range of that kind of replacement'that's j 13 being considered in the industry.
5/
% 14 Some of them are replacement of a pin. recorder 15 with a digital recorder, the replacement of a digital 16 meter with a -- I mean analog meter with a digital meter 17 that you can escalate up to a meter that's got an actuated.
18 contact in it. And then you can go on to systems.
19 And I think one of the things we say there is 20 if you look at what's in the industry guideline that's 21 been developed in that area, they have tried to lay out 4
22' some examples where they have looked at these gradations 23 and apply specific levels of review with lesser, more -- i i
24 other way around -- more detailed, more indepth reviews of r
(_,) 25 systems and system applications like reactor protection NEAL R. GROSS COURT REPORTERS AND TRANSCRIBERS 1323 RHODE ISLAND AVE., N.W.
(202) 234-4433 WASHINGTON, D.C. 20005-3701 (202) 234-4433
. . . _- . _ _ _ -- ~ - . . - - . . - - - _ _ -. - , _ -
349 1 systems.and the like, and a lesser,'more tailored, focused 2 review for digital meters or recorders.
3 So-I think-in that one, again, it's difficult 4' to define complexity. We would agree with that. But I ,
5 don't think it paralyzes you from making practical and 6 fairly well founded engineering judgements as to what you 7 ought to do in some of those' cases.
8 The human-machine' interface one, I guess' I 9 would ask Christine to give me some help on that one. I 10' remember those words.
11 VICE CHAIRMAN POWERS: As you explore this 12 human-machine interface, there's a recommendation in the 13 report also to look at classic human-system interaction 14 design problems. But the report doesn't define.those 15 classic problems. In fact, the report goes on to say that 16- the NRC should search out and find these classic problems. l 17 I wonder if you had any idea what they should 18 be looking for here? I was a little at a loss what to do .
19 there. 1 I
20 MR. CHAPIN: We had an interaction with one j 1'
21 particular presenters, for example, who has a list of 22 those kinds of things. And I think in the human factors 23 community, some of these are well known.
l 24 Chris, let me turn that over to you.
( f 25 MS. MITCHELL: Let me answer the second thing NEAL R. GROSS COURT REPORTERS AND TRANSCRIBERS 1323 RHODE ISLAND AVE., N W.
(202) 2344433 WASHINGTON. D C. 20005-3701 (202) 234-4433
350 1 first because that's probably easier. One of the people 2 who spoke with us was Dr. David Woods from Ohio State.
I,,_ i
\'~'/ 3 And he and his colleagues have done a set of studies in 4 complex systems that -- most of which are safety critical.
5 And they include nuclear systems, aviation systems, and 6 medical applications.
7 And in a recent book which is cited in the 8 references, they have identified persistently occurring 9 errors. And one example is mode errors that we've known 10 in the community for a long time that to design a modal 11 interface for users or human operators, whether it's a 12 text editor or it's in a control room, creates the
~
13 potential or the context where problems may occur and
(_) 14 forces from the design point of view the issue of whether 15 the interface supports understanding what mode the system 16 is in for the operator.
17 So this notion of persistently reoccurring or 18 persistently occurring errors I think is a real one in 19 terms of understanding what other industries have seen 20 that are examples of bad design. And again, I'd refer you 21 to that book. I think that's probably the best summary 22 piece on that.
23 With respect to commensurate, what I was 24 trying to bound -- and I guess it was commensurate in the n 25 context of what reasonable people might reasonably agree
( ,/
NEAL R. GROSS COURT REPORTERS AND TRANSCRIBERS 1323 RHODE ISLAND AVE., N W.
(202) 234-4433 WASHINGTON, D C. 20005-3701 (202) 234-4433
. , - . . - - ..... - ~ - - - - . . - . - - - - . - . . - - . - ..... - - . . . - ,
351 1
1 on. :If what was being changed in a control room was a one 2 for one analog to digital meter or' display device, what 7s U 3 specifically wasn't being recommended was a complete, full 4 scale, empirical evaluation that~might go on for 200 hours0.00231 days <br />0.0556 hours <br />3.306878e-4 weeks <br />7.61e-5 months <br /> 5 to ensure the human factors implications of that change.
6 I mean, I think the terminology that was there l
7 said that anything that affected the human interaction 8 ought to be seriously evaluated; but the extent of how 9 much it was evaluated had to be bounded by the degree of 10 impact that it would have on a human operator.
11 VICE CHAIPMAN POWERS: I'm delighted with the 12 example you've chosen because with the Advanced Test 13 Reactor, I got to participate in a review where a one to T/ s . 24 une flop had been made between analog to digital and 15 promptly created an unexpected scram and probably took us 16 well over 200 hours0.00231 days <br />0.0556 hours <br />3.306878e-4 weeks <br />7.61e-5 months <br /> to understand why it had occurred and ,
17 to come up with a way to prevent it in the future.
18 Sometimes very simple changes. When you work i I
19 on the man-machine interface, especially when you go to 1 1
)
20 digital, it's formidable. !
21- MS. MITCHELL: You know, there's no question l l
22 I that you can find five sigma out examples of catastrophes.
23 But the notion is that if you're going to proceed with 24 digital changes, that -- and you're going to be rigorous
( 25 about any of it, that you want to design studies that are NEAL R. GROSS 1 COURT REPORTERS AND TRANSCRIBERS '
1323 RHODE ISLAND AVE., N.W.
(202) 234-4433 WASHINGTON, D C. 20005-3701 (202) 234-4433
1 352 l l
1 bounded by the context of the kinds of changes you're i
, - ~ ;.
_ 2 making.
f
\v' 3 MEMBER APOSTOLAKIS: This issue of complexity, 4 if I can pursue it a little bit -- there is a widespread 5 belief among nuclear people that the digital software we 6 are using or about to start using are simple. Simpler 7 than the shuttle's software or the Arriane that had the l l
8 prot i t.m recently and so on.
9 Is it true that there is no widely accepted 10 measure of complexity? Is it just like we'll recognize it ;
11 when we see it?
12 MR. CHAPIN: My answer to that would be yes.
13 There may be others on the committee that want to do that.
'N ' 14 I worked in the ALWR program for a number of years and 15 helped write some of the fundamental documents. And we 16 have a requirement in there, and one of tha precepts of l
17 that whole problem -- of that whole project was to 18 simplify the existing machine so that the future machines 19 were much simpler to operate and maintain.
20 And the committee itself, the ALWR steering 21 committee, commissioned a study of how do we measure 22 complexity. And I think there was a project given to MIT 23 and they came back and there were log integrals of 24 information contained in systems and states and that sort f m.
() 25 of thing.
NEAL R. GROSS COURT REPORTERS AND TRANSCRIBERS 1323 RHODE ISLAND AVE., N.W.
(202) 234-4433 WASHINGTON, D C. 20005 3701 (202) 234-4433
1 353 j 1 To my knowledge, none of that resulted in any
, ~x 2 quantitative measure. It was more in looking at systems v)
(
3 and saying don't assign multiple functions to systems 4 because of concerns like Christine has just mentioned.
5 It's very hard in the BWR to sort out is this the 6 injection system or the RHR system or the RCSI system or 7 whatever when it's all the same pumps and valves being 8 used for various things.
9 Similar examples for PWR. So my experience 10 would tell me almost independent of this committee that I 11 don't think there is any specific measure of complexity.
12 MEMBER APOSTOLAKIS: The problem -- the p,
13 practical problem that I see there is that in the -- I i ;
i# 14 mean, this agency's moving towards risk informed 15 regulation, as most of you probably know. And of course, 16 that means that there is a serious effort to quantify the 17 failure probabilities of as many elements as we can.
18 And of course, hardware is probably easier 19 to -- hardware failures are easier to quantify in terms of 20 frequencies than perhaps the human error. We've make some 21 significant strides in the last 15, 20 years. And now we 22 have the software. We're still in the dark.
23 And one of the things that always comes up is 24 that you try loing certain things, you know, to bound the 7s
(_ ,/ 25 probability perhaps or to say something about it. And NEAL R. GROSS COURT REPORTERS AND TRANSCRIBERS 1323 RHODE ISLAND AVE., N.W.
(202) 234-4433 WASHINGTON, D C. 20005-3701 (202) 234-4433
354 1 then typically, a computer scientist will give you a 73 2 strange occurrence that happened to the shuttle some time
) .
1
\/
3 ago or somewhere else, and you really scratch your head i 4 and say but does this really apply here? l 5 I've tested the things to death. I've applied 6 other methods. I've done this and that. Is there any way 7 we can say certain things happen because that system was 8 very complex and they don't apply to other things, or is 9 it really impossible to eliminate some of these instances 10 which really lead to paralysis.
11 Because every time you say something -- yeah, 12 but this happened; would you have thought of that? No, I 13 couldn't. Well, forget it then. I don't know. And maybe )
,ey
- s
- 14 before the reactor safety study was published in the late 15 1960's, there were a lot of people who were saying the 16 same thing about quantifying the risks from reactors.
17 So is there any guidance there as to maybe not 18 trying to inc'.ude all these strange things that have 19 happened to all these complex systems to very simple 20 actuation systems?
21 MR. CHAPIN: George, there are several people 22 who want to talk. Let me sort of -- we started at one !
23 point in the committee with the hope on behalf of several 24 of us that we could conclude that digital I&C systems at I'T l
(,/ 25 nuclear power plants were so much simpler than ?
NEAL R. GROSS COURT REPORTERS AND TRANSCRIBERS 1323 RHODE ISLAND AVE., N.W.
(202) 2344433 WASHINGTON, D C. 20005-3701 (202) 234-4433
355 l
1 applications like 777 landing systems or the NASA systems,
,, 2 that there was some quantum difference such that we could l I
~'
3 say the systems were testable and we could provide some 4 quantitative demonstration.
5 l We did do an exhaustive test of that. Several 6 people on the committee looked at examples of software 7 that was written. And we concluded that simpler was 8 better, but it wasn't so simple that we could just draw a 9 line and say we're on this side of the line and they're on 10 that side of the line.
11 Now with that as introduction, I think Larry 12 wants to talk. And I guess the other guy I would like to 13 have say something about this is Dave Hill, f
q f
\/ '
14 MR. DAMON: Yes, we have to look -- realize 15 that the committee was charged with lookir.g at a spectrum 16 of applications. In operating reactors, the systems we 17 have for safety are relatively simple. They're not -- the 18 controls that you exercise are not modulating controls.
19 These systems are essentially cocked and ready 20 to go to completion. And once they're engaged, there's no 21 return. It goes to completion. This is a design 22 requirement. So the systems that have to monitor a 23 relatively few number of variables -- few in terms of a l 24 dozen -- if any two or three of those go out of line, they i
(~~
j () 25 engage, and that starts the process.
1 NEAL R. GROSS COURT REPORTERS AND TRANSCRIBERS 1323 RHODE ISLAND AVE., N W.
(202) 234 4433 WASHINGTON, D.C. 20005-3701 (202) 234-4433
356 1 Digital or analog, that's a very simple 73 2 control process. It's bang, bang; on, off; go, stop, N] '~'
3 control. Software for that kind of thing, if it were 4 modular, it is pretty simply constructed. But when you 5 get to where you're going to go like with the advanced 6 reactors where it's a total digital control, things 7 change.
8 The odd thing about it is in advanced 9 reactors, which approach inherent safety, the dependencies 10 that you have on these safety systems is substantially 11 diminished. So now the advanced reactors have more 12 complex control systems, it's all digital; but the 13 consequences are somewhat subdued because of inherent
' )
I 14 safety.
15 So it's -- talking about complexity is a very 16 complex subject. What do you really mean? It's kind of a 17 Heisenberg thing. Once you define it a little bit, it 18 kind of gets away from you. There has to be a judgement 19 process. Certainly for operating reactors, my feeling 20 is -- and this is personal -- that digital controls solve 21 an awful lot of difficulties, and they are essentially in 22 simple systems that go from initiation to completion, and 23 the software requirements that we've got in front of us 24 now are sufficient to make them so that we can go forward rN
() 25 with an increased level of safety.
4 l
NIDAL R. GIU3SS l COURT REPORTERS AND TRANSCRIBERS l 323 RHODE ISLAND AVE., N.W.
(202) 234-4433 ASHINGTON, D C. 20005-3701 (202) 234-4433 l
357 1 When we get into where we're doing completely 73 2 new controls and completely new plants, advanced reactors i i i
^) 3 depending on displays, then the complexity comes into not 4 looking at meters, but looking at human factors 5 interaction with the video screen, whether that video 6 screen is clear and cogent in the displays or some 7 computer jockey has gotten a lot of power techniques into 8 it and it's difficult to sort out.
9 That complexity is different. The report, I l
10 think, tried to embrace both ends by just addressing the 1
11 subject as something that really has to be defined in 12 terms of the present application. And that's the 13 commensurate element. Commensurate with what you're
( i
'w/ 14 really doing. Look at complexity in that context.
15 MR. HILL: I have a little to add on the 16 subject of complexity except that I became convinced ,
17 during the course of this process that I didn't know how 18 to define complexity and come up with anything that was 19 meaningful in terms of bounding probabilities on that 20 basis.
21 What was simple in terms of software -- I was 22 rapidly disabused of the notion by the software experts 23 that simple software existed. I have more to say on 24 probability, but I don't want to take us there unless we fs
(_) 25 finish the subject of complexity. So I would defer to the NEAL R. GROSS COURT REPORTERS AND TRANSCRIBERS 1323 RHODE ISLAND AVE., N W.
(202) 2344433 WASHINGTON, D.C. 20005-3701 (202) 234-4433
358 1 committee first.
fy 2 MEMBER MILLER: Any further questions on v) 3 complexity?
l l
4 MEMBER KRESS: Well, I'm not sure it's on 5 complexity. It may be related. I'm particularly 6 interested in and intrigued by one of your conclusions and i
7 a related part of it. And that is that the reliability of I 8 software / hardware systems could be quantified for us in a )
1 9 PRA which may re]. ate to this complexity issue.
10 And that if you wanted to -- a second part of 11 that, if you wanted to achieve a particular high 12 reliability, that diversity is not -- in software and 13 hardware is not the way to do it. I wonder if you could N
- )
14 comment on that issue. That seems to relate to the 15 complexity in some way.
16 MR. HILL: It does. It also relates to 17 something that Doug mentioned in his presentation, and 18 that was the language barriers. I don't think I'd be 19 giving away any secrets if I said that this was a much 20 discussed subject in the committee.
21 And the actual wordings of the conclusions and 22 the recommendation was difficult. One wanted to be 23 precise in what one could and couldn't do, and different 24 people were coming from different positions. Software n
f
\. )
m 25 reliability -- you used reliability. That to a software NEAL R. GROSS COURT REPORTERS AND TRANSCRIBERS 1323 RHODE ISLAND AVE.. N W.
(202) P34-4433 WASHINGTON D C. 20005-3701 (202) 234-4433 1
i 359 l 1
1 engineer, I believe -- and John may correct me here -- has 73 2 a specific meaning which may not be the way you meant it.
( ')
3 When I hear software reliability, being a 4 nuclear person first, I think -- I translate that mentally 5 into the probability of a software failure given certain i
6 conditions leading to the failure of the system. It's )
l 7 shorthand.
8 MEMBER KRESS: That's the definition of -- j l
9 MR. HILL: Yes, but it's not their definition. l k
10 The image that's invoked, and maybe John could help me l
11 here, is some quantifiable reliability through multiple l 12 passes through a string of code. You know, this is an l l
13 example of the language barrier. Well, it took us a long
's '4 14 way to get there.
15 So to come back to the conclusion of the 16 committee on communication -- so I am now talking about 17 the probability of software failure leading to a system 18 failure with all the appropriate conditional statements 19 which you might use in doing a PRA. And I think once we 20 overcame that and defined more carefully what we were 21 talking about, that we were able to gain consensus that 22 such concept for evaluating digital systems and the risk 23 to any engineered system that relies on those systems was l 24 in principle doable.
(m) a 25 Now whether you can as a matter of practice NEAL R. GROSS l
COURT REPORTERS AND TRANSCRIBERS l 1323 RHODE ISLAND AVE., N.W.
l (202) 2344433 WASHINGTON, D.C. 20005-3701 (202) 234-4433 1
360 1 quantify it is an open question. And the best example of 7, 2 that -- and Christine or John may like to -- Christine, in
('"' )
3 this case, may like to add, it's human reliability. I 4 think the closest analogy we have today in w'lat we do is 5 human reliability.
6 You have debated for some decades now. We 7 have debated what you can say about human reliability, 8 whether it's meaningful er context you phrase it in, and 9 what does it tell you about the contribution to the 10 system. The creation of software and the application of 11 software is a human endeavor and the two are interlinked, I
12 in my personal opinion.
13 And it can be treated in the same way to some
(~)
\- / 14 degree. I was careful to add my opinion as distinct from l
15 the committee's opinion. That may be the committee's 1
16 opinion, but we didn't actually say that explicitly. We l
17 do believe that you can use PRA concepts including the 18 probability of software failure or software reliability 19 again with all the appropriate conditional phrases to make 20 sure it's a precise statement to what you mean to evaluate 21 the impact of a software failure, inverted commas, on the 22 risk of a system.
23 Does Christine or John want to add to this?
, 24 MR. CHAPIN: Let me give you this one
,m
( ) 25 anecdotal story. I know in one of the ALWR designs which NEAL R. GROSS COURT REPORTERS AND TRANSCRIBERS 1323 RHODE ISLAND AVE., N W.
(202) 234-4433 WASHINGTON, D.C. 20005-3701 (202) 234-4433
. _ ~ _ _ _ _ - _ _ . _ . _ . _ . _ . - _ _ . . . . _ _ _ _ _ . _
361 1- I was involved in that we had a system which we wanted-to
-2 act in some cases and not act in others. And one went 3 through and concluded what backed ~out of that, if-you 4 will, the desired reliability numbers for what that 5 implied'with regard to the reliability of the digital 6 control system which was going to control that piece.of 7 the plant.
-8 And the numbers required very high reliability 9 for that system, and the design decision was made that's i
10- inappropriate because we're not that confident that we- )
- 11. know all the things that might go on, and so the design of l l
l 12 the system was changed. )
i 13 So even without assuming that I know exactly-O 14 what the reliability is, one can use the classic PRA 15 process to identify weaknesses, to identify points where- 1
-16 you lean on the importance of the software or on the 17 programming. And if you're demanding very high 18 reliability, then you go back to, you know, gee, should I 19 da this, can I do this, can one really assure -- you .
1 20 picked the number, ten to the minus some digit, that he's 21 going to be willing to depend on that digital system.
22 We have the advantage in the nuclear business 23 that we generally have a safe state. And, as Larry talks 24 about, what these digital systems do is they detect we're
() 25 approaching the boundary of where we would like to be, and NEAL R. GROSS COURT REPORTERS AND TRANSCRIBERS 1323 RHODE ISLAND AVE., N.W.
(202) 234-4433 WASHINGTON, D.C. 20005-3701 (202) 234-4433
362 1 we put the plant on the safe state. I would point out to
,fyi 2 you that somebody like the aircraft industry does not have V
3 that luxury.
4 And Mike DeWalt, who's not here today, 5 reminded us several times that, you know, if landing 6 systems in these planes land with large -- these large 7 aircraft and are automatic digital mode and there is no 8 safe state to put the plane in when you're 100 feet off 9 the ground.
1 10 So, you know, we -- one can look at the l l
11 judgements that he has to make with regard to the !
12 reliability of the software even without a precisely 13 quantitative number with regard to what that may be.
' 14 Now let me go down here and -- I believe, John 15 or Christine, did either of you have anything?
16 DR. LEWIS: I wonder if I could add some 17 confusion to this situation. There's room -- there's a 18 little bit of room. First of all, language, which we were 19 talking about, we really do have to be careful because the 20 word complexity has a very precise meaning to computer 21 scientists, and we're using it in a way here which is sort 22 of a vulgarian way which is just fine for this 23 conversation.
24 But if we go into a community of -- I'm sure fm.
() ,
25 you'll back this up -- go into a community of computer NEAL R. GROSS COURT REPORTERS AND TRANSCRIBERS 1323 RHODE ISLAND AVE., N.W.
(202) 234-4433 WASHINGTON D C. 20005-3701 (202) 234-4433
363 1 scientists, they really know what they mean by complexity.
, >3 2 They write books about it, they write papers about it.
C 3 It's a well defined subject. It's not an easy subject.
4 It's complex in the other sense of the word.
5 The same thing is true for software failures.
6 There's really no such thing, broadly speaking, as a 7 software failure. My wife -- not my wife. My daughter 8 once told me that if you push the keys on your computer 9 and it gives a result you don't want, don't do it again 10 because it will do the same thing.
11 And you know, we all fall into this trap of 12 doing it again in the hope that somehow it will come out 13 better, but it doesn't come out better. It does what it's
/
)
14 supposed to do, and it does not fail. The electrons do 15 not wear out. Now there are slight exceptions. You know, 16 in the nuclear weapons business, there are cosmic rays 17 that interweave bits into a thing.
18 Those are minor exceptions. Those are really 19 out of the 78 sigma level; although it's not Gaussian, so 20 we shouldn't say 78 sigma. But the real failures in 21 software come when the software program is challenged by 22 initial conditions or states, if you want to call it, 23 which were not anticipated in the design.
24 I came east on a 777 last night, and my seat
('T
( ,) 25 mate was horrified when I told her there was nothing NEAL R. GROSS COURT REPORTERS AND TRANSCRIBERS 1323 RHODE ISLAND AVE., N.W.
(202) 234 4433 WASHINGTON, D.C. 20005-3701 (202) 234-4433
364 i l
1 except wires connecting the cockpit with those wings. And 1 1
- 2 you know, I ignored the steel and aluminum and that sort
/ i i !
3 of thing. She was horrified. You know, I told her it's 1 1
4 okay, it's been done before. l 5 (Laughter.)
6 MEMBER KRESS: Did you believe what you just 7 said?
1 8 DR. LEWIS: She let it go at that, but she 9 wasn't -- she was a very nice woman, but she was a little 10 bit uneducated.
11 But the point is that software failures result 12 from things which are originally in the software 23 challenged by events which have not been confronted by the m
t
\
'N._/' .4 software before. And I mentioned to John a moment ago 15 that with great personal pain, I've switched my personal 16 computer to Windows NT.
17 It's, you know, a rotten operating system, but 18 it's what there is, you know. Same reason my wife married 19 me.
20 (Laughter.)
21 And I just loaded a service pak for Windows i
22 NT, Service Pack II, which was 15 megabytes zipped, and 23 Lord knows what it is when it was unzipped; and it had
]
24 literally thousands of bugs that were present in the em 1 (v ) 25 original release of Windows NT, none of which have ever j 3
NEAL R. GROSS COURT REPORTERS AND TRANSCRIBERS 1323 RHODE ISLAND AVE., N W.
(202) 234-4433 WASHINGTON, D C. 20005-3701 (202) 234-4433 l 1
365 1 bothered me.
, _s 2 So there's essentially no operating experience r ;
' 3 on my part that there were any bugs at all, but there were 4 in fact thousands of them and there will be more thousands 5 found which will prevent people using other printers, 6 other operating systems, doing other problems dealing with 7 them.
8 So the probability you're dealing with has 9 nothing to do with the failure of probability, and that's 10 confusing a lot of the people in the nuclear industry who 11 are using the word failure as if it were the same as a 12 valve failure or something lile that. The probability 13 you're dealing with is the probability that you initiate a es I )
1/ 14 -- you push the combination of buttons that trigger this 15 lurking failure.
16 Now the only exception I know -- a software 17 program is a functional translator. It takes an input and 18 it provides an output. And if it's digital, you can even 19 count them both. If it's analog, you would count them 20 both too -- not that big a deal to round off your numbers.
21 But the only case I know in which you can 22 actually guarantee that there will be no unwarranted input 23 that leads to a surprising output is the cases of which 24 you can use formal methods. And I'm surprised that the
,r~
( ,) 25 light treatment that you've given to formal methods --
NEAL R. GROSS COURT REPORTERS AND TRANSCRIBERS 1323 RHODE ISLAND AVE., N.W.
(202) 234-4433 WASHINGTON, D C. 20005-3701 (202) 234-4433
366 1 this is what I was leading up to in this report.
- 2 Because it's the only case in which the kind
() 3 of vague concerns that have been around the table can be 4 raised. My friends to whom I speak about these matters, 5 because I'm not a software engineer -- not that there's 6 such a -- and I'm not a computer scientist either. ,
7 My friends tell me that the simplicity, to 8 just use a word, of the nuclear business is such that it 9 probably is within reasonable -- and you've probably 10 discussed this at length among yourselves -- is pretty 11 close to being addressable by formal methods.
12 In which case, not all problems -- obviously 13 we still have the human interface problem. But at least a t\
's / 14 certain class of problems, the problems going into a 15 number of unexpected state and getting an unexpected --
16 can be erased. And I wonder why you're not prodding 17 people to push in that direction?
18 I'm sorry that was such a long speech, but I 19 didn't -- always mottling it up.
20 MR. CHAPIN: Okay, yes; we do talk about 21 formal methods. It was something that we discussed 22 internally at some length. It's been a subject, frankly, 23 of several of the review comments that we received from 24 outside reviewers as to what we -- what our feelings were d 25 with regard to formal methods.
NEAL R. GROSS COURT REPORTERS AND TRANSCRIBERS 1323 RHODE ISLAND AVE., N.W.
(202) 234-4433 WASHINGTON, D.C. 20005-3701 (202) 234 4433
367 1 And I think that -- again, I won't open the 2 report and quote it, but I think the sense of the report l 5
)
3 is that we think they have merit and they ought to be 4 applied, but they're not a silver bullet. And with that 1
5 as sort of an introduction, let me let John go ahead.
6 MR. GANNON: Yes, one of the things that the j 7 committee debated was a recommendation that says, you 8 know, thou shalt use formal methods before you can field 1
1 9 digital systems. And I don't think anyone was willing to 10 go that far. We think there are areas where they're 11 particularly effective on the requirement side early in 12 the process where you're trying to make guarantees about 13 properties the resulting system will have.
p) x> 14 And we do make a recommendation that more 15 precise requirements be written -- requirements be written 16 in a precise language, something that is amenable to 17 analysis. So I think there is a direction to take for 18 work in formal methods on systems.
19 MEMBEh APOSTOLAKIS: And you are familiar with 20 the work Ontario Hydro has done on that front?
21 MR. GANNON: Yes, my own research is in model 22 checking and verification of requirements, so I'm very 23 familiar with Dave Parnassus' work, 24 MR. CHAPIN: One of the things that came out
/~N
(,) 25 of the business of thou shalt use formal methods, I think NEAL R. GROSS COURT REPORTERS AND TRANSCRIBERS 1323 RHODE ISLANL, AVE., N.W.
(202) 234-4433 WASHINGTON, D C. 200G5-3701 (202) 234-4433
368 1 one of the things that caused some of this pause was this 7y 2 issue of language. That if you use the formal methods, i
(/
3 one typically ends up in a highly mathematical language.
4 And in the words of -- you know, scientific 5 words, the domain experts, the operators, the maintainers, 6 the engineers are usually not expert in that language.
7 And so I think even Ontario Hydro's experience or the like 8 is that this communication difficulty is very difficult.
9 There's also some mathematical things you can do, and the 10 proof may have an error.
11 There's some difficulty in translating word 12 requirements into the formal requirements. There's --
13 I've even heard a paper by a man who is a digital I&C man i 1
\/ 14 at one of the utilities, the nuclear utilities, who is 15 also an English major, and he uses English language type 16 approach where he tries to -- says that specification to 17 be structured in certain formal ways in a language sense.
18 And I think that was one of the problems that 19 we have. And we say gee, you just ought to use these 20 formal methods, bang. It's not clear that in all cases 21 you would be able to effectively get the domain experts, 22 if you will, involved in that so that you get the full 23 checking.
24 MEMBER MILLER: Can I pick up on this?
,3 l \
\ ,/ 25 DR. LEWIS: One word in response to that.
NEAL R. GROSS COURT REPORTERS AND TRANSCRIBERS j 1323 RHODE ISLAND AVE., N W.
(202) 234-4433 WASHINGTON, D C. 20005-3701 (202) 234-4433 l
1
_ _ . _ . _ _ _ _ . .~._ __. _ - _ _ . _ . _ . . . _ _ ._ _. .__ -._.._ _ _
=369 l
1 There are lots of things in our. society in which the. users-4
- 2. don't have the foggiest notion how they work. Juni you
- - 3 don't have to communicate to users how things work.
People who watch sitcoms on TV have no idea that those 1
4
)
5- rays.come through the air and are collected by an antenna.
j 6. There are various levels. You know, the
- - 7 technicians know -- who repair them know a little bit 8 more. But generally, they don't know what a comb filter-i I 9 is. It's just a thing you take out'and you put in if it l 10 improves the picture.. So there are degrees in which you
- 11- can-implement higher level technologies without requiring i
12 that'all the way down the chain of use, people understand l J e
13 them.
ls 14 They need to know enough to respond to unusual 1,
15 conditions within their jobs, but they don't need to know 1
16 everything. And I put formal methods in that category.
I 17 That is, I'm not that much of an advocate of the -- or you j i
18 may be misunderstanding what I'm saying.
19 But I don't think they should be ignored as a 20- possible solution to real problems within this business.
21 MR. CHAPIN: No, and the committee does not i 22 ignore -- maybe somebody can find the words in the report.
23 But I think what we say is that it's -- even in a footnote 24 or something -- that the effect that the formal methods
() 25 aspect -- and we see that the formal methods have benefit, NEAL R. GROSS COURT REPORTERS AND TRANSCRIBERS 1323 RHODE ISLAND AVE., N.W.
(202) 234 4433 WASHINGTON, D.C. 20005-3701 (202) 234-4433
370 1 particularly in this front end definition.
,3, 2 But the language -- I agree with you that it's l
! i 3 not important that the users understand all the way dowr. I 4 the line. But so much of the domain exuertise as to what 5 the hazards are and what are the problems that have to be 6 dealt with then effectively translated into software, that ]
i 7 domain expertise really does probably reside largely with 1
8 people who are not able to read the code, if you'll allow !
9 me to use that expression. I 10 So that there is a real concern, I think, to 11 make sure that the domain expertise -- the guy who says q 12 gee, if this set of conditions comes up and I don't want l ll 13 this to happen, he can tell you that, then you want him to
/~N
)
4 5
' - 14 judge at the other end that the software really -- or the 15 system really does do that. !
I 16 He's not one of these low level users. He's a 17 high level specifier of the system and a judge that the !
l l
18 requirements have in fact been met. So we don't disabuse !
l 19 the use of formal methods. We say they're important. We i 20 think they ought to be considered. l 21 But, as John said, we have a real problem with 22 an endorsement that said, you know, by golly, you 23 shouldn't use digital systems in nuclear plants unless you 24 use formal methods to review it.
f~N 25 DR. LEWIS: I don't disagree, (u)
NEAL R. GROSS COURT REPORTERS AND TRANSCRIBERS 1323 RHODE ISLAND AVE., N.W.
(202) 234-4433 WASHINGTON, D.C. 20005-3701 (202) 234-4433
371 1 MEMBER MILLER: Can I pick up on this issue?
,f 3 2 I agree that it's the requirements phase of software
( )
3 development or system and software development is 4 certainly a major area where errors will kind of creep in.
5 And you did allude to it in your software quality 6 assurance section.
7 But I was surprised that you didn't go into 8 more detail of the type you just said verbally here on 9 some recommendations how software requirements could be 10 more effectively implemented. In addition to formal 11 methods, there are many other areas -- ways in which it 12 could be implemented effectively over and above.
13 You have one statement basically under fA}
K/ 14 recommendation two in that area which says system 15 requirements should be written in a language with a 16 precise meaning, so on and so forth. There are other ways 17 that that could be implemented other than just that 18 recommendation.
19 I was surprised you didn't have more meat on 20 that recommendation than you did. Maybe it's in there 21 somewhere.
22 MR. CHAPIN: Okay, John again can help me.
23 But again, I think one of the things we said is that there 24 are a lot of people in the software community. We talked
,m I,s_ -) 25 to a number of individuals. If you talk to Mr. Formal NEAL R. GROSS COURT REPORTERS AND TRANSCRIBERS 1323 RHODE ISLAND AVE., N.W.
(202) 234-4433 WASHINGTON, D C. 20005-3701 (202) 234-4433
372 1 Methods, he'll tell you he's got the silver bullet right
,3 2 there in his hand.
( )
v 3 (Laughter.)
4 MEMBER MILLER: What I'm really saying -- I 5 don't think there is a silver bullet. I think there's a 6 number of different ways with each one having less than a 7 silver bullet.
8 MR. CHAPIN: Yeah, and I think that's what we 9 said. What we're saying is there is -- there is a lot of 10 discussion in the software community with different 11 experts who say they have a way which will dramatically 12 increase things. And what we said was it's not clear that 13 any one of those is, you know, sort of inherently superior c 4
/ 14 to the others and that it's difficult from our point of 15 view -- I think we're sort of concerned with the Good 16 Housekeeping seal.
17 You know, if we say this is the one or this is l 18 the way, then that implied to us that we had to make a l l
19 detailed review ourselves and to conclude that that was a l
20 method or a detailed approach that should be used. I'm l 21 very rapidly getting off the edge of my --
22 MR. GANNON: I'd just say conclusion five in 23 the report outlines a number of steps that we think are 24 part of a good assessment process. We didn't actually
,m, i j 25 move those into the recommendations, but there's certainly l
NEAL R. GROSS COURT REPORTERS AND TRANSCRIBERS 1323 RHODE ISLAND AVE., N.W.
(202) 234-4433 WASHINGTON, D.C. 20005-3701 (202) 234-4433
373 1 some detail there, g
-N 2 MEMBER APOSTOLAKIS: Would you though 3
Q,)
3 encourage the staff, the NRC staff, the U.S. NRC staff, to 4 pursue all these different techniques in a research mode 5 in the next couple of years until finally they understand 6 better the advantages and disadvantages of each and how 7 they can be adapted to the problems we're having? l 8 MR. CHAPIN: I'll let John give you a more !
9 precise maybe answer with regard to the software. I think 10 it's an important topic, but I think our main guidance to 11 the research folks is that they need to sit down and 12 identify all of the "important topics" that they need to i
_ 13 work on which I personally would certainly put this on the
[s'l
\# 14 list. -
15 Then they need to look at their resources and j i
16 who they can interact with, their cooperators in the 0
17 field, and put together a structured plan which is the l t
18 best "se of those resources.
. I'm not prepared to tell 19 you, George, because I haven't done that trade, that if I 20 did that, this item would be up in the list that says by 21 golly, we ought to go after this one.
22 But it should certainly be on the stalking 2
23 horse list for me that says this is something we ought to 24 consider and let's put it in the hopper.
(y
( ,)
25 John?
NEAL R. GROSS COURT REPORTERS AND TRANSCRIBERS 1323 RHODE ISLAND AVE., N.W.
(202) 234-4433 WASHINGTON, D C. 20005-3701 (202) 2344 433
374 l' MR. GANNON: You sound great.
- 2 MEMBER-APOSTOLAKIS: Now I have to come back l(
.\
3 to --
4 MR. CHAPIN: I'd like credit for that one.
5 MEMBER APOSTOLAKIS: -- what David Hill said !
6 earlier regarding the probability of failure'of software. 1 7 And I fully endorse what-David said, that it-is meaningful 8 to talk.about the probability of failure of a system 9 including its software; but it is not meaningful to talk
]
10 about the probability of failure of software in isolation 11 because, as Professor Lewis rightly pointed out, given a 1 1
12 set of inputs, the software will always do the same thing, 13 at least with a very, very high probability.
( 14' 'So I must say that I agree with Professor 15 Leveson when she does not concur with your conclusion 16 three on page 108 because that conclusion three is not 17 stated the way David Hill explained to us the probability 18 of failure of software.
19 If I read conclusion number.three on page 108 20 without knowing any better, I would also conclude that it 21 refers to probabilities of failure of -- first of all, it
'22 starts out with a very wrong statement, in my opinion.
23 The assignment of probabilities of failure of software --
- 24. for software are more generally for digital systems is not
.( 25 substantially different from the handling of many of the NEAL R. GROSS COURT REPORTERS AND TRANSCRIBERS 1323 RHODE ISLAND AVE., N.W.
WASHINGTON, D.C. 20005-3701 f (202) 234-4433 (202) 234 4433
375 1 probabilities of rare events.
,_s 2 Well, it is because of what Dr. Lewis said.
/ \
3 We don't have any other events in PRA's that always do the 4 same thing given the same inputs. And I must say I cannot 5 say anything about human reliability. I've been advised 6 by the staff I have a conflict of interest because I 7 talked to you about it.
8 So anything -- when you say human reliability, 9 I have to shut up.
10 MR. CHAPIN: But George, the anecdotal i I
11 evidence that Professor Lewis gave us, I have personal l l
12 experience where it does work to put the same keys back 1
13 into the computer, and it will in fact print that time.
/~N
(_,) 14 And the reason is the system is so complex and there's so 15 many other things running in the background that the state 16 of the thing really has changed.
17 And so it really does have some of the 18 characteristics of a probablistic system. So when we talk 19 about the failure of the software, I would agree with you 20 if we had one piece of software and the only thing was 21 read this key when I press it, that that's not what the 22 software does.
23 The software says what's the state of the 24 disk, was I talking to the network, what's today's date, O) v 25 what time is it, what other packages are loaded up on the NEAL R. GROSS COURT REPORTERS AND TRANSCRIBERS 1323 RHODE ISLAND AVE., N.W.
(202) 234-4433 WASHINGTON, D.C. 20005-3701 (202) 234-4433
376 1 system. And so there's a context that even --
- 2 MEMBER APOSTOLAKIS: Yes.
'~'
3 MR. CHAPIN: And I think what Nancy's -- we 4 ought to -- unfortunately, Nancy's not here. She's not l l
5 avoiding you guys. She has exams or something with her l
6 students and she couldn't come. But I think Nancy's 7 objection precisely comes from the argument that's made 8 that in the software sense it's completely deterministic.
9 And I think where the committee is coming from 10 is that these are systems even in the sense of software. l 11 And there's so many different states and conditions that 12 one can make a judgement.
13 MEMBER APOSTOLAKIS: But isn't it really what ry i
k/m 14 we're after is the probability of failure of the system?
15 Let's say failure to scram, failure of high pressure 16 injection, and so on. Now, 20 years ago when PRA was 17 still in its infancy, most people would look at the 18 hardware -- in fact, they would even do what they used to 19 call random independent failure analysis which drove 20 people crazy because they produced probabilities like 10-'
21 and 12 and so on.
22 Then we learned. Then we added common cause 23 failures, hardware common cause failures. Then, as we 24 move along, we learn more. And we added possible human
,r
(,3) 25 errors. We never talked about probabilities of common NEAL R. GROSS COURT REPORTERS AND TRANSCRIBERS 1323 RHODE ISLAND AVE., N W.
(202) 234-4433 WASHINGTON, D C. 20005-3701 (202) 234-4433
377 l l
1 cause failures or human errors independently of the l l
.x 2 system. l l I
\~j 3 Now we are forced to think about digital 4 software that are possibly part of the system. So we will 5 add that to the system. But the ultimate goal is always l
6 to produce a probability of failure of the system. In l
7 that sense, I think software, human-ware, whatever-ware l 8 should be part of the analysis of the system. l l
9 And I believe by separating out the software l l
10 and saying, you know, what is the probability of failure 11 of the software, all you do is you start an endless debate l
12 as to what that means. And I think what David said is l 13 perfectly reasonable. And maybe that was the intent of i r~s '
(
's
)
/ 14 conclusion number three, but it doesn't come across.
15 And if I read it like this without talking to 16 any of you, I have to concur with Professor Leveson.
17 That's my opinion, of course. I mean, --
18 MR. CHAPIN: You're certainly entitled, you 19 know, to your opinion.
20 Dave, do you want to -- I don't want to argue 21 with you, so I'll let Dave go ahead.
22 MR. HILL: I don't really want to argue 23 either. The sense of it is the way I described it before, 24 and I don't propose to repeat that. The way I interpreted f3
( ,) 25 the -- not substantially different from handling many NEAL R. GROSS COURT REPORTERS AND TRANSCRIBERS 1323 RHODE ISLAND AVE., N.W.
(202) 2344433 WASHINGTON, D C. 20005 3701 (202) 234-4433
I 378 l l
1 probabilities of rare events, is in the sense that it is 2 not a testable, repeatable -- you know, you just run a 1 / l 3 zillion trials and measure.
4 In the end, it relies on judgement, 5 subjectivity. That was the sense.
6 MEMBER APOSTOLAKIS: And I think you're -- you 7 know, that's a reasonable thing to say.
8 MR. CHAPIN: And I would also pick a little 9 bit with your conclusion that we don't look at these other 10 kinds of failures like human failures and isolation. I I 11 know of a number of PRA's not only from nuclear systems 12 but from other systems in which the failure of the 13 operator to explicitly do something is counted in the PRA, s
(. )
\/ 14 and one makes judgements as to what those numbers are.
15 So in the sense there's a box that says, you 16 know, success is here and one way to get to it is for the 17 hardware to fail, and another way to get to it is for the 18 operator to punch the wrong button, and a third box is 19 that the software that is supervising this process also 20 fails, I would assert that for engineering design purposes 21 -- and I think the committee senses the same -- one can 22 use that kind of an analysis with a box that says software 23 in it to assess these success states.
24 That's not different than human or these other
.O
(,/ 25 rare events.
NEAL R. GROSS COURT REPORTERS AND TRANSCRIBERS 1323 RHODE ISLAND AVE., N W.
(202) 234-4433 WASHINGTON, O C. 20005-3731 (202) 234-4433
379 l
1 MEMBER APOSTOLAKIS: ' No , but even'in that case l
.- 2' with a human, you look at isolated actions in the context 3 of what'the system is doing at that time and the 4- information the operator has at that time. So if you want 5 to talk about software-failure in a local sense, given 6 theseLthings have failed and so on, that's the came thing-l 7 then.
8 But I don't.think that's what most people mean 9' by software failure.
t 10 MR. CHAPIN: But I think that's what it says.
11 MEMBER APOSTOLAKIS: They mean the whole 12 program.
t 13 , MR. CHAPIN: It says, as in other PRA 14 confrontation, --
15 MEMBER APOSTOLAKIS: That's a different story,
- 16- yes. l 17 MR. CHAPIN: -- bounded estimates for software l 18 failure probabilities can be obtained by processes that 19 include valid random testing and expert judgement.
20 MEMBER APOSTOLAKIS: And this is general ;
21 enough to accommodate both points'of view. And as I say, 22 I subscribe to what David said,- but I can see how this can 23 be misinterpreted. Because I can't imagine that Leveson 24 disagreed with what David said, but then I can't imagine a A
g 25 lot of things.
NEAL R. GROSS COURT REPORTERS AND TRANSCRIBERS 1323 RHODE ISLAND AVE., N.W.
(202) 234-4433 WASHINGTON, D.C. 20005-3701 (202) 234-4433
380 1 MR. CHAPIN: For example, she has a lot of 73 2 concern with the idea that testing can be used to
/ \
l l 3 determine a number that is sufficiently accurate, for l
4 example.
5 MEMBER APOSTOLAKIS: And I agree with that. l 6 MEMBER MILLER: But your statement there did 7 not assume that.
8 MR. HILL: It may be the bounds are zero and 9 one.
10 CHAIRMAN SEALE: It sounds like a PRA.
11 MR. HILL: Sometimes that's all you can say.
12 MEMBER MILLER: You very carefully said a 13 bounded estimate.
('
(_/
) 14 MR. HILL: But a parenthetical comment. What 15 you're seeing in microcosm here is the debate we have 16 partly about language and terms.
17 MEMBER APOSTOLAKIS: Yes, I agree.
18 MR. HILL: I can't emphasize enough how 19 difficult this was for us.
20 MEMBER APOSTOLAKIS: The equivalent statement 21 for human would be to say this operator has a probability 22 of .2 of doing something wrong, period.
23 MR. HILL: Nobody can make that statement.
24 MEMBER APOSTOLAKIS: They would never make
(() 25 that statement.
NEAL R. GROSS COURT REPORTERS AND TRANSCRIBERS 1323 RHODE ISLAND AVE., N.W.
(202) 2344 433 WASHINGTON, D.C. 20005-3701 (202) 2344433
381 1 MR. HILL: Exactly.
,-~, 2 MEMBER APOSTOLAKIS: You will say, you know,
! l
\ /
'~'
3 in this context and these conditions and -- and I think 4 you have to do the same thing with the software. I don't 5 know.
6 Have I crossed the line, Mike, here? I said l
7 human. ,
I l
8 MR. MARKLEY: George, I'm not knowledgeable of l l
9 all the details of what you talked to them about, so you 10 have --
11 MEMBER APOSTOLAKIS: They didn't listen l
12 anyway.
13 (Laughter.)
)
/ 14 MR. MARKLEY: Can we keep this within bounds 15 here?
16 MEMBER APOSTOLAKIS: I'm sorry. We have to 17 stay away from human --
18 MEMBER MILLER: Maybe we'd like to go to other 19 members of the committee.
20 Bill, do you have a --
21 MEMBER SHACK: Just let me -- I want to follow 22 up a comment that Tom made just because -- maybe an 23 example again of communication. He made the comment that 24 he thought the reports came to the conclusion that
,r -y
( ,) 25 diversity was not the best way to assure software NEAL R. GROSS COURT REPORTERS AND TRANSCRIBERS 1323 RHODE ISLAND AVE., N.W.
(202) 234-4433 WASHINGTON, D C. 20005-3701 (202) 234 4433
382 1 reliability.
s 2 And I actually thought one of the things that t i
\)
3 struck me most about the report, considering the diversity 4 with Dr. Lewis here, it's been an extremely controversial 5 topic that you came out with a rather ringing endorsement 6 of diversity. And yet, Dr. Kress seems to get a different 7 reading.
8 MEMBER KRESS: I based that on the common mode 9 failure chapter which I think actually said that, that the 10 NRC programs on diversity were directed and ought to be 11 redirected more towards developing high reliability by 12 other ways because diversity's not the way to achieve this 13 reliability.
a
,k/ 14 MEMBER SHACK: See, I read the conclusion on 15 page 95 that says the U.S. NRC position with respect to 16 diversity is appropriate.
17 MEMBER MILLER: Maybe we should have the ;
l 18 committee speak to this issue. 1 l
19 MEMBER KRESS: Yeah, why don't we. l l
20 DR. CATTON: We have somebody over here. !
l 21 MR. CHAPIN: The diversity discussion about -- l l
22 I want to make sure you're -- and remind the committee l 23 that you have to be -- you have to read diversity. We had 24 so much trouble with it. We actually defined diversity in
( ,) 25 this chapter as design diversity and functional diversity NEAL R. GROSS COURT REPORTERS AND TRANSCRIBERS 1323 RHODE ISLAND AVE., N.W.
(202) 234-4433 WASHINGTON, D C. 20005-3701 (202) 234-4433
383 1 and-some other slices. ,
1 2 'So when we talk diversity, we're going to drag
- 3 you onto our turf. So again, we may be on one of these 4
4 issues.where'we're all using the word diversity and it
{ 5 means different things, i
j .6. Tom?
1-7 MEMBER'KRESS: Yes, my remarks were aimed at l; 8 the design diversity part of it.
i 9' MR. CHAPIN: And I think the conclusion there i
10 that you interpret is valid. We think that if in terms of 4,
4 l 11 design diversity, that it's very difficult to create a i
12 piece of software by design which does the same job which 7
s l 13 is diverse from some other piece of software.
14 MS. MITCHELL: And again, I think examples
} 15 help to clarify that. I would have very little confidence
! 16 that two pieces of software that were simply implemented i :
l 17 in different languages based on the same requirements and i-
!- 18 specifications were in fact functionally diverse. And so ,
{ 19 I think that the statement that diversity at the l.
j 20 functional level as opposed to diversity at the software i
21 level depending on what was -- what the index to diversity
{.
22 was.
23 So we didn't think using different software j 24 language constituted a diverse acceptance criteria or a 25 sufficiency criteria. ,
- NEAL R. GROSS
, COURT REPORTERS AND TRANSCRIBERS i l 1323 RHODE ISLAND AVE., N.W. I (202) 234 4433 WASHINGTON, D.C. 20005 3701 (202) 234-4433 l
)
384 1 MEMBER MILLER: That goes to the assumption 2 that the error is built into the requirements stage, 7.
( )
3 right?
4 MS. MITCHELL: I think that's something many 5 of us --
6 MEMBER MILLER: Which I agree with.
7 MS. MITCHELL: Yes, yes.
8 MR. CHAPIN: It may be. Or, I think --
9 MEMBER MILLER: Software diversity buys you 10 nothing.
11 MR. CHAPIN: Or that even with the same ,
1 l
12 requirements in which people are working in different 13 languages, there are certain hard spots in solving any
,s l t
(_/ 14 problem. And so people tend to trip over the same spots.
I 15 And I think this is what some of Nancy 16 Leveson's research tends to show, that if you give people 17 three points and ask them to write a program that gets 18 them to define a line, then they will forget the cases I 19 where they're all three the same point or they're all 20 three in the same straight line and which direction, 21 there's 180 , those sorts of things.
22 And so there are not only problems in 23 misspecifying the requirements, but there are hard spots 24 in actually applying the requirements which are difficult 7
( , 25 for humans to overcome. So the fact that I've had two NEAL R. GROSS l COURT REPORTERS AND TRANSCRIBERS 1323 RHODE ISLAND AVE., N W.
(202) 234-4433 WASHINGTON, D.C. 20005-3701 (202) 2344433 l
385 l
1 different programs write this in two different langeages 1
1
,_s 2 is not a good -- it's also not, on the other hand, proof l
/
t
\i l 3 that they're not diverse.
4 They may well be. But the -- to assume that l l
5 if I have two pieces of software written by two different 6 people in two different languages to the same functional 1
7 requirements that they will be diverse is probably not 8 correct. So in that sense, we don't believe diversity in 9 that sense is valid. ,
10 MEMBER MILLER: In that case, it's a matter of 11 there is a hard spot where you translate the requirements 12 into the software design requirements, and those same -- a 13 group of -- or three different groups of people will make 7s ;
( ./ 14 the same arror.
15 MR. CHAPIN: Yes. And then I think the part 16 that says we endorse diversity position goes back to what 17 Larry spoke about, which was that we believe that the NRC 18 has taken a position that says you can have these common 19 mode software failures, that we are not able to convince 20 ourselves that we can make systems sufficiently simple or 21 to test them in some way that we can absolutely conclude 22 that they're not susceptible to theses things.
23 So the NRC's position that one should do the 24 best he can to keep those kinds of things out and then
,,m (v ) 25 still impose some diverse means downstream of that to deal NEAL R. GROSS COURT REPORTERS AND TRANSCRIBERS 1323 RHODE ISLAND AVE., N.W.
(202) 234-4433 WASHINGTON, D.C. 20005-3701 (202) 234-4433
_ _ _ . . . _ . . _ _ . . _ . _ _ . ._. . . _ . ~ . - . . _ . _ _ . _ . _ _ _ _ _ . . . . _ . _ . . _ _ . _ .
1 4
386 I 1
~
j 1 with the consequences of that event, we think that's an 2' appropriate proce'ure. d
! 3- MEMBER MILLER: And then I go on -- you have a l'
s r 4 recommendation, it's number three.under common mode-5- software. 'It says the U.S. .NRC should revisit its
~
f -6 guidelines'in assessing whether adequate diversity exists. .j '
I i 7 MR. CHAPIN: Yes.
8 MEMBER MILLER: So that recommendation falls- ]
1 9 upon what you've just discussed here. l i
10 MR. CHAPIN: Then you read the next' couple of 11 sentences. For example,.we say that they should not place f 12 reliance'on different programming languages or different
- 13 design approaches that are intended to meet the same i
14 functional requirement, that the use_of different. teams is j i
15 not a good guarantee of diversity, and the use of 16 different vendor's equipment -- that is, nameplate i 17 diversity.
18 And we don't find any of those things 19 particularly convincing with regard to diversity. And 20 those are things which the NRC presently considers in one 21 form or another in their criteria. Some of them are 22 "anded" together. They say if you do this and this and 23 this and this, then we accept sufficient diversity.
24 And what we are saying is these particular 25 ones you should go back and take a hard look at'because NEAL R. GROSS COURT REPORTERS AND TRANSCRIBERS 1323 RHODE ISLAND AVE., N.W.
(202) 234 4433 WASHINGTON, D.C. 20005-3701 (202) 234-4433
367 1 we're not convinced they guarantee you the kind of 2 diversity that you're looking for.
3 DR. LEWIS: This is a deep questior.. I think 4 if I could be allowed a couple of words. It's a deep 5 question because diversity is not an end. It's a means to 6 an end. And it's a means to the end of reliability. And 7 you want diversity if you lack confidence in the 8 reliability of a single system.
9 And sometimes you should, and sometimes you 10 shouldn't. It depends very much on the quantitative 11 probability of failure of the system you're talking about.
12 I'm old enough to recall when jet engines were first 13 introduced for commercial passenger aircraft. The FAA did
/, N
& i
'/
14 not require that you carry a piston engine on one wing and 15 a jet engine on the other wing.
16 It would have been a really stupid thing to 17 do , but it would have satisfied the NRC, I think, because i 18 it would have been diverse. But instead, it turned out
)
19 that the jet engines which were unfamiliar -- you know, 20 they had been used in war time but hadn't been used j i
\
21 commercially -- turned out that they were simply very much 22 more reliable than piston engines and there was almost an 23 immediate complete switch to jet engines, and you hardly 24 over rur. into engine failures nowadays.
(~)h
(_ 25 So the reason is that the jet engines were 1
l NEAL R. GROSS COURT REPORTERS AND TRANSCRIBERS 1323 RHODE ISLAND AVE., N W.
(202) 234-4433 WASHINGTON, D.C. 20005-3701 (202) 234-4433
388 1 much more reliable. And the same thing is true for 2 diversity requirements. I have no problem whatever with a g S,
<j 3 single system non-diverse or even with redundancy in a 4 system -- single system without diversity if the systems 5 involved are more reliable than the alternatives.
6 But putting in a less reliable system for the 7 purpose of diversity which -- it went five years ago when 8 I was looking at these things, which NRC was pushing 9 toward, seemed to me, you know, ineffably dumb.
10 MR. CHAPIN: Yeah, and I agree with you.
11 There are simpler -- people talk about putting in, you 12 know, a gas turbine and a diesel. And to me, the most 13 likely failure in that mode is the guys who get the fuels i
'/
14 mixed up or they'll maintain it with a different technical 15 manual or that sort of thing.
16 But the NRC's position is a little more 17 sophisticated, and I think we ought to give them credit 18 for that. It's basically at the bottom of page 85 in the 19 report. And what it says is not if you want to have a 20 diverse system period, stop. What it says is put together 21 this package if you have a digital I&C system and assume 22 you have a common mode failure, and then look and see how 23 bad it is.
24 And if it's a problem for you and you can't
(_) 25 deal with it in the acceptable way, then, you know, you do NEAL R. GROSS COURT REPORTERS AND TRANSCRIBERS l
1323 RHODE ISLAND AVE., N.W.
l (202) 234-4433 WASHINGTON, D C. 20005-3701 (202) 234-4433
389 1 these diversity analyses and we want you to hunt for some s 2 diverse way to do it. And in a philosophical sense, they
/ i t
' '/
~
3 accept diversity of there's some other way to get the trip 4 or there's an operator action, that sort of thing.
5 So there -- I would argue that their position 6 which we endorse is not one which simply says we don't 7 have sufficient trust of the system; therefore, add 8 another one. What we're saying is don't assume the system 9 will always work. Assume it fails, and then explain to me 10 how you're going to deal with that.
11 And out of that will come your diversity. And 12 the reason that becomes important is what Larry talked 13 about, that it then becomes important to decide what is
,~
\
\~/ 14 adequate diversity in these systems. And that's when 15 you're led to the next level of the argument which has to 16 do with languages and nameplate diversity and trying to 17 design different software.
18 MR. DAMON: I think that in looking at 3, diversity, one of the things that we keep having 20 difficulty with is we start getting way, way far out 21 compared to what we've already thought and what we're 22 trying to improve. If you look at the present analog 23 systems that are in the plants today, you don't have 24 diversity.
rx (ss ) 25 There are common channels on both sides, NEAL R. GROSS COURT REPORTERS AND TRANSCRIBERS 1323 RHODE ISLAND AVE., N.W.
(202) 234 4433 WASHINGTON, D C. 20005-3701 (202) 234-4433
390 1 common equipment. Now if we go into putting in software
- 2 driven systems, you -- to get those in there because of p
\' )
3 common mode software failure, you start engaging the issue 4 of diversity.
5 But I think what you have to do is when you i
6 determine adequacy of diversity is how far you want to go.
7 What's really adequate? The staff has used NUREG-6303, l l
8 the work put out by the Lawrence Livermore Laboratory, 9 that is four echelons and -- it's a tremendous document 10 for advanced reactors defining diversity.
11 This goes back to complexity. How far would 12 you want to go in 6303 to define event diversity for a 13 relatively simple system comparing it, of course, to the p
E./ 14 thing that you're replacing which is already not diverse 15 inherently?
16 It's a judgement call. My belief is that you 17 don't have to go very far in diversity if you have code 18 that's pretty simple, modular; a system that's simple, not 19 modulating, but just engages and goes to completion; there 20 is redundancy; and then there is the operator standing 21 behind all of that to make sure everything comes off all 22 right.
23 If you bring diversity out of the theoretical 24 and academic back to the practical, it doesn't become
/"N
( ) 25 quite a big problem. But if you let diversity go too far, x
HEAL R. GROSS COURT REPORTERS AND TRANSCRIBERS 1323 RHODE ISLAND AVE., N W.
(202) 234-4433 WASHINGTON, D C. 20005-3701 (202) 234-4433
391 1 you have the tail wagging the dog and you're going to end 73 2 up with something that's theoretically perfect and
(
%) 3 practically useless.
4 And there's that judgement thing that has to 5 come into play here. We've got to keep looking at this 6 diversity issue in context of what's being replaced.
7 DR. CATTON: Common sense.
8 MEMBER MILLER: Do we have other questions on
'9 diversity? Everybody happy? Other questions by other 10 members?
11 I have a question I'm going to hold near the 12 end here that I really want to have this committee speak 13 to, but other questions of -- I'd like to go to the
( )
N/ 14 commercial -- or the case by case licensing process. And 15 I knew I'd get a smile from Mr. Damon over there because 16 I'm certain he had part in writing this.
17 You have a conclusion and recommendation. The 18 conclusion basically says the U.S. NRC now has sufficient 19 experience with safety related systems that they can 20 establish a generically applicable regulatory regime that 21 would govern review and approval of such upgrades.
22 And recommendation six I'll read verbatim.
23 "The U.S. NRC should revisit the systems level issue in 24 Generic Letter 95-02 which endorses EPRI report TR-(q ,/ 25 1002348. The committee strongly endorses maintaining and NEAL R. GROSS COURT REPORTERS AND TRANSCRIBERS 1323 RHODE ISLAND AVE., N.W.
(202) 234-4433 WASHINGTON. O C. 20005 3701 (202) 234 4433
392 1 formalizing distinct from major and minor system upgrades g3 2 containing digital technology."
N] And I could go on and -- basically the issue 3
4 is the NRC's position is the system's defined by the 5 system you're installing and the guideline says the system 6 is -- the system in which this digital system is being 1
l 7 installed.
l 8 I'd like to have the committee maybe speak 9 more eloquently or more -- elaborate on your pcsition )
10 there.
l 11 MR. CHAPIN: Okay, the other thing that enters !
l l
12 into this is we had a number of discussions with the staff l
13 members who came and talked to us. And our understanding x- 14 is that -- and we're going to talk to them again this 15 afternoon after we talk to you, and maybe at some point 16 they would like to comment on this as well.
17 What I think we saw was that the staff 18 position changed somewhat over the period of the time that 19 we were talking to them and we were conducting the study 20 over the period of a year and a half. And I think at the 21 outset, our concern was that there was very little 22 distinction given to the major or minorness of the 23 application.
24 And I think by the time we came to the end of n
( ,) 25 the study and our final interactions with the staff, that NEAl. R. GROSS COURT REPORTERS AND TRANSCRIBERS 1323 RHODE ISLAND AVE., N.W.
(202) 234-4433 WASHINGTON, D C. 20005-3701 (202) 234-4433
393 1 the staff was taking a -- what I would consider a more y 2 pragmatic point of view, and that they regarded this
/
3 system level distinction to be more applied to system 4 being a major system defined as something like safeguards 5 actuation or reactor protection system.
6 so I think there has been a migration where we 7 have come fairly close together. And I think Larry has 8 been deeply involved in this. I'd be glad to have him 9 make some remarks. And Jerry Wermiel is sitting back 10 there, and if he wants to say something at this point, you 11 know, --
12 MEMBER MILLER: I'd be happy to listen to both 13 of you at once.
/
- - 14 MR. CHAPIN: Okay, Larry?
15 CHAIRMAN SEALE: A capella?
16 MR. DAMON: Well, this particular chapter was 17 a hard one. In the second phase, we broke up into groups 18 and each of us took one of the chapters. I originally was 19 writing this chapter up, and we went through quite a bit 20 of difficulty on it. And then Jim Curtiss took over and 21 rephrased an awful lot of what I had originally written, i 1
22 and it came to something which is in front of us here l 23 which is acceptable.
24 It's a committee document. It's a committee
!o)'
25 consensus. This particular issue on a systems level goes l
NEAL R. GROSS l COURT REPORTERS AND TRANSCRIBERS ;
1323 RHODE ISLAND AVE., N W.
(202) 234-4433 WASHINGTON, D C. 20005 3701 (202) 234 4433 i
- .. - . . _ . _ _ . . . _ . _ . _ - .__ _ . . ~ . _ . _ . _ _ _ . _ . _ _ _ _ _ _ . _ _ _ . = . _ . _. . _ . .
j- 394 1 back to the -- what's determined to be the" system that.
1 2 you're dealing with. And the exceptions taken to the EPRI
{- 3~ position is a subject of continuing controversy. What 4' really defines the. system level that we're looking at, the 5 one that's-being replaced or the one that goes to complete 6 satisfaction of the' function?
'7 It is not' resolved. And in the draft review 8 plan,.the predecision draft, it still comes out that it is 9 ont fully resolved. Our recommendation'here is that this 10 has ramifications beyond just_ case by case licensing.
11 This has ramifications into the areas of commercial grade' 12 dedication.
~
13 And an awful lot of what we're going to be 14 facing is how hard or how cleanly we can come to 15 resolution to preclude the need for case by case licensing 16 depends upon redefining this term or this systems level 17 application.
18 The way that we've looked at case by case 19 licensing is you cannot use case by case licensing as an
. 20 excuse to have fuzzy acceptance standards and work them 21 out on a case by case basis because the industry will say 22 look, I want to know what I'm getting into before I get 23 into it.
24 So case by case licensing cannot --
it has to 25 be used as the last alternative; the first alternative NEAL R. GROSS COURT REPORTERS AND TRANSCRIBERS 1323 RHODE ISU,ND AVE., N.W.
(202) 234-4433 WASHINGTON, D.C 20005-3701 (202) 234-4433
395 1 being clear and crisp guidelines. One of the difficulties 2 we had on human factors is the clear and crisp guidelines 7_
' '-~' ')
3 are never there. It's always empirical and that's a 4 judgement call, se it's a hard thing to deal with.
5 So if we -- we felt essentially this. We 6 heard both from the EPRI standpoint and from the staff's 7 standpoint on systems level and the dichotomy that it 8 presents. Our feeling was you're making this 9 recommendation. And Jim, correct me if I'm wrong. If 10 they can get in and work this thing out, clearly you 11 probably are never going to have to worry about case by 12 case licensing again.
13 And that will mean that 50.59 is back doing k/ 14 the job that it should be doing. And that is really 15 important because 50.59 transcends digital systems. It 16 goes to the basic ability to change without every time 17 having to go to docket.
18 MEMBER MILLER: So you're saying that working 19 this issue out, --
20 MR. DAMON: Very important.
21 MEMBER MILLER: -- which ironically is one of 22 the first things I got involved in when I got in this 23 committee, is a very critical issue --
24 MR. DAMON: Very critical.
,r x
( ,) 25 MEMBER MILLER: -- with the use of digital NEAL R. GROSS COURT REPORTERS AND TRANSCRIBERS 1323 RHODE ISLAND AVE., N W.
(202) 234-4433 WASHINGTON, D.C. 20005-3701 (202) 234-4433
396 1 both for upgrades and also commercial off the shelf and so l
l 2 forth and we should place a high priority on it? l
!, s.
\'~') 3 MR. DAMON: Yeah, it's a very high priority.
4 And we have to also look at, I mean, commercial off the i
5 shelf systems by what's occurred. Commercial off the l
6 shelf equipment dedicated to the nuclear industry is 1 I
7 probably going to be what's necessary. Because I don't 4 1
1 8 think manufacturers for the limited market are going to I l
9 get into all of the things that we've got in front of us 10 here. l l
l 11 So how this impacts COTS is very important.
1 12 MEMBER MILLER: He said Jim had some input on l l
13 this. l
(- '
(_/ 14 Jim, can I put you on the spot? Do you want ;
15 to make further comments maybe more from the regulatory i 1
16 viewpoint?
17 MR. CURTISS: Well, I share everything that 18 Larry has said. This was a difficult chapter for us to 19 write. It reflected, I think, a lot of input of the full 20 committee because this does cut across several of the 21 issues and we've labeled it as a strategic issue for that 22 reason.
23 I think there were two or three things that we 24 were trying to get at here in this chapter, all of which
(,)
25 Larry has touched upon. We believe it's appropriate.
NEAL R. GROSS COURT REPORTERS AND TRANSCRIBERS 1323 RHODE ISLAND AVE., N.W.
(202) 234-4433 WASHINGTON, D C. 20005-3701 (202) 2344433
397 1 We've reached the time where we can move to the point
/
,- s
\
2 where the process has largely been focused on a case by
\
'~'
l 3 case review to know incorporate that experience in some 4 broader, more generically applicable set of standards.
5 And the SRP would be the vehicle for doing 6 that. We did not endorse the substance of the SRP in this 7 review, 8 Secondly, I think we found historically that, 9 particularly focusing on the 50.59 question, that the i 10 early decision to declare all digital upgrades in the i
11 safety arena to be unresolved safety questions was in I 1
12 retrospect probably an overreaction.
~
13 I think we'd all agree, including the staff,
/'_ T
'M 14 and correct us if we're wrong, that it's an appropriate 15 framework within which utilities can evaluate upgrading to 16 digital, but it has its own well defined criteria for 17 defining whether an issue poses a USQ and you need to go 18 through that evaluation. That's an appropriate vehicle 19 today.
20 The issue with respect to the systems level !
21 analysis is set out in the document. I don't have 22 anything to add to that. It's described. I think it's an 23 issue that the staff needs to come to grips with. Some j 24 confusion I think early on in the process.
r~~'T
(,) 25 In fact, it preceded my participation in phase NEAL R. GROSS COURT REPORTERS AND TRANSCRIBERS i 1323 RHODE ISLAND AVE., N.W.
(202) 234-4433 WASHINGTON, D.C. 20005-3701 (202) 234-4433 1
398 1 two as to whether the result that was reached in reviewing
,_s 2 EPRI TR-1002348 was legally required or simply a matter of
( )
3 policy discretion from a technical standpoint. We were 4 advised in phase two that it was not a legal result that 5 50.59 dictated in reviewing the EPRI document.
6 And I think that leaves it in the staff's 7 technical corner to evaluate that issue. I think we'll 8 hear perhaps more this afternoon as to where they are on 9 that issue. But I think Larry has generally accurately 10 characterized the gist of this section.
11 MEMBER MILLER: What time is that meeting 12 going to be? Probably won't let me go.
13 MR. CHAPIN: 2:00 this afternoon. Jerry was
)
(
x_/ 14 supposed to get with you.
15 MEMBER MILLER: Actually, this committee will 16 meet with the staff on this and many other issues on April 17 17th and 18th. So I'm certain we'll --
18 MEMBER 7; OSTOLAKIS : Subcommittee or full i
19 committee? I l
20 MEMBER MILLER: I'm sorry, subcommittee. I'm )
1 21 sorry. We'll have a subcommittee meeting on the 17th and l l
22 18th, and I'm certain this will be -- I know from the I l l 23 committee chair's perspective this will certainly be an 24 issue on my front burner since this ras the first chapter 73 i,, ,) 25 I read in the report -- kind of read backwards, NEAL R. GROSS COURT REPORTERS AND TRANSCRIBERS 1323 RHODE ISLAND AVE., N.W.
(202) 234-4433 WASHINGTON, D C. 20005-3701 (202) 234 4433
399 l 1 So I'd be anxious to hear what comes out of p_ 2 the meeting this afternoon.
1! )
\~ 3 MR. CHAPIN: Yes, we worked hard on this 4 chapter. We struggled to come to consensus. And as I 5 say, I think part of the things that happened in this 6 study was that the staff's position changed somewhat along 7 the line -- along the time line.
8 So, you know, this reflects where we are.
9 We're interested in how they're going to come out too.
10 But this is a study with a finite length, and so we took 11 it to where we were and we stopped.
12 MEMBER MILLER: Do we have any additional 13 questions from committee members?
)
(_/ 14 VICE CHAIRMAN POWERS: Maybe I could ask just 15 a question on the research programs.
16 You comment several points in the research 17 programs and characterize it generally. You found the 18 research programs wanting. But I got the impression that 19 you did not take an attempt to define a research program 20 for the NRC. You address specific ones and made 21 suggestions on those.
22 I wonder if the committee has some views other 23 than those or in addition to those that are expressed in 24 the document on what kind of research NRC should be doing n
( ,) 25 as opposed to the kind of research that the licensees and NEAL R. GROSS COURT REPORTERS AND TRANSCRIBERS 1323 RHODE ISLAND AVE., N W.
(202) 234-4433 WASHINGTON, D.C. 20005-3701 (202) 234 4433
400 1 the applicants should be doing.
,_ 2 MR. CHAPIN: Jim White, do you want to comment 3 on that?
4 MR. WHITE: Yes, I will.
5 One thing that we struggled with on this 6 chapter is trying to determine what is the strategic goal 7 of the NRC research in these areas. Since we are not 8 certain we understand the goal, it's hard for us to 9 recommend in view of all the funding constraints that we 10 think we understand what the best research plan would be.
11 Certain elements it would have is certainly 12 trying to leverage the scarce funds that are likely to be 13 available. We think the research should be long term
$ \
K_) 14 enough that it would establish some credibility, 15 particularly in the software and human factors technical 16 communities.
17 That is to say multi-year programs much like, ,
18 l
say, the pressurized -- the pressure vessel work that the ;
I 19 NRC has funded for, oh, a decade or so and now is regarded l l
20 as an important research program.
21 The way the work is done now where most j
i 22 frequently you need an endorsement from NRR for the RES l 23 work makes it a little difficult to have these long term i i
24 plans which -- I mean long term projects which establish l
()
25 some technical credibility.
i NEAL R. GROSS COURT REPORTERS AND TRANSCRIBERS 1323 RHODJ ? AND AVE., N.W.
(202) 234-4433 WASHINGTON, D.C. 2000G3701 (202) 234-4433
401 1 So, you know, if the goal of the research s 2 program is to prevent mistakes in the application of
\
'~'
3 digital I&C to nuclear power, you can formulate, I think, 4 a pretty good program. But if the goal is to help the 5 industry move forward in a responsible fashion, then the 6 research program will have a different flavor, if you 7 understand what I mean.
8 So it was hard for us to do that because I 9 think we need some more feedback on what should be the 10 strategic goal of the research.
11 MR. CHAPIN: One thing that was of interest to 12 me in this area was that in recent NRC strategic issues 13 affecting and reflecting NRC's, you know, preliminary
. i N/ 14 review of what had happened, one of the things that we did 15 was we struggled in our report-and with some of the review 16 comments we got about the issue of confirmatory versus i
17 exploratory research and that we needed -- you know, we i 18 had a hard -- we were cautioned that we could not tell the l 19 NRC research staff that they should be looking long term 20 and that they should be doing more exploratory work 21 because that was not in their charter.
22 I copied this out of -- or one of the bullets 23 out of the thing is that the research program should 24 include elements of both confirmatory and exploratory
(\
( ,) 25 research balanced so that both currently and potentially NEAL R. GROSS COURT REPORTERS AND TRANSCRIBERS 1323 RHODE ISLAND AVE., N.W.
(202) 234-4433 WASHINGTON, D.C. 20005-3701 (202) 234-4433
402 1 emerging issues are addressed. I think that sounds real g 2 good to me personally, particularly in the context of this i ;,
N. s '
3 report.
4 So if there is a little bit of a wind that's 1
5 saying to look at something other than the immediate i 6 problems that NRR has, I think that's a real benefit to 7 the research program.
l 8 VICE CHAIRMAN POWERS: I think you don't want l l
9 to assume that simply because of the user needs approach l 10 and NRR asking for things that that necessarily translates 11 into short term much of the materials research that you l
12 lauded was in fact part of a users need request. It's 13 really short term versus long term, not why it's short
/,%
14 term, and where the request comes from that's the concern.
1 15 And I think we understand that, and we I 16 understand the difference between exploratory and 17 confirmatory. One other area that you commented on that 18 relates to research is you were generally concerned about 19 the participation in which you called the public forum.
l 20 And that was a puzzlement to me because j l
21 certainly when the staff comes before us, they come before l 22 us with credentials that speak to participating in l 23 enumerable standard setting committees and things like l 24 that and seem to have quite good credentials with learned
(- )
(, 25 societies and they come with contractors that have reams NEAL R. GROSS COURT REPORTERS AND TRANSCRIBERS 1323 RHODE ISLAND AVE., N.W.
(202) 234 4433 WASHINGTON, D.C. 20005-3701 (202) 234-4433
403 1 of publications.
7- 2 And I struggled with what does it take to I l
\ /
3 satisfy you in that area.
4 (Laughter.)
5 MR. CHAPIN: Well, in this area -- let me say 6 that when I sat down in this, I had the same sort of 7 reaction when we started out. And I can particularly send 8 in my other colleagues who are not so nuclear connected 9 speak to that, but it was striking that the literature 10 that is being used in the software community and in the 11 human factors community and in the digital computer design 12 community and the design of robust systems and that sort 13 of thing is -- none of that cites the NRC's work.
rN l
(
E/ 14 Many of the people in that area don't have any 15 knowledge of what the NRC is doing. There seems to be 16 little explicit participation between, you know, the heavy 17 hitting journals and the meetings and the like. So that ;
l 18 the message I thought came through loud and clear to the 1 19 committee, and in my own experience -- you know, I've been 20 in this business over 30 years -- somewhat reflects that.
21 We are somewhat insulated. And let me get 22 John and Christine who both come to this with, you know, 23 high powered credentials in the digital I&C area, but not 24 particularly connected to the nuclear area. Let me get
(~\
(,) 25 them to comment on that.
N EAl. R. GROSS COURT REPORTERS AND TRANSCRIBERS 1323 RHODE ISLAND AVE., N.W.
(202) 234-4433 WASHINGTON, D.C. 20005-3701 (202) 234-4433
404 !
1- .MR. GANNON: There are a large number of 2 reports that were presented to us as representative of 3 research results that had been. commissioned by -- and I 4 won't get the letters right because I'm not part of the 5 community -- whloever. Those reports don't turn into 6 presentations at prestigious conferences. They don't turn 7 into journal publications.
8 And as a result, they don't have much impact 9 on the way that software is developed.
10 MR. CHAPIN: , Yeah, the peer review -- would 11 you comment on the peer review, because that was something i
12 we talked about a lot. )
i 13 MR. GANNON: Yeah, and the -- for me, a
.14 technical report -- they looked like technical reports to 15 me which is -- if I want to put out a technical report or 16 one of my students do, we just get a cover and we put the i
17 cover on and put a technical report number on it and maybe 18 somebody else reads it and maybe they. don't.
19 VICE CHAIRMAN POWERS: Certainly the ones that !
20 comes out with the national laboratories aren't so 21 cavalierly done. !
22 MR. GANNON: Yes, I certainly won't comment on 23 that. But they tend -- a large, thick document that isn't 24 boiled down and isn't peer reviewed by other people who 25 are doing research on the same topic just isn't going to NEAL R. GROSS COURT REPORTERS AND TRANSCRIBERS 1323 RHODE ISLAND AVE., N.W.
(202) 234-4433 WASHINGTON, D.C. 20005-3701 (202) 234-4433
405 1 have the impact that you will if you publish in, say, 7- 2 Transactions on Software Engineering or the ACM Journal on
(' -) 3 Software Engineering and Software Methods.
4 Or if you present you work at the 5 International Conference on Software Engineering. These 6 pieces aren't having a large impact, I think.
7 MEMBER APOSTOLAKIS: One quick question 8 though. One issue is whether these documents have an 9 impact. Another issue is if you look at the list of 10 references of these documents, are you satisfied that they 11 are representative of the current thinking?
12 MR. CHAPIN: Let me speak as someone who I 13 think sort of had an epiphany or whose eyes were opened.
,\
'w. 14 I've looked at many NUREG's and national lab reports and 15 documents produced by the NRC. They do have references.
16 But if you look at the references in the light of what 17 I've seen in the last 18 months, it's pretty insular and 18 it's pretty isolated, and it's not as broad a net as most 19 of these other people would have thrown in their own work.
20 Personal opinion.
21 Christine?
22 MS. MITCHELL: I think I agree with everything 23 John and Doug said, that if you're talking about research, 24 there is -- or there are forums both nationally and
/)N
(, 25 internationally. And a peer review -- a public peer NEAL R. GROSS COURT REPORTERS AND TRANSCRIBERS 1323 RHODE ISLAND AVE., N.W.
(202) 234-4433 WASHINGTON, D.C. 20005-3701 (202) 234-4433
406 1 review, not where you hire people, bring them together, s 2 and then you yourself synthesize their opinions, it is
! \
3 really a different process.
4 So for example, during the course of this 5 committee, the human-system interaction guidelines and --
6 they could tell you what the number was -- appeared in 7 press and was available for comment for 12 months. And I 8 don't know that there was anybody on this committee 9 besides me that read it.
10 And I said Doug, has your staff commented on 11 this, have they read it, do they know it's out there? And 12 I guess I'm concerned if the industry is not reading those p_
13 and I know for a fact that the research community as
(\~ ') 14 defined certainly by the academics and a lot of other 15 government agencies isn't reading them, who is?
16 And so Mike DeWalt is not here, but he was on 17 one of the software engineering review committees, and he 18 said well, I certainly didn't endorse this. But, of 19 course, in that review process, that didn't come out.
20 What was said is, it was peer reviewed. It didn't have a 21 vote as to how many people -- I mean, I don't know what 22 that process even should look like.
23 So when you say that the national lab process 24 isn't the same as John's process, well it costs more money p
(_,) 25 and it involves more people; but I'm not sure you get more NEAL R. GROSS COURT REPORTERS AND TRANSCRIBERS 1323 RHODE ISLAND AVE., N.W.
(202) 234-4433 WASHINGTON, D C. 20005-3701 (202) 2344 433
407 1 for your dollars. I'm not sure you don't either. I'm
-- 2 just not -- no one has convinced me that that results in 3 qualitatively better opinions.
4 I mean, I know how a public review process 5 works.
6 VICE CHAIRMAN POWERS: Most of the committee 7 here knows that I do not share any of your opinions on 8 that.
9 MEMBER APOSTOLAKIS: No, but I think there is 10 an important issue here though. You talked about the 11 impact that this work or that these NUREG's could have on 12 the state of the art. And it seems to me that that's not p, 13 really important because this is not a research agency, t )
'- 14 This is a regulatory agency.
15 What's more important in my opinion is that 16 these documents are current -- that they reflect the 17 current state of the art. That when a NUREG is issued 18 that maybe is used in the regulations and you look at the 19 list of references, you say yes, these people do know what 20 is going on out there.
21 So I think that's really the important thing.
22 Not that the NUREG will have an impact on the field 23 because that's not their job. Their job is not to advance 24 the state of the art.
's_) 25 MR. CHAPIN: No, but I think the impact --
NEAL R. GROSS COURT REPORTERS AND TRANSCRIBERS 1323 RHODE ISLAND AVE., N W.
(202) 234-4433 WASHINGTON, D C. 20005-3701 (202) 234-4433
408 1 it's got to go both ways. What I said in my opening 7-2 remarks in the communication, for example, was that you t i i
~
3 need to put before the larger community the issues and 4 constraints that you have in the nuclear business so that I l
5 you can get people like Nancy Leveson and John Gannon and l l
6 Joanne Dugan who are all experts in the technology which 7 you're supporting to be focused on and thinking about your 8 problems.
1 9 And from my perspective, ladies and gentlemen, I 10 we have not done a good job of cluing them in to what our l i
11 problems are. The other thing is, it brings their '
12 expertise to bare. And right now, I don't think we're 13 doing a good job of that either. So I think the benefits
- i kl 14 of participating in this wider forum are it puts some of 15 our hard problems up in front of a different group of 16 people, and maybe some other researchers would be 17 interested.
18 And the other thing is, we might get some
! 19 answers from those guys. And right now, I think we're 20 doing a little too much of talking to ourselves.
21 MEMBER APOSTOLAKIS: Proba>ly true.
l l 22 MR. DAMON: In this chapter in which we 23 address the research issue is one of the adequacy of the 24 technical infrastructure. And I don't want to contend
( ,) 25 that necessarily research failed to give the staff the NEAL R. GROSS COURT REPORTERS AND TRANSCRIBERS 1323 RHODE ISLAND AVE., N.W.
(202) 234-4433 WASHINGTON, D.C. 20005 3701 (202) 234-4433
409 1 heads up that was required, but we have to look at what
,s 2 has happened.
t ) l 3 For 20 years, we've been using digital 4 technology and humming along doing a lot of IEEE upgrades 5 and standards and things like that. NUREG's were written.
6 And u..'n up to the point in time that they wanted to 7 implement the reactor protection system -- digital system 8 at Haddam Neck, everybody thought everything was okay.
9 As a matter of fact, the staff's own onsite 10 reviewer thought everything was going along okay. The 1
11 equipment was already approved by complying with all of l 1
12 the standards. Everything looked all right. And then 13 suddenly, out of the clear blue sky was the idea whoa, l
/}
\- 14 wait a minute; I don't think the standards are enough 15 anymore.
16 There was no heads up for that surprise. That I i
17 'was a real showstopper. Essentially the industry hasn't l 18 implemented a reactor protection system upgrade for four 19 or five years while we resolved this thing and worked it 20 out and this committee did their thing and the staff did 21 their thing on the Standard Review Plan.
22 The question is, in an infrastructure that's 23 technically robust and put together including the research 24 element, why did this happen? Why wasn't there a heads up p
( j 25 that could have prevented the jarring effect of what NEAL R. GROSS COURT REPORTERS AND TRANSCRIBERS 1323 RHODE ISLAND AVE., N.W.
(202) 234-4433 WASHINGTON, D C. 20005-3701 (202) 234-4433
. - _ . - . . _ . - - . - . . ~ _ = . - - - . ~ . - - . . _ . - . . . . - . - . . . . . . _ . . .
i
! 410 I'
1 occurred'in the industry just after Haddam heck and 2 through Zion and sequoyah and all that?
>O 3 Our feeling, I think, was that if t here was a 4 separate check on it through the research group that was 5 not just doing investigations fcr NRR but was looking at l
l 1 6 things-from a different perspective, the staff may have 7 had a heads up. The industry might have had something i
8 tlrc could have headed all of this off.
9 We can't look into it -- was it a. management 10 structure, a research problem, a staff technical awareness 11 problem, an industry problem, the lethargy of developing <
12 standards? Perhaps it's a mix of all of these things, ,
13 But this has been a very jarring experience that's cost an 14 awful lot of people and an awful lot of time and an awful 15 lot of money that should have been able to be resolved ;
16 much more elegantly than this indicates.
t 17 And possibly, research may have he.'. pad t.
18 Dr. Forest once told me that research is nothing tare 19 than a continuous sanity check. Looking heads up, looking i 20 forward. So maybe there was that kind of thing.
21 MEMBER MILLER: You know, this has been a very 22 interesting morning. Unfortunately, the morning is moving 23 into lunch time, and I'm being encouraged to bring this 24 all to a close. And I'll do that by first thanking the O
V 25 committee for your time in coming to this meeting and your NEAL R. GROSS COURT REPORTERS AND TRANSCRIBERS 1323 RHODE ISLAND AVE., N.W.
(202) 234 4433 WASHINGTON, D.C. 20005-3701 (202) 234-4433
411 1 putting up with our questions and us putting up with your 3 2 answers, I guess.
L ) l 3 I think it's been very illuminating for all of 4 us. I think it's clarified a couple of key points in the 5 report. We could probably spend a few more hours, but I
6 hopefully you'll spend hours with the staff and then we'll l l
7 get that feedback and we'll get it all together.
1 8 I'd also like to thank Dr. Lewis for coming 9 and his comments were very illuminating and stimulating to l
10 the committee as well as this committee. :
l l
11 And with that, I would say it's time -- unless 12 there's any more for the good of the order, I'd say it's 13 time for lunch. And the ACRS will reconvene -- I'll turn
! \ \
L 1
'/ 14 it over to the Chairman, I'm sorry. Turn it over to I l
15 Chairman Seale and he'll reconvene at his leisure. l l
16 CHAIRMAN SEALE: We'll reconvene at a quarter 17 of 2:00 since many members have things to do.
18 (Whereupon, the proceedings recessed for lunch 19 at 12:46 p.m.)
20 21 22 23 24 i .C')
(_) 25 NEAL R. GROSS COURT REPORTERS AND TRANSCRIBERS 1323 RHODE ISLAND AVE., N W.
(202) 234-4433 WASHINGTON, D.C. 20005-3701 (202) 234-4433
412 1 A-F-T-E-R-N-0-0-N P-R-O-C-E-E-D-I-N-G-S f; 2 (1:47 p.m.)
'w/
3 CHAIRMAN SEALE: Well, it's time for us to get 4 started again. Our presentation this afternoon is on the 5 Department of Energy Program, a proposal for tritium 6 production and Dr. Powers is our subcommittee chair on 7 this one, so Dana, I'll drop it in your lap. j 8 MEMBER POWERS: Thank you, Bob. We're going 9 to hear, I guess, first an overview and then some details 10 about the Department of Energy's efforts to develop 11 technology for producing tritium for the nuclear deterrent 12 force. They have several options for doing that, one of )
- 13 which involves use of the commercial power.
s ;
14 The reactor fuel subcommittee has had a chance 15 to look at some of the draft documentation on this 16 proposals to use the commercial nuclear reactor and the 17 committee has before them some comments from the committee 18 on the lead test assembly that's being proposed which I 19 guess will be described to us in a fair amount of detail 20 today.
21 I think this is not advanced to the point that 22 the staff is going to comment on it, so this is really an 23 information briefing, not one that we will prepare a 24 letter.
f'~T
( ,/ 25 I guess Steve, are you going to give us the !
NEAL R. GROSS COURT REPORTERS AND TRANSCRIBERS 1323 RHODE ISLAND AVE., N.W. I (202) 234-4433 WASHINGTON, D C. 20005-3701 (202) 234-4433 1
413 1 overview and other speakers will follow you.
,~
, 2 Our first speaker is Steve Schinki and he is
'~'] 3 from the Department of Energy and I can never remember the 4 acronym, CLWR, it's Civilian Light Water Reactor Project 5 Office, that's what it is. Anyway, he's out of Defense 6 Programs.
7 MR. SOIG NA T Good afternoon, gentlemen. I'm 8 Steve Schinki. As Dana said, I'm the Director of the 9 Office of Commercial Light Water Reactor Production within 10 the Department's Tritium Project Office. This is an 11 office that was set up, I guess, a year ago November 12 following, just preceding actually the Secretary of 13 Energy's decision on a long time tritium supply source.
r~N f I
(_/ 14 I wanted to give you just a five minute 15 overview of how we got where we are and then Dr. Jerry 16 Ethridge who is sitting behind you from Pacific Northwest l 17 National . Labs is our project manager at the laboratory and 18 responsible for the design of the tritium producing 19 burnable absorber rods, will give you a technical briefing 20 on the rods and the lead test assembly process.
l 21 First of all, I'm not going to talk to you 22 about the physics of tritium. You people know the physics i
23 of tritium much better than I do. The only point I wanted 24 to make is that every weapon in the United States s-
) 25 stockpile requires tritium to function as designed and -- i i
NEAL R. GROSS COURT REPORTERS AND TRANSCRIBERS 1323 RHODE ISLAND AVE., N.W.
(202) 234-4433 WASHINGTON, D C. 20005-3701 (202) 234-4433
414 1 1
1 MEMBER POWERS: We don't have any unboosted
,_s 2 weapons?
- \
( /
3 MR. SOHINKI: What was that?
4 MEMBER POWERS: We don't have any unbooster 5 weapons?
6 MR. SOHINKI: That's correct. Obviously Fat 7 Man and Little Boy were unboosted, but one of the things 8 that tritium allows you to do is put a larger yield in a 9 smaller package. And so all of the current weapons use 10 tritium.
11 We made tritium down at the Savannah River ;
12 site for several decades in heavy water reactors down s
13 there. The last of those were shut down in 1988 for
(-
(_) 14 safety and environmental reasons. And since then we have 15 not made tritium in the United States.
16 We have been living off of tritium that has l
17 been mined from weapons that have been retired from the l 18 weapons stockpile as a result of arms limitation l 19 agreements, but because it decays, obviously, we haven't 20 figured out a way to stop that yet.
21 MEMBER POWERS: Steve, your reactor capability 22 at Savannah River is shut down, I think it's worthwhile 23 pointing out to the committee that you have a state-of-24 the-art tritium handling facility and recovery facility f3 l
V) 25 there. ;
NEAl. R. GROSS COURT REPORTERS AND TRANSCRIBERS 1323 RHODE ISLAND AVE., N.W.
(202) 234-4433 WASHINGTON, D C. 20005-3701 (202) 234-4433
415 1 MR. SOHINKI: That's right. The replacement
,7 2 tritium facility opened, I believe, in 1992 and that is a
.i i
/
3 state-of-the-art facility for recycle of tritium.
4 MEMLER POWERS: If you come before us again I 5 think this committee would be just stunned to see the ;
6 technology that's applied there.
7 MR. SOHINKI: It's fairly sophisticated. So 8 we have been relying on recycled tritium. The President l 9 every year issues a classified dccument called " Nuclear l 10 Weapons Stockpile Plan" with an associated Presidential 11 decision directive. It contains the types of weapons and 12 the numbers of each type that will be in the stockpile for 13 the following, at least, five years and then each year
,m I \
N. / 14 that's modified according to the current requirements 15 after discussions with both the Department of Energy and 16 the Department of Defense.
17 The current Nuclear Weapons Stockpile Plan and 18 Presidential decision directive require the Department to 19 have a source of tritium as early as 2005 and if we're 20 talking 2005, it would be the commercial reactor option.
21 CHAIRMAN SEALE: Excuse me, is an element of 22 this plan that if a reactor is used it will be a 23 commercial reactor?
24 MR. SOHINKI: Yes, the dual past strategy was
[">
( ,
) 25 either com.e cial reactor or an accelerator. Now one NEAL R. GROSS COURT REPORTERS AND TRANSCRIBERS 1323 RHODE ISLAND AVE., N.W.
(202) 234-4433 WASHINGTON, D.C. 20005 3701 (202) 234-4433
. - - . ~ . - . ~ . . - ~ . - . ~ . . , . . - - ~ . - . - . - . - . ~ . . . . . - . - - . . . - . - . . - . . . .
'416 ,
1- caveat to that: for the past almost a year, the 2 Department has done a reexamination of the fast flux test lO 3. facility out at'Hanford as a result of'some political 4 intrigue that we need not go:into. We had originally --
l 5' we had done an environmental impact statement on tritium 6 alternatives which actually is a good segue'into this next.
7 slide which analyzed a number of alternatives for tritium t
c 8 production including the accelerator, commercial reactors, l
l 9 a new modular high temperature gas reactor, new. heavy i
10 water reactor and a new advanced light water reactor. In
- 11 fact, we've been looking at tritium alternatives since l-12 1988 when the Savannah River reactor shut down. ,
. 13- In that environmental impact statement, we had O' . 14 dismissed FFTF as a reasonable alternative because it f
15 couldn't make the required amount of tritium and because 16 we said we didn't want to rely on it until the middle of 17 the next century as a tritium producer. But the main 18 reason really had to do with the capability of the i
f 19 machine. There were then questions raised about the 20 capability of the machine. There was a private group out 21 at Hanford who wanted to_come in and make a proposal to l
22 privatize the facility. A congressman from Washington
! 23 prevailed.upon the Secretary on the eve, by the way, of
(
24 the draining of the sodium loop which would have k 25 effectively killed the reactor and we've been looking at NEAL R. GROSS
- COURT REPORTERS AND TRANSCRIBERS I 1323 RHODE ISLAND AVE., N.W.
I (202) 234-4433 WASHINGTON, D C. 20005-3701 (202) 234 4433
417 1 it ever since.
,S 2 Secretary O' Leary, just before she left, made (v) 3 a decision to keep the reactor in stand-by, kind of as an 4 insurance policy, a low cost insurance policy so that we 5 had additional assurance of meeting our tritium 6 requirements. It's unclear how the issue of the final 7 disposition of FFTF is going to be resolved and Secretary 8 Designate Peha has committed to the Congress to take 9 another look at the status of the facility later this 10 year. So we'll see what happenc.
11 In any case, in December of 1995, in her 12 record of decisions, Secretary O' Leary decided that we 13 were going to pursue what she called the dual past
\
'- 14 strategy for tritium, the first of which was to design, 15 build and test components of an accelerator which is now 16 on-going. The two sides of the tritium project office, l
17 I'm heading commercial reactors side. There's another 18 gentleman who heads the accelerator work.
19 And then we were going to investigate further ;
1 20 either the purchase of a commercial reactor or the 1 21 purchase of irradiation services from a commercial reactor l 22 for the production of tritium.
23 And then the Secretary said by late 1998, the 24 Department expected to make a decision on which of those 4i I
(_) 25 paths would be the long-term tritium supply path and which NEAL R. GROSS COURT REPORTERS AND TRANSCRIBERS 1323 RHODE ISLAND AVE., N.W.
(202) 234-4433 WASHINGTON, D C. 20005-3701 (202) 234-4433
418 1- would constitute a backup capability.
1 2 One caveat to that, the Congress in last 'I p
~O- . 3 year's Defense Author 3 2 & tion Act required ' the Department .
1 4 to make:the decision this year or explain to them why we .
I 5 couldn't make it this year. So in the~next few months )
6' we'll have to prepare a report, either telling them we've -
F 7 made a decision or providing the reasons.why we can't make I
t 8 a decision.
1 9- And then part of that record of decision and
)
10 subsequent events mandated that in any case, regardless of ;
11 whether the accelerator or the commercial reactor were 12 selected as the primary long-term supply option, the l 13 commercial reactor would be continued as a contingency i k 14 option and that means to us that we would continue all.of 15 the activities with the NRC up to and including
.i
.16 qualification of the target for use in a reactor, i 17 fabricating the first core loaded targets, constructing a 18 new extraction facility down at the Savannah River site l
! 19 and what we'd like to do is we'd like to negotiate options !
20 with utilities to use their' reactor if it becomes !
I 21 necessary.
22 MEMBER POWERS: There's nothing unusual about 23 having dual capability for the production of tritium.
l 24 MR. SOHINKI: That's true.
25 MEMBER POWERS: I mean historically, we've l' NEAL R. GROSS COURT REPORTERS AND TRANSCRIBERS 1323 RHODE ISLAND AVE., N.W.
(202) 234-4433 WASHINGTON, D.C. 20005-3701 (202) 234-4433 1
419 1 done it that way. j l
2
,.. MR. SOHINKI: Now the only major points, it's
' \
\ '/ 3 kind of a busy slide, but the only major point of this l
l 4 slide is that all of the activities in green are 5 activities that would be undertaken regardless of which 6 path, accelerator or commercial reactor is selected as the l l
7 primary long-term supply option. So that for the i 1
8 commercial reactor, you do all of the things I've just 9 suggested. For the accelerator, if commercial reactors 1
10 are chosen, the accelerator would be taken to the I
11 completion of the engineering design to be put on the 12 shelf and then if needed, you'd start construction. l 13 MR. LEWIS: I confess to be a little bit out l
,/ 3
, 14 of touch. Is there a one senterce summary of the 15 advantages of accelerator production?
16 MR. SOHINKI: Well, yes, it's not a reactor.
17 MR. LEWIS: That answers the question.
18 MR. SOHINKI: The Assistant Secretary for 19 Defense Programs has used that sentence and what he means 20 by that is for the people who are sensitive about creating 21 an additional source of spent fuel or worrying about 22 fissile materials, you don't have those concerns with an 23 accelerator. So perception-wise, the accelec; Lor has that 24 advantage. There's no nuclear criticality, etcetera, rN
( ) 25 etcetera. So I would say in a nutshell, that's probably NEAL R. GROSS COURT REPORTERS AND TRANSCRIBERS 1323 RHODE ISLAND AVE., N.W.
(202) 234-4433 WASHINGTON, D C. 20005-3701 (202) 234 4433
420 1 the advantage of the accelerator.
,~.s 2 MR. LEWIS: It also has the potential to not
( \
3 produce tritium in which case is a contribution to arms 4 control.
5 MR. SOHINKI: That's true, although if you 6 were to talk to Dr. Bishop, I mean in the accelerator's 7 defense, if you were to talk to Dr. Bishop who is the head 8 of the accelerator project in our office, he would tell 9 you about a number of technology developments which 10 provide greater confidence to him that the accelerator is 11 a go for tritium production.
12 MR. LEWIS: I'm not trying to heckle you.
13 MR. SOHINKI: I know. From my standpoint, I'd
's / 14 be perfectly happy if the commercial reactor option were 15 selected.
16 MR. LEWIS: I've been on that side for 15 17 years.
18 MR. SOHINKI: Obviously, by the way, this is 19 somewhat obvious, but the advantages of the reactor that 20 caused us to make it part of the dual path were the 21 financial advantages as compared to the construction of a 22 multi-billion facility and in addition, if we're talking 23 about an operating reactor, it had the lowest incremental 24 environmen*.al impact of all of the machines. So there are
,s q
( ,) 25 some advantagea to both. You can argue about the weight NEAL R. GROSS COURT REPORTERS AND TRANSCRIBERS 1323 RHODE ISLAND AVE., N.W.
(202) 234-4433 WASHINGTON, D.C. 20005-3701 (202) 234-4433
421 1 that ought to be given to those advantages.
,f 3 2 MEMBER MILLER: How much power is this
) l 3 accelerator going to require?
4 MR. SOHINKI: Depending on whether it's room 5 temperature or superconducting and whether they use i 6 clistrans or a new development, clistodes, in the worse 7 case you'd be talking about 500 to 550 mege.wetts to run l
I 8 the accelerator.
l 9 MEMBER MILLER: So we need to build an AP600 10 just to power it?
11 MR. SOHINKI: That was jokingly suggested.
12 Maybe not so jokingly, as a matter of fact.
13 MEMBER MILLER: I assume that was part of the I I
\- ' 14 environmental impact with the accelerator --
15 MR. SOHINKI: Yes, we took into the account 16 the impacts of the power source that you have to run to 17 power the accelerator.
18 MEMBER KRESS: I guess there's some debate 19 about whether an accelerator for producing tritium is a 20 reactor or not. It's almost a reactor, just a little 21 subcritical, right?
22 MR. SOHINKI: I've heard those kinds of 23 arguments. Yes. Particularly in conjunction with the 24 idea of using an accelerator for transmutation of waste l
(3) _,
25 which is another use that's been suggested.
NEAL R. GROSS COURT REPORTERS AND TRANSCRIBERS 1323 RHODE ISLAND AVE., N W.
(202) 234-4433 WASHINGTON, D C. 20005-3701 (AC: P34-4433
. - - . - - . . - - . . . - - - - . _ . - - - - . . - . . . - . - . ~ . - _ ~ . - . - .~
422 1 But this is simply a cartoon that we created L
2 to try'to describe the system for production of tritium in
[ O.
i 3 a commercial' reactor in one slide.
~
It starts with the
.I -
1 0 4 burnable tritium producing burnable absorber rod 1-4 . .
S' manufacturer for the lead test assembly activity which Dr.
t i 6- Ethridge will describe. Pacific Northwest National' Labs 1
l 7 will be the assembler of these rods, with components that i
j 8- are now being fabricated by private companies. ,
p ,
' 1
(
9 When we get to the production scenario we'll l
i l
l 10 have to have a manufacturer who can handle in the n
. )
3 11 thousands of rods. We haven't decided yet whether that's
, 12 going to be done inside the-DOE complex or by private i
j 13 company. We're going to do a-make-buy analysis'.to.see
'V b 14 which of those would be more cost effective. That has not I
- 15 been done yet.
?.
j 16 But basically, the rods would be shipped by 4
2
- 17 the rod manufacturer to the fuel fabricator who would l 18 install them in standard reactor fuel assemblies. Dr.
i j 19 Ethridge will describe that. And then they'd be l
20 irradiated for whatever cycle is the normal cycle for the j 21 utility and these rods can accommodate anything between
! 22 one and two years and even if you monkey around with it, 23 you can fool around with that period as well.
l 24 After the normal cycle for the reactor -- and g 25 that depends somewhat by the way on whether we own the NEAL R. GROSS COURT REPORTERS AND TRANSCRIBERS 1323 RHODE ISLAND AVE., N.W.
(202) 234-4d h WASHINGTON, D.C. 20005-3701 (202) 234 4433 ll
i 423 1- reactor or we purchased-the reactor or whether-we're -j i
2 talking irradiation services, because for the irradiation O 3 services scenario,.the goal is to make this invisible to.
4 reactor operation, so we want to impact the operation of 5 the reactor as little as possible, 6 After irradiation for the' cycle, DOE, as'the 7 shipper, would pick up the rods at the reactor site. They 8 would have been loaded into casks by.the utility'. They 9 will be shipped to Savannah River. We have a new 10 extraction facility being designed, as I-mentioned earlier
'11 and then the extraction facility, as currently envisioned, 12 would be adjacent to the replacement tritium facility and 13 the gas would be sent over there and processed.
14 MEMBER POWERS: Do you have a feeling for the 15 amount of storage time after offload that the irradiated 16' rods would have to be on-site?
17 MR. SOHINKI: They don't really require a long 18 period for cooling and so what we're envisioning is we 19 don't want to impact the outage for the utilities so what 20 we envision is they remove the assembly, store them in the 21 spent fuel pool and then after they're done with the 22 outage and they have some time, they'll pack them in the 23 casks.
24 MEMBER POWERS: The spent fuel pool itself and f 25 not their own storage facility?
NEAL R. GROSS COURT REPORTERS AND TRANSCRIBERS 1323 RHODE ISLAND AVE., N.W.
(202) 234-4433 WASHINGTON, D.C. 20005-3'01 (202) 234 4 433
I l
424 1 MR. SOHINKI: Not in a --
/2 MEMBER POWERS: Not-in their own facility, but 3 actually in the spent fuel pool? .
I 4 MR. SOHINKI: That's what we envision, yes.
}
5 MEMBER MILLER: If you used the FFTF, how much 6 will-it change the system approach, besides the physics?
7 MR. SOHINKI: Well, the reactor is different 8 so they don't accommodate 12-foot rods so the rod length 9 will have to be different and I don't know, Max, you want 10 to address that? This is Max Clausen, my deputy, who 11 spent some time with this.
12 MEMBER POWERS: Would you just give your name ;
13 at the beginning?
14 MR. CLAUSEN: I'm Max Clausen with the 15 Department of Energy. The FFTF system would be 16 substantially different for a number of reasons. One, we 17 would expect to use the fuel and materials examination 18 facility which is adjacent to the FFTF for the fabrication 19 of the target rods. It also has large hot cells which may 20 be attractive to build an on-site extraction facility. ,So 21 we would essentially eliminate all the transportation off-22 site until we were ready to perhips transport tritium on Z
-23 beds or uranium beds.
24 There are some other serious changes in the
() 25 physics which are driven, that gets into a fairly complex NEAL R. GROSS COURT REPORTERS AND TRANSCRIBERS 1323 RHODE ISLAND AVE., N.W. '
(202) 2344 433 WASHINGTON, D.C. 20005-3701 (202) 234-4433
425 1 thing. It takes much higher plutonium concentration fuel,
,7 2 very different targets in terms of the enrichment of the
(' '*3)
~'
3 lithium and there are some interesting challenges there.
4 We have a couple of reports we'd be happy to 5 share with anybody that's interested to talk about that at 6 some length.
7 MR. SOHINKI: They assert that up to a point 8 they can use the rame target that we're using and then 9 they have to go to a lithium oxide target instead of a 10 lithium aluminate, but -- okay, that's briefly an overview 11 of how we got where we are and I guess I can turn it over 12 to Dr. Ethridge now who will take you through the target 13 design in some detail.
,f
\s / 14 MR. ETHRIDGE: Okay. Good afternoon, my name 15 is Jerry Ethridge from Pacific Northwest National Lab. As 16 Mr. Sohinki indicated, I'm a project manager there at the 17 lab for the target technology aspects of his project. So 18 I'll be talking a little bit about the technical aspects 19 of the target.
20 The first diagram that you have, I assume you 21 have slides, fairly simple. The physics, take a Lithium-22 6, bombard it with a neutron. It generates Helium-4 and 23 tritium; tritium has about a 12.5 decay half life and it 24 goes to Helium-3.
/3
(,) 25 The target then takes advantage of that by NEAL R. GROSS COURT REPORTERS AND TRANSCRIBERS l 1323 RHODE ISLAND AVE., N.W. l (202) 234-4433 WASHINGTON, D C. 20005-3701 (202) 234-4433 l l
l
i i
l 426 l 1 placing Lithium-6 in the presence of a neutron reactor, if fm s 2 you will, and what I've done is taken the target and
\
\/
3 peeled off each layer so we can go through and describe 4 each of the major components of the target.
5 Let me start where the meat of the target is 6 with the lithium aluminate pellets, high temperature l l
4 7 ceramic material, white in color, very strong, high !
i 8 temperature properties and we'll provide the vehicle, if 1
9 you will, for the Lithium-6. l 10 Outside of that then is a getter, which is 11 zirc material that has been nickel plated. Nickel plating l
12 is designed to protect the zirc so that it doesn't oxidize 13 and then fail in its gettering function.
4
\)
~ 14 On the inside then is a zirc liner. It is not l 15 plated and its oxidation is exactly what we want to have 1
16 it do and then on the outside to capture all of this is a
)
l 17 piece of stainless steel target material that has had a i l
18 coating placed on the inner surface to minimize the amount 19 of diffusion of tritium that goes through the cladding.
20 So this is where the action is. Once the 21 tritium atom is born here, it will migrate out of the 22 pellet. It will pick up extra oxygen. It will be 23 released as TiO, triated water. That will react with the 24 liner and the liner will crack the TiO and give you pure f) 25 tritium and zirc oxide. That tritium will then go through l
(/
NEAL R. GROSS COURT REPORTERS AND TRANSCRIBERS 1323 RHODE ISLAND AVE., N.W.
(202) 234-4433 WASHINGTON, D.C. 20005-3701 (202) 234-4433
427 1 this nickel window and be gettered, if you will, on the f~s 2 zirc getter.
t, ,n 3 And while it's doing that, this coating is 4 just kind of an additional defense tor that tritium to 5 prevent it from diffusing out of the steel into the 6 reactor coolant.
7 The next page then talks about -- let me do 8 the following. I do have a model that I'd like to pass 9 around and you'll see where it's got a section pulled out 10 for each of the major components and that's obviously 11 foreshortened, but they will be full length, i.e., 12 feet 12 and will replace standard burnable poison assembly and a 13 fuel assembly and I'll show a picture of that in a moment.
A/ 14 Please, if you have any questions about that as it's going 15 around, please stop.
16 This is what it looks like axially where we 17 have a spring at the top to kind of hold things in place, 18 very similar to a fuel rod. And the pellets are annular 19 in geometry. This liner is then on the inside. The 20 getter is here on the outside with the coined end to kind 21 of hold these pellets in place and the liner and this is 22 what we call a pencil. It's about 12 inches in length and 23 then the next one goes below that and so you just stack 24 these pencils inside this cladding.
[~'N
() 25 So we've got essentially 12 of these stacked NEAL R. GROSS COURT REPORTERS AND TRANSCRIBERS 1323 RHODE ISLAND AVE., N.W.
(202) 234 4433 WASHINGTON, D C. 20005-3701 (202) 234-4433
- . - . . . . . . ~ . - . - . - . - . . - - . . _ _ . - . - _ _ . . ~ . - . . - _ . . - . . _ . , . . _ _
428 1; on top'of one-another. A getter plug here at the top to
.g 2 prevent inadvertent diffusion of the tritium up through Q -
13- here and cut the top, so we've got a number of'fcatures 4 then that are intended to capture as much as the tritium 5 as possible.
6 This says it a lot better than what-I just l
did, but describes each of the major components, the
.7 8 purpose of the nickel plate, again, to protect the getter l J
9 material. During irradiation and after irradiation the' ;
I 10 tritium is held in the ceramic material and very tightly 11 held in this getter. Now many - I should point out that
!- 12 many of the aspects of the design are clasoified as you l
13 can appreciate, but what I'm discussing here is I l r J l 14 unclassified and'again, I can talk a little bit about the i-15 getter, but not how it's manufactured or how well it works 16 other than to indicate that in previous testin'g that l 17 getter has retained a significant amount of the tritium 18 that was produced, i.e., it does work very well.
19 It is held here as I indicated in the getter, 7
20 the liner and the pellet and all of that then is released i L
l 21 when we take the rods out of the reactor and place it into 22 a very high temperature furnace, in excess of 1000 C. is !
1 23 what it takes to drive any of the tritium off of these l 24 components, 25 This is looking at it a little bit NEAL R. GROSS COURT REPORTERS AND TRANSCRIBERS 1323 RHODE ISLAND AVE., N.W.
(202) 234-4433 WASHINGTON, D.C. 20005-3701 (202) 234-4433
429 1 differently. Here's the reactions that I've talked about,
,s 2 the liner where it cracks the tritiated water, the getter
( )
\'
3 material that getters the tritium and you end up with high 4
4 temperature hydride. I'll talk a little bit about the 5 barrier. I have tritium in all three of these components.
6 The number that is in the getter is well in excess of 50 7 percent. The balance is basically in the liner and in the 8 pellets and the free gas that we -- when we -- once we 9 punctured some irradiation tests showed virtually no 10 tritium at all and we were able to count for it all in 11 these components in solid form.
12 MEMBER ?OWERS: Do you exceed saturation in 13 your getter?
p.
4
'N_/ 14 MR. ETHRIDGE: I'm sorry?
15 MEMPER POWERS: Do you exceed saturation in 16 your getter?
17 MR. ETHRIDGE: No, we do not. In fact, we're 18 quite a ways from saturation.
19 This barrier, we've coined a new word, if you 20 would, permeation reduction factor and it basically 21 represents a gauge of the ability of that barrier to 22 prevent diffusion through the stainless as compared to 23 pure stainless. So that when I end up with a high PRF or 24 permeation reduction factor, it's basically the im i ) 25 diffusivity of the tritium through the barrier material NEAL R. GROSS COUe4T REPORTERS AND TRANSCRIBERS 1323 RHODE ISLAND AVE., N W.
(202) 2344433 WASHINGTON, D.C. 20005-3701 (202) 234-4433
430 1 like that of stainless steel. So high PRFs mean very low
.s 2. tritium diffusion through the stainless.
I i'^'/
3 Yes sir?
4 MEMBER FONTANA: Is the ratio on the bottom 5 line correct?
6 MR. ETHRIDGE: Pardon? Yes, it's the ratio of 7 the permeability in the bare stainless steel to that of 8 the coating. High PRF means low --
9 MEMBER FONTANA: See what I mean?
10 CHAIRMAN SEALE: Stainless steel to coating? l l
11 MR. ETHRIDGE: The stainless steel has high -- l l
12 MEMBER FONTANA: High permeability. j 13 MR. ETHRIDGE: High permeability. l
/~'T l k /
N' 14 MEMBER FONTANA: Ah, gotcha. l 15 MR. ETHRIDGE: Right.
16 MEMBER FONTANA: I've got you.
17 MR. ETHRIDGE: Not to be outdone with 18 cartoons, I have my own and this is -- I'm going to 19 describe the purpose of the test in a moment, but this is 20 the lead test assembly cycle, if you will, very similar to 21 what you saw before. We'll be taking parts from industry l 22 manufactured to very tight specifications. We're using 23 Appendix B type quality requirements. We're assembling 24 them here at PNNL. They'll be transported to Columbia
('~x,
(,) 25 which turns out to be where Westinghouse has their fuel l NEAL R. GROSS l COURT REPORTERS AND TRANSCRIBERS
- 1323 RHODE ISLAND AVE., N.W.
(202) 234-4433 WASHINGTON, D.C. 20005 3701 (202) 234-4433
(- - -. . ~ . . . - - . . . . - . . . ~ ~ ~ . - - . . - .
I 431 l l
E 1~ supply and_they.are the fuel supplier for the Watts Bar 2 plant which will be where this lead test assembly is
'3 irradiated. They will attach the rods to the standard i 4 base plate. They'll install that into the fuel assembly.
5 Those fuel assemblies then will be shipped via normal 1
6 shipment to Watts Bar and irradiated in the reactor for 18 7' months. Their total cycle is 18 months in length. I 8 Following their shutdown, be removed from 9 the -- they will be removed from the fuel assembly. Those l
l 10 fuel assemblies will most likely go back into the reactor 11 for the normal second and third burns. These~will be
'12 stored until the shutdown is over and they have restarted l
l 13 and then-they will be shipped in a.special fixture that l
l
) 14 will probably look very similar to a fuel assembly-in a 1
15 cask back to the laboratory where we have hot cells and 16 we'll do a post-irradiation examination.
17 MEMBER POWERS: Is there any limit, did you 18 put any limit on the burnup you can take the fuel rods to?
19 MR. ETHRIDGE: Yes, we do. We do have, based 20 upon the test data that we have, we have limited much of 21 our performance parameters to that test data base and I
l 22 that, in fact, results in a maximum life time in the 23 reactor.
24 Now where we can fluctuate that a little bit i
25 relates to lithium enrichment, what types of burnups do we j NEAL R. GROSS COURT REPORTERS AND TRANSCRIBERS 1323 RHODE ISLAND AVE., N W.
(202) 234-4433 WASHINGTON, D.C. 20005-3701 (202) 234 4433
432 1 achieve in the rod?
g 2 I'm sorry you don't have these in your vu-tv) 3 graph. I just pulled them out of what Steve took out of 4 his. I do have copies if one would like them, but these 5 are the purposes and plans for the lead test assembly.
6 They replace four of these lead test assemblies into the 7 Watts Bar. This will be their first refueling and operate 8 that for about 18 months. They'll be removed at the end 9 of the cycle and then taken to the laboratory.
10 In each of these four assemblies we will have 11 eight rods for a total of about 32 rods placed 12 symmetrically around the core, so you don't mess up the 13 core analysis. They typically use somewhere on the order i
'/'-
14 of about 1000 BPAs. We're replacing 32 of those. And we 15 really are not looking upon this as an experiment because 16 none of what we're testing here is without precedence and 17 without a data set and I'll go through that in a moment 18 for you.
19 This is to kind of revalidate that purpose, 20 not an experiment, but really to confirm the results that 21 we had from prior testing. And this is more to insure 22 that those like the NRC and the utilities and other 23 stakeholders in this process. This is not something 24 really new, goofy, unexplainable. It's been done before.
(,) 25 Here's the standard Westinghouse hold down NEAL R. GROSS COURT REPORTERS AND TRANSCRIBERS 1323 RHODE ISLAND AVE., N.W.
(202) 2344433 WASHINGTON, D.C. 20005-3701 (202) 234-4433
i
. 433
! 1 assembly. It has 24 slots on it for either BPA' assemble' t
J 12 plugs or in this case.we'll be replacing eight of them O 3 for, with the TP bars, tritium-producing burnable absorber
. 4 assembly. I can't say that very'welluso I-say'TB bars.
5 Again, those will be installed at Columbia, go through :
- 6 acceptance testing and then installed in the fuel i l
7 assemblies and the number of thimble plugs will be l i
j 8 tailored to insure the flow rate through the assembly-is 9 what it needs to be from an individual assembly-
[ 10 perspective, as well as the balance of plant perspective.
11 So from the outside, not knowing what's
[ 12 inside, virtually no difference between this assembly and
.13 a standard fuel assembly that has burnable poison O 14 assemblies in them.
15 MEMBER FONTANA: In production will these 16 replace all the burnable poison rods?
17 MR. ETHRIDGE: Yes and no. We will if the 18 tritium demands require that and really where the answer ,
l 19 lies is how many reactors are they going to buy or rent.
20 We can replace all of them or we can replace a limited 21 number such that you balance that out, if you have more 22 than one core and the demand for tritium requires that 23 amount.
l 24 MEMBER FONTANA: In principle, you could
() 25 replace them all.
NEAL R. GROSS COURT REPORTERS AND TRANSCRIBERS 1323 RHODE ISLAND AVE., N W.
(202) 234-4433 WASHINGTON, D.C. 20005-3701 (202) 234-4433
434 1 MR. ETHRIDGE: We could replace them all.
2 Yes. Here's just a comparison to the BPAs. Basically 7-.
l I i
/
l 3 serves the same function. We're simply replacing the 4 boron with lithium, same dimensions. They look the same.
5 MEMBER POWERS: Is there a difference in 6 weight?
7 MR. ETHRIDGE: There is a slight different in 8 weight, 10 percent or so.
l 9 The only difference here again is the Lithium-10 6 versus boron and obviously no uranium, plutonium, no l
11 fissile material at all. I 12 Let me go into some confirmative testing that 13 was done. This was done in a program that Mr. Clausen led
/\ 4
\./ 14 several years ago as part of the new production reactor. I 1
15 The light water option was kind of the black sheet of the 16 three options that were considered at the time, wouldn't 17 work. You know, who would ever do that sort of thing and 18 in reality it's worked out quite nicely. But that's part 19 of that program. There were three tests done, three 20 capsule tests. These were static capsule, static water in 21 the advance test reactor at Idaho. There was also a loop 22 test done to simulate a series of rods attached together.
23 It had flowing water, an entire separate system from the 24 advanced test reactor. So in virtually every aspect it D).
s , 25 did simulate a group of eight rods together.
NEAL R. GROSS COURT REPORTERS AND TRANSCRIBERS 1323 RHODE ISLAND AVE., N.W.
(202) 234-4433 WASHINGTON, D C. 20005-3701 (202) 234-4433
435 1 MEMBER POWERS: Does that include the 7- 2 operating temperature?
(.'~') 3 MR. ETHRIDGE: Operating temperatures, 4 pressures, water chemistry, and so forth. The only l 5 difference is for those of you who know ATR is only a four 6 foot core, not an eight foot, so there is a length 7 difference as well. But the diameters and those sorts of 1
8 things are very similar to what we're considering here.
9 MEMBER POWERS: There's a spectrum difference 10 and there's a --
^
11 MR. ETHRIDGE: Yes, and I'll show that in a 12 moment, spectral difference.
13 We did measure on a periodic basis when the
(
t
'% - 14 reactor was shut down. An ATR typically had a 30 to 40 15 day cycle and then it was shut down for putting Navy type 16 experiments into it, but we took that opportunity to flush 17 the capsule, measure the tritium in the capsule, in the 18 water, and do an analysis on that and then we were able to 19 measure continuously on the loop. It had again the 20 separate system that went through a tritium analysis loop.
21 These are those data compared all on the same 22 plot and this is a little bit misleading, but I'll go 23 through that. This is the time in days that the reactor 24 or the rods were under neutron flux and what I've done is p)
(_, 25 to translate these data into a 12 foot rod, basically NEAL R. GROSS COURT REPORTERS AND TRANSCRIBERS 1323 RHODE ISLAND AVE., N.W.
(202) 234-4433 WASHINGTON, D C. 20005-3701 (202) 234 4433
436 1 multiplied them by 3. Okay? And then what data I've l
7- 2 shown here are in millicuries now the amount of tritium
( )
\ j 3 that was accumulated in either the continuous measurements 4 that were taken on the loop or the discrete measurements l
5 that were taken in the capsule, i 6 What I've done is compared this against a kind 7 of typically amount of tritium that could be released from 8 a reactor, 20,000 curies per year and to try and give a 9 flavor for -- we're really in the noise here, as far as l 10 the amount of tritium that might be released from these ,
1 11 capsules under normal operations, recognizing that some of l
12 these were static capsule and the chemistry wasn't all 13 that well controlled, but the loop really does, I think,
,em
\ ,
Nl 14 reflect that quite nicely.
15 The WC-1, water capsules is where that came 16 from, is here. This is the one that happened to have gone 17 the most in days. If I were to replot this versus burnup, 18 these data would look a lot closer to one another, so the 19 fact that this is deviating a little bit is reflective of 20 the fact that the B-1 test was under -- in a different 21 position in the core and was accumulating burnup at a 22 higher rate than the others. So that's the difference 23 there.
24 But before the program was shut down in 1992, (h
( ,) 25 only the WC-1 capsule went through post-irradiation NEAL R. GROSS COURT REPORTERS AND TRANSCRIBERS 1323 RHODE ISLAND AVE., N.W.
(202) 234 4433 WASHINGTON, D.C. 20005-3701 (202) 234-4433
437 1 examination and I want to share some of those results for
,m 2 you.
( l x/
3 This is the amount of helium, tritium and 4 tritiated water that was predicted based on our models and 5 what was measured in that water, I'm sorry, not water, but 6 in the gas volume inside WC-1 when it was post-irradiation 7 examined. Models look pretty nice. The amount of tritium 8 in the free gas, as I indicated, is virtually nonexistent, 9 very small amount. The amount of tritiated water, we 10 predicted a little bit higher, found virtually none there, 11 tritium improves the models a little bit. And again, the 12 total amount of tritium, this is a -- I just noticed 13 that -- this is millicuries, not megacuries, if you'd make
( ;
k/ 14 that change, very small amount of total tritium activity.
l 15 These are not in your handout either but
]
16 there's some pictures that I pulled from some of our i 17 files. This is an example of one of the pellets from WC-18 1. This is in a hot cell, of course. This chip on the 19 side is from getting the pellet out, but these pellets --
20 crush tests were done. These were very stout pellets l 21 after all this irradiation. Really no indication at all l 22 that the stability of that pellet is compromised in any l l
23 way as a result of that irradiation.
24 If I look then at some of the structural i
(_) 25 materials, what I have is a micrograph of the cladding at i NEAL R. GROSS
! COURT REPORTERS AND TRANSCRIBERS
! 1323 RHODE ISLAND AVE., N.W.
(202) 234-4433 WASHINGTON O C. 20005-3701 (202) 234 4433
438 1 150X here and I also have a micrograph of the getter. The 2 cladding shows that the interbarrier is unaffected
' 7s i'
\'j 3 essentially by the irradiation. It's still there. It's 4 nice, intact and did a very good job in preventing any 5 tritium from getting out of the cladding.
6 This particular experiment did have a barrier 7 on the outside and what we discovered is that one, we did 8 not need that; and two, it's gone virtually after 100 days 9 or so anyway.
10 So we decided that as a result of that, that 11 we would remove that outer barrier from the LTA, that's 12 not needed. The getter material, here's the nickel 13 plating on both the ID and OD and as is characteristic of g
kl 14 a very highly hydrided material the getter worked and 15 absorbed a lot of the tritium.
16 MEMBER POWERS: You'll discuss some of your 17 accident analysis?
18 MR. ETHRIDGE: Yes. Toward the end.
19 MEMBER POWERS: When you get there will you be 20 able to explain how these intermetallic reactions between 21 these coatings and the meat affect the accident?
22 MR. ETHRIDGE: I'll try to do that. Again, 23 Watts Bar is where the lead test assembly will go. I 24 don't know how many of you have been down there. This is
(
(),j 25 what it looks like when it's reproduced from a computer.
NEAL R. GROSS COURT REPORTERS AND TRANSCRIBERS 1323 RHODE ISLAND AVE., N.W.
(202) 234-4433 WASHINGTON, D C. 20005-3701 (202) 234-4433
439 1 Most of the time one of these has steam out of it, but as
,_s 2 Jim indicated the plant happens to be down right now.
( )
\# 3 Spring City, Tennessee, right? Home Sweet 4 Home. MEMBER POWERS: You're bringing 5 tears to Dr. Kress's eyes.
6 MR. ETHRIDGE: Let me apologize then.
7 MEMBER APOSTOLAKIS: Why didn't he mention --
8 MEMBER FONTANA: He did.
9 MEMBER APOSTOLAKIS: You include it here?
10 MR. ETHRIDGE: It's not bringing tears to your 11 eyes.
12 (Laughter.)
13 In fact, he's smiling. You're a transplant
,~
, i s_ / 14 from the north. You're a Yankee. l i
15 MEMBER APOSTOLAKIS: I'm naturalized. l 16 MEMBER POWERS: But there is the problem of 17 prolonged shutdowns in commercial reactors for various 18 reasons, especially if you're using it for strictly 19 irradiation services.
20 MR. ETHRIDGE: A lot of it has to do with 21 happens in this building too.
22 MEMBER POWERS: Some.
23 MR. ETHRIDGE: Occasionally.
24 MEMBER POWERS: Right. And have you borne r ^ N.
U) 25 that in mind in your plant?
NEAL R. GROSS COURT REPORTERS AND TRANSCRIBERS 1323 RHODE ISLAND AVE., N.W.
(202) 234-4433 WASHINGTON, D C. 20005-3701 (202) 234 4433
440 1 MR. ETHRIDGE: Yes, we have defined our 2 success criteria for a given number of days of irradiation 7.
( )
^^
3 to be able to make some of the -- draw some of the 4 conclusions that we need to and that is far short of a '
5 typical 18-month cycle. Now if they were down for the 6 whole cycle and not able to start up for a year, that 1
7 would cause us some programmatic problems, but certainly 8 not any safety issues.
9 MEMBER POWERS: The build up of HCliam-3 10 wouldn't be a problem?
11 MR. ETHRIDGE: Not for 32 rods.
l 12 MEMBER POWERS: Yes, but for a larger load?
13 MR. ETHRIDGE: For a larger load that would be .
/~~% l
</ 14 an issue. And we're accounting for that in the reload 15 methodology and the codes that are being used for those .
I i
16 reload analyses will incorporate that Helium-3 build up.
i 17 MEMBER POWERS: In the analysis that we've !
i 1
18 seen there were no Helium-3 cross sections. (
I 19 MR. ETHRIDGE: Right.
20 MEMBER POWERS: So you have Helium-3 cross 1
21 sections now? i 1
22 MR. ETHRIDGE: Yes. Right. It was helpful, I I
23 think, to go to much more detail for both the utility's 24 sake, the project's sake and obviously for the NRC's sake n
(,) 25 to document this in a form that could be used by the NEAL R. GROSS COURT REPORTERS AND TRANSCRIBERS 1323 RHODE ISLAND AVE., N.W.
(202) 234 4433 WASHINGTON, D C. 20005-3701 (202) 234-4433
441 1 utility in some of its preparations for irradiation of 73 2 this lead test assembly and so we have documented this in
! \
'~'
3 a lead test assembly report that is in the final version 4 is in draft form right now. There are a variety of drafts 5 available here at the NRC already that describes the 6 design and some of the fabrication requirements for the 7 lead test accemblies and the TP bars. It evaluates the 8 impact on operation of these four LTAs and really has used 9 the 0800 recipe, if you will, as guidelines for the types 10 of things that we fe " were important that other people 11 understand and know about, so we followed that recipe and 12 this report talks, as I said, about the design. It goes 13 into the nuclear design and these are the standard k -) 14 chapters from the review plan, the thermal hydraulic 15 design, the materials issues, the impacts of irradiation 16 which covers a variety of chapters, the QA section, 17 safeguards and security as I mentioned earlier. We have 18 some classified aspects of the target so it wasn't clear 19 where that might fit into this, but that certainly is 20 there. And then obviously the regulatory issues 21 associated with that.
22 What I'd like to do is to highlight the 23 results that are in the draft report in each of these 24 areas, related to safety. I don't spend a lot of time on n
() 25 what's in the report from the QA or this aspect, but HEAL R. GROSS COURT REPORTERS AND TRANSCRIBERS 1323 RHODE ISLAND AVE., N W.
(202) 234-4433 WASHINGTON, D C. 20005-3701 (202) 234-4433
442 1 really focus on the design, nuclear design, thermal fS 2 hydraulics, materials and the impacts of the irradiation L
)
3 and this is what I'll try and address unfortunately only 4 at a high level in the metallic issues.
5 These are the components and I'll go through 6 each one and then basically the high level summary of the 7 results of that analyses that are documented in that 8 report.
9 For the cladding and the end plugs, the 10 structural integrity is maintained under all conditions 11 except in an Appendix K type analysis. We do show failure 12 in a large break LOCA and we tal).ed about that in the 13 report and my version of what's in the report is when that
\ l N/ 14 happens in a lot of other things you've got to be worrying 15 about besides a little amount of tritium, but that's 16 documented in the report.
17 Acceptable stress and strain under all of 18 these conditions, again, with the exception of the LOCA, 19 no cladding collapse when we've taken the maximum 20 manufacturing tolerances of ovality and all those things 21 into account.
22 MEMBER POWERS: Even in a thermal excursion?
23 MR. ETHRIDGE: Pardon?
24 MEMBER POWERS: Even in a thermal excursion
' _ ,/ 25 for a large break LOCA, you don't get triggering of inter-NEAL R. GROSS COURT REPORTERS AND TRANSCHIBERS 1323 RHODE ISLAND AVE , N W.
(202) 234-4433 WASHINGTON, D C. 20005 3701 (202) 234 4433
443 1 metallic reaction?
7- 2 MR. ETHRIDGE: The temperatures do not get
- i
\' ,)
3 high enough to do that.
4 MEMBER POWERS: What temperature does it take 5 to trigger zirc nickel?
6 MR. ETHRIDGE: I don't know the answer to 7 that.
8 MEMBER POWERS: I don't think it takes very 9 much. It's an isothermic / exothermic reaction.
10 MR. ETHRIDGE: As I recall, the cladding which 11 is your barrier is not zirc. That's stainless. Okay, so 12 the boundary that prevents any tritium release from the 13 rod is not going to form that, react in that way like your
( )
~-
14 nickel.
15 MEMBER POWERS: On the clad, you've got 16 stainless steel aluminum.
17 MR. ETHRIDGE: Very thick.
i 18 MEMBER POWERS: And we can be pretty sure that l l
19 that one is going to trigger at relatively modest 20 temperatures. l 21 MR. ETHRIDGE: I just don't know the answer to 22 the temperature question, I'm sorry. It is in the report, 23 however, and if more interest is there, the report's i
24 available.
( ,) 25 MEMBER POWERS: Is it in the report that I N EAl. R. GROSS COURT REPORTERS AND TRANSCRIBERS 1323 RHODE ISLAND AVE., N W.
(202) 234-4433 WASHINGTON, O C. 20005-3701 (202) 234-4433
1
, l l 444 l
1 have?
l
,~ 2 MR. ETHRIDGE: I'm not sure I could tell you l
\ ,
\ '~) 3 right now.
4 MEMBER POWERS: Because I didn't see a 5 discussion of intermetallic reactions, i
1 6 MR. ETHRIDGE: Probably part of the reason was )
7 because you only have access to the unclassified report 1 1
8 and the classified report is --
l 9 MR. SORENSEN: Jerry, I'm not sure that issue 10 is discussed.
11 MR. ETHRIDGE: At all. Okay. That's something 1: we need to note.
l
'3
_ Pellets, we put these in high temperature j g
(.)
6 4 14 "ater before. Again, very stable. They don't dissolve.
15 They don't break apart. Your absorbing material, a 16 neutron absorbing material is not going to go anywhere in j 17 the event if a cladding failure and I'll talk about a l l
18 large break LOCA phenomena in a moment which -- that still 19 holds, but -- and the stability to these pellets as I l 20 demonstrated in WC-1 is quite good up to what we're 21 calling a gas volume ratio which is our calculation of the 22 burnup. As far as the pellet behavior and so forth, the 23 helium and the tritium are important and so we calculated 24 what's called the gas volume ratio to account for that as fo,
) 25 opposed to just simply percent burnup. That only gives >
NEAL R. GROSS ;
COURT REPORTERS AND TRANSCRIBERS l 1323 RHODE ISLAND AVE., N W.
(202) 234 4433 WASHINGTON, D.C. 20005-3701 (202) 2344433
445 1 you have the story.
,s 2 MEMBER KRESS: Lithium water, lithium is
,'~')
3 notorious for acting very vigorously in wate_
4 MR. ETHRIDGE: You bet.
5 MEMBER KRESS: What suppresses that reaction 1
6 with these pellets? l l
7 MR. ETHRIDGE: Because this is not lithium. l 8 MEMBER KRESS: It's lithium aluminate? ;
l 9 MR. ETHRIDGE: Yes. It's not -- l l
10 MEMBER KRESS: Would that stay, that l 11 nonreactivity be true if you upped the temperature 12 considerably? l l
l 13 Let's say, would it react with steam, for
,Oi
\~ l 14 example?
15 MEMBER POWERS: There's really nothing for it 16 to react to. You might volatilize and look at it.
17 MEMBER KRESS: That's what I'm thinking about, 18 volatilizing and then you've got lithium volatile lithium 19 which can react to steam.
20 MEMBER POWERS: It would volatilize as lithium 21 hydroxide.
22 MEMBER KRESS: Oh. We've already got the 23 hydroxide?
24 MEMBER POWERS: Well, you just suck the -- the
(\
(_,) 25 exposed pellet exposed to the high pressure steam would NEAL R. GROSS COURT REPORTERS AND TRANSCRIBERS 1323 RHODE ISLAND AVE., N.W.
(202) 234-4433 WASHINGTON, D C. 20005-3701 (202) 234-4433
l 446 1 1 create a vapor pressure of lithium hydroxide. I'm sure l l
em 2 I'm guessing. Correct me if I'm wrong and you transport l
[ ) l 3 lithium hydroxide since you've got lots of lithium in the l 4 coolant water anyway, it's not going to make any l
5 difference. l 6 MEMBER KRESS: You're right. l l
7 MR. ETHRIDGE: Going further, there are not l l
8 related to safety although they certainly have.
9 programmatic impacts and certainly the getter has an l
10 impact as well. There's a typo here. l 1
11 Under the -- again, assuming nonfailure of the l l
12 pellet, of the cladding, there's nothing to react with the 13 zirc and the nickel and we have in our design in the gaps
/~N L
\ 'i 14 and so forth accounted for the swelling that will occur 15 witn the hydriding and the springs are fairly standard 16 design.
17 MEMBER POWERS: As I read your documentation, 18 you discuss a lot about the pellets not cracking during i 1
19 the irradiation in ATR and not getting any pellet clad 20 loadings. What I didn't understand was whether they had 21 gone through the same sort of thermal cycling that you 22 would get in a commercial reactor and you'd be putting the 23 same kind of thermal stresses on it.
24 MR. ETHRIDGE: Well, how I might respond to
\
('~'/
(_, 25 that is again, ATR went down every 40 days and that i l
NEAL R. GROSS COURT REPORTERS AND TRANSCRIBERS 1323 RHODE ISLAND AVE., N W.
(202) 234 4433 WASHINGTON, D C. 20005-3701 (202) 234-4433
447 4' 1 happened to be the one experimerit hat we did do post-w 2
[ irradiation examination on, so assuming we're not going 3 through those types of cycles and even if we had, the WC-1 4 results do not suggest those thermal cycles'would create a
- l 5 significant. amount of cracking of those pellets.
1 i _6 MEMBER POWERS'. You really haven't looked at i 7 thermal shock'and thermal stress analyses'on these pellets
) 8 then? )
{. 9 MR. ETHRIDGE: Bill, can you answer that? :
10 MR. SORENSEN: Yes, I'm not sure that I have i 11 the information on that. Gerald Sorensen from' Pacific
-12 Northwest National Laboratory, but I don' t: know that we j i
13 have looked at that. . Somebody may have done, but I'm not !
14- aware of what was done there.
15 MR. ETHRIDGE: Okay, from the nuclear design 16 then, looking at the impacts of these relative to burnable 17 poison assemblies, again we've got a very limited number.
18- I'm not talking production here, just the lead test 19 assembly, limited number, limited number per assembly.
20 These will not be placed iri any sort of limiting l 21 condition. The Lithium-6 then will be set to insure that 22 1.2 grams of tritium per rod less than that amount is
.23' produced to -- this is somewhat of our limit based upon 24 the saturation of the getter, based upon our experimental
\ 25 data base for burnup, etcetera, etcetera. This is a "not NEAL R. GROSS COURT REPORTERS AND TRANSCRIBERS 1323 RHODE ISLAND AVE., N.W.
(202) 234 4433 WASHINGTON, D C. 20005-3701 (202) 234 4433
448 1 to exceed" number. So we've taken this number nd then
- 7. 2 backed off of that accounting for the reactor i i
'~
3 uncertainties, the thermal hydraulic uncertainties and 4 basically the hot channel factor. So under nominal 5 conditions we're quite a ways from this, probably on the 6 order of .8, .9 grams per rod.
7 As I mentioned earlier, the fuel supplier for 8 Watts Bar is modifying their standard reload design code 9 to account for these targets.
10 MEMBER POWERS: From your documentation I 11 couldn't tell exactly what neutronics had been done, 12 accident neutronics had been done. You're looking at a 13 rod drop accident with variable fuel burnups?
g, i )
k/ 14 MR. ETHRIDGE: What we did is to look at all 15 of the FSAR type accidents and determine -- I can't 16 describe for you how, but determine that the large break 17 LOCA was our most limiting type condition. That was 18 looked at and then as far as all of the conditions 1 19 through 4, those were locked at as well. I don't know if 1
1 20 that answers your question, but -- l 1
1 21 MEMBER POWERS: You're talking about using i I
22 Watts Bar which has relatively young fuel in it, so the ,
l 23 analyses that you've done are probably well within the {
1 24 bounds of known technology.
(a) ,
25 As you go to more, to higher burnup on the I
NEAL R. GROSS COURT REPORTERS AND TRANSCRIBERS 1323 RHODE ISLAND AVE., N W.
(202) 234-4433 WASHINGTON, D C. 20005 3701 (202) 234-4433
449 1 fuel, we start pushing the limits of that, how well we
,_x 2 understand that envelope.
I \
3 MR. ETHRIDGE: The strategy at this point is 4 to only put thermal poison assemblies in fresh fuel.
5 MEMBER POWERS: It's the surrounding fuel 6 elements that are going to be the problem.
7 MR. ETHRIDGE: Power distribution, again we've 8 tailored these to try and not deviate too far from the 9 BPAs, the control requirements, we're not going to 10 impacting those in any way. The reload analysis is being 11 done with Westinghouse PHOENIX . They're in the process of 12 finalizing, I think, or completed the validation, 13 verification of that and will be documenting that and 7,
ss 14 their normal path for modifying that methodology.
15 I'm not going to go into a lot of depth here, 16 but this is absent some proprietary information, a 17 comparison between Westinghouse WABA rod and burnable 18 poison assemblies and the TBAR and the things that you'll 19 notice most often with the exception of some material 20 differences is the similarities between certainly this 21 one. The differences I can say here are in this case none 22 and in this case none, but again some slight material 23 differences.
24 This goes on here, some of the nuclear fluxes,
,o (w/ j 25 a little bit lower than what we saw in the loop rod. This NEAL R. GROSS COURT REPORTERS AND TRANSCRIBERS 1323 RHODE ISLAND AVE.. N W.
(202) 2344 433 WASHINGTON, D C. 20005-3701 (202) 234-4433
1 1
450 1 1
1 1 was tailored for, it had hafnium shields and a series of
,- 2 other things to try and tailor this as much as possible in (s) 3 the ATR to achieve typical PWR type conditions. i l
4 The fluence ended up being right. Peak to I 5 average for the thermal was close. This is not far off.
6 Again, there was quite a bit that had to be done to try l 7 and tailor this nuclear distribution to match what we I i
8 would see in a TPBAR.
9 And the -- certainly the kilowatts per foot 1
10 was enveloped by what was done there as well, heat rate. l 11 There's more here. I might point out our peak gas volume i
12 ratio does appear to exceed what's been done in the l
13 previous test. There are other classified data that O
\~ l 14 exceed what we're going to see here and they would also 15 indicate that going to this is no problem, going to that 16 level of burnup.
17 Looking at the thermal hydraulics design, i 1
18 cladding temperature met all of our criteria. Tech specs 19 would not allow this to be put in the limiting position so 20 we're not going to overtemperature it or affect the core 21 in any way.
22 Thermal hydraulics design from the core 23 perspective there are so few of them, essentially no 24 impact. That analysis will look very very similar to a
,e s ,j 25 standard or their standard reload.
NEAL R. GROSS COURT REPORTERS AND TRANSCRIBERS 1323 RHODE ISLAND AVE., N.W.
(202) 234 4433 WASHINGTON, D C. 20005-3701 (202) 2344433
451 1 From a thermal hydraulics perspective then, 7w 2 this is even less eventful. All of this is all saying the V
3 same and we're not changing coolant temperatures, flow, 4 those sorts of things. There are just too few of them in 5 the core.
6 When we go to a full production core, of 7 course that will likely change. No change here. Peak l 1
8 power in the pins, peak power in the assemblies, average. l 9 None of those are going to change.
10 We will be seeing some of our peak absorber 11 temperatures here. I might add that with the last column, 12 last table in this column as well, these are not best l
13 estimates. These are assuming all of the uncertainties and i (D i L
A/ 14 assuming all of the factors that place a lot of these in 15 their worse light, so we've accounted for those 16 uncertainties in those data.
17 The materials with exception of the utectic 18 that you mentioned earlier which we need to go look at, 19 negligible corrosion on the stainless, stress corrosion 20 cracking we've looked at that extensively. The conditions 21 are just not there, either in the reactor or in the spent 22 fuel pool to have this go into play. It is in the reactor 23 and in the pcol such a short time and of course, the 24 chemistry is not there to have that be a problem as well.
tO
(. ) 25 The weld which might be another vulnerability NEAL R. GROSS COURT REPORTERS AND TRANSCRIBERS 1323 RHODE ISLAND AVE., N.W.
(202) 234-4433 WASHINGTON, D C. 20005-3701 (202) 234-4433
i 452 1 point, we're going to go through a thorough weld i
2 qualification and pass through all the hoops that 7]
> l 3 Westinghouse is forced to go through on their fuel. We'll 4 do a similar exercise with their oversight on the rods.
5 Okay, operational impacts again, looking at 6 this from a global reactor perspective, normal operations, 7 essentially insignificant amount released in reactor, I 8 showed you the data from the previous test. They look 9 very much like the standard burnable poison assembly, so 10 there's no special tooling required.
I 11 DOE then will be the shipper of record from j 12 the reactor to the DOE site so there's not a
,_ 13 transportation impact as far as the NRC is concerned. And i ,
\# 14 only in the worse case, assuming everything, although even 15 under normal conditions, assuming all the uncertainty 16 factore go the wrong way, less than a few millirem 17 additional occupational exposure.
)
18 Off-normal, none of these are predicted to 19 fail under any of these conditions. We'll need to look at 20 the technic as I indicated. I do have a -- let me see if ]
21 I have it here -- a slide that I'll show in a minute that )
22 talks about the LOCA. We do show under the worse case I 23 condition in a LOCA the loss of one 12-inch pencil and I 24 I'll show you that in a moment. Cladding defects, l
/N f
( ,) 25 assuming not a large LOCA type event, minimum impact to ;
i NEAL R. GROSS COURT REPORTERS AND TRANSCRIBERS .
1323 RHODE ISLAND AVE., N W.
(202) 234 4433 WASHINGTON, D C. 20005-3701 (202) 234-4433
453 l
1 the core, very little impact to the chemistry and the off- l
~~s 2 site doses are insignificant.
/ ,
L
/
3 MEMBER POWERS: Do I take it that under the 4 LOCA conditions you don't expect cladding, clad ballooning 5 and rupture.
6 MR. ETHRIDGE: Yes, there's a little bit of 7 that. Under the LOCA then, like the fuel, again, under 8 appendix K type calculation, we cannot show using all the 9 ASME type requirements that integrity is maintained 10 although we do show that there's adequate cooling 11 maintained. There's no increase in the calculated i
12 radiological doses off-site and the negligible 13 contribution to the combustible gas inventory, given this,
(~ %!
(
s_/ 14 tritium being a combustible gas, and when you look at it 15 it turns out being less than an equiv-lent number of the 16 BPAs. So from that perspective it's leoa of an issue.
17 This is not intended to scare anybody, but we 18 did do some burst tests. This is the old program and does 19 show on a short segment that had a relatively typical 20 although in that amount of length you can't see it, 21 profile. What we were attempting to demonstrate here was 22 that under some of these very, very severe type 23 conditions, that we're not getting a rip up the length of 24 the cladding. Okay? And losing all of that neutron
) 25 absorbing material. In fact, it's fairly localized and NEAL R. GROSS CoVRT REPORTERS AND TRANSCRIBERS 1323 RHODE ISLAND AVE., N W.
(202) 234-4433 WASHINGTON, D C. 20005-3701 (202) 234-4433
j 454 i
, 1. this is why when we look at these sort of data and i 2 translate this:to a 12-footer, we're not expecting this to l' 3 grow, this part to grow in any amount because in fact, 4 during this period of -- or in this length we had a fairly 5 typical temperature profile, but we did lose some material I i j 6- and in a conservative estimate we did expand that to what f' 7 you'might expect in a 12-footer. So that's why we say in 1 8 a conservative estimate that in a LOCA that-as/much as one
- 9 pencil'might be lost.
1 1
10 Now there is a little bit of a safeguard in-11 this, that these are segmented, they are in zirc, they are 12 in pencils and there would-be a lot of that at high 13 temperature and that does play into how much is estimated f~n km 14 to be lost.
15 MEMBER-POWERS: Have you compared the 16 experience the people have had with testing short rod 17 assemblies of any type versus doing actual 12-footers like ;
18 in the FLHT tests? l 19 MR. ETHRIDGE: We have asked INEL, their LOFT 20 facility, to look at the differences between the short and 21 the 12-footers and the results from that analysis was that 22 we would not have to go through a full 12-foot length type 23 burst test that we could use these data for this analysis.
24 That's as much as I guess I can say about that.
25 MEMBER POWERS: I'm not really sure my memory NEAL R. GROSS COURT REPORTER". AND TRANSCRIBERS 1323 RHODE ISLAND AVE., N.W.
(202) 234-4433 WASHINGTON, D.C. 20005-3701 (202) 234-4433
455 1 fails, but my recollection was when they did the FLHT
, ~ .s 2 tests on the NRU reactor they found substantial i \
i
\!
3 differences between what 12-foot rods and what short t
4 segments did.
5 I don't know whether it's pertinent or not.
6 MR. ETHRIDGE: We did a lot of that testing 7 for NRC.
8 MEMBER POWERS: So you know all about it.
9 MR. ETHRIDGE: In fact, it was my department 10 that did some of that.
11 MEMBER POWERS: When you do the Appendix K i
12 type analyses on these rods, at least in your 13 documentation, you did not seem to include the hydrogen ,
l r~'s 4 l
's /
L 14 production from the steam reactions with the clad?
15 MR. ETHRIDGE: Hydrogen production, in what 16 way?
17 MEMBER POWERS: Metal water reaction with the 18 clad. l l
19 MR. ETHRIDGE: That, I understand, but I guess 20 I'm trying to understand where that might be an issue in 21 an Appendix K type, just the additional zire that you !
22 have?
23 MEMBER POWERS: Just the additional hydrogen j 24 production. You went through a fairly elaborate looking
/~h 1
(_) 25 at sources of hydrogen during accident situations for I
NEAL R. GIUDSS I COURT REPORTERS AND TRANSCRIBERS 1323 RHODE ISLAND AVE,, N W.
(202) 234-4433 WASHINGTON, D C. 20005-3701 (202) 234-4433
456 1 these rods, but it didn't seem to include that reactions
. 2 with the clad?
e 3 MR. ETHRIDGE: The clad stainless again?
4 MEMBER POWERS: That's right.
5 MR. ETHRIDGE: Okay. So we're not going to 6 have a zirc.
7 MEMBER KRESS: ATO is just as good as zirc.
8 MEMBER POWERS: At these temperatures.
9 MEMBER KRESS: Yes. Mixes hydrogen pretty 10 well. MR. ETHRIDGE: I think part of where 11 -- I know the specific answer to your question, but I 12 think where we may have come on this is that again we were 13 consistently comparing them to a BPA, all right? BPA is 5/ 14 in some cases a clad or stainless clad and so the ;
15 reactions there -- we need to say this, okay, the 16 reactions there or here should be no worse than they were 17 in that case.
18 MEMBER POWERS: I'm familiar 'th data bases l 19 on steam reactions with 304 stainless a a couple of 20 other stainless, but I have never seen any data on 316.
21 MR. ETHRIDGE: Okay.
22 MEMBER POWERS: One would guess that it would 23 be pretty close to 304, but there is that moly that you 24 i have to -- it's going to act funny.
O
( ,) 25 MR. ETHRIDGE: That might be something else NEAL R. GROSS COURT REPORTERS AND TRANSCRIBERS 1323 RHODE ISLAND AVE., N W.
(202) 234-4433 WASHINGTON, D C. 20005-3701 (202) 234-4433
457 1 that we note that we would look at.
,c3 2 Let me take kind of all of his and condense it
)
in the conclusions that are drawn in the report.
~
3 Again, 4 what we've attempted to do was to place this on a playing 5 field or on a slate that is comparing the BPAs with what 6 we're doing here and highlighting the differences.
7 The off-site releases under a varir?.y of 8 conditions are insignificant. The tritium is just really 9 held solidly in the matrix and you really have to go to 10 extreme temperatures to even get it out and so we have 11 confidence that the component in and of themselves in
?
12 these off or abnormal type events will hold that tritium.
13 They don't fail under the analyzed accidents
( s
's / 14 except the Appendix K type thing and the consequences of a 15 non-LOCA release are inconsequential to the requirements.
16 They've been demonstrated because they ere 17 specifically designed that way to be consistent with fuel 18 assemblies, the handling tools and so forth. All the 19 materials, I need to check the 316 versus 304, but have 20 been irradiated before either in test reactors or 21 commercial reactors. We showed basically no change from 22 the introduction of these on the thermal hydraulic aspects 23 of the core and essentially no impact on the reactor 24 neutronics. There's just too few of them to make a big
( ,) 25 difference.
NEAL R. GROSS COURT REPORTERS AND TRANSCRIBERS 1323 RHODE ISLAND AVE., N W.
(202) 234-4433 WASHINGTON, D.C. 20005-3701 (202) 234-4433
458 1 We don't make a big difference in the overall fs 2 results of a LOCA, the off-site consequences even if we I ) 3 dump all the tritium. You've got so much else to worry 4 about there that your increases are essentially not 5 existence.
6 TVA has taken a very careful look at their 7 technical specifications and we don't violate any of 8 those. In fact, we meet all of those and it's these and a 9 variety of other issues that give us confidence that with 10 the assistance of the utility that we can go and irradiate 11 this very small task for confirmatory LTA under 10 CFR 12 50.59.
13 MEMBER BARTON: You don't think there's any
,-)
(
\_/ 14 unreviewed safety question?
15 MR. ETHRIDGE: That's light 16 MEMBER BARTON: Good luck.
17 MEMBER POWERS: It would surprise the reactor 18 fuel subcommittee if there were no --
19 MEMBER BARTON: It would surprise the hell out 20 of me.
21 MR. ETHRIDGE: Well, I guess if that's really 22 the case you mentioned earlier Dana that there were some 23 additional comments or questions. We'd sure like to have 24 those.
(_,) 25 MEMBER BARTON: I think the way the Commission NEAL R. GROSS COURT REPORTERS AND TRANSCRIBERS 1323 RHODE ISLAND AVE., N W.
(202) 2344 433 WASHINGTON, D C. 20005-3701 (202) 234-4433 1
_ . . _ . . _ . _ _ _ . ~ . . . - _ . - . - . _ _ . - . - - . . - . . _ - - - - . .
)
459 l 1 is looking at 50.59 today I think you're going to be hard-a ;
2 pres' sed to try to get through the license amendment. I i 3-3- have.to look more and more at the process but my initial 4
l- 4 reaction.is that's going to be tough. _
I think the 5 shipping of irradiated stuff off site is going to be an 6 another issue, even though using approved DOE, casks, 1 7- etcetera etcetera, it's going to be a --
8 MR. ETHRIDGE: That is not an NRC issue.
- 9 That's a DOE shipment. I I I
! 10 MEMBER BARTON: How about the public?.
J 11' MR. ETHRIDGE: Well -- l l
i 12 MEMBER BARTON: How about making weapons I
- 13 material in a commercial reactor. That whole issue, I l 14 mean, it's no problem?
I
- 15 MR. ETHRIDGE
- I'm not saying that. There are
- 16 issues there and there are strategies in place to handle 17 that. And whether that's really an NRC issue is also in 18 question, I guess.
- 19 MEMBER POWERS
- My chairman is reminding me s
20 that I'm letting the schedule get out of hand, but I
.21 certainly appreciate --
22 MR. ETHRIDGE: Well, let me assist in helping i
i 23 to get back on schedule. 1 s i 24 MEMBER POWERS: You see how much we appreciate --
25 this is kind of' fun and exciting to see this new NEAL R. GROSS COURT REPORTERS AND TRANSCRIBERS ;
1323 RHODE ISLAND AVE., N.W. '
.(202) 234-4433 WASHINGTON, D C. 20005 3701 (202) 234 4433
460 1 innovative approaches to the use of commercial reactor.
s 2 Well, just the nuclear technology and in general I find it 1
('~'/ 3 very exciting and I have to admit that I was quite l 4
4 impressed with the documentation that I've looked at so l 5 far, as far as the thoroughness with which you have 6 scoured the regulations to make sure you've touched as l l
7 many bases as you could, at least at this stage in the 8 process.
1 9 I certainly appreciate you taking the time to 10 address us and keep us informed of what goes on here. We )
11 very much appreciated your oversight and introduction and 12 wondered and if you had any closing comments that you l
13 wanted to make or are we covered from soup to nuts here?
/~S..
r l
l k/ 14 MR. SOHINKI: I think there are a couple of 15 questions we need to look at and I appreciate your taking 16 the time for the presentation.
17 MEMBER POWERS: We very much appreciate you 18 guys 'oming down and talking to us and look forward, some 19 time maybe even come down and tell us the whole breadth 20 and scope and bring some of your competitor sister 21 organizations along and talk to us a little bit. These 22 are --
23 MR. ETHRIDGE: Some of the accelerator people 24 would love the break.
(-) 25 MEMBER POWERS: Mr. Chairman, I think I can --
NEAL R. GROSS COURT REPORTERS AND TRANSCRIBERS 1323 RHODE ISLAND AVE., N W.
(202) 234 4433 WASHINGTON, D C. 20005-3701 (202) 234-4433
461 1 unless there -- if the Committee has any questions now
.- 2 they'd like to pose? I think I can turn it over to you, 3 Mr. Chairman.
l 4 CHAIRMAN SEALE: Okay, I can't promise that 5 we'll be in communication with you, but then I can't l
6 promise you that you might not hear from some of our j l
7 members either. But thank you very much.
8 It's 3 o' clock. Somehow I think we could all )
i 9 use about a 15-minute walk and then be back here and we 10 will be discussing plant-specific applications of safety 11 issues.
i i
12 (Whereupon, the proceedings went off the l 13 record briefly at 3:00 p.m. and resumed at h i l V 14 3:15 p.m.) l i
15 CHAIRMAN SEALE: The next thing we want to !
1 16 talk about is Tom's letter while we still have a i l
17 transcript of the meeting available. We're told we're to l 18 do this because we didn't have a specific identification 19 of this item in the " Federal Register" notice. And so we l 20 need a good record so that we can ensure people that we l l
21 haven't tried to do anything sub ROSA.
1 1
22 MEMBER FONTANA: Wrong word. ROSA?
23 VICE CHAIRMAN POWERS: He meant to say sub 24 OSU, but --
() 25 CHAIRMAN SEALE: Okay. Tom, do we have a NEAL R. GROSS COURT REPORTERS AND TRANSCRIBERS 1323 RHODE ISLAND AVE., N W.
(202) 234-4433 WASHINGTON, O C. 20005-3701 (202) 234-4433
462 1 letter on risk-based --
y3 2 MEMBER KRESS: It's white. I don't know. Do
( \
3 we have one out here?
4 CHAIRMAN SEALE: It's buff.
5 MEMBER KRESS: It's green? Sorry. Buff?
j 6 CHAIRMAN SEALE: Buff.
7 MEMBER KRESS: Green, green. Which one are we 8 working on?
9 CHAIRMAN SEALE: I'm sorry. I beg your 10 pardon. Plant-specific application of safety goals.
11 That*s the buff one.
12 MEMBER KRESS: It's green. It's green. Oh, 13 it's buff. Okay.
I\
\- 14 CHAIRMAN SEALE: It's buff. Yes.
15 MEMBER KRESS: Buff. Now, how do you want to l 16 approach this, Bob? Do you want me to do this the new 1
17 process, where I tell you what's in it, or do you want me 18 to read it?
l 19 CHAIRMAN SEALE: Well, I think we did this 20 once before, didn't we?
21 MEMBER KRESS: We did, yes, but it's been a 22 long time.
23 CHAIRMAN SEALE: All right. Well, why don't 24 you give us a --
O
( ,) 25 MEMBER KRESS: I could read it. It's NEAL R. GROSS COURT REPORTERS AND TRANSCRIBERS 1323 RHODE ISLAND AVE., N W.
(202) 234 4433 WASHINGTON. D C. 20005-3701 (202) 234-4433
463 1 basically the same stuff redistributed and incorporating e
-, 2 some of the comments that were made by some of the people.
s
!'# I 3 CHAIRMAN SEALE: Okay. Go ahead and read it.
4 MEMBER KRESS: Do you want me to just read it?
5 That would be the best way. Okay. Now, you might hear 6 some slight differences in your buffed one, but don't pay 7 much attention to those. The title is different, 8 " Risk-Based Regulatory Acceptance Criteria for
?
9 Plant-Specific Application of Safety Goals."
10 "During the 439th meeting of the Advisory 11 Committee on Reactor Safeguards, March the 6th through 12 8th, 1997," -- this is not in yours; just pay attention --
13 'we continued our deliberations on the issue of p
k_ 14 plant-specific application of safety goals.
15 "In an earlier ACRS report on this subject 16 dated November 18th, 1996, we stated that 'The safety 17 goals can and should be used to derive guidance for 18 plant-specific application.'
19 "We also noted that, instead of using Level 3 20 PRA evaluations to determine plant status with respect to 21 the quantitative health objectives, QH0s, it would be a 22 pragmatic approach to use lower-tier criteria on core 23 damage frequency, CDF, and a large early release 24 frequency, LERF, or a conditional containment failure gs
() ,
25 probability, CCFP, so long as these are chosen in such a NEAL R. GROSS COURT REPORTERS AND TRANSCRIBERS 1323 RHODE ISLAND AVE., N W.
(202) 234-4433 WASHINGTON, D.C. 20005-3701 (202) 234-4433
I 464 1 way as to ensure that the OHOs are satisfied.
2 "This report further discusses the need for
/,_,T
'~' 3 plant-specific application of risk acceptance criteria and 4 the appropriateness of these being derived from the safety 5 goal QHO on early fatalities.
6 "The attachments to this report are intended 7 to provide detailed guidance for quantifying such l 8 lower-tier acceptance criteria in terms of CDF and LERF or 9 CCFP that will ensure that the early fatality QHO is met 10 at each site. These methods can be used by the staff to 11 evaluate the appropriateness of their choice of LERF as 12 diecussed in the draft regulatory guide and draft standard 13 r3 view plans in support of risk-informed regulation.
(, l 14 "Need for Plant-Specific Application. The 15 safety goal policy statement makes it clear that the QHOs 16 and the ancillary goal on CDF were intended only to 17 provide standards for the NRC to use in order to judge the 18 overall effectiveness of its regulatory system. The 19 policy statement specifically precludes enforcement of 20 such goals on a plant-specific basis. !
21 "As part of the preparation of the draft l l
22 regulatory guide and the standard review plan in support {
i 23 of risk-informed regulation, the staff has found it !
24 necessary to develop risk acceptance guidelines that can n
( ,) 25 be applied on a plant-specific basis. These are used NEAL R. GROSS COURT REPORTERS AND TRANSCRIBERS !
1323 RHODE ISLAND AVE., N.W. l (202) 234-4433 WASHINGTON, D.C. 20005-3701 (202) 234-4433
465 1 along with other considerations and input for making
,s 2 judgments on the acceptability of requested changes to a i)
3 licensee's current licensing basis. Allowing such 4 plant-specific application would be a positive action 5 towards implementing the move towards risk-informed 6 regulation.
7 "We also note that as the state of the art of 8 PRA improves in the long term, the Commission may want to 4 9 consider having a quantified acceptable level of risk to 10 replace the basically unquantified ' adequate level of 11 protection.' This could eventually serve as a risk 12 acceptance criterion for enforcement decisions.
13 " Discussion. Risk-Informed Regulation." That
/)i 14 word " Discussion" shouldn't be in there. " Risk-informed 15 regulation. The Commission has directed the staff to 16 increase its use of PRA in a more risk-informed regulatory 17 climate.
18 "The ACRS has endorsed this because we believe 19 that a risk-informed regulatory approach will lead to 20 increased coherence in the regulatory system, enhance 21 decision-making ability and into a technically defensible 22 basis for granting regulatory relief.
23 "A risk-informed regulatory system cannot be 24 implemented without the existence of top-level risk
(
\ ,) 25 acceptance criteria. The obvious choice for these is the NEAL R. GROSS COURT REPORTERS AND TRANSCRIBERS 1323 RHODE ISLAND AVE., N.W.
(202) 2344 433 WASHINGTON, D C. 20005-3701 (202) 234-4433
466 1 NRC's safety goal quantitative health objectives. As it
,s 2 is NRC's responsibility to license and assure adequate
(' ')
~
3 protection for individual plants, then there seems to be 4 no alternative for plant-specific application.
5 " Relationship Between Adequate Protection and 6 the Safety Goals. Currently licensing acceptance criteria 7 are embodied in the concept of ' adequate level of 8 protection.' With this concept, a given plant is deemed 9 to provide adequate protection; that is, acceptable risk, 10 if it complies fully with all the rules and regulations.
11 "As there are many ways plants can be designed 12 and operated within the confines of the regulations, the 13 natural result is a spectrum of risk levels across the
\/
- 14 population of operating plants. It is apparent, then, 15 that an absolute risk level to be associated with adequate 16 protection is basically unquantified.
17 "Since each plant must, each licensed plant 18 must, by definition provide adequate protection, then the 19 licensed plant that poses the highest level of risk puts a 20 least upper bound on a single value level of risk to be 21 associated with adequate protection.
22 "Within the spectrum of risks, it is likely 23 that there are plants with risk levels above the safety 24 goals and other plants with risk levels below. If this is
(
(),) 25 indeed the case, a single risk level to be associated with NEAL R. GROSS COURT REPORTERS AND TRANSCRIBERS 1323 RHODE l$ LAND AVE., N W.
(202) 234-4433 WASHINGTON, D.C. 20005-3701 (202) 234-4433
-467 d
j 1 adequate' protection would be a' risk level; greater than the
- 2 safety goals. '
t,O-
) 3' "For those' plants with risk levels below the
- 4 safety goals, the difference in risk level can be viewed 5 as margin. It is from-the unnecessary portion of such 6 margin that plant-specific regulatory relief can be 7 granted.
~
l -8= "For those plants with risk levels greater
- 9 than the safety' goals, the challenge will be to eventually i
f 10 reduce their. risk to below the safety goal level within 4
11 the confines of the backfit rule.
12 " Regulatory Transparency. -The unquantified
, 13 adequate protection concept is not well-understood by the O
Tw/ 14 general public because it is unfamiliar with the 15 regulatory process in the entirety of the body of nuclear 16 regulations with their underlying technical bases.
-17 "We believe a well-articulated and quantified 18 acceptable level of risk that would replace the adequate 19 protection concept would enhance the public's 20 understanding and acceptance of the regulatory process and 21 if enforced would lead to a more uniform level of 22 protection to all individuals living in the vicinity of 23 the various licensed nuclear plants.
24 " Safety Goals as Risk Acceptance Criteria. It
) 25 is our opinion that the appropriate choice for an NEAL R. GROSS COURT REPORTERS AND TRANSCRIBERS 1323 RHODE ISLAND Ai'E., N.W.
(202) 234-4433 WASHINGTON, D C. 20005 3701 (202) 234-4433
468' 1 acceptable risk criterion for plant-specific application
,-~s 2 is the safety goal QHO for early fatalities. In addition, i \
3 we believe the ancillary CDF goal should be elevated to 4 the status of a fundamental goal. We have several reasons 5 for these beliefs.
6 "One, the safety goals were the expression by 7 NRC for: How safe is safe enough? We think this is what 8 risk acceptance criteria ought to be.
9 "Two, the early fatality QHO generally 10 controls the risk if it is met and the QHO on latent 11 effects is also met except under very special 12 circumstances. Elevating the CDF ancillary goal to the 13 status of a fundamental goal can be considered as a
,O
's / 14 defense-in-depth principle that provides balance between 15 prevention and mitigation.
16 " Lower-Tier Acceptance Criteria. Attachment 1 17 is a discussica paper by ACRS Member T. S. Kress that 18 provides his proposal on how lower-tier criteria in terms 19 of LERF and CCFP can be derived directly from the early 1
20 fatality QHD.
21 " Attachment 2, developed by ACRS Fellow 22 Richard Sherry, provides an alternativ definition for 23 LERF.'
24 MEMBER SHACK: I guess I have a problem. I (3
(,) 25 think the concept of adequate protection is not only not l NEAL R. GROSS COURT REPORTERS AND TRANSCRIBERS 4 1323 RHODE ISLAND AVE., N.W. l (202) 234-4433 WASHINGTON, D.C. 20005-3701 (202) 234-4433 l l
469 l
1 understood by the general public. I'm not sure it's l
l
,- 2 understood by this Committee.
i /
('~')
3 I thought Joe Murphy's theory was that 4 adequate protection was what you had through any cost.
5 MEMBER KRESS: Well, that concept and this 6 concept are not in any --
7 MEMBER SHACK: I think it is, though, because 1
8 I think he would argue that what you get by satisfying all )
l 9 the rules and regulations ir some level of protection l 10 above adequate protection. !
l l
11 MEMBER KRESS: That's not in conflict with 12 what I say.
13 MEMBER APOSTOLAKIS: Well, last time we had a
,n,
- 14 disagreement on that, didn't we?
15 MEMBER SHACK: Yes.
16 MEMBER KRESS: I changed it.
17 MEMBER APOSTOLAKIS: Adequate protection means 18 complying with the regulations, does it not?
19 MEMBER SHACK: No. I thought adequate 20 protection meant the level at which you could be forced to 21 achieve regardless of any economic cost, 22 MEMBER APOSTOLAKIS: Yes. And I thought that 23 such a concept did not exist.
l l 24 VICE CHAIRMAN POWERS: I mean, the situation (G) ,
25 was that the courts ruled that the NRC was obligated to NEAL R. GROSS COURT REPORTERS AND TRANSCRIBERS 1323 RHODE ISLAND AVE., N.W.
(202) 234-4433 WASHINGTON, D.C. 20005-3701 (202) 234-4433
470 i 1 assure adequate protection without regard to cost. But to f_ 2 do that, they had to define what adequate protection was.
i 3 The NRC came along and said adequate 4 protection is what Tom says: compliance with the rules 5 and regulations. Now, I question the " fully" because 6 there are exemptions granted and things like that --
7 MEMBER KRESS: Well, that's --
8 VICE CHAIRMAN POWERS: -- within that context.
9 MEMBER KRESS: -- why I feel --
10 VICE CHAIRMAN POWERS: I think that's 11 understood.
12 MEMBER APOSTOLAKIS: But there's no concept 13 there that this is the absolute minimum.
14 MEMBER SHACK: But you apply things through 15 the backfit rule.
16 VICE CHAIRMAN POWERS: If a plant does not 17 provide adequate protection, it is shut down.
18 MEMBER APOSTOLAKIS: Right. But the notion 19 was: Here is a bunch of regulations. If you meet them, i
20 then we have provided adequate protection to the public.
21 MEMBER KRESS: That's correct.
22 MEMBER APOSTOLAKIS: And in deciding whether l
23 to meet them or not, cost is not a factor. l 24 VICE CHAIRMAN POWERS: Exactly. l
,r 3
(,) 25 MEMBER APOSTOLAKIS: This is very different NEAL R. GROISS l COURT REPORTERS AND TRANSCRIBERS 1323 RHODE ISLAND AVE., N.W. l (202) 234 4433 WASHINGTON, D C. 20005-3701 (202) 234-4433
471 l 1 from saying adequate protection is the absolute minimum l
f3 2 that you have to do, regardless of cost, and then the
- )
'~'
3 license depends on something else, which I think is what 4 Bill is saying.
5 And last tirte we had this debate, there isn't 6 such a thing. If you meet my rules --
7 MEMBER KRESS: You have provided --
8 MEMBER APOSTOLAKIS: -- you've provided 9 adequate protection.
10 MEMBER KRESS: That's what I'm saying.
11 MEMBER APOSTOLAKIS: If I change my rules 12 tomorrow, then you change what you're doing. You still 13 provide adequate protection. And I think that's
(\ ') 14 consistent --
15 MEMBER SHACK: I certainly agree you need all 16 the rules that you have adequate protection.
17 MEMBER APOSTOLAKIS: Right.
18 MEMBER SHACK: What I'm not sure is that the 19 converse is true that in order to provide adequate 20 protection, you have to fully comply with all the rules 21 and regulations --
22 VICE CHAIRMAN POWERS: Yes. I think you --
23 MEMBER SHACK: -- because I'm not sure all the 24 rules and regulations could be imposed if their cost was rx
,) 25 infinite.
NEAL R. GROSS COURT REPORTERS AND TRANSCRIBERS 1323 RHODE ISLAND AVE., N.W.
(202) 234-4433 W/ ohlNGTON, D C. 20005-3701 (202) 234-4433
472 1 VICE CHAIRMAN POWERS: Well, I think it's a f3 2 tautology. And, consequently, the statement and its l k ]
3 converse or contrapositive work. That is, I mean, any 4 plant that doesn't have adequate protection within the 5 existing set of rules, they impose a new rule. And they 6 do that through the licensee commitments, which are part 7 of the rules and regulations.
8 MEMBER SHACK: But you can now impose through 9 the backfit rule a regulation if it meets a certain 10 economic cost.
11 VICE CHAIRMAN POWERS: That's right.
12 MEMBER SHACK: That same regulation if it
,_ 13 costs mcre would not be deemed necessary. How can that be i
)
\/ 14 part of adequate protection, then, if adequate protection 15 has to be applied regardless of cost?
16 MEMBER APOSTOLAKIS: Well, that's why the 17 system is incoherent.
18 VICE CHAIRMAN POWERS: No. The situation is 19 this, Bill, that if I am a plant, a particular plant, and 20 the new rule you come along with costs more than the 21 benefit I particularly derive, tough because it's done on 22 the fleet of plants. Now, I can apply for an exemption, 23 and I can do a lot of things.
24 If I'm a plant that has not been built and I r^%
k._,) 25 come along and say, "What rules do I have to comply to?";
NEAL R. GROSS COURT REPORTERS AND TRANSCRIBERS 1323 RHODE ISLAND AVE.. N W.
(202) 234-4433 WASHINGTON, D C. 20005-3701 (202) 2344433
473 1 I have to comply to that rule that was a backfit, s 2 regardless of cost. I don't even get a vote in this
\
i '
L./
3 anymore.
4 MEMBER SHACK: Okay. Well, I guess I'm still 5 not convinced that the two are exactly the same, but --
6 MEMBER KRESS: Yes.
7 CHAIRMAN SEALE: Yes, Noel?
8 MR. DUDLEY: If a plant is able to comply with 9 the safety goals without meeting all of the regulatory 10 requirements, is it meeting adequate protection 11 requirements?
12 MEMBER APOSTOLAKIS: Right now, no. No.
13 There has to be a new policy statement.
x-) 14 MEMBER SHACK: Well, but Murphy's diagram 15 showed adequate protection well below the safety goals.
16 MEMBER KRESS: No, no.
17 VICE CHAIRMAN POWERS: But I think it is 18 absolutely --
19 MEMBER KRESS: l You're talking about safety l
20 now. Be sure. l 21 MEMBER APOSTOLAKIS: Yes. It was the reverse.
22 You remember?
23 MEMBER KRESS: It's one over risk.
24 MEMBER APOSTOLAKIS: It's one over.
rs
() 25 MEMBER SHACK: Okay. But the point is --
NEAL R. GROSS COURT REPORTERS AND TRANSCRIBERS 1323 RHODE ISLAND AVE., N.W.
(202) 234 4433 WASHINGTON, D C. 20005-3701 (202) 234-4433
474 1 MEMBER KRESS: I'm talking risk here. So you
.s 2 want to invert his diagram.
(, \ \
I
\
3 MEMBER SHACK: Okay. If I inverted it, it l 1
4 would be well above.
5 MEMBER KRESS: Right. And that's not 6 inconsistent with what I say. l 7 VICE CHAIRMAN POWERS: But I think Noel is i l
l 8 absolutely correct that I could design a plant --
9 MEMBER KRESS: To meet the safety goals --
10 VICE CHAIRMAN POWERS: -- if it was in 11 violation of 90 percent of the rules and still satisfy the 12 safety goals.
l 13 MEMBER KRESS: I agree with that. And it does !
y-~n
- 14 not -- then you're not providing adequate protection.
15 MEMBER APOSTOLAKIS: No. You're not. Well, 16 no.
17 (Laughter.)
18 MEMBER APOSTOLAKIS: If you license the plant, 19 you do. See, that's the thing, that tautology. If the 20 NRC says --
21 VICE CHAIRMAN POWERS: No, no. The plant 22 would never get licensed. But I could still design one 23 such that that was the case.
24 MEMBER APOSTOLAKIS: I can give the plant a
(~.,
() 25 large number of exemptions.
NEAL R. GROSS COURT REPORTERS AND TRANSCRIBERS 1323 RHODE ISLAND AVE., N W.
(202) 234-4433 WASHINGTON, D.C. 20005-3701 (202) 234-4433
1 475 1 MEMBER KRESS: A huge number.
,y 2 VICE CHAIRMAN POWERS: And I would apply for
( )
'~'
3 those exemptions through the risk-informed regulation 4 process.
5 MEMBER APOSTOLAKIS: Yes. Then it will have l
6 adequate protection. Yes.
7 VICE CHAIRMAN POWERS: Then it would magically l 8 become adequately protected, -- )
l 9 MEMBER APOSTOLAKIS: Exactly. l 10 VICE CHAIRMAN POWERS: -- even though I --
11 MEMBER APOSTOLAKIS: When the NRC says you are 12 licensed to operate --
13 MEMBER KRESS: I think you guys have got it
,x r ;
k/ 14 right. You've got it right.
15 MEMBER APOSTOLAKIS: Yes.
16 MEMBER KRESS: That's the concept, yes. I 17 agree. And that's certainly not --
18 MEMBER APOSTOLAKIS: No. It's not 19 inconsistent with what you say.
20 So what are we recommending the Commission to 21 do: to abandon the whole concept of adequate protection 22 and --
23 MEMBER KRESS: I very carefully say they ought 24 to think about that as a long-term goal.
/
(N) 25 MEMBER APOSTOLAKIS: What was the title again, NEAL R. GROSS COURT REPORTERS AND TRANSCRIBERS 1323 RHODE ISLAND AVE., N W.
(202) 234-4433 WASHINGTON, D.C. 20005-3701 (202) 234-4433
l 476 1 the new title, because it's not the same as here?
g3 2 MEMBER KRESS: The reason the title changed,
'^'
3 Noel changed that because it just so happens that "The l 1
l Plant-Specific Application of Safety Goals" was the title i 5 of an old letter we had.
6 MEMBER APOSTOLAKIS: Yes.
7 MEMBER KRESS: And we didn't want to reproduce 8 the same title because the only thing that distinguishes 9 the two letters, then, is the date. So we wanted a new --
10 VICE CHAIRMAN POWERS: Well, you could put it 11 "The Sequel" or " Number Two."
12 MEMBER KRESS: Yes, "The Sequel" or something 13 like that. So the title in my mind is open for discussion
/~N f )
\/ 14 in what everyone --
15 VICE CHAIRMAN POWERS: Could I ask --
16 MEMBER APOSTOLAKIS: There's another question, 17 then. What question from the Commission are we answering?
18 Is there an SRM?
19 VICE CHAIRMAN POWERS: No. We're answering 20 the question of: Tom, where the hell are your risk 21 acceptance criteria?
22 MEMBER KRESS: That's pretty much it, yes.
23 MEMBER APOSTOLAKIS: But, I mean, I would like l 24 to hear the actual question from the Commission.
?
,,3
( ,) 25 MEMBER KRESS: They've asked us two things.
NEAL R. GROSS COURT REPORTERS AND TRANSCRIBERS 1323 RHODE ISLAND AVE., N.W.
(202) 234-4433 WASHINGTON, D.C. 20005-3701 (202) 234-4433
477 1 One is during our meeting, I told them that the LERF and j i
I 73 2 the CDF could be derived from the QHOs. And they said, j N-] 3 "All right. Show us how." That's Question Number 1.
4 Question Number 2, they've asked us since in a l f
I 5 different context as to what we think adequate protection )
6 is with respect to the safety goals. What are their l t
7 relationships? l I
8 MEMBER APOSTOLAKIS: Yes. !
i 9 MEMBER KRESS: So it answers that question. t I
10 And then we've talked about: Why should they be -- if you l 11 had these lower-tier acceptance criteria, why should they
)
12 be applied on a plant-specific basis? And so they've j i
i p_ 13 asked us about that. So it's basically those three things j (x -) 14 intertwined.
i 15 MEMBER APOSTOLAKIS: I would like the j 16 introduction to the letter to actually quote the l 17 questions. What do you think?
t 18 MEMBER KRESS: I wouldn't like that because ;
19 you're going to have a hell of a time finding something to i
20 quote it on. !
21 MEMBER APOSTOLAKIS: They must be in specific j 22 SRMs.
l 23 MR. DUDLEY: From a January 14th, 1997 SRM, it ;
24 identifies the issues that were discussed at the December (g
(_j 25 meeting with the Commission, one of those items being use NEAL R. GROSS '
COURT REPORTERS AND TRANSCRIBERS 1323 RHODE ISLAND AVE., N.W.
(202) 2344433 WASHINGTON, D.C. 20005-3701 (202) 234 4433
478 1 of safety goals on a plant-specific basis. The paragraph
,- 2 that makes --
(
3 MEMBER APOSTOLAKIS: What exactly does that 4 mean?
5 MR. DUDLEY: Okay. This is the paragraph that 6 asks the question.
7 MEMBER APOSTOLAKIS: Okay.
8 MR. DUDLEY: "The ACRS should continue to be 9 forward-looking to bring developing concerns to the 10 Commission's attention and continue follow-up on issues 11 such as digital I&C and use of safety goals for regulatory 12 purposes. In this regard, the Commission would be 13 interested in the ACRS view on the relationship between
,/m 5
)
\_./ 14 the concept of adequate protection as used in the NRC 15 regulations and the NRC safety goals from the stanup lau 16 of level of risk."
17 MEMBER KRESS: That's one thing.
18 MEMBER APOSTOLAKIS: Yes. So why don't we 19 take that quote and put it up front?
20 MEMBER KRESS: Because you've got three 21 others.
22 MEMBER KRESS: Well, we'll put those, too.
23 MEMBER BARTON: You could reference this 24 document.
,m
(/)
s_
25 MEMBER APOSTOLAKIS: I think we should, yes.
NEAL R. GROSS COURT REPORTERS AND TRANSCRIBERS 1323 RHODE ISLAND AVE., N W.
(202) 234-4433 WASHINGTON, D C. 20005-3701 (202) 234-4433
479 1 MEMBER KRESS: Well, we could reference it as !
- 7. s 2 a document, sure.
l }
3 MEMBER APCSTOLAKIS: I propose that we --
4 MEMBER KRESS: We also had a Commission 5 meeting where I told them we could derive the lower-tier 6 acceptance criteria. And they said, "Sure. Go do it."
7 And I don't know where that --
8 MEMBER APOSTOLAKIS: Well, there must be a 9 transcript.
10 MR. DUDLEY: That came out of the -- that SRM l
11 came out of the same -- l 12 EXECUTIVE DIRECTOR LARKINS: That was in the 13 earlier meeting. It was posited, the question, from i ') I
\_/ 14 Rogers. 1 1
15 MR. DURAISWAMY: Yes. Also, that was in the l 16 same meeting from -- i l
17 MEMBER KRESS: What are you going to quote, 18 the transcripts?
19 MR. DURAISWAMY: Yes. It's in the transcript, 20 yes.
21 MEMBER KRESS: See, that's my problem.
22 MEMBER APOSTOLAKIS: Why is that a problem?
23 MEMBER KRESS: Quoting the transcripts? I 24 don't think we ever do that.
,i m (w/ ) 25 MEMBER APOSTOLAKIS: If you are responding to NEAL R. GROSS COURT REPORTERS AND TRANSCRIBERS 1323 RHODE ISLAND AVE., N W.
(202) 234-4433 WASHINGTON, D.O 20005-3701 (202) 234-4433
480 1 something that came up then --
7s 2 EXECUTIVE DIRECTOR LARKINS: You could say i 2
=
i/
3 questions that came up during meetings.
4 VICE CHAIRMAN POWERS: I'm not unsympathetic 5 with George's requent, but I'd like him to articulate 6 exactly why he wants this specificity of the quotation.
7 MEMBER APOSTOLAKIS: Because I thin'c it 8 completes the letter. In other words, you start out by 9 saying, "Here is what you asked us to do, and we're doing 10 it." ,
l 11 Right now if you read the first two l l
12 paragraphs, you really don't know why the rest folliws. I l 13 mean, it says, "This report further discusses the need for i p-s ,
Esl 14 plant-specific application" and so on.
s 15 VICE CHAIRMAN POWERS: But wouldn't he satisfy 16 that with a little more collegial introduction? I think 17 this Commission has asked us to be more collegial. And 18 he's responding in a collegial fashion.
19 MEMBER APOSTOLAKIS: I thought that we always 20 made sure that in the introductory paragraph we stated 21 what we were responding to.
22 CHAIRMAN SEALE: Certainly it's desirable to 23 cut off at the pass, if you will, any suggestions that 24 we're fishing in somebody else's lake. And I can see
/^5
( ,
) 25 where some people might figure: Why in the world is the NEAL R. GROSS COURT REPORTERS AND TRANSCRIBERS 1323 RHODE ISLAND AVE., N W.
(202) 2344433 WASHINGTON, D C. 20005-3701 (202) 234-4433
481 1 ACRS worrying about this?
f-~s> 2 MEMBER APOSTOLAKIS: Yes, yes. l
( ! l
'~'
3 CHAIRMAN SEALE: And the answer is they asked l 4 us to.
5 MEMBER APOSTOLAKIS: We were specifically l
6 asked to do it.
7 MEMBER BARTON: You can be more general in 8 response to questions asked by the Commission or something 1
9 in --
10 MEMBER APOSTOLAKIS: Right. I mean, I found i 11 what Noel just read very enlightening. This is what they 12 want.
13 MEM.BER KRESS: The only reason they want --
f- 3 I
's / 14 VICE CHAIRMAN POWERS: But I'm not sure that 15 it needs to be a quotation. I think I understand the 16 difficulty because the first part of the question, not the 17 adequate protection part of the queetion, but the first 18 part, was given in conversational language. If you quote 19 it exactly, it's going to look like --
20 MEMBER APOSTOLAKIS: Well, no. I understand 21 that. And maybe that part, we can say, "Furthermore, 22 during our meeting with the commission in December, the 23 question was asked," and paraphrase it. But if we have an 24 SRM, it seems to me it would be best to flow and not quote (3
) 25 from the transcript. I will agree with you on tnat.
NEAL R. GROSS COURT REPORTERS AND TRANSCRIBERS 1323 RHODE ISLAND AVE., N W.
(202) 234-4433 WASHINGTON, D C. 20005-3701 (202) 234-4433
l 482 1 MEMBER KRESS: Okay.
7 e' 2 MEMBER APOSTOLAKIS: And you just say, "During
( )
3 the meeting, we were asked" because I would like these 4 documents to be as self-contained as possible. And I 5 think it helps if you know why we are writing this l
6 particular report. i 1
7 So I think you very clearly answered the l
8 question of the relationship between adequate protection ;
l 9 and risk. Now, another concern now that comes to my mind l 10 is -- you see, that's why I want an exact quote.
l 11 What else did they want? They wanted to know 12 what -- Noel, would you do me a favor and read again the 13 relevant passage there?
!k '! 14 MR. DUDLEY: Certainly. "In this regard, the 15 Commission would be interested in the ACRS views on the 16 relationship between the concept of adequate protection as 17 used in the NRC regulations and the NRC safety goals from 18 the standpoint of level of risk."
19 MEMBER APOSTOLAKIS: Okay. That we have 20 answered, yes.
21 VICE CHAIRMAN POWERS: Tom, while he considers 22 that, let me ask you one question. In here you have a 23 sentence that says, " risk-informed regulatory system 24 cannot be implemented without the existence of top-level
,3
(,/ 25 risk acceptance criteria."
NEAL R. GROSS COURT REPORTERS AND TRANSCRIBERS 1323 RHODE ISLAND AVE., N.W.
(202) 234-4433 WASHINGTON, D.C. 20005-3701 (202) 234-4433
483 1 It seems to me that I could do that. I could
, 2 say, "Look, I will formulate all of these prescriptive
('"')
3 regulations in the areas that risk analysis tells me are 4 important." And I don't have to implement any risk 5 acceptance criteria.
6 MEMBER KRESS: Well, you don't know how 7 important they are unless you have some sort of risk 8 acceptance criteria.
9 VICE CHAIRMAN POWERS: What you're saying is 10 that there will always be a riskiest aspect of a plant and 11 I could always impose regulations on it, even though it 12 was orders of magnitude below any rational goal?
13 CHAIRMAN SEALE: Yes. The thing with risk --
,a V 14 VICE CHAIRMAN POWERS: Could I suggest, then, 15 it is that the risk-informed regulatory system ought not 16 be implemented without --
17 MEMBER KRESS: I don't think it can, but I 18 would be willing to put in an "ought not." I think 19 "cannot" is right because I don't think it would be truly 20 risk-informed otherwise.
1 21 MEMBER APOSTOLAKIS: That's correct. l l
I 22 MEMBER KRESS: But I would be willing to put l l
l 23 "ought not" if that's what the Committee wants.
l 24 MEMBER SHACK: Go back to this " risk-informed"
,f 3
) 25 and " risk-based." I think Dana has it right. In a NEAL R. GROSS COURT REPORTERS AND TRANSCRIBERS 1323 RHODE (SLAND AVE., N.W.
(202) 234-4433 WASHINGTON, D C. 20005-3701 (202) 234-4433
484 1 risk-informed, you ought not, but you could.
7.s 2 MEMBER MILLER: Risk-based, it should not.
! i 3 CHAIRMAN SEALE: Yes. It's the sole question 4 of: How safe is safe enough again?
5 MEMBER APOSTOLAKIS: But, you see, that's 6 related to one of my concerns. That's why I really wanted 7 to know exactly what question they asked.
8 CHAIRMAN SEALE: Yes.
9 MEMBER APOSTOLAKIS: Does the Commission 10 expect us to say something about what they should do if 11 the QHOs do become risk acceptance criteria and you have a 1
12 number of plants that are above the goal? That's a l l
13 serious policy issue for them. And this data says nothing !
r"N, l
-- 14 about it.
15 MEMBER KRESS: What? l l
16 MEMBER APOSTOLAKIS: Does it? Where?
17 MEMBER KRESS: Absolutely.
18 MEMBER APOSTOLAKIS: Where?
19 MEMBER MILLER: Near the end here.
20 MEMBER KRESS: Read this paragraph on the --
21 MEMBER APOSTOLAKIS: Lines?
22 MEMBER KRESS: Line 78.
23 MEMBER APOSTOLAKIS: Okay. Seventy-eight?
24 No. You have a different letter.
,r x (x,,) 25 MEMBER KRESS: Line 75.
NEAL R. GROSS COURT REPORTERS AND TRANSCRIBERS 1323 RHODE ISLAND AVE., N.W.
(202) 234-4433 WASHINGTON, D.C. 20005-3701 (202) 234 4433
l 485 l
1 MEMBER APOSTOLAKIS: Seventy-five.
s 2 MEMBER KRESS: Through 78.
\
'"J 3 MEMBER APOSTOLAKIS: Okay. "It is from the 4 unnecessary portion of such margin that plant-specific 5 regulatory relief can be granted.
6 "For those plants with risk levels greater, 1
7 the challenge will be to eventually reduce." Yes. And ,
1 8 they might come back and say, "Yes. We realize we have a l 9 challenge, and we want your opinion on how to do it." l 10 See, that ; why I want to know precisely what they said.
11 MEMBER KRESS: Okay. They didn't say anything 12 about that. But we shouldn't confine our advice to things 13 they asked us. Also read Lines as soon as I can find it, t'
' l 14 35 through 39.
l 15 MEMBER APOSTOLAKIS: Okay. " Quantified 16 acceptable level of risk to replace the basically." Yes, 17 sure. I agree with that, but let's think about it 18 logically, though.
19 On the one hand, as a Committee, we're saying 20 you really cannot have a risk-informed system without risk 21 acceptance criteria. So you must make the goals risk 22 acceptance criteria.
23 Now, the moment you do that, you have a 24 problem with the plants that violate it.
p.
(_) 25 MEMBER KRESS: Right.
NEAL R. GROSS COURT REPORTERS AND TRANSCRIBERS 1323 RHODE ISLAND AVE., N W.
(202) 234-4433 WASHINGTON, D C. 20005-3701 (202) 234-4433
486 1 MEMBER APOSTOLAKIS: And all we're saying
,y 2 there, it's a challenge. Well, I mean, you can't have it
\ l 3 both ways because if they violate it, you shut them down.
l 4 MEMBER KRESS: No, you don't. That's just the l 5 point. You make --
6 MEMBER APOSTOLAKIS: Why?
i 7 MEMBER KRESS: Because you still have an 1 8 adequate protection regulatory system, and you're not 9 going to replace that all at once. I'm saying in the long l l
10 term, it would be nice to force everybody below some l l
11 risk-acceptable level. You cannot do it by saying --
l 1
12 MEMBER BARTON: You can't do it by throwing a l
13 light switch here.
/, T r ;
14 MEMBER KRESS: No. And so you can't just do 15 that all at once. I'm saying, "Well, that's the goal in 16 the long run," but you ought to think about it. And you 17 ought to figure out how to do it. And it's a challenge to 18 do it within the backfit rule because we're never going to l 1
19 change that rule.
20 MEMBER SHACK: Actually, it looks like you can 21 because if I read 50.109, I have an out here that says I 22 can -- the regulatory action involves defining or 23 redefining what level of protection should be regarded as 24 adequate.
,o
( ,)
25 MEMBER KRESS: Yes. So you could redefine the NEAL R. GROSS COURT REPORTERS AND TRANSCRIBERS 1323 RHODE ISLAND AVE., N.W.
(202) 234-4433 WASHINGTON, D C. 20005-3701 (202) 234-4433
.. . - . . . . . . - - - . . .. , -. - - - .. ~ . - - .. . . . ~ - ....-
i '
487 '
}
1 level of protection. Then you've got i t' m a d e'. !
)
2 . MEMBER SHACK: I've got it made. ,
- 3 MEMBER APOSTOLAKIS
- -See, that's my point. -
1 J
But.if they say that QHOs-are on the adequate'l'evel, .then 4 l t
- j. 5 the others are inadequate.
6- MEMBER KRESS: That's right. Then you can
'7 force'them to do it.
j 8 MEMBER APOSTOLAKIS: That's precisely.my 4
9 question. Are they asking this Committee to give them 10 advice regarding that whole issue?
11 MEMBER KRESS: They asked what our feelings i
4 12 were about adequate protection and the safety goals in i i
1 i 13 general. 4 14 CHAIRMAN SEALE: You can't interpret it in any ;
. 15 other way. i a
l 16 MEMBER KRESS: We don't know what they want, 17 but that's --
18 MEMBER APOSTOLAKIS: But all we're saying is l
19 that is a challenge. I mean, we'd better make sure that I
20 we understand that that's not what they want. I'm not '
21 saying that that's what they want, but I want to see the 22 quotation because if they say, "Look,-you're our advisory 23 committee. We know we have a big problem if we establish 24 these acceptance criteria with the violators" and we come
.OV 25 back and say, "That's going to be a challenge" --
NEAL R. GROSS I COURT REPORTERS AND TRANSCRIBERS 1323 RHODE ISLAND AVE., N.W.
(202) 2344433 WASHINGTON, D.C.- 20005-3701 (202) 234 4433
l 488 l 1
1 MEMBER KRESS: Did you see a quotation in our 1 1
,_ 2 research letter where the Commission asked us what our --
(
/
3 MEMBER APOSTOLAKIS: No. I'm trying to see 4 whether they want --
5 MEMBER KRESS: I was going to say that's not a 6 --
7 MEMBER APOSTOLAKIS: In the SRM, they didn't.
8 MEMBER KRESS: That's not a standard we should 9 go by. We should not have in our letters direct 10 relationships to Commission guidelines. That's not a i 11 standard we should -- I 12 MEMBER APOSTOLAKIS: I thought we always said, 13 "This la in response to SRM" such and such.
(~'g O 14 MEMBER KRESS: No. We do that if you --
15 CHAIRMAN SEALE: When it is. l l
16 MEMBER KRESS: When it is. l 17 MEMBER APOSTOLAKIS: It is. Here it is.
18 MEMBER KRESS: But here it is --
19 CHAIRMAN SEALE: We have been asked also to 20 think out of the box.
21 MEMBER APOSTOLAKIS: And that's what I'm 22 trying to figure out. What exactly were we asked during 23 that discussion?
24 CHAIRMAN SEALE: To think out of the box on (n) w,' 25 the question of adequate protection and --
NEAL R. GROSS COURT REPORTERS AND TRANSCRIBERS 1323 RHODE ISLAND AVE., N.W.
(202) 234 4433 WASHINGTON, D.C. 20005-3701 (202) 234 4433
l 489 l 1 MEMBER APOSTOLAKIS: Do you have a transcript?
! I 2 MR. DUDLEY: That's what we're looking for.
h is !
- ' 3 MEMBER APOSTOLAKIS: Okay. Okay. Good. No. ;
4 All I'm saying is I'm not saying you're not addressing it.
5 I'm saying --
6 MEMBER KRESS: But I don't -- l 7 MEMBER APOSTOLAKIS: What did they ask?
8 MEMBER KRESS: I don't think it's appropriate 9 to always tie back what we talked about in a letter to 10 what they asked. I don't think that's a standard we ought 11 to adopt.
12 MEMBER APOSTOLAKIS: No. You may go beyond 13 what they ask.
\_/ 14 MEMBER KRESS: That's what we're doing in 15 here.
16 MEMBER APOSTOLAKIS: But the adequate 17 protection is to answer what they asked you as a minimum.
10 Right?
19 MEMBER KRESS: You already heard what the 20 minimum request was on adequate protection.
21 MEMBER APOSTOLAKIS: Yes. That was in the 22 SRM.
23 MEMBER KRESS: Yes.
24 MEMBER APOSTOLAKIS: That was in the SRM. But
/~T
( l 25 we don't know what they actually asked of --
- N_./
NEAL R. GROSS COURT REPORTERS AND TRANSCRIBERS 1323 RHODE ISLAND AVE., N.W.
(202) 234-4433 WASHINGTON, D.C. 20005-3701 (202) 234-4433
490 1 MEMBER KRESS: Well, they just wanted our
, ~ ,
_ 2 thoughts on the OHOs and --
I'~'h 3 MEMBER APOSTOLAKIS: Anyway, let's look at 4 something else.
5 MEMBER KRESS: Yes. They just wanted our 6 thoughts on QHOs and acceptance criteria. j s
7 MEMBER APOSTOLAKIS: And what I'm saying is:
8 Did they want our thoughts also as to what to do with the 9 plants that would be above the QHOs? That's all I'm i
10 saying.
11 MEMBER KRESS: They didn't specifically ask i l
12 for that. I l
13 MEMBER APOSTOLAKIS: Okay. If they didn't, l
(~~'s \
Y/ 14 they didn't. !
l 15 CHAIRMAN SEALE: But it's a logical 26 consequence of doing it.
17 MEMBER APOSTOLAKIS: If they didn't, they 18 didn't. I'll go along. I'm just -- l 19 MEMBER KRESS: You've heard the entirety of 20 what they asked on that.
l 21 CHAIRMAN SEALE: It's where the chips fall.
22 MEMBER APOSTOLAKIS: But that was in the SRM. ,
1 23 MEMBER KRESS: Yes. But that's all I'm ;
24 saying. ,
,r'T l
() 25 CHAIRMAN SEALE: That's all it said.
NEAL R. GROSS I COURT REPORTERS AND TRANSCRIBERS 1323 RHODE ISLAND AVE., N.W.
(202) 234-4433 WASHINGTON, D C. 20005-3701 (202) 2344433
l 491 1 MEMBER APOSTOLAKIS: Okay. Let me ask one
,s 2 more question. This issue of elevating the CDF ancillary
(
\
)
'~'/
3 goal to the status of a fundamental goal, we're saying 4 here this is a defense-in-depth principle. That's ngt 5 really why we're doing it.
6 As a matter of fact, one could very well 7 attach the calculations that Rick had done showing how one 8 could derive a CDF goal from the QHOs; correct? You've 9 done that.
10 And then we saw that that was too high. And 11 both the staff and we agreed that it would be wise for the 1
12 industry to have a 10" goal or for other reasons, i
13 fundamentally the reason being that if you have a core
,n i )
\/ 14 damage incident, the whole industry is gone. l l
15 MEMBER KRESS: You're mixing apples and I 16 oranges again because what it specifically says is the 17 ancillary goal in the safety goal policy statement, which 18 is 10" per reactor year. Bringing that up as a 19 fundamental goal, not just CDF whatever, bringing that 20 value up, is a defense-in-depth principle because it's 21 been deemed as the appropriate balance between prevention 22 and mitigation.
23 You can't use that as a QHO at all because 24 it's just not going to do it. The reason it has that Q,iq 25 value --
NEAL R. GROSS COURT REPORTERS AND TRANSCRIBERS 1323 RHODE ISLAND AVE., N.W.
(202) 234-4433 WASHINGTON, D.C. 20005-3701 (202) 234-4433
492 1 MEMBER APOSTOLAKIS: Wait a minute. Wait a fx 2 minute. We recommended to the Commission in one of our
'N ) i 3 earlier letters that they elevate CDF to a fundamental ;
4 level.
5 CHAIRMAN SEALE: That's right. And now the <
l 6 staff has also coopted that.
7 MEMBER APOSTOLAKIS: Yes, right. And the 8 reason was not defense-in-depth, as I recall. The reason 9 was Rick showed us his calculations, and we agreed that 10 the number was too high. Right? And we said that 10 is 11 much more meaningful and reasonable.
12 So what I'm saying is --
13 MEMBER SHACK: That may be why we did it, but
)
x' 14 we always justify it from a principle, which is 15 defense-in-depth.
16 MEMBER APOSTOLAKIS: No. We never said 17 defense-in-depth.
18 MEMBER KRESS: Let Rick. Rick has --
19 MR. SHERRY: I think you can view it as 20 defense-in-depth because what is occurring is that the 21 performance of containment in mitigating releases and 22 off-site protection actions are sufficiently reducing the 23 off-site consequences that you could have a very high CDF.
24 So you could view lowering the CDF to 10-4 as a
(_,) 25 defense-in-depth feature.
NEAL R. GROSS COURT REPORTERS AND TRANSCRIBERS 1323 RHODE ISLAND AVE., N.W.
(202) 234-4433 WASHINGTON, D.C. 20005-3701 (202) 234-4433
493 1 MEMBER KRESS: You certainly can view it. And 2 that's the way I view it.
3 MEMBER APOSTOLAKIS: Well, what does it mean 4 to elevate the CDF to a fundamental goal? What does that 5 mean in practice?
6 MEMBER KRESS: It means to have it as one of 7 the independent acceptance criteria.
8 MEMBER APOSTOLAKIS: Right. So --
9 CHAIRMAN SEALE: And what is --
10 MEMBER KRESS: It has to be met in addition to 11 the other ones.
12 MEMBER APOSTOLAKIS: Okay. So it's not 13 derivable?
p k- 14 MEMBER KRESS: Not derivable.
15 MEMBER APOSTOLAKIS: Okay.
16 MEMBER KRESS: You fix it.
17 MEMBER APOSTOLAKIS: But, I mean, if I read l
18 the last paragraph that says Attachment 1 tells me how to i 1
19 derive a LERF or a CCFP, isn't it a natural question: Why 20 isn't there an attachment telling me how to derive a CDF?
21 MEMBER KRESS: Because you can't derive a CDF.
22 MEMBER SHACK: Once we've decided to make it a 23 fundamental goal, you can't.
24 MEMBER KRESS: Well, there's something missing rm
(, ) 25 from that. I'm sorry. There was a "CDF" in there. It NEAL R. GROSS COURT REPORTERS AND TRANSCRIBERS 1323 RHODE ISLAND AVE., N.W.
(202) 234-4433 WASHINGTON, D C. 20005-3701 (202) 234-4433
. . _ . . . _ . _ - . _ ___. _ . _ ,- _ ~. _ ._.__._ - _.-. _ . _ ._ _ _ _._ _... _ -.-. _ .
. 494 i: 1 got lost. Where's the original letter?
i'
! 2 MEMBER APOSTOLAKIS: Here it says only LERF r \.
. 3 and --
I 4 MEMBER KRESS: That wasn't the original.
5 CHAIRMAN SEALE: Paragraph 3 on the line --
! 6- the paragraph before the last page.
i j .7 MEMBER APOSTOLAKIS: Paragraph 3? Yes. - What I
! 8 I'_m saying is there ought to be an acknowledgement 9 somewhere here that one can also derive a CDF goal from ,
10 the'QHOs. But we put-that aside, and we elevated the 11 fundamental goal, e
12 For example, Item 3 could start by saying, 13 "Even though one could derive a CDF" --
14 MEMBER KRESS: Here's CDF right here.
15 MEMBER APOSTOLAKIS: -- " criterion from the ;
16 QHOs, we recommend f.nat 10-4, " actually, -- that's the f
17 accepted value -- "be elevated to the level of a 18 fundamental goal because of" -- I think we can find the 19l right words. 1 20 MEMBER KRESS: CDF and LERF or that's supposed 21 to-be --
'22 MEMBER APOSTOLAKIS: These are questions that 23 come to mind, but you see the last paragraph.
24 MEMBER SHACK: I don't think you can. You 25 can't derive two criteria'from one without an application NEAL R. GROSS COURT REPORTERS AND TRANSCRIBERS 1323 RHODE ISLAND AVE., N.W.
(202) 234 4433 WASHINGTON, D.C. 20005-3701 (202) 234-4433
-1
- i
495 1 of additional principle.
_s 2 MEMBER APOSTOLAKIS: No because you can say, s }
3 "This is my QHO. I want to go back to LERF. Roughly what l
4 numbers do I have between here and there?"
I 5 MEMBER SHACK: But if I allow both LERF and l 6 CDF to float, I can always adjust the LERF and CDF 7 together. I 8 VICE CHAIRMAN POWERS: But LERF is correlated l
9 with CDF.
10 MEMBER KRESS: They're correlated.
1 1
11 MEMBER SHACK: They're correlated. i l
12 VICE CHAIRMAN POWERS: Tom?
1 13 MEMBER KRESS: Yes?
',a x/ 14 VICE CHAIRMAN POWERS: Can I take you into a 15 different subject just --
16 MEMBER APOSTOLAKIS: Well, did we resolve the 17 other subject?
1 18 VICE CHAIRMAN POWERS: -- on this derivation 19 business? l 20 MEMBER KRESS: Right.
21 VICE CHAIRMAN POWERS: You derive or prompt l
22 fatalities because you assert they're limiting.
23 MEMBER KRESS: In most cases.
24 VICE CHAIRMAN POWERS: Yesterday and the day (3
(,,) 25 before yesterday we saw risk analyses in connection with i NEAL R. GROSS COURT REPORTERS AND TRANSCRIBERS l 1323 RHODE ISLAND AVE., N W. l (202) 234-4433 WASHINGTON, D C. 20005-3701 (202) 234-4433 1
1
496 1 the steam generator rule where they seem to feel that the 7 -~. 2 latents were limiting.
(
'~ )
3 MEMBER KRESS: I think they're talking about 4 dollars. I'm not so sure which is limiting in terms of 5 dollars, in terms of number of fatalities or in terms of 6 risk, in terms of fatalities per year.
7 VICE CHAIRMAN POWERS: Your language -- by the 8 way, I think your letter is excellent, by the way.
9 MEMBER KRESS: Thank you.
10 VICE CHAIRMAN POWERS: Your language begins by 11 saying that the prompt fatalities are generally limiting.
12 And then you go on to say except in very extreme 13 circumstances.
,7 s t
's / 14 MEMBER KRESS: There were some circumstances 15 that Rick Sherry outlined where the latent fatalities 16 might call for a lower LERF than the early fatalities.
17 And I forget what. They had to do with specific site 18 characteristics that are low population within one mile 19 and something.
20 What were they, Rick?
21 MR. SHERRY: A couple of things could lead to 22 the situation where the latent fatality QHO may control.
23 One is for sequences where your protective actions are 24 very successful in moving the near-end people away from
(_) 25 the plant so that you have greatly reduced your early NEAL R. GROSS COURT REPORTERS AND TRANSCRIBERS 1323 RHODE ISLAND AVE., N W.
(202) 234-4433 WASHINGTON, D C. 20005-3701 (202) 234-4433
497 1 f atalit y r t.sk. Then the latent cancer fatality risk may u 2 dominate.
( )
v 3 The other is, similarly, for shutdown 4 sequences, where you have a very different mix of 5 radionuclides and evacuation.
6 VICE CHAIRMAN POWERS: And it's a relatively 7 non-intense source term.
8 MEMBER KRESS: Yes.
9 VICE CHAIRMAN POWERS: So that your societal 10 risk dominates. At the risk of being accused of riding a 11 hobbyhorse, which I am, I submit for your consideration 12 saying something to the effect derivable if we had 13 adequate risk understanding of shutdown accidents.
/_
$' ')
14 MEMBER KRESS: I would be perfectly willing.
15 Where would that go again, Dana?
16 VICE CHAIRMAN POWERS: I'm not sure exactly, 17 Tom.
18 MEMBER KRESS: That sort of expands on the 19 special circumstances, and I wouldn't mind putting that in l
20 somewhere if we can figure out where to put it.
21 George, on the last page, George, --
22 MEMBER APOSTOLAKIS: Yes.
23 MEMBER KRESS: -- there is supposed to be -- ;
l 24 it got left out. When you read that where it says, "The l r~x l l ) l
's_/ 2E Attachment 1 is the discussion paper that provides NEAL R. GROSS COURT REPORTERS AND TRANSCRIBERS 1323 RHODE ISLAND AVE., N.W.
(202) 234-4433 WASHINGTON, D C. 20005-3701 (202) 234-4433
498 2
1 lower-tier criteria in terms of," that should read."CDF ,
- 2 and/or CCFP." It got left out.
L O 3 MEMBER APOSTOLAKIS: Okay. That's very f 4' important.
5 MEMBER KRESS: That just got dropped somehow.
i 6 MR. DUDLEY: Line 108. .i '
l 7 MEMBER KRESS: Yes.
8 MEMBER APOSTOLAKIS: Okay. That's why I I 9 raised the question.
i 10 MEMBER KRESS: I didn't understand your
) 11 question because I didn't know it wasn't in there. It l
l 12 just got lost somehow.
- _ 13 CHAIRMAN SEALE
- Tom, is your problem with the 14 earlier concern about referencing where all this thing i
j 15 came from the difficulty in finding an adequate summary
! -16 listing of what the challenge was?
17 MEMBER KRESS: It's that plus we understood 1
18 what they said by the exchange of verbal information. It l
19 doesn't come through well in citing something like
- f. 20 transcripts, and not only that.
I.
!- 21 I don't think we should make a policy to have
- 22. our letters always tied to something directly that they i~
'23 asked us to do. And if we give them more than they want 24 or something slightly different than they asked us for, I l
25 And I just think it messes il (/
think that's part of our role.
- NEAL R. GROSS i COURT REPORTERS AND TRANSCRIBERS 1323 RHODE ISLAND AVE., N.W.
l i (202) 234 4433 WASHINGTON. D.C. 20005-3701 (202) 234 4433 j l
i i
- .-_-.--._ . . _ . .- _ r _ . . _ , , . . ..m._ . , , _ . , , , , _ . _ . , , ,.,4 ,m.-
499 1 up the water to always try to wedge things'in there that 2- don't necessarily have any real relevance to the whole
() 3 thing.
l 4 And so-I've got a basic antipathy to the 5 concept. Plus, I think in this case, it's going to be 6 difficult to find the right quotes from the right sources 7 because we've got quotes two or three different times at 8 two or three different venues.
9 They're not always written down. They may_be 10 in transcripts and may be as a result of a verbal exchange 11 between us and one of our meetings. Just it didn't seem 12 like a good place to do that.
13 MEMBER APOSTOLAKIS: But it's not really clear 14 what Rogers is asking.
i l
15 MEMBER KRESS: Yes. !
l 4
16 MEMBER APOSTOLAKIS: There are lots of 17 questions, he says. How the safety goals relate to an 18 adequate protection standard. And we're saying that it's
- . 19 up and down; right?
!~
- 20 What's the relationship between these things?
! i
- 21 If you want to start using safety goals for regulatory
, 22 purposes, what does that really mean if we already have an 23 adequate protection-standard that is being met?
24 MEMBER KRESS: Good questions, I think.
f'( ) 25 MEMBER APOSTOLAKIS: But what does the i NEAL R. GROSS
- COURT REPORTERS AND TRANSCRIBERS t 1323 RHODE ISLAND AVE., N.W.
.(202) 234-4433 WASHINGTON, D.C. 20005-3701 (202) 2344433
500 1 question mean? I don't understand. Are you making the
,3 2 regulations more stringent, then, for some plants?
~) 3 CHAIRMAN SEALE: It depends on what the status 4 of the old --
5 MEMBER APOSTOLAKIS: Let's say it's above.
6 Let's say it's above.
7 CHAIRMAN SEALE: No. It depends upon what the 8 status of the present adequate protection concept is once 9 you adopt a safety goal approach.
10 MEMBER KRESS: It's a policy more than 11 anything.
12 CHAIRMAN SEALE: Yes.
13 MEMBER APOSTOLAKIS: I don't --
,a I i
\j 14 CHAIRMAN SEALE: See, you have it one way now.
15 You have adequate protection.
16 MEMBER APOSTOLAKIS: Yes.
17 MEMBER KRESS: And then if you want to go some 18 other way --
1 19 CHAIRMAN SEALE: If you go to the safety l 20 goals, you can't have it both ways.
21 MEMBER APOSTOLAKIS: Yes.
22 CRAIRMAN SEALE: That's what Tom is saying.
23 MEMBER KRESS: It's pretty much what I'm l l
24 saying. We don't want to make precipitous changes in the
/^\
(_j 25 regulatory structure.
NEAL R. GROSS COURT REPORTERS AND TRANSCRIBERS 1323 RHODE ISLAND AVE., N.W.
(202) 234-4433 WASHINGTON. D C. 20005-3701 (202) 234-4433
._ - _ ._- __ _ __ _ ....._ .. ._.-. .....__ _ . ___ . . . _ . . _ _ _ m. . . ._ _ _ _ _ . _
501 1- MEMBER APOSTOLAKIS: What does that really c 2 mean if we already have an adequate protection standard?
I
'3 CHAIRMAN SEALE: Well, what it means is that l
4 --
l 5 MEMBER APOSTOLAKIS: That you're' forcing them l
l 6 slowly to come down to'the goal?
V.
l
! 7 CHAIRMAN SEALE: No. 'I think basically what l l
8 it means is that they are in a position where if they need l_
9 adequate protection in what it really means, they are 10 almost immune from any arguments based on risk.
11 MEMBER APOSTOLAKIS: That's not what has 12 happened here. And I'll tell you something that was told 13- to me a week ago. An IPEEE is about to be submitted to
, h D 14 the agency. And I guess a utility thought that'it would 15 be the honest thing to do.
16 They had gotten an exemption from Appendix R.
17 They did their fire risk analysis. They found a core 18 damage frequency 10-3 per year. They met all the l 19 regulations. The exemption was a legal thing; right?
20 On their own, they decided to change a few l 21 things and reduce that number down to 10-4 So it's hard 22 for me to see that plants that do meet the criteria for 23 adequate protection and find that they exceed the goal, 24 especially the CDF, which the utilities really believe is 25 real, by an order of magnitude. They're going to say, i.
NEAL R. GROSS COURT REPORTERS AND TRANSCRIBERS 1323 RHODE ISLAND AVE., N.W.
(202) 234-4433 WASHINGTON, D.C. 20005 3701 (202) 234-4433
502 1 "Well, leave me alone because I have adequate protection."
g-) 2 CHAIRMAN SEALE: Well, George, I don't know
]
3 what utility that is. And I don't really want to know.
4 MEMBER APOSTOLAKIS: Well, you will know soon.
5 CHAIRMAN SEALE: Okay. But we also know of 6 cases where utilities have been pretty flagrant in cutting 7 back on various kinds of maintenance programs and that 8 sort of thing under financial pressure.
9 And even more so where they begin an l
10 overwhelming need, like the requirement to do a 11 full-service steam generator tube check, they rob other 12 areas in order to get the resources to pay for it. j
,_ 13 MEMBER APOSTOLAKIS: I agree, but I think --
l l
}
\/ 14 CHAIRMAN SEALE: Now, in those cases, I'm not 15 sure that those utilities are always going to respond-in 16 the self-enlightened way that this particular --
17 MEMBER APOSTOLAKIS: I agree with that. And I 18 think in a lot of these instances that you mentioned, it's 19 really very hard to see the quantitative impact of these 20 changes. I mean, you reduce your maintenance activities 21 and so on.
22 But in my experience, I think there are two l
23 issues here. One is I don't think the utilities view a 24 core damage frequency criterion the same way they view the
(,,) 25 quantitative health objectives. They take core damage
- NEAL R. GROSS l COURT REPORTERS AND TRANSCRIBERS 1323 RHODE ISLAND AVE., N W.
(202) 234-4433 WASHINGTON, D C. 20005-3701 (202) 234-4433
503 1 frequency much more' seriously.
~
2 Second, unless they are above the core damage
~
3 frequency. goal, say, by a factor of two or three at the 4 most, in every single instance that I've had direct 5 experience with, they did-something to the plant to bring 6 it down to the goal. So the licensees themselves are.
7 responding to what they perceive as being real. What
- 8. bothers them is the OHos. They don't think they are real.
9 Okay?
10 So things are happening. That doesn't 11 necessarily have to be in our letter, but I think that if 12 the Commission then comes out and says, "Now this becomes 13 a risk acceptance criterion," it seems to me they will 14 take it much more seriously than then are, which is 15 already pretty serious already.
16 So, anyway, the reason why I say this is 17 because I would like to know what questions we're 18 answering. And I think we should quote the SRM, but not 19 the transcript. It would just say that, "Also during our 20 discussions, we were encouraged to think about these -l 21 issues. And here is the letter."
22 But if there is an SRM with specific 23 questions, it seems to me it should be quoted,-not the 24: transcript.
25 MR. DUDLEY: The SRM is intended to clarify NEAL R. GROSS COURT REPORTERS AND TRANSCRIBERS 1323 RHODE ISLAND AVE., N.W.
(202) 234-4433 WASHINGTON, D.C. 200 6 3701 (202) 234-4433
j- 504 1 the' questions that were intended to'be asked-during the 2 meeting.
!O 4
] 3 MEMBER APOSTOLAKIS: Yes. But the SRM is kind f 4 of an official document. I mean, you can cite it --
] 5 'MR. DUDLEY: Correct.
j- 6 MEMBER APOSTOLAKIS: -- and give the date.
7, Then you really know what you're answering.
8 MEMBER KRESS: Okay. We'll get it-into the l i- .
, 9 boilerplate.
$ .10- MEMBER APOSTOLAKIS: Now, one last --
11 MEMBER MILLER: About a half-hour ago you said 12 you had one more.
13 MEMBER APOSTOLAKIS: Well, I'm sorry, but O 14 these letters are not --
15 MEMBER KRESS: They're not trivial, and this 16 is good, i c: j 17 MEMBER SHACK: It just ratchets like other 1
18 systems. It just ratchets.
19 MEMBER APOSTOLAKIS: Since we are saying in 20 the last paragraph that one can derive CDF and LERF and so 21 on from the QHOs and Item 3 says that "We are elevating 22 the CDF ancillary goal to the status of a fundamental .
1 23 goal," I think it would'be clearer to somebody who is new 24 to this business if we made a connection between the two.
25 I mean, I understand what you're saying, but NEAL R. GROSS COURT REPORTERS AND TRANSCRIBERS 1323 RHODE ISLAND AVE. N.W.
(202) 234-4433 WASHINGTON, D.C. 20005-3701 (202) 234-4433
..-,-. ...,- ~ . - -
505 1 I've been in the business a long time. I don't know how
'x7-
\
2 Commissioner Diaz or Dykus --
/
3 MEMBER KRESS: Well, there's an attachment 4 that goes with this.
l 5 MEMBER APOSTOLAKIS: No. The attachment shows l 6 how to do it.
7 MEMBER KRESS: Well, it makes very clear what l 1
8 we're talking about in terms of definitions, how they're 9 related to each other. And if you try to gather that in a 1 10 report -- ;
11 MEMBER APOSTOLAKIS: No, no, no. I'm not 12 talking about that. I'm saying if I do what you say in 13 the last paragraph, then I can derive a CDF.
,/'
> 4
\w 14 MEMBER KRESS: No.
15 MEMBER APOSTOLAKIS: I can't?
16 MEMBER KRESS: No. You cannot derive a CDP.
17 You can derive lower-tier acceptance criteria in terms of 18 a CDF and LERF, It's got an "and" in there. It's not an 19 "or "
20 CHAIRMAN SEAI.E: Derive a pair.
21 MEMBER APOSTOLAKIS: Okay. But I can do it 22 the way Rick did it, can't I? Didn't you do it when you 23 had a presentation here?
24 MEMBER KRESS: It makes it very clear that the (p,,/ 25 QHO and a LERF are a one-to-one correlation. It makes it NEAL R. GROSS COURT REPORTERS AND TRANSCRIBERS 1323 RHODE ISLAND AVE., N W.
(202) 234-4433 WASHINGTON, D.C. 20005-3701 (202) 234-4433
506 1
1 very clear. If you're going to raise the CDF to a 10",
, -~3 2 then you have what I would call just a constraint on the U 3 arrangement between CDF and CCFP that go into making this 4 LERF. j 5 And if you're going to try to explain all of 6 that here, in the first place, you're going to preempt 7 everything I'm going to read in the Attachment 1. If they 8 want to know all of that, they ought to read Attachment 1. j 9 That's the way I feel about it.
10 MEMBER APOSTOLAKIS: I'm lost.
11 MEMBER KRESS: I want to give them some ;
12 incentive to read Attachment 1 because the first thing 13 they're going to do is look at that and say --
/ \
\/ 14 MEMBER APOSTOLAKIS: What did you do, Rick, l
15 when you made the presentation here? How did you derive 16 the CDF?
17 .M R . SHERRY: The CDF and the LERF? Well, the 18 CDF was derived based on relationships coming from 1150, 19 MEMBER APOSTOLAKIS: Right.
20 MR. SHERRY: Right. And in my initial 21 presentation, LERF was also derived that way. Okay? But 22 then subsequently I worked on a methodology for deriving 23 the LERF using the exposure index and things like that. l 24 Okay?
(~;
(_j/ 25 MEMBER APOSTOLAKIS: No. The CDF --
NEAL R. GROSS COURT REPORTERS AND TRANSCRIBERS 1323 RHODE ISLAND AVE.. N.W.
(202) 234-4433 WASHINGTON, D.C. 20005-3701 (202) 234 4433
l 507 j l
1 MR. SHERRY: But not the CDF.
l l
, ~ 2 MEMBER APOSTOLAKIS: Can I take the QHOs and
/ .
3 the knowledge that 1150 has accumulated there and derive a 4 CDF?
5 MR. SHERRY: You can make an estimate of CDF, 6 yes. l 7 MEMBER APOSTOLAKIS: Yes. And you did that.
l' 8 MR. SHERRY: Yes.
I 9 MEMBER APOSTOLAKIS: And you came up with --
10 MEMBER KRESS: Whoa. Hold it. An estimate of 11 CDF to do what? l l
12 MEMBER APOSTOLAKIS: If you are to meet the 1
13 QHOs and if you take as typical values the estimates of I'h
\'
l
~ 14 NUREG-1150 regarding containment performance -- l l
15 MEMBER KRESS: Oh, you did: t say that. l l
4 16 MEMBER APOSTOLAKIS: Right. That's what I'm 1 17 saying.
l 18 MEMBER KRESS: Sure, sure.
19 MEMBER APOSTOLAKIS: He did it.
20 MEMBER KRESS: Yes.
21 MEMBER APOSTOLAKIS: Then you can derive a 22 CDF.
23 MEMBER KRESS: Sure. But that's a dumb thing 24 to do if you're talking about acceptance criteria. If f~%
( ,) 25 you're building into it existing CCFPs, which is --
NEAL R. GROSS COURT REPORTERS AND TRANSCRIBERS 1323 RHODE ISLAND AVE., N.W.
(202) 234 4 433 WASHINGTON, D.C. 20005 3701 (202) 2344433
508 1 MR. SHERRY: Maybe just a sentence that says 2 that, that CDF values derive from the QHOs based on a
( l
'~'
3 containment performance as exhibited in the 1150 would 4 lead to fairly high values with --
5 MEMBER KRESS: But that's a dumb thing to do 6 that I don't want to tell somebody, " Don't do a dumb i
7 thing." That's --
~
z ,
8 MR. SHERRY: No, no, no, no, no. I'm saying 9 these are just you're presenting evidence from past 10 studies, which indicates that if you would do this to 11 derive a CDF, it would be very high and we would judge to 12 be unacceptably high. Hence, we recommend that --
13 MEMBER APOSTOLAKIS: But we've said that k 14 already.
15 MEMBER KRESS: It's a straw man that you throw 16 up to knock down. And I don't want to throw up straw men 17 and confuse the issue. It's very clear in Attachment 1 18 what I have in mind. And I don't want to confuse it by 19 throwing in these ifs, ands, and buts, or other ways i 1
20 people that might think to do it, which are straw men. ;
21 And you're going to knock them down anyway.
22 MEMBER APOSTOLAKIS: In fact, I think we've 23 already said that in an earlier report.
24 MEMBER KRESS: If you have --
y-( ) 25 MEMBER APOSTOLAKIS: We said these things.
NEAL R. GROSS COURT REPORTERS AND TRANSCRIBERS 1323 RHODE ISLAND AVE., N.W.
(202) 234-4433 WASHINGTON, D C. 20005-3701 (202) 234-4433
509 1 MEMBER KRESS: Yes, yes. And I'm not i ,,- . _ .
2 inconsistent with what was said in the past. In fact,
(
.)
3 that's one of the reasons this letter exists, because we 4 said that in the past. And they said, "Well, how do you 5 do that?" This is a response to the "How do you do that?"
6 That's the whole reason for Attachment 1, because we said 7 in a letter that this could be done.
8 CHAIRMAN SEALE: I don't think we want to sit 9 down and edit the text of 1.
10 MEMBER KRESS: I would leave it. It's under 11 my name, and it's mine. So I don't think you want to do 12 anything to it.
13 CHAIRMAN SEALE: Yes.
,s (s 14 MEMBER KRESS: That's the reason I did it that 15 way.
16 CHAIRMAN SEALE: Well, I think we ought to try 17 to resolve our conundrum at this point.
18 MEMBER KRESS: Noel is going to add a little 19 boilerplate statement quoting the SRM, I think.
20 MR. DUDLEY: What my recommendation is is to 21 use a boilerplate, "During the 439th meeting of the ACRS, 22 we continued our deliberation on the issue of 23 plant-specific applications of the NRC safety goals and 24 the relationship between the concept of adequate
,m 25 protection in the safety goals," and then place the SRM as (ss)
NEAL R. GROSS COURT REPORTERS AND TRANSCRIBERS 1323 RHODE ISLAND AVE., N.W.
(202) 234-4433 WASHINGTON D C. 20005-3701 (202) 234-4433
510 1 a reference.
2 MEMBER KRESS: That would be good for me.
7
\ )
3 MR. DUDLEY. And the phraseology relationship 4 between the concept of adequate protection and the safety 5 goals is lified directly out of the SRM.
6 MEMBER KRESS: That would suit me.
7 CHAIRMAN SEALE: That would make the 8 connection, then.
9 MR. DUDLEY: And then go right into your 10 let te' .
11 MEMBER KRESS: Yes. That would be good, about 12 as unobtrusive as you could make it. I appreciate that.
13 CHAIRMAN SEALE: Okay. Do we want to do a 7
s \
'El 14 line by line?
15 MEMBER KRESS: I would like to get this letter 16 out.
17 CHAIRMAN SEALE: I would, too.
18 MEMBER KRESS: Okay.
19 MEMBER APOSTOLAKIS: So, again -- no, not a 20 line by line.
21 CHAIRMAN SEALE: All right. What do you want 22 to do, George? Do you want to fight about it some more?
23 MEMBER APOSTOLAKIS: No, I'm not fighting.
24 I'm trying to understand it.
/~~T t,.
) 25 So this letter, "This report further discusses NEAL R. GROSS COURT REPORTERS AND TRANSCRIBERS 1323 RHODE ISLAND AVE., N.W.
(202) 234 4433 WASHINGTON, D.C. 20005-3701 (202) 234-4433 ll
511 1 the need for plant-specific application of risk acceptance 73 2 criteria." Fine. "And the appropriateness of these being
> t V
3 derived from the safety goal QHO on early fatalities.
4 "The attachments are intended to provide 5 detailed guidance for quantifying such lower-tier 6 acceptance criteria in terms of CDF and LERF or CCFP." So 7 lower-tier criteria are the CDF --
l 8 CRAIRMAN SEALE: And LERF. l 9 MEMBER APOSTOLAKIS: Okay.
10 MEMBER KRESS: Those are lower than the QHO. l l
11 MEMBER APOSTOLAKIS: CDF is not lower anymore.
12 CDF is up there with everything else.
13 MEMBER KRESS: It's lower than the QHO.
,,~
-- 14 MEMBER APOSTOLAKIS: It's lower in the sense 15 of a PRA.
16 CHAIRMAN SEALE: Yes.
17 MEMBER APOSTOLAKIS: But in terms of 18 importance, it's up there.
19 MEMBER KRESS: Oh, yes. I just meant 20 lower-tier than they already found in QHO.
21 MEMBER APOSTOLAKIS: But in the attachments, 22 you're not producing any other lower-tier criteria.
23 MEMBER KRESS: No. Just CDF.
24 MEMBER APOSTOLAKIS: For functions or --
/#'N
(/ 25 MEMBER KRESS: No, no.
NEAL R. GROSS COURT REPORTERS AND TRANSCRIBERS 1323 RHODE ISLAND AVE., N.W.
(202) 234 4 433 WASHINGTON, D.C. 20005 3701 (202) 234-4433
512 1 MEMBER APOSTOLAKIS: Okay.
73 2 MEMBER KRESS: No. That's possible in the
~
3 long term, but I'm not --
4 MEMBER APOSTOLAKIS: Okay. Okay.
5 MEMBER KRESS: I don't even know how to do 6 that right now. That's CDF and LERF.
7 MEMBER APOSTOLAKIS: Okay. Let's go line by 8 line.
9 CHAIRMAN SEALE: Okay.
10 MEMBER KRESS: Paul, do you want to give her 11 the corrected lines there?
12 CHAIRMAN SEALE: I think it's already there.
13 DR. BOEHNERT: All except the phrase.
I i
/
'- / 14 MEMBER KRESS: All except the phrase that goes l 15 in after the first sentence. l 16 MEMBER APOSTOLAKIS: Now, I understand the 17 staff is now getting away from using conditional 18 containment failure probability. Is that correct?
19 MEMBER KRESS: Yes.
20 MEMBER APOSTOLAKIS: I think this is --
21 MEMBER KRESS: I think they're interrelated, 22 though. If you really want to know, I like the LERP 23 better myself, but I don't think that necessarily leaves 24 it out.
,f )
i ,) 25 MEMBER APOSTOLAKIS: It's not really clear NEAL R. GROSS COURT REPORTERS AND TRANSCRIBERS 1323 RHODE ISLAND AVE., N.W.
(202) 234-4433 WASHINGTON, D C. 20005-3701 (202) 234-4433
1 513 1
1 what Rogers wanted.
.s 2 MEMBER KRESS: No, but --
I i 3 MEMBER APOSTOLAKIS: If you look at the l
4 questions -- l l
5 CHAIRMAN SEALE: Tom had a more detailed I 6 description of them verbally. I l
l 7 MEMBER APOSTOLAKIS: Oh, he did? '
8 CHAIRMAN SEALE: Yes. i l
1 9 VICE CHAIRMAN POWERS: Orally. -
10 CHAIRMAN SEALE: Orally. Okay. Verbally.
11 VICE CHAIRMAN POWERS: Writing or spoken, it's 12 going to ha verbal.
13 CHAIRMAN SEALE: Okay. You're right. Okay. I
,s t )
V 14 Orally. Did you notice this morning that we say 15 " analytic"?
l 16 MEMBER APOSTOLAKIS: Can you have a discussion 17 in writing?
18 CHAIRMAN SEALE: We never say "vertic"?
19 MEMBER KRESS: Yes. Yes, you can, especially 20 on the --
21 MEMBER APOSTOLAKIS: You won't call it a 4
22 discussion.
23 CHAIRMAN SEALE: Analytical, vertical.
24 MEMBER KRESS: Chat room. We even have e-mail n
f ) 25 conferences, where we discuss all sorts of --
NEAL R. GROSS COURT REPORTERS AND TRAN9CRIBERS 1323 RHODE ISLAND AVd., N W.
(202) 234-4433 WASHINGTON, D.C. 20005-3701 (202) 234-4433
I 514 1 CHAIRMAN SEALE: You say " vertical." You 2 don't say "vertic."
/ N
\' 3 MR. DUDLEY: RULENET is an example. They had j l
4 conferences. '
5 CHAIRMAN SEALE: Okay. Now we've got it.
i 6 MEMBER SHACK: How about our " deliberations"? {
l 7 CHAIRMAN SEALE: Yes. l 8 MEMBER APOSTOLAKIS: There is one other thing :
9 I forgot to raise. t 10 CHAIRMAN SEALE: As we get to it. Okay?
11 MEMBER APOSTOLAKIS: Okay.
12 CHAIRMAN SEALE: All right. Okay. You can 13 drop down a ways there, I think.
() 14 MEMBER APOSTOLAKIS: Is it also plant-specific 15 application or applications? I don't remember now.
16 Plural?
17 MEMBER KRESS: Singular. Singular. It's a 18 concept. It's not how many of them you do. It's 19 singular.
20 VICE CHAIRMAN POWERS: It's the act that's 21 being discussed and not the --
22 CHAIRMAN SEALE: Not the actions, yes.
23 "It would be a pragmatic approach." Why not 24 just say, "It would be pragmatic"?
) 25 MEMBER KRESS: Okay. Where is that?
NEAL R. GROSS COURT REPORTERS AND TRANSCRIBERS 1323 RHODE ISLAND AVE,, N.W.
(202) 234-4433 WASHINGTON, D C. 20005-3701 (202) 234-4433
. _ _. _ . _ . _ _ .. _ _ _.__ _ .. _ _ . _ ____.._m_._ . . _ . . _ . . . _ . .
515 i
1- CHAIRMAN.SEALE: Fifth line from the bottom.
^
2 VICE CHAIRMAN POWERS: I'm stunned that you
- O V i
, 3 people can read this thing on that screen. i
! 1
- 1 4 TECHNICAL SECRETARY SUMMERS
- Well, if you
)
o 5 turned your lights down a bit, it would help.
6 VICE CHAIRMAN POWERS: It doesn't help. Trust 1
7 me. l
, 1
.I 8 MEMBER SHACK: "It would be more pragmatic."
1
'9 MEMBER KRESS: That would be better. Okay. I i ]
i
- 10 like that.
- I
- (
f 3:
11 MEMBER MILLER: More than what? '
I l
} 12 MEMBER KRESS: More than just using h OHC. -
5 13 CHAIRMAN SEALE: Than just using the QHO.
4 i
[ 14 MEMBER MILLER: Okay.
15 CHAIRMAN SEALE: I think we're generally in 16 agreement.
I i
- j. 17 MEMBER APOSTOLAKIS: How about if you say, the 3 18 last sentence, instead, "so long as these are chosen in
] 19 such a way as to ensure that the QHOs are satisfied," to 1
}' '20 say, "as long as these are consistent with the QHO" l
I- 21 because you don't necessarily guarantee-that QHOs are s
- 22 satisfied if you meet the lower-level criteria?
, 23 MEMBER KRESS: Yes, you do if you do it my 1 i
4
- 24 way. l 4
\
8 25 MEMBER APOSTOLAKIS: You do?
1-NEAL R. GROSS COURT REPORTERS AND TRANSCRIBERS I 1323 RHODE ISLAND AVE., N.W.
WASHINGTON. D.C. 20005-3701 (202) 234-4433 (202) 234-4433
516 1 MEMBER KRESS: Yes. But " consistent" is not
,- 2 inconsistent with that. Whichever way you guys want it is
( )
3 fine with me, 4 MEMBER SHACK: I think there's some notion 5 that they're chosen, though, I think has to sort of be in 6 there.
7 MEMBER KRESS: Yes. They are chosen.
8 MEMBER APOSTOLAKIS: 'They are chosen to be 9 consistent with the QHos.
l 10 MEMBER SHACK: That's fine. I 11 MEMBER KRESS: I don't really know what l 12 " consistent" means. I know what " satisfying the QHO" 13 means. Well, I say it in such a way that the QHOs are
,m, f J
\/ 14 satisfied.
15 MEMBER FONTANA: You say some go over and some 16 --
17 MEMBER KRESS: Okay. Pardon?
18 MEMBER FONTANA: Later on you say some of them 19 go over and some of them --
20 MEMBER KRESS: All I want to do is talk about 21 acceptance criteria, not where the plants sit in 22 relationship to that. I don't like "to be consistent 23 with" because it's such a vague term it doesn't mean 24 anything.
,r
(,,/ 25 MEMBER MILLER: Why don't you say " chosen to NEAL R. GROSS COURT REPORTERS AND TRANSCRIBERS 1323 RHODE ISLAND AVE., N.W.
(202) 234-4433 WASHINGTON, D.C. 20005-3701 (202) 234-4433
517 1 satisfy the OHO"?
7 2 MEMBER KRESS: Okay. That would be better.
(j 3 It's saying what I say in a little better way.
4 MEMBER SHACK: Well, I actually think that 5 " ensure" is better because, in fact, --
6 MEMBER KRESS: That's really what I'm doing.
7 MEMBER SHACK: -- that's what you're going to 8 do. I mean, he's --
9 MEMBER KRESS: Because I'm going to --
10 MEMBER SHACK: -- not going to satisfy the 11 QHOs. He's just going to make sure they're satisfied.
12 MEMBER KRESS: Yes. Let's go back to the 13 " ensure" because that's basically what I'm doing.
'w I 14 CHAIRMAN SEALE: Next paragraph.
15 MEMBER KRESS: I have a little problem with 16 the "such lower-tier acceptance criteria" when we haven't 17 mentioned lower-tier acceptance criteria before. Halfway 18 down, it says that --
19 MEMBER SHACK: From the first paragraph, we 20 have " lower-tier criteria on core damage frequency."
21 MEMBER KRESS: Yes. But here we change a 22 paragraph, and we say, "such lower-tier acceptance 23 criteria." And it ought to be more like "The lower-tier 24 acceptance criteria mentioned above" or something like C'i 1,,,) 25 that because once you get out of the paragraph, the "such" NEAL R. GROSS COURT REPORTERS AND TRANSCRIBERS 1323 RHODE ISLAND AVE., N.W.
(202) 234-4433 WASHINGTON, D.C. 20005 3701 (202) 234-4433
1 518 1 has to refer back to something.
7- 2 CHAIRMAN SEALE: How about for quantifying --
( i
~~
3 MEMBER SHACK: Just leave the "such" out.
4 MEMBER KRESS: Okay. Leave the "such" out.
5 CHAIRMAN SEALE: Yes.
6 MEMBER KRESS: Okay. That would do it for me.
7 MEMBER APOSTOLAKIS: Does this sentence there, 8 "The attachments to this report are intended," and so on, 9 "These methods can be used" -- do these imply an 10 endorsement by the ACRS?
11 MEMBER KRESS: Yes.
12 MEMBER APOSTOLAKIS: Shouldn't we, then, 13 review those attachments?
i I i/ 14 MEMBER KRESS: Yes. You should have already 15 had them three or four weeks ago and looked at them all 16 and decided what I did and decided --
17 MEMBER APOSTOLAKIS: We never discussed them 18 because they were under your name. And they're still 19 stated as under your name. ]
l 20 CHAIRMAN SEALE: Well, one of them is, and one 21 of them is under Rick's name.
22 MEMBER APOSTOLAKIS: Well, either a part of i
1 23 the ACRS report or they're not. If they're under Tom 24 Kress' name, we should make sure here that we're saying
,\
Q 25 that.
NEAL R. GROSS COURT REPORTERS AND TRANSCRIBERS 1323 RHODE ISLAND AVE., N.W.
(202) 234 4433 WASHINGTON, D C. 20005-3701 (202) 234-4433
- ..... . .. - ..~.-- - ....- .- -.. - - - . _ . .-. -~. .~. _
'519
- 1. MEMBER MILLER: As suggested by Tom Kress.and 2 --
O V 3 MEMBER APOSTOLAKIS: I don't know. We can u.
4 find..--
5 MEMBER KRESS: You could say, "These methods 6 were suggested by Tom Kress as a means to evaluate the 7 appropriateness of the staff's" --
l l
8 MEMBER APOSTOLAKIS: Or we.can say, "The j i
i 9 attachments'to this report provide a means,of" -- 'l I
10 MEMBER MILLER: In some way that implies we're 1
11 not endorsing the -- l 12 MEMBER APOSTOLAKIS: Exactly. I. don't mind 13 endorsing them, but then we should open it up and discuss O 14 it here.
15 MEMBER MILLER: You might want to say, "Use i
16 them at your own risk."
17 MEMBER SHACK: Provide examples of --
18 MEMBER APOSTOLAKIS: Or "If you don't like it, 19 go after Kress."
20 MEMBER SHACK: -
approaches that can be used 21 to quantify? Provide examples of approaches that can be 22 used to quantify?
23 MEMBER KRESS: Because, actually, there is a 24 slight difference between this approach and what they used 25 in the standard review plan. And, actually --
NEAL R. GROSS COURT REPORTERS AND TRANSCRIBERS 1323 RHODE ISLAND AVE., N.W.
(202) 2344 433 WASHINGTON, D.C. 20005-3701 (202) 234 4433
520 1 MEMBER SHACK: Besides, we give them actually p- 2 two ways to do it anyway.
i )
3 MEMBER KRESS: Right. Well, I like that 4 change. You need something --
5 CHAIRMAN SEALE: There's a " provide" after 6 " report" on the line above.
7 MEMBER KRESS: Okay.
8 MEMBER APOSTOLAKIS: So what does it say now, 9 "The attachments provide examples of approaches"? Yes.
10 CHAIRMAN SEALE: Okay. Does that take care of 11 your problem, George?
12 MEMBER APOSTOLAKIS: Yes.
13 CHAIRMAN SEALE: Okay. Let's --
,a
\)- 14 MEMBER SHACK: It's to quantify --
15 CHAIRMAN SEALE: No. I think your point is 16 well-taken, yes.
17 MEMBER SHACK: "In terms of" I think is 18 awkward. I mean, to quantify CDF and LERF, we've got to 1
19 get rid ot that "in terms of." l l
l 20 CHAIRMAN SEALE: No.
1 21 MEMBER APOSTOLAKIS: That's what concerned me. l l
22 That's why I asked the question of whether you derive any 23 other lower-tier criteria because I thought you started 24 with a CDF and LERF and then you derive that, but that was
/N, I
( ,) 25 not your intent. '
1 NEAL R. GI0OSS COURT REPORTERS AND TRANSCRIBERS 1323 RHODE ISLAND AVE., N.W.
(202) 234-4433 WASHINGTON, D.C. 20005-3/01 (232) 234-4433
l 521 l 1 MEMBER KRESS: Well, it's pretty clear. It 2 says, " quantifying lower-tier acceptance criteria in terms 7s 3 of CDF and LERF or CCFP that will ensure that the early I 4 fatality QHO is met." I don't see any ambiguity or any --
l 5 MEMBER APOSTOLAKIS: Well, let me tell you how 1
1 6 I interpreted that.
7 MEMBER KRESS: Okay. It can only be 1
8 interpreted one way. l 9 MEMBER APOSTOLAKIS: You start out with your 1
10 CDF and LERF, and you derive lower --
l 11 MEMBER SHACK: That's what it's saying now.
l 12 That's not what you mean, but that's what it's saying. l 13 MEMBER APOSTOLAKIS: Right. But Tom thinks V- 14 there is no interpretation. !
15 MEMBER KRESS: I don't think it says that. It 16 says that the " lower-tier acceptance criteria in terms of 17 CDF and LERF."
18 MEMBER APOSTOLAKIS: No. That's not the way I 19 understand it.
20 MEMBER KRESS: That's what it says.
21 MEMBER SHACK: Yes. But I interpret that to 22 mean that you're quantifying some fractions of CDF or 23 LERF.
24 MEMBER APOSTOLAKIS: Right.
,a
) 25 MEMBER KRESS: Well, let's change it, then.
NEAL R. GROSS COURT REPORTERS AND TRANSCRIBERS 1323 RHODE ISLAND AVE., N W.
(202) 234-4433 WASHINGTON, D C. 20005-3701 (202) 234-4433
. . . . _ . . . . _ . _ _ . _ . . . _ - _ . ~ . . . . _ ..m._..___. _. ._, .
i I 522
~
- l. 1-' MEMBER SHACK: Just.take out "in terms of." )
.]
[' -2 MEMBER APOSTOLAKIS: Yes. Take it out, i 3 MEMBER KRESS: Put a comma there, then?
I 4 MEMBER SHACK: Yes. You could just say' comma 5 --
i 6 CHAIRMAN SEALE: "E.g." )
l l I 7 MEMBER KRESS: Comma? l l
( 8 MEMBER FONTANA: "I.e."
l l
9 MEMBER KRESS: "I.e."
10' ' MEMBER APOSTOLAKIS: Yes. That --.
11 MEMBER KRESS: Okay. And you need another 12 comma after "CCFP."
13 CHAIRMAN SEALE: All right?
14 MEMBER KRESS: Okay.
15 CHAIRMAN SEALE: Fine.
16 MEMBER MILLER: Is there a way we could limit 17 one or both of these "the appropriateness of"? It seems-18 like we're using words --
19 MEMBER KRESS: Which one?
20 MEMBER MILLER: " Appropriateness."
21 EXECUTIVE DIRECTOR LARKINS: From an unbiased 22 observer, it looks like it's saying CDF and LERF or CCFP.
23 MEMBER APOSTOLAKIS: Why don't you say "CDF 24 and LERF," repeat CDF?
( 25 EXECUTIVE DIRECTOR LARKINS: Or you could put NEAL R. GROSS COURT REPORTERS AND TRANSCRIBERS 1323 RHODE ISLAND AVE., N.W.
(202) 2344433 WASHINGTON, D.C. 20005-3701 (202) 234-4433
523 1 "CDF" --
,7 s 2 MEMBER KRESS- Put "or CDF and CCFP."
s
('~)
3 MEMBER APOSTOLAKIS: Yes. That's what I'm 4 saying.
5 MEMBER SHACK: Why don't we just forget CCFP?
6 MEMBER KRESS: Well, because if you go to 7 Attachment 1, it gives you the option to do it either way.
8 MR. SHERRY: You could say, "CDF and LERF" and 9 then put a parentheses and then "or CCFP."
10 MEMBER KRESS: Do it that way. Do it that 11 way. That's better than a comma.
12 MEMBER SHACK: Yes. That's much better. ,
13 MR. SHERRY: Parentheses.
,g 14 MEMBER KRESS: Yes. Okay.
15 MEMBER APOSTOLAKIS: Do we really need the 16 last sentence?
17 MEMBER SHACK: Yes.
18 MEMBER KRESS: That was the --
19 CHAIRMAN SEALE: Well, it certainly --
20 MEMBER APOSTOLAKIS: Well, it's obvious.
21 CHAIRMAN SEALE: -- makes the -- well --
22 MEMBER APOSTOLAKIS: And also I have noticed 23 that in our letters, we tend to avoid telling the staff 24 what to do, what method to use. We tell them what to do
~../
! 25 but never how to do it.
NEAL R. GROSS COURT REPORTERS AND TRANSCRIBERS 1323 RHODE ISLAND AVE., N.W.
(202) 234 4433 WASHINGTON, D.C. 20005-3701 (202) 234-4433
524 1 MEMBER KRESS: Well, what I'm interested in
,s _
2 is: How good is the 10'5?
( )
3 MEMBER APOSTOLAKIS: Yes, sure.
4 MEMBER KRESS: And they justified it on the 5 basis of taking a few plants with sticking them on there 6 and drawing a line that showed, "See, they meet the QHO,"
7 these few plants. This was a method by which they can do 8 that better.
9 MEMBER APOSTOLAKIS: Sure. But the message is 10 --
11 MEMBER KRESS: And I wanted to flag that for 12 them to say, " Hey, you guys. Here's a way you can go in 13 and justify that a little better than what you did."
\
\v-.) 14 MEMBER APOSTOLAKIS: But isn't the message 15 there already?
16 MEMBER KRESS: No. ,
1 17 MEMBER APOSTOLAKIS: I mean, here are two 18 attachments. They tell you how to do it. You don't have 19 to beat it to death and say, "Now go and do it."
20 MEMBER KRESS: Well, one reason that it ought 21 to be in there is because we're talking first to the 22 commissioners and then to the staff. This is sort of a 23 side message to the staff.
24 MEMBER APOSTOLAKIS: I thought we never --
,r3
. (v) 25 MEMBER KRESS: I don't think it hurts to be NEAL R. GROSS COURT REPORTERS AND TRANSCRIBERS 1323 RHODE ISLAND AVE., N W.
(202) 234-4433 WASHINGTON, D C. 20005-3701 (202) 234-4433
.- . .-- . - _ .-. . _ - . ~ .
p 525 i
I '1' overly clear about~what you'mean if you're a'little 2 redundant. I don't think it hurts to be a little O 3 redundant if it makes things clearer about what you mean.
I 4- . MEMBER APOSTOLAKIS: Now, I'm really confused' i-
- . 5 here, though. And that's independent of your letter, Tom.
! 6 If I have done some work and I have published a paper in a
7 " Nuclear Technology," I am forbidden to talk about it i 8 here.
i ,
j 9' 1md here is a letter that says Tom Kress did i
j 10 something in the appendix. Go ahead and use it. Where i
l- 11 the hell is-the conflict of interest here?
12 MEMBER KRESS: I did it specifically --
- - 13 MEMBER APOSTOLAKIS
- There is no conflict ,
4 .
1 i
14 here? 1 i
1 d 15 MEMBER KRESS: I did it specifically --
i
- - 16 MR. SAVIO
- Even your paper, " Nuclear i
- 17 Technology" would --
}
4 18 MEMBER APOSTOLAKIS: You guys have told me any i
4 19 time anything that I have participated in comes in here, I
.i 20 should shut up immediately. And here we have an I 4
i j 21 attachment written by a member, and we're telling the l
22 staff, "Go use it."
23 MR. SAVIO: If he had done it for somebody for 24 pay --
f) 25 TECHNICAL SECRETARY SUMMERS: Dick, get the NEAL R. GROSS COURT REPORTERS AND TRANSCRIBERS l 1323 RHODE ISLAND AVE., N.W.
(202) 234-4433 WASHINGTON, D.C. 20005-3701 (202) 234-4433 ;
526 microphone.
.~ 1-2- MEMBER APOSTOLAKIS: Oh. Well, if it is.for.
iJ\
4 -
3 pay --
1 i 4 MR. SAVIO: Dick Savio, l
{
.5 'Somebody other then the federal government.
i 6 MEMBER APOSTOLAKIS: -I have done the work l 1
)
i 7 using. federal funds, NRC funds. And, yet, I'm told that l
8' any time I have' participated in anything, I have a big l j--. 9 conflict of interest, I shouldn't tell the staff to do it.
. l
.10 MEMBER FONTANA: Unless you publish them, you- i s
11 can't-talk anymcre?
12 MEMBER APOSTOLAKIS: No. !
13 MEMBER FONTANA: No, no , no. That can't be O- 14 right.
15 MEMBER APOSTOLAKIS: I should let other people 16 read it, appreciate it, and bring it up.
17 MEMBER FONTANA: That's baloney.
18 MEMBER APOSTOLAKIS: It's real life.
19 MEMBER KRESS: You're misinterpreting what the 20 staff tells you there.
21 EXECUTIVE DIRECTOR LARKINS: George, I think 22 you're talking about if you do some work and it comes back 23 up in discussion in terms of review or commenting, that 24 you're supposed to refrain from commenting on your own 25 work.
NEAL R. GROS $
COURT REPORTERS AND TRANSCRIBERS 1323 RHODE ISLAND AVE., N.W.
(202) 234 4433 WASHINGTON, D.C. 20005-3701 (202) 234-4433
527 1 MEMBER APOSTOLAKIS: That's not the way I r's 2 understood it. But I've been here sitting quiet when Pat A
3 Baranowski says there is no method to handle 4 organizational factors. A.7d I published white papers on 5 that telling them how to do it. And I thought it was my 6 duty to keep quiet.
7 EXECUTIVE DIRECTOR LARKINS: No. You always 8 have the responsibility to correct factual mistakes and 9 factual comments. I said that from day one.
10 MEMBER APOSTOLAKIS: No. I'm sorry, but I 11 thought I had a conflict.
12 MEMBER APOSTOLAKIS: Only if you are quoting
,_ 13 or citing your own work.
(') 14 MEMBER SHACK: Only if she says " George's work 15 is rubbish."
16 MEMBER APOSTOLAKIS: No. The man said there 17 is no method. I mean, that's only one example.
18 VICE CHAIRMAN POWERS: What he meant, of 19 course, was that there was no useful method. l 20 MEMBER KRESS: No acceptable, good, usable 21 method.
22 MEMBER APOSTOLAKIS: Another example is when 23 Theofanous complains so bitterly that we ignore him. I 24 was told explicitly that this Committee never endorses
(~'N
(_,/ 25 methods. And here we are telling them, "Go and use this."
NEAL R. GROSS COURT REPORTERS AND TRANSCRIBERS l 1323 RHODE ISLAND AVE., N W.
(202) 234-4433 WASHINGTON, D.C. 20005 3701 (202) 234 4433
528 1 MEMBER KRESS: Sure, we do. We endorse
,s 2 methods all the time.
/ i
~-
3 VICE CHAIRMAN POWERS: I mean, we just had the 4 DOE come in and give us a presentation where they brought 5 a guy in from Livermore. We're not going to write a 6 letter that says, " Gosh. Gee. Thanks for dropping by, 7 guys." But think what favor they will occur on this when 8 --
9 MEMBER APOSTOLAKIS: I have misunderstood, 10 then, the rules of conflict of interest because I thought 11 we should never tell them to use something if any of us 12 had anything to do with it unless somebody else says that.
13 MEMBER KRESS: I will guarantee you I don't (3 t
\_/ 14 have a conflict of interest on this.
15 CHAIRMAN SEALE: Well, I think --
16 MEMBER APOSTOLAKIS: Anyway, I move that the 17 last sentence be dropped. If anybody else wants it --
18 MEMBER SHACK: I'll second it.
19 VICE CHAIRMAN POWERS: But understand that 20 that's only because you drop any sentence that comes --
21 (Laughter.)
22 VICE CHAIRMAN POWERS: There's no l
l 23 philosophical business here.
24 CHAIRMAN SEALE: I will stipulate that we have (3
) 25 been told that Tom does not have a conflict in referring NEAL R. GROSS l COURT REPORTERS AND TRANSCRIBERS j 1323 RHODE ISLAND AVE., N.W.
(202) 234-4433 WASHINGTON, D.C. 20005-3701 (202) 234 4 433 l
b 529 1- to his report in.this. letter. We may wish to later s-2 revisit that issue in terms of George's concerns and l l! .O iV 3' probably should, but let's not do that now.
j 4 Any more discussion on the motion to drop this-p 5 sentence?-
+
j 6 MEMBER KRESS: I believe this is the message
- 7 that ought.to be given and clear and stated in the letter, i
8' the message that's needed. And I would like to see it-i 1
9 retained.
i 10 MEMBER APOSTOLAKIS: Counter to that, I.think 11 the message has already been given in the preceding 1 12 sentence.
) -13 MEMBER KRESS: Where? Where does it say that?
i (/ 14 MEMBER APOSTOLAKIS: Well, it says, "The
, 15 attachments to this report provide examples of )
16 approaches." It says, you know -- ;
, i 17 MEMBER SHACK: We didn't think it would be
- 18 used by the average seventh grader in Montgomery County.
19 MEMBER APOSTOLAKIS: Yes. I mean --
20 VICE CHAIRMAN POWERS: But let me understand 21 Tom. Your concern arises because right now they have a 22 capricious and arbitrarily chosen LERF.
23 MEMBER KRESS: Yes.
I. l 24 CHAIRMAN SEALE: That's right.
t 4
25 MEMBER KRESS: Well, it's not all that
, b i
NEAL R. GROSS 4
COURT REPORTERS AND TRANSCRIBERS
! 1323 RHODE ISLAND AVE., N.W.
(202) 234-4433 WASH!NGTON, D.C. 20005-3701 (202) 234-4433 i
m _ , _ . _ _ _ . - . . ~ _ _ . ._.. _.._. ..__. ~ . .- .. _ . - . _ . _ _ _ _ . _ . _ . . _ _ _ . . _ . _ .
, 530 .,
1 capricious and arbitrary. In fact --
f
. 2 VICE CHAIRMAN POWERS: Capricious, arbitrary, 5h i ~3 and useless' definition of LERF unrelated to'the important j: ,
4 aspects of risk --
5 MEMBER KRESS: No, it doesn't --
6 VICE CHAIRMAN POWERS: -- because it only 7 deals with operational risk. ,
8 MEMBER KRESS: Well, I agree thatLthat is an .
9 issue, yes. But what I feel is that they have a value for 10 an acceptance criteria for LERF, which hac nothing to do 11 with shutdown, low power, or whatever because you can 12 assume -- I mean, that could be part of the definition.
13 This acceptance criteria, this 10~5 value for i
14 LERF, was justified on the basis of comparing the actual 15 risk of a number of plants and showing that if you use 16 this, then you will bound those risks for that number of ,
l 17 plants. That's not arbitrary and capricious. It's a l 18 relatively good way. It's a limited sample. )
l 19 The concept I give them gives them a more I l
20 technically defensible, detailed way to do it for every 21 site in the U.S. and show that their value chosen is 22 appropriate or not for every site. And that's the reason 23 I want to put it in there because it makes that clear.
24 VICE CHAIRMAN POWERS: What you want them to 25 do is to take something like the siting study or something l
NEAL R. GROSS COURT REPORTERS AND TRANSCRIBERS j 1323 RHODE ISLAND AVE., N.W. '
(202) 234-4433 WASHINGTON, D.C. 20005 3701 (202) 234-4433
_ . _ . ..-...__...m .. _.m .. . _ _ . _ . . _ _ . _ . _ _ _ - . . . _ . . _ . . _ . _ _
j 531
.\
l' like that --
2 MEMBER KRESS: That's exactly right.
!O i
I 3 VICE CHAIRMAN POWERS: .-- and calculate 110 1 I j
. 4 values of-LERF,--
]
2
-5 MEMBER KRESS: That's right. And show how .!;
,\
l 1
6 many of:them are above their value and how many below --- i j-7 VICE CHAIRMAN POWERS: And then --
i 8. MEMBER KRESS: -- and then show that, well, I 9 maybe ours doesn't bound all of them, but it bounds 99 j .10 percent of them or some such thing.
) - 11 VICE CHAIRMAN POWERS: Falls right in the 12 middleLor wherever it is.
13 MEMBER KRESS: Falls right in the middle or 14 where. That's what I'm saying here, yes. That's what I 15 mean by the-word " appropriateness" here. )
16 MEMBER APOSTOLAKIS: You guys are discussing 17 the technical merits of what Tom is proposing. That's not 18 the issue I'm raising.
19 MEMBER KRESS: Okay.
20 MEMBER APOSTOLAKIS: I'm raising the issue of 21 --
22 MEMBER SHACK: We've already said it.
23 MEMBER APOSTOLAKIS: First, you've already 24 said it. Second, is it appropriate in our own letters to 25 promote our own methods?
NEAL R. GROSS COURT REPORTERS AND TRANSCRIBERS 1323 RHODE ISLAND AVE., N.W.
. (202) 2344433 WASHINGTON, D.C. 20005-3701 (202) 2344433
532 1 MEMBER SHACK: But we have already done that g3 ,
2 with the first sentence, George.
- a LJ 3 CHAIRMAN SEALE
- We're going to --
4 MEMBER APOSTOLAKIS: I'm really confused about 5 the whole --
6 VICE CHAIRMAN POWERS: Tom, is it the 7 situation that you would prefer if the last sentence said 8 one or more of these methods should be used by the staff 9 to evaluate the appropriateness of their choice of LERF?
10 MEMBER KRESS: That would be good. That would 11 be all right. I'm not directing them to do it. I think 12 that they should.
13 VICE CHAIRMAN POWERS: The LERF that they use
(~%
(, )
's> 14 in the regulations that they're developing, et cetera, et 15 cetera, et cetera.
16 MEMBER KRESS: Right. That's a good 17 phraseology.
18 MEMBER APOSTOLAKIS: I think the preceding l l
19 sentence is appropriate. There are examples, hint, hint, 20 hint, period. But to go out and explicitly say, "Now go 21 and do it" seems to me --
22 MEMBER KRESS: It doesn't say to do it. It 23 says it can be used.
24 VICE CHAIRMAN POWERS: I'd like to change it g\
C) 25 to "At least one or more of these methods should be used."
NEAL R. GROSS COURT REPORTERS AND TRANSCRIBERS
, 1323 RHODE ISLAND AVE., N W.
(202) 234-4433 WASHINGTON, D C. 20005 3701 (202) 234-4433 l
533 1 MEMBER KRESS: I would like that. That's even
- 2 stronger than "can be."
t )
3 VICE CHAIRMAN POWERS: Since you and I are the 4 ones who are going to vote in favor of this sentence, we 5 might as well get the --
6 MEMBER KRESS: Yes.
7 VICE CHAIRMAN POWERS: -- the sentence might 8 as well go down for what we want it to be.
9 MEMBER KRESS: I like your change. I like 10 your changa, Dana. Make the change Dana said because I 11 think it's --
12 MEMBER FONTANA: What is it?
13 CHAIRMAN SEALE: One or more, r'N.,
- 4 s/ 14 VICE CHAIRMAN POWERS
- "One or more of these 15 methods should be used to evaluate the appropriateness of" 16 --
17 MEMBER MILLER: Now you're really --
18 MEMBER APOSTOLAKIS: That's all right.
19 MEMBER MILLER: Now you're endorsing thic.
20 MEMBER APOSTOLAKIS: Why don't we make the 21 attachments part of the letter, then, and review them like 22 we review everything else?
23 VICE CHAIRMAN POWERS: Their choice of LERF.
24 CHAIRMAN SEALE: las. There we go. Okay,
,r
(,)/ 25 Now, we're going to vote on the motion to drop the last HEAL R. GROSS COURT REPORTERS AND TRANSCRIBERS 1323 RHODE ISLAND AVE., N W.
(202) 2344433 WASHINGTON, D.C. 20005-3701 (202) 234 4433
I 534 1 sentence in its modified form, I assume. Those in favor
~~s 2 of dropping the sentence, raise your hand.
I i
'# 3 (Whereupon, there was a show of hands.)
4 CHAIRMAN SEALE: One, two, three.
5 MEMBER APOSTOLAKIS: Four.
6 VICE CHAIRMAN POWERS: It looked like you had 7 four.
I 8 CHAIRMAN SEALE: Four. Oh, God.
l 9 Those in favor of keeping the last sentence? )
l 10 (Whereupon, there was a show of hands.) ;
11 MEMBER FONTANA: According to the rules, it i
12 doesn't pass. !
i l
13 MEMBER KRESS: Not according to our bylaws. j
/T i w/ 14 CHAIRMAN SEALE: No.
15 MEMBER KRESS: According to our bylaws, the ;
J 1
16 Chairman is the tie-breaker. l 17 MEMBER APOSTOLAKIS: I'll have to think about i
18 added comments if it stays there.
19 CHAIRMAN SEALE: Well, the alternative is that !
20 we could take this sentence and make it an added comment.
l 21 VICE CHAIRMAN POWERS: Let's understand what >
22 your added comment would be, George.
23 MEMBER KRESS: Yes, before we --
24 MEMBER APOSTOLAKIS: Right now I'm having a
/m
/
a )
25 big problem understanding what the hell conflict of NEAL R. GROSS COURT REPORTERS AND TRANSCRIBERS 1323 RHODE ISLAND AVE., N.W.
(202) 234 4433 WASHINGTON. D C. 20005-3701 (202) 234-4433
535 1 Interest means in this item.
-2 MEMBER MILLER: That's not relevant to the --
tO 1
b 3 ' MEMBER KRESS: Conflict of interest has
- . 4~ nothing to do with-it.
! 5 MEMBER APOSTOLAKIS: It is very relevant. ;
i 6 Here is something that the Committee has not' reviewed 7 written by a member of this committee. And we're telling l
8 the staff, "Go use it" when I keep getting the advice, 9 " Don't you ever.dare talk about anything you had anything 10 to do with." I mean, I'm utterly confused.
11 MEMBER MILLER: If the word "should" --
12 MEMBER APOSTOLAKIS: Even today, for heaven's 13 sake, I just made a' presentation to the Committee. And I 14 was advised to be careful not to talk about anything I 15 talked about then.
16' MEMBER MILLER: "Can" I would go along with. .
17 MEMBER KRESS: I hear a comment over here if 18 we change the word "should" to "can," that that would make 19 it more powerful.
20 VICE CHAIRMAN POWERS: That "can."
21 MEMBER SHACK: How about a sentence that said 22 something like " Estimates of acceptable LERF values at all 23 sites are needed to ensure appropriateness'of" --
24 VICE CHAIRMAN POWERS: That would make you 25 happy? That would make me happy.
NEAL R. GROSS COURT REPORTERS AND TRANSCRIBERS 1323 RHODE ISLAND AVE., N.W.
(202) 234-4433 WASHINGTON, D.C. 20005-3701 (202) 234-4433
536 1 MEMBER KRESS: That doesn't make me happy 7s 2 because you don't estimate acceptable LERF. You set
! )
\
3 acceptable LERF values. You estimate actual levels.
4 MEMBER SHACK: Okay. " Estimates of LERFs 5 consistent with the QHOs at each site." No.
6 MEMBER KRESS: No. " Estimates of LERFs 7 consistent with what the LERF is at that plant." Each 8 plant has a LERF.
9 MEMBER SHACK: Right.
I 10 VICE CHAIRMAN POWERS: Well, Tom, the question i
11 is, it seems to me Bill is struggling and trying to i 12 articulate a cappella here, but what he's saying is: Can j 13 you change the last sentence so that this hint that people vs 14 speak of in the previous sentences is driven home without 15 saying, "Use one of these two methods"?
16 MEMBER BARTON: Right, right.
17 VICE CHAIRMAN POWERS: What he's saying is 18 that we say up above, -- I forget what it is -- " examples 19 of approaches that can be used to quantify lower-tier 20 acceptance criteria to ensure," et cetera. You can say 21 that you need to do that to have a technically defensible 22 value of LERF --
23 MEMBER SHACK: LERFs at each site need to be 24 evaluated --
f~% t i
(_) 25 VICE CHAIRMAN POWERS: -- as used in --
NEAL R. GROSS COURT REPORTERS AND TRANSCRIBERS 1r3 RHODE ISMND AVE., N.W.
(202) 234-4433 WASHINGTON, D.C. 20005-3701 (202) 234-4433
537 1 MEMBER SHACK: -- in order to --
7g 2 VICE CHAIRMAN POWERS: -- as used in the draft
~s' l 3 regulatory guide.
1 4 MEMBER KRESS: Okay. l 5 VICE CHAIRMAN POWERS: I mean, that would be 6 acceptable to me because --
7 CRAIRMAN SEALE: Bill, would you give us --
8 VICE CHAIRMAN POWERS: It's very broad and --
9 CHAIRMAN SEALE: Modify that sentence.
10 MEMBER SHACK: Just "LERFs at each site are 11 needed to ensure the appropriateness of." l 12 MEMBER KRESS: Yes.
13 TECHNICAL SECRETARY SUMMERS: You don't want f~\ l
\-) 14 to start with "LERF," I don't think.
1 l
15 MEMBER KRESS: "Quantification of the LERFs." l l
l 16 MEMBER SHACK: Yes. i 17 CHAIRMAN SEALE: And that's a small "s."
i 1
18 Okay.
19 TECHNICAL SECRETARY SUMMERS: One LERF at each l l
20 site; right?
21 CHAIRMAN SEALE: That's right.
22 MEMBER KRESS: It is the LERF at each site --
23 CHAIRMAN SEALE: You're right. You --
24 MEMBER KRESS: -- is needed. l
/~'
(_,
25 MR. SHERRY: With regard to the value of 10-s l l
NIL 4L R. GROSS l COURT REPORTERS AND TRANSCRIBERS l 1323 RHODE ISI.AND AVE., N.W.
(202) 2344 433 WASHINGTON, D C. 20005-3701 (202) 234 4433 1
I 538 1 for the LERF, you'may have noted that the physical i 2 definition of the LERF has changed in this last --
,f3 iD 3 . MEMBER KRESS: Yes. And that --
I 1 4' MR. SHERRY: -- edition of the regulatory. l j'
- .5 guide. However, the acceptance criteria did not change.
6' . MEMBER KRESS: That's right.
i l
- j. 7 .MR. SHERRY: And that doesn't seem to make a I 8 lot of sense because the definition --
l~
i -9 MEMBER SHACK: This definition is what they i 10 always really meant anyway.
i l 11 CHAIRMAN SEALE: Yes.* I e
j ~12 MEMBER SHACK: They just said it.
- i i
j 13 MR. SHERRY: Okay.
! 14 MEMBER'KRESS: They.put in the choice --
i -
i 15 MR. SHERRY: Well, as long as I understand.
I i 16 MEMBER KRESS: Yes.
i l 17 CHAIRMAN SEALE: Okay. Now, is this modified 1
18 sentence acceptable?
! 19 MEMBER SHACK: I'll buy it.
-20 CHAIRMAN SEALE: George?
l
- 21 MEMBER APOSTOLAKIS
- I'm looking at the i
i 22 preceding sentence. Are you really quantifying CDF, Tom?
23 You said that was taken as a given, didn't you? To 24 quantify lower-tier acceptance criteria, that is CDF and I-l 25 LERF.
- NEAL R. GROSS COURT REPORTERS AND TRANSCRIBERS 1323 RHODE ISLAND AVE., N.W.
- (202) 234-4433 WASHINGTON, D.C. 20005-3701 (202) 234-4433 i .~. - _
. ~ ~ - . . - - . . - ~ . . . . - ... -.. - - -
'539 1 MEMBER KRESS: I'm not quantifying. I'm 2 taking --
.O.
,V
- 3. MEMBER APOSTOLAKIS: You're taking'--
~
4 MEMBER KRESS: -- 10** as a given.
5 MEMBER APOSTOLAKIS: As a given. So the i l
6 sentence is not quite accurate, is it? You're really 7 addressing yourself to LERF. l 1 8 MEMBER KRESS: Or you could say that is --
1 9 MEMBER APOSTOLAKIS: So you have'already j 10 elevated a CDF to a fundamental --
l 11' MEMBER KRESS: It's LERF or CCFP given CDF as -l 12 one --
13 MEMBER APOSTOLAKIS: So this sentence is not l 14- quite accurate?
15 MEMBER KRESS: Not quite. ;
16' MEMBER APOSTOLAKIS: You have already elevated 1
17 the CDF to the fundamental level. QHO plus CDF is used 18 now using the attachments derived LERF or CCFP. That's l
19 really the accurate --
20 MEMBER SHACK: That's why the changes made to 21 108 didn't make sense.
22 MEMBER KRESS: The accuracy here is that the 23 lower-tier acceptance criteria is LERF, --
24 MEMBER APOSTOLAKIS: Right.
25 MEMBER KRESS: -- not CDF. It's LERF --
NEAL R. GROSS COURT REPORTERS AND TRANSCRIBERS I 1323 RHODE ISLAND AVE., N.W. )
(202) 234-4433 WASHINGTON, D.C. 20005-3701 (202) 2344433 -
540 1 ~ MEMBER APOSTOLAKIS: That's really --
2 MEMBER KRESS: -- or CDF and CCFP.
f-- S 3 MEMBER APOSTOLAKIS: Right. But the CDF is 4 already taken as a fundamental goal, as a fundamental 5 given.
6 MEMBER KRESS: Right. But if you're going to 7 use CCFP, you have to use the --
8 MEMBER APOSTOLAKIS: So to derive the LERF, 9 you don't need the CDF? Yes, you do. No.
10 MEMBER KRESS: Well, you do need it, but LERF 11 includes the CDF.
12 MEMBER APOSTOLAKIS: That's right.
13 MEMBER KRESS: LERF includes the CDF.
/y I
V) 14 MEMBER APOSTOLAKIS: That's right. Read the 15 sentence, though. It says, "The attachments provide 16 examples of lower-tier acceptance criteria. That is CDF j 17 and LERF. That gives the impression that -- l l
18 MR. SHERRY: Should "CDF" be moved inside the j l
19 quotes? I 20 MEMBER KRESS: Yes. l l
21 MR. SHERRY: Should it say "LERF" --
22 MEMBER KRESS: It should say -- l 23 MR. SHERRY: - say, "CDF and CCFP"? ,
l 24 MEMBER APOSTOLAKIS: "Or CDF and CCFP."
p 1
) 25 MEMBER KRESS: Right.
, NEAL R. GROSS COURT REPORTERS AND TRANSCRIBERS 1323 RHODE ISLAND AVE., N W.
(202) 234-4433 WASHINGTON, D.C. 20005-3701 (202) 234-4433
541 1 MEMBER APOSTOLAKIS: Okay. Okay. That's more
,- 2 accurate. I
!, 1 3 MEMBER KRESS: Right.
4 MEMBER APOSTOLAKIS: Now we talk about the 5 other one, which is a value judgment.
l 6 MEMBER KRESS: Okay.
7 MEMBER APOSTOLAKIS: "Or, CDF and CCFP." Put 8 a comma there.
9 MEMBER KRESS: Parentheses you --
10 MEMBER APOSTOLAKIS: Well, the parentheses are 11 still needed; right?
12 MEMBER KRESS: No.
13 MEMBER APOSTOLAKIS: No?
s ;
%> 14 MEMBER KRESS: You don't need the comma.
15 MEMBER APOSTOLAKIS: Oh, okay.
16 TECHNICAL SECRETARY SUMMERS: What do you want 17 now?
18 MEMBER KRESS: We want parentheses, instead of 19 the commas.
20 MR. SHERRY: Yes. Better use parentheses.
21 TECHNICAL SECRETARY SUMMERS: But the "or,"
22 though, has to go with it.
23 MEMBER APOSTOLAKIS: Yes.
24 TECHNICAL SECRETARY SUMMERS: Then the
,c\
() 25 parentheses?
NEAL R. GROSS COURT REPORTERS AND TRANSCRIBERS 1323 RHODE ISLAND AVE., N.W.
(202) 234-4433 WASHINGTON, D.C. 20005-3701 (202) 234-4433
542 1 MEMBER KRESS: Right.
s 2 MEMBER APOSTOLAKIS: Yes.
(
'% ))
3 MEMBER KRESS: Right.
4 MEMBER APOSTOLAKIS: So what does the last 1
5 sentence say now, "The quantification of the LERF is l 6 needed"?
7 MEMBER KRESS: Well, that's supposed to say, 8 "The appropriateness of the choice of the LERF acceptance 9 criteria."
10 MEMBER APOSTOLAKIS: Yes, yes.
11 MEMBER KRESS: That's a fixed number. The 12 othere are all over the place.
13 MEMBER SHACK: Can we leave out the "as I
)
14 discussed"?
15 MEMBER KRESS: Well, I don't know. We ought 16 to think about that because --
17 MEMBER APOSTOLAKIS: As proposed.
18 MEMBER KRESS: -- because that's where it's 19 proposed.
20 MEMBER SHACK: I was just going to say --
21 CHAIRMAN SEALE: Just say " proposed in the" --
22 MEMBER APOSTOLAKIS: " Proposed."
23 MEMBER KRESS: Okay. Okay. " Proposed" would 24 be good.
[b
( ,) 25 MEMBER APOSTOLAKIS: Instead of " discussed."
NEAL R. GROSS COURT REPORTERS AND TRANSCRIBERS 1323 RHODE ISLAND AVE., N.W.
(202) 234-4433 WASHINGTON, D C. 20005-3701 (202) 234-4433
. . . . . --~,. . . - . - - . . . . . . - . . . . - . ~ . . - . -- ~ . . - - . . . .- - . . . . - - . - .. - ..
4 4
543 1 MEMBER'KRESS: " Proposed" would be better.
. 2 MEMBER SHACK: Or just say " proposed,." rather -
, 3 than "as proposed."
a j 4 MEMBER KRESS: " Proposed."
Okay.
a 5 CHAIRMAN SEALE: Okay. We've beaten this poor 6 paragraph to its. utter --
I. ~7 MEMBER SHACK: Yes. Where's our junkyard dog l- 8 to tell us, "What the hell are we doing here?"
~
9 VICE CHAIRMAN POWERS: He is allowing the J
I 10 discussion to proceed until logical resolution, i
j il 11 MEMBER KRESS: Very good.
i.
I. 12 CHAIRMAN SEALE: Okay. Next paragraph. Any h 13 factual problem with that paragraph?
14 MEMBER SHACK: Well,'I'm not sure I can.-- I
, 15 don't enforce goals, do I? How about applications of such 4
. 16 goals? Do I enforce goals? I enforce criteria maybe, but i
1
- 17 I apply goals.
i 18 VICE CHAIRMAN POWERS: It's my impression that 19 Bill's statement is correct that the prohibition that the 20 ACRS sent up was application of the goals on a 21 plant-specific basis.
22 MEMBER FONTANA: That's a Commission policy 23 statement, isn't it?
24 MEMBER KRESS: The policy statement 25 specifically precludes enforcement of the safety goals on NEAL R. GROSS COURT REPORTERS AND TRANSCRIBERS 1323 RHODE ISLAND AVE., N W.
(202) 2344433 WASHINGTON, D.C. 20005-3701 (202) 2344433
544 1 a plant-specific basis, i 2 CHAIRMAN SEALE: Okay. Fine.
- ,__3 3 VICE CHAIRMAN POWERS: Okay.
Fair enough.
l 4 MEMBER FONTANA: Is that past tense or still l 5 " precludes"?
l 6 MEMBER KRESS: It's still that.
7 CHAIRMAN SEALE: No. It's " precludes."
8 MEMBER FONTANA: It's " precludes"? l 9 CHAIRMAN SEALE: Yes. l l
10 MEMBER KRESS: And it ought to be capitalized.
1 11 CHAIRMAN SEALE: Okay. Next paragraph. I 12 MEMBER SHACK: How about "In the development 13 of the draft regulatory guide"?
/~1 l
', l 14 MEMBER APOSTOLAKIS: Should we actually use l l
15 the numbers?
16 MEMBER SHACK: It~ we know them. Draft Reg l l
17 Guide 1061, yes.
18 MEMBER APOSTOLAKIS: Is it only 1061 that uses l
19 it or all of them?
20 MEMBER KRESS: All of them, but --
21 MEMBER SHACK: Since they're all subsidiary 22 reg guides.
23 TECHNICAL SECRETARY SUMMERS: What is it?
24 What's the number?
g l ,/ l 25 MEMBER BARTON: Ten sixty-one.
NEAL R. GROSS COURT REPORTERS AND TRANSCRISERS 1323 RHODE ISLAND AVE., N W.
(202) 234 4433 WASHINGTON, D C. 20005-3701 (202) 234 4433
545 1 MEMBER APOSTOLAKIS: No. It's one number.
s 2 It's actually DG-1061, DG-1061.
I
'~'b 3 MEMBER KRESS: Yes, but the DG stands for 4 Draft Guide.
5 MEMBER APOSTOLAKIS: Yes. But that's the 6 official number right now.
7 MEMBER KRESS: And you said what at the front, 8 Bill?
9 MEMBER SHACK: "In the development."
10 MEMBER KRESS: Okay. Instead of "As part of 11 the" --
12 MEMBER SHACK: Instead of "As part of the 13 preparation of."
/~N N- 14 MEMBER KRESS: That sounds better.
15 TECHNICAL SECRETARY SUMMERS: "In the 16 development"?
17 MEMBER KRESS: Yes.
18 VICE CHAIRMAN POWERS: I hate to bring up a 19 philosophical point at this point, but just as a question 20 of clarification, Tom.
21 MEMBER KRESS: Okay.
22 VICE CHAIRMAN POWERS: When we think about all 23 of the regulations, we think about regulations that have 24 to do with accidents. We also think about regulations
( ,) 25 that have to do with safeguards and security. They are NEAL R. GROSS COURT REPORTERS AND TRANSCRIBERS 1323 RHODE ISLAND AVE., N W.
(202) 234-4433 WASHINGTON, D.C. 20005-3701 (202) 234-4433
_ . , _ _ . . . _ _ _ _ - _ . ~ . . . ~ . . . _ _ _ . _ _ - _ _ . _ . . _ . . . _ - _ _ . . _ _ . . _ . _
546 f' 1- not treated in risk assessments.
. 2 So when we think about adequate protection,
'i .
h 3 we're thinking about adequate protection not only from 4 stochastic events but also from events of sabotage and
- . 5. insurrection.
6 MEMBER KRESS: .I think that has to be assumed, 7 yes.
- 8. VICE CHAIRMAN POWERS: There's no measure of 9 that within a risk criterion. Do we need to put a caveat?
10 MEMBER KRESS: We may because in replacing the 11 words " adequate protection" -- if the concept of replacing.
12 adequate protection by the safety goal is a concept, the 13 safety goals were intended not'to include safeguards. So 14 we have a disconnect. And I think there must be some sort 15 of caveat in there that says you have to maintain that 16 part of adequate protection in some way.
17 And so I don't know where to put that, but you 18 have a real good philosophical point.
19 VICE CHAIRMAN POWERS: When we discuss the 20 relationship of adequate protection to the risk levels, --
21 CHAIRMAN SEALE: That's the place to bring 22 that up.
23 VICE CHAIRMAN POWERS: -- could we put that 24 caveat in there and say, "This has to be set aside because 25 it's outside the current limit of technology for" --
NEAL R. GROSS COURT REPORTERS AND TRANSCRIBERS 1323 RHODE ISLAND AVE., N.W.
(202) 234 4433 WASHINGTON, D.C. 20005-3701 (202) 234-4433
.. - - ~ . -~
547 1 CHAIRMAN SEALE: "To treat these on a risk i
2- basis."
t 3 VICE CHAIRMAN POWERS: "On'a risk basis." I I
) 4 mean,'whatever the words are in.the policy. statement.
b
. 5 MEMBER KRESS: I think that.would be a good
'6 place to put it. Wh'en we get there, let's see.if we can 7 work the words in --
$ 8 CHAIRMAN SEALE: Yes, yes.
l l 9 MEMBER KRESS: -- because I think i 10 philosophically you're exactly right.
) 11 CHAIRMAN SEALE: Any other problem with that 4
l- 12 paragraph?
13 MEMBER APOSTOLAKIS: What's the message of
- N, i,
j- 14 this~ paragraph, Tom?
?
1
{ 15 MEMBER KRESS: That they should -- we've l
l 16 already said it in another --
i l 17 MEMBER APOSTOLAKIS: That they needn't --
i 1
18 MEMBER KRESS: That they should feel good
- l 19 about allowing the staff to move forward with
- 20- plant-specific application and safety goals in the context I
l 21 of requested changes to a licensee's current license. ;
- l l 22 It's an --
2 j- 23 MEMBER APOSTOLAKIS: So how about " Allowing s'
24 such plant-specific application is a positive action," --
25 MEMBER KRESS: "Is."
!. NEAL R. GROSS
- COURT REPORTERS AND TRANSCRIBERS d~ 1323 RHODE ISLAND AVE., N.W.
, .(202) 234-4433 WASHINGTON, D.C. 20005-3701 (202) 2344433 a
548 1 MEMBER APOSTOLAKIS: -- rather than "would f 2 be"?
? \
', )
3 MEMBER KRESS: "Is" would be better.
4 CHAIRMAN SEALE: "Is." Very good. Last 5 sentence.
6 MEMBER SHACK: How about "necessary to 7 propose," just because we've got " development" up there?
8 MEMBER KRESS: Let's get the "is" first. l l
9 MEMBER SHACK: Oh, I'm sorry.
10 MEMBER KRESS: Now, where were you?
l 11 MEMBER SHACK: "It's necessary to develop."
i 12 Since we have "in the development of," I was just going to 1
13 make it "to propose."
,m, 14 MEMBER KRESS: Okay. Yes. That's better. l 15 Good.
16 MEMBER APOSTOLAKIS: Yes. Okay.
l l
17 CHAIRMAN SEALE: Fine. I think we can move on 18 to the next paragraph.
19 MEMBER KRESS: Good.
20 TECHNICAL SECRETARY SUMMERS: If they're "just 21 propose," is it correct to say "These are used" or "These 22 would be used" or "They should be used" or --
23 MEMBER KRESS: That's a good point.
24 MEMBER SHACK: I guess "These would be used"
/"~N
() 25 because, I mean, the whole thing is draft at the moment.
NEAL R. GROSS COURT REPORTERS AND TRANSCRIBERS 1323 RMODE ISLAND AVE., N W.
(202) 234-4433 WASHINGTON, D.C. 20005-3701 (202) 234 4433
549 1 CHAIRMAN SEALE: 'le s .
,, 2 MEMBER KRESS: That's good, Roxie. I thank c
$ 3 you.
4 CHAIRMAN SEALE: That's fine. Okay. Now next 5 paragraph.
6 MEMBER FONTANA: -- rather than "Is"?
7 MEMBER KRESS: No.
8 MEMBER APOSTOLAKIS: Are we on the next 9 paragraph?
10 MEMBER KRESS: No, no. It's "Is" because 11 we're talking about a fait accompli.
12 MEMBER SHACK: Right.
13 MEMBER APOSTOLAKIS: I'm having a --
O
\._ 14 MEMBER KRESS: Yes. The Commission has said 15 they can do it, SRM.
16 MEMBER FONTANA: Without waiting for us to 17 tell them?
18 MEMBER KRESS: No. We told them. We told 19 them.
20 CHAIRMAN SEALE: Earlier.
21 MEMBER KRESS: Earlier. We told them again.
22 MEMBER APOSTOLAKIS: Are we on this paragraph?
23 CHAIRMAN SEALE: Yes, sir.
24 MEMBER APOSTOLAKIS: Now, this paragraph
,s t,, ) 25 implies that it's the state of the art of PRA that NEAL R. GROSS COURT REPORTERS AND TRANSCRIBERS 1o3 RHODE ISLAND AVE., N W.
(202) 234-4433 WASHINGTON, D.C. 20005-3701 (202) 2344433
1 550 l l
1 inhibits the Commission from replacing the adequate level 7- 2 of protection by quantified acceptable risk levels. j
! l '
3 MEMBER KRESS: I think that -- l l
4 MEMBER APOSTOLAKIS: Is that something we want )
5 to say? ;
1 6 MEMBER KRESS: I don't know. Good question. I I
7 MEMBER APOSTOLAKIS: I think that -- i 1
8 MEMBER KRESS: It's what the staff told us -- I 9 MEMBER APOSTOLAKIS: Oh , but the staff is -- l l
10 MEMBER KRESS: -- several times.
l
'l CHAIRMAN SEALE: Well, but do they mean that 12 --
l l
13 MEMBER APOSTOLAKIS: I think the real reason n
i i
'A s' 14 is different. The real reason is that we know that there l l
15 are plants that violate the goals and we are not prepared 16 to face that. It has nothing to do with the state of the 17 art in PRA in my opinion.
18 In fact, we saw the IPE results that Druin l
19 presented here. There are plants that violate the core 20 damage frequency goal. There are many plants, the 21 majority, violate CCFP.
22 And the Commission is not ready to declare, "I 23 have this level now, A. All of you guys do something."
24 MEMBER KRESS: Suppose we start --
(,,) 25 MEMBER APUATOLAKIS: I think that's the real I NEAL R. GROSS
, COURT REPCRTERS AND TRANSCRIBERS 1323 RHODE ISLAND AVE., N.W.
(202) 234-4433 WASHINGTON, D C. 20005-3701 (202) 234-4433
551 1- reason, j l
.e 2 MEMBER KRESS: Suppose.we start that sentence 3- out with "In the long term." l l
I 4 MEMBER SHACK:
"In the longer term," yes.
5 MEMBER KRESS: "In the longer-term."
6 MEMBER APOSTOLAKIS: And drop.the "as the 7 state of the art."
8 MEMBER KRESS: Ye,- That would suit me. 1 9 MEMBER APOSTOLAKIS: Yes.
- 10 CHAIRMAN SEALE: Yes.
11 MEMBER APOSTOLAKIS: That is much more --
12 CHAIRMAN SEALE: Yes.
13 . MEMBER APOSTOLAKIS: -- meaningful to me, b 14 MEMBER SHACK: Now, as long as we're quoting, 15-' isn't the quote really " adequate protection," rather than 16 " adequate level of protection"?
17 MEMBER APOSTOLAKIS: .I'm not even sure you 18 need the quote anymore. Why don't we just --
1 19 MEMBER SHACK: Well, I mean, we need " adequate i 20 protection" here as_ kind of a term.of the art, j 21 MEMBER KRESS: Yes. Let's keep the term. I ;
'22 guess we could drop the quotes since we've-quoted it and 23 we've talked about it up front.
24 CHAla AN SEALE: Yes.
25 MEMBER SHACK: Oh , I'd keep it in quotes NEAL R. GROSS COURT REPORTERS AND TRANSCRIBERS 1323 RHODE ISLAND AVE., N.W.
(202) 234-4433 WASHINGTON, D.C. 20005 3701 (202) 234-4433
- , - a.
552 1 because you really mean it as a --
2 MEMBER KRESS: Exactly.
.S i G
3 CHAIRMAN SEALE: It's a diad.
4 MEMBER KRESS: Well, that's a good word. I'm 5 glad you threw that one in. I like that. It is a diad.
6 MEMBER SHACK: As contrasted to a quaternium?
7 MEMBER KRESS: Yes.
8 CHAIRMAN SEALE: All right, you guys. Next 9 paragraph.
10 MEMBER FONTANA: Looks good to me.
11 TECHNICAL SECRETARY SUMMERS: You either want 12 to repeat 2 here or you want to take out the 2 in the next 13 one.
(
k 14 MEMBER KRESS: Let's put it in each one of l
15 them.
16 CHAIRMAN SEALE: Good.
17 MEMBER KRESS: Okay. Next paragraph.
18 MEMBER FONTANA: Moving on.
19 MEMBER KRESS: Good title.
20 TECHNICAL SECRETARY SUMMERS: You don't really 21 need the "then."
22 CHAIRMAN SEALE: Yes, that's right.
23 MEMBER KRESS: Mark it out, then.
24 MEMBER SHACK: How about "As it is the
,g ft,) 25 responsibility of the NRC"?
NEAL R. GROSS COURT REPORTERS AND TRANSCRIBERS 1323 RHODE ISLAND AVE., N.W.
(202) 234-4433 WASHINGTON. D.C. 20005-3701 (202) 234-4433
, ._ _..._. __._ -._ _ _..__--_-_____,.- _ _ _ m ._._ _ _ m - m_ . -__m.
o 553 .;
l 1 MEMBER KRESS: Good. 1
- f. 2 TECHNICAL SECRETARY SUMMERS: That's better.
lV-Q i 3 MEMBER KRESS: Yes. That's very good. .Gets 4 through this possessive. 1 5 MEMBER SHACK: And I think it's-just the NRC i- .6 safety goal QHOs.
I I
i 7 MEMBER KRESS: Okay. That's! good, good 1 !
8 improvement.
- 9 MEMBER SHACK
- Now, should we have " adequate >
I i 10 protection" in quotes again?- I mean, we either ought to l 1
$' 11 take them all.in or put them all out. Consistency, s
l 12 MEMBER KRESS: Put it in quotes. !
) _
13 CHAIRMAN SEALE: Make it " diad."
[
14 MEMBER KRESS: Make it in quotes.
i l 15 MEMBER FONTANA: Do you really want to say 1
j ]6 there seems to be "no alternative" or "We have identified i
L7 no alternatives"? Forget it.
(- i
- 18 CHAIRMAN SEALE
- Yes. .1
]
l 19 MEMBER FONTANA: Forget that comment.
- i. 20 CHAIRMAN SEALE: Next paragraph. 1 l'
i 21 MEMBER APOSTOLAKIS: So you will not give us a 22 break until we finish the letter?
i.
j- 23 CHAIRMAN SEALE: Yes.
i f' 24 MEMBER APOSTOLAKIS: Yes?
4 25 CHAIRMAN SEALE: I won't.
J
! NEAL R. GROSS i COURT REPORTERS AND TRANSCRIBERS I.
1323 RHODE ISLAND AVE., N.W.
, (202) 234-4433 WASHINGTON, D.C. 20005-3701 (202) 2344433 l.
554 1 (Laughter.)
2 CHAIRMAN SEALE: That's my junkyard dog.
I j 3 MEMBER APOSTOLAKIS: This is too much.
1 1
4 MEMBER SHACK: Can I come back to that .
1 1
5 paragraph? I'm not sure that one logically follows from G the other. I mean, I might agree with it. l 7 MEMBER KRESS: Oh, no. They're separate 8 statements. They're not related.
l 9 MEMBER FONTAI.A: They're not? It doesn't l 10 follow.
11 MEMBER KRESS: Well, they're not supposed to. l l
12 They're two separate statements.
l l
I 13 CHAIRMAN SEALE: Does that change the sentence
('8
() 14 in any way? j l
l 15 MEMBER KRESS: It doesn't to me.
16 MEMBER FONTANA: I think the sentence that 1
17 would follow is that "they must." "As it is their 18 responsibility, they must use plant-specific 19 applications." It's a non-sequitur the way it is.
20 MEMBER SHACK: They haven't done it so far.
21 You mean they've been derelict in their duty all these 22 years?
23 MEMBER KRESS: Oh, no. They always -- with 24 the regulatory system we have now of adequate protection, rh
(,a) 25 it is enforced on a plant-specific basis.
'NEAL R. GROSS COURT REPORTERS AND TRANSCRIBERS 1323 RHODE ISLAND AVE., N.W.
(202) 234-4433 WASHINGTON, D.C. 20005-3701 (202) 234-4433
555 1 CHAIRMAN SEALE: That's right.
,-s 2 MEMBER KRESS: If you're going to do it
- \ j<
3 another way, you're still going to have to do it on a 4 plant-specific basis because that's their --
5 CHAIRMAN SEALE: Yes, yes. Let's go on to the 6 next paragraph. !
7 MEMBER KRESS: I don't think " adequate 8 protection" ought to be in quotes on this line, though. j i
9 MEMBER SHACK: Yes. I agree.
10 MEMBER KRESS: That's where we need to drop 11 the quotes. )
l 12 CHAIRMAN SEALE: All right.
l 13 MEMBER APOSTOLAKIS: Oh, yes. There's a i
(s!
i l
x/ 14 difference. I don't think it's logical. I 15 MEMBER FONTANA: Oh, yes. There's a big 16 difference.
17 MEMBER KRESS: You can't read it from over l l
18 there?
19 CHAIRMAN SEALE: Yes. It's a side angle.
20 MEMBER SHACK: It's easier from there.
21 CHAIRMAN SEALE: These wide screens are not 22 side viewable?
23 MEMBER KRESS: That should be " adequate 24 protection," instead of " adequate level of."
,/ 3
( ,) 25 CHAIRMAN SEALE: Yes. Just say " adequate NEAL R. GROSS COURT REPORTERS AND TRANSCRIBERS 1323 RHODE ISLAND AVE., N W.
(202) 234-4433 WASHINGTON, D C. 20005-3701 (202) 234-4433
556 l
1 protection." '
,y 2 MEMBER APOSTOLAKIS: So we drop the quotes? I i
~'
3 think it's time to do that.
l 4 MEMBER KRESS: Yes. You can -- l 5 MEMBER APOSTOLAKIS: We've used it many times.
6 CHAIRMAN SEALE: Yes. Let's drop it.
7 MEMBER KRESS: Well, you need it there, but 8 you don't need it on the next one.
9 CHAIRMAN SEALE: Yes. l l
l 10 VICE CHAIRMAN POWERS: Tom, the words I handed l 1
11 you I thought went logically at the end of the word 1 12 " regulations" on Line 59.
l 13 MEMBER KRESS: Line 59.
,rN.
s- 14 CHAIRMAN SEALE: That's right there.
15 MEMBER KRESS: Okay. Now let me -- okay.
16 MEMBER SHACK: Okay. I give up. I still 17 don't believe this is the right definition for adequate 18 protection, but I --
19 tiEMBER APOSTOLAKIS: What definition is that?
20 MEMBER SHACK: Does this mean adequate 21 protection increases every time we pass a rule and 22 regulation?
23 VICE CHAIRMAN POWERS: Yes.
24 MEMBER APOSTOLAKIS: Changes.
( ,) 25 MEMBER KRESS: Not necessarily.
NEAL R. GROSS COURT REPORTERS AND TRANSCRIBERS 1323 RHODE ISLAND AVE., N.W.
(202) 234-4433 WASHINGTON, D.C. 20005-3701 (202) 234-4433
557 1 MEMBER APOSTOLAKIS: It changes.
73 2 MEMBER SHACK: I don't believe it does myself.
- )
~~'
3 MEMBER KRESS: Well, let me tell you how it 4 may --
5 MEMBER SHACK: But that's what this says.
6 MEMBER KRESS: Let me tell you why it may not.
7 Later on down there, you'll see that I say that the risk 8 posed by the plant with the highest level of risk provides 9 a least upper bound through adequate protection. That 10 doesn't mean it couldn't be a lot higher.
11 And every time you change the rules, you're 12 only changing the least upper bound one. Adequate 13 protection could be way up there. I don't know where it
/i 14 is because it's not quantified. And that's why I say it's 15 basically unquantified. There is only the possibility at 16 the moment of providing a least upper bound to it.
17 VICE CHAIRMAN POWERS: I think every time we 18 change the rule, we're getting a least upper bound on the 19 risk.
20 CHAIRMAN SEALE: We should be anyway.
21 VICE CHAIRMAN POWERS: We're reducing that 22 least upper bound or most upper bound rule.
23 CHAIRMAN SEALE: Do you want to add that 24 sentence that you've got or that paragraph or that little N ,) 25 --
NEAL R. GROSS COURT REPORTERS AND TRANSCR!BERS 1323 RHODE ISLAND AVE., N W.
(202) 234 4433 WASHINGTON, D.C. 20005-3701 (202) 234 4433
1 558 1 MEMBER KRESS: Yes. Let me give --
,~ 2 CHAIRMAN SEALE: -- couple of lines here?
i i
~
3 MEMBER KRESS: -- just give it to Roxanne and 4 --
5 MEMBER APOSTOLAKIS: Do we really want the 1
6 emphasis there; "i.e., acceptable risk"? I mean, it sort l
7 of casually says that adequate protection and acceptable 8 risk are the same thing. And does it serve any purpose?
9 Tom, what do you think? Do we need the "(i.e., acceptable 10 risk)"?
11 MEMBER KRESS: Well, what do you mean by 12 " adequate protection"?
13 MEMBER APOSTOLAKIS: What you say. If it
\I 14 complies fully with the rules and regulations.
15 MEMBER KRESS: And that gives acceptable risk.
16 We've never said that in --
17 MEMBER APOSTOLAKIS: I don't know about that.
18 MEMBER KRESS: But we've never said that 19 anywhere else, q n
20 MEMBER APOSTOLAKIS: Nobody, not even the 21 Commission. They've never said that. Nobody has ever j 22 said that adequate protection specifies acceptable risk at i 23 that -- !
24 MEMBER KRESS: I think the charter of NRC says
/s !
(_,) 25 that they will provide --
l l
NEAL R. GROSS COURT REPORTERS AND TRANSCRIBERS 13M RHODE ISMND AVE., N W.
(202) 234-4433 WASHINGTON. D.C. 20005-3701 (202) 234-4433
i 559 1 MEMBER APOSTOLAKIS: Adequate protection.
7.s 2 MEMBER KRESS: No. Read the charter.
! )
3 MEMBER APOSTOLAKIS: There will be no undue 4 risk.
5 MEMBER KRESS: No. It says, " risk, undue 6 risk."
7 MEMBER APOSTOLAKIS: Undue risk, but not 8 acceptable risk.
9 MEMBER KRESS: Okay. Say "no undue risk,"
10 then.
11 MEMBER APOSTOLAKIS: That's different, yes.
12 Then you quote --
13 MEMBER KRESS: You're splitting hairs. "No
,o I )
(/ 14 undue risk" means acceptable risk.
15 MEMBER APOSTOLAKIS: Does it?
16 MEMBER KRESS: Yes.
17 VICE CHAIRMAN POWERS: I'm not sure it does.
18 MEMBER APOSTOLAKIS: I don't think so.
19 VICE CHAIRMAN POWERS: No. I think George is 20 absolutely correct. Undue risk and acceptable risk are 21 two different things.
22 MEMBER APOSTOLAKIS: They're two different 23 things.
24 VICE CHAIRMAN POWERS: I sit in my living
,~
( ,/ 25 room. I expect that to be acceptable risk. I go out on a l l
NEAL R. G500SS l COURT REPORTERS AND TRANSCRIBERS 1323 RHODE ISLAND AVE., N W.
(202) 234-4433 WASHINGTON, D.C. 20005-3701 (202) 234-4433
560 1 highway. That's no undue risk.
. 2 MEMBER KRESS: It's acceptable to you, though. l
^
3 VICE CHAIRMAN POWERS: Acceptable personally 4 to me. It may not be acceptable to anybody else. l 5 MEMBER KRESS: Anybody that does it.
l 6 VICE CHAIRMAN POWERS: Anybody that does it 7 it's acceptable to.
8 MEMBER APOSTOLAKIS: Anyway, the letters of l l
l 9 the ACRS used to end by saying, "There is no undue risk," l 10 not " acceptable risk." So we might as well use that.
11 MEMBER KRESS: Well, I'd just as soon leave it l
12 off if you're going to do that to it.
l 13 MEMBER APOSTOLAKIS: Or leave it off. I
^'
/' i !
i !
\/ 14 MEMBER SHACK: Okay. We're making progress.
15 MEMBER APOSTOLAKIS: That's fine with me.
16 TECHNICAL SECRETARY SUMMERS: So we're leaving l
17 this off?
18 MEMBER APOSTOLAKIS: No. The part -- ,
i 19 MEMBER BARTON: "I.e., acceptable risk." Take 20 that out. There you go.
21 MEMBER APOSTOLAKIS: Should you put a comma 22 after " protection"?
23 CHAIRMAN SEALE: No.
24 MEMBER KRESS: No.
r's
( ,), 25 CHAIRMAN SEALE: Okay. Be sure to read that NEAL R. GROSS COURT REPORTERS AND TRANSCRIBERS 1323 RHODE ISLAND AVE., N W.
(202) 234-4433 WASHINGTON, D.C. 20005-3701 (202) 234-4433
561 1 next sentence there because that's the new one, fx 2 MEMBER'KRESS: That "s" should be an "a" 3 there.
4 MEMBER SHACK: That should be an "a"?
5 CHAIRMAN SEALE: Right above the cursor.
6 MEMBER KRESS: There.
7 CRAIRMAN SEALE: That's the sentence we're 8 talking about.
9 MEMBER APOSTOLAKIS: It's two sentences.
10 CHAIRMAN SEALE: Yes. Okay. We've alerted 11 them, then, that we have this gap.
12 MEMBER APOSTOLAKIS: Why are they set aside?
13 I don't understand it.
\->' 14 CHAIRMAN SEALE: Because we're saying 15 everything is going to be in terms of those things that 16 you can quantify.
17 VICE CHAIRMAN POWERS: We can't do anything 18 about sabotage or materials diversion.
19 MEMBER APOSTOLAKIS: Right. Well, those, too, 20 but, I mean, the integration phase of what the staff 21 proposed is supposed to take care of everything.
22 VICE CHAIRMAN POWERS: What we're going to 23 discuss now is how adequate protection relates to the 24 quantitative measures of risk.
((3,) 25 MEMBER KRESS: With this part missing.
NEAL R. GROSS COURT REPORTERS AND TRANSCRIBERS 1323 RHODE ISLAND AVE., N.W.
(202) 234-4433 WASHINGTON, D.C. 20005-3701 (202) 234-4433
562 1 CHAIRMAN SEALE: In the following discussion.
73 2 Okay. ;
) l 3 MEMBER APOSTOLAKIS: Now, do you want to cite 4 this NUREG where you say, "the natural result is a
]
l 5 spectrum of risk levels across the population of operating 6 plants" as the recent IPE studies have demonstrated? I !
7 mean, that's a major study from the staff.
8 MEMBER KRESS: If you'd like to put that in, 9 it would be all right. You can --
10 MEMBER APOSTOLAKIS: Because there is more to 11 it if you look at some of the figures here. I mean, they 12 do violate the goal.
13 MEMBER KRESS: You could put that in there if 14 you'd like.
15 MEMBER APOSTOLAKIS: I mean, this is a major 16 effort on the part of the staff.
17 CHAIRMAN SEALE: So you'd put " population of 18 operating plants"?
19 MEMBER APOSTOLAKIS: It's the IPE status.
20 Yes.
21 VICE CHAIRMAN POWERS: It's actually quite 22 interesting.
23 MEMBER APOSTOLAKIS: It's quite interesting.
24 MEMBER KRESS: It wouldn't hurt to give them a
("\
( ,) 25 little bitty here.
NEAL R. GROSS COURT REPORTERS AND TRANSCRIBERS 1323 RHODE ISLAND AVE., N W.
(202) 234-4433 WASHINGTON D C. 20005-3701 (202) 234-4433
563 1 MEMBER APOSTOLAKIS: "Across the population of s 2 operating plants, as the recent IPE studies have
)
3 demonstrated."
4 MEMBER KRESS: Okay. I think that --
5 MEMBER APOSTOLAKIS: And then cite, remember 6 to cite, this NUREG, the IPE NUREG.
7 MEMBER KRESS: That's IPE.
8 MR. DUDLEY: Has everybody in the Committee 9 had an access to this NUREG --
10 MEMBER APOSTOLAKIS: Oh, yes. We all have it.
11 We all have it. It's NUREG-1560.
1 12 VICE CEAIRMAN POWERS: However, in their 13 problem, they don't do risk in the IPE. l f^^)
i
's_/ 14 MEMBER APOSTOLAKIS: Yes. It's CDF and l l
15 containment. l 16 MEMBER KRESS: Yes. So you really don't --
17 MEMBER APOSTOLAKIS: It's not risk levels.
18 You're right.
19 MEMBER KRESS: So you really can't ascertain 20 this spectrum out of there. Maybe you ought to leave it 21 off.
22 MEMBER APOSTOLAKIS: Well, if you have 23 different CDFs and different CCFPs.
24 MEMBER KRESS: If you have a CDF of 1 times
(~~)\
( 25 10'3, you could probably infer that they --
NEAL R. GROSS COURT REPORTERS AND TRANSCRIBERS 1323 RHODE ISLAND AVE., N W.
(202) 234-4433 WASHINGTON, D C. 20005-3701 (202) 234-4433
, . _ . _ _ . . _ . . _ . _ _ _ _ . . ~
564 1 VICE CHAIRMAN POWERS: And there are some that 2 come pretty close, by the way,
.b-3 MEMBER KRESS: That's right.
4' MEMBER APOSTOLAKIS: Well, we can cite them l 5 later when we say some of the plants --
6 MEMBER SHACK: Although in a water reactor l
, 7 safety meeting based on this, they actually give you a l-j' 8 list of plants that are above safety goals, 9 -MEMBER KRESS: Safety goals.
1
! 10 MEMBER APOSTOLAKIS: Yes,- but they're 11 subsidiary goals, not --
l
, 12 MEMBER SHACK: No, no. l l l l i 13 MEMBER APOSTOLAKIS: The actual goal? I ym '
14 MEMBER SHACK: Safety goal. They estimate 15 that they are above the safety goals.
i 16 VICE CHAIRMAN POWERS: I haven't seen that 17 paper.
18 MEMBER KRESS: I hadn't seen that one, but 19 that's interesting.
20 MEMBER SHACK: That's the one where Palo Verde 21 gets listed. That's the one that always bothers --
22 MEMBER KRESS: Well, you could say the reason L 23 IPE indicates, studies indicate" -- you don't have to say l
24 the --
25 VICE CHAIRMAN POWERS: Should I have seen this NEAL R. GROSS COURT REPORTERS AND TRANSCRIBERS 1323 RHODE ISLAND AVE., N.W.
(202) 2344433 WASHINGTON, D.C. 20005-3701 (202) 234-4433
.. _ . _ .._~ _
565 1 paper?
2 CHAIRMAN SEALE: Yes.
'o i )
~~
3 MEMBER APOSTOLAKIS: I think later on when you 4 talk about the relationship between adequate protection 5 and safety goals, this will have to be --
6 MEMBER KRESS: What will happen?
7 CHAIRMAN SEALE: The IPE studies.
8 MEMBER KRESS: I don't cite it here. I don't 9 cite it somewhere else.
10 MEMBER SHACK: It's the dude from CMD.
11 MEMBER APOSTOLAKIS: IPE studies indicate, 12 studies indicate.
13 VICE CHAIRMAN POWERS: That's why I don't y s
'k/) 14 know.
15 CHAIRMAN SEALE: Okay. Bottom half of this 16 paragraph, any problems?
17 MEMBER APOSTOLAKIS: "Is basically 18 unquantifiable."
19 MEMBER KRESS: No. "Unquantified."
20 MEMBER APOSTOLAKIS: "Unquantified"?
21 MEMBER KRESS: Yes. I can quantify it. It's 22 not unquantifiable. I can get a value for it.
23 MEMBER APOSTOLAKIS: I don't know.
24 VICE CHAIRMAN POWERS: Well, Tom, do you want 25 to -- I think this basically unquantified" is a bit of a
()
NEAL R. GROSS COURT REPORTERS AND TRANSCRIBERS 1323 RHODE ISLAND AVE., N.W.
(202) 234-4433 WASHINGTON, D C. 20005-3701 (202) 234-4433
566 1 colloquialism. Don't you want to say "is to date
,y 2 unquantified"?
N]'^^
3 MEMBER KRESS: Okay.
4 MEMBER APOSTOLAKIS: Wait a minute now.
5 Adequate protection means --
6 MEMBER KRESS: Just how about "is 7 unquantified."
8 VICE CHAIRMAN POWERS: "Is unquantified."
9 Yes.
10 MEMBER APOSTOLAKIS: Adequate protection means 11 comply with the rules and regulations. Rules and 12 regu'ations cover a wide spectrum of things, including 13 programmatic requirements, including all sorts of things.
t
's-)s 14 VICE CHAIRMAN POWERS: Which we have set ,
15 aside.
16 MEMBER APOSTOLAKIS: And all your --
17 VICE CHAIRMAN POWERS: We've set aside 18 anything that was not quantifiable.
19 MEMBER APOSTOLAKIS: I thought we set aside -- l 20 what is it?
21 MEMBER SHACK: Those are examples.
22 MEMBER APOSTOLAKIS: We just set aside 23 sabotage and nuclear materials.
24 VICE CHAIRMAN POWERS: Those are two examples.
p
( ,) 25 MEMBER SHACK: Such as?
NEAL R. GROSS COURT REPORTERS AND TRANSCRIBERS 1323 RHODE ISLAND AVE., N.W.
(202) 234 4433 WASHINGTON, D.C. 20005-3701 (202) 234-4433
567 1 MEMBER APOSTOLAKIS: I don't know. It's up to
, _ . 2 you, not me. So even if you leave those out, you think
\
3 that an absolute risk level is not quantified at this 4 time, but it will be at some time in the future?
5 VICE CHAIRMAN POWERS: No. It'c a situation 6 of it's not quantified now, but he's going to quantify it.
7 MEMBER KRESS: I can quantify it.
8 VICE CHAIRMAN POWERS: He can.
9 MEMBER KRESS: I don't intend to.
10 MEMBER SHACK: Do we need the least --
11 MEMBER APOSTOLAKIS: Anything I can not 12 quantify, I will not include. Therefore, everything else 13 I can quantify. I mean, what does that mean?
(
, 4 L ,/ 14 VICE CHAIRMAN POWERS: It means we've got all 15 the bases covered. You cannot find fault with this 16 phraseology here. We have covered the bases.
17 CHAIRMAN SEALE: If it ain't there, it ain't.
18 VICE CHAIRMAN POWERS: What I have difficulty 19 quantifying, I put it up in the upper category.
20 MEMBER APOSTOLAKIS: It seems to me adequate 21 level of protection is unquantifiable, no matter what the 22 state of the art is, because it refers to programs. It 23 refers to the good practice. It refers to a lot of things 24 that we cannot quantify.
25 (J ) VICE CHAIRMAN POWERS: And we've set all of NEAL R. GROSS COURT REPORTERS AND TRANSCRIBERS 1323 RHODE ISLAND AVE., N.W.
(202) 234 4433 WASHINGTON, D C. 20005-3iO1 (202) 234-4433 I
568 1 those aside.
7, 2 TECHNICAL SECRETARY SUMMERS: Are you making a
)
3 distinction between the non-quantifiable and the l 4 unquantifiable or unquantified?
5 VICE CHAIRMAN POWERS: The non-quantified --
6 everything is non-quantified. Some fraction of it is 7 unquantifiable in the current state of technology.
8 TECHNICAL SECRETARY SUMMERS: You've got 9 "non-quantifiable" up here and "non-quantified" down here.
10 VICE CRAIRMAN POWERS: Yes. And we've used 11 them as we want to.
12 MEMBER FONTANA: Wait a minute. If you're 13 doing a PRA on an existing plant, you're quantifying gy
\_s 14 something. i 15 MEMBER KRESS: That's part of the risk. q 16 VICE CHAIRMAN POWERS: That part of the risk 17 that's quantifiable.
18 MEMBER FONTANA: Right.
19 VICE CHAIRMAN POWERS: And that part that's 20 not, we don't, 21 MEMBER APOSTOLAKIS: The main message is the 22 last sentence, is it not?
23 MEMBER SHACK: Yes.
24 MEMBER APOSTOLAKIS: "Since each licensed
,es 25 plant" and so on?
() Isn't that the last sentence?
HEAL R. GROSS COURT REPORTERS AND TRANSCRIBERS 1323 RHODE ISLAND AVE., N W.
(202) 234 4433 WASHINGTON, D C. 20005-3701 (202) 234-4433
569 1 VICE CHAIRMAN POWERS: That's the main
,,, 2 message.
! \
\~j 3 MEMBER APOSTOLAKIS: Drop the sentence 4 beforehand.
5 MEMBER KRESS: It puts it in context.
6 MEMBER APOSTOLAKIS: Drop the sentence before, 7 the last --
8 MEMBER KRESS: It puts it in context, though.
9 MEMBER APOSTOLAKIS: But it confuses the hell 10 out of me.
11 VICE CHAIRMAN POWERS: He finally despaired of 12 looking through the --
13 MEMBER APOSTOLAKIS: There are many plants
. 14 "the natural result is a spectrum of risk levels. Since 15 each licensed plant must by definition provide adequate 16 ,
protection, then the licensed plant that poses the highest 17 level of risk" is the "least upper bound." There is some 18 logic to it. This penultimate sentence there is 19 confusing. And it is not needed in the argument.
20 MEMBER SHACK: I'll buy that with George, 21 especially now that we've got the range of plants as the 22 IPE indicates.
23 MEMBER APOSTOLAKIS: Yes.
24 MEMBER SHACK: That sentence --
ym
) 25 MEMBER KRESS: Zap it. Zap it.
NEAL R. GROSS COURT REPORTERS AND TRANSCRIBERS 1323 RHODE ISLAND AVE., N.W.
(202) 234-4433 WASHINGTON, D C. 20005-3701 (202) 234-4433
570 l
\
1 MEMBER APOSTOLAKIS: Okay.
l l
,s
, 2 VICE CHAIRMAN POWERS: Your credibility for l
'~
3 dropping the sentence is --
4 MEMBER APOSTOLAKIS: Is a very useful second 5 --
6 VICE CHAIRMAN POWERS: If you ever want to add 7 one, we take you seriously. But dropping one just carries 8 no weight whatsoever.
9 CHAIRMAN SEALE: That's kind of a hurting i 10 statement. Okay. Have we beaten this poor paragraph to l l
11 death?
l 12 MEMBER SHACK: "Least upper bound." How about 13 just put the "least upper bound" on the level of risk?
r~N ,
N'_- 14 MEMBER KRESS: Well, it is quite possible that l
15 adequate protection might not be a similar value. l l
16 MEMBER SHACK: By definition, the highest !
l 17 level of risk, you can have the least upper bound on it. i l
18 MR. SHERRY: Adequate protection can't have a l
19 value. For example, you have a wide range in your tech l 20 specs of what you can do legally. For example, a plant 21 could take a piece of equipment out for two weeks and in 22 two weeks and in two weeks and two weeks still meet the 23 regulations that would be within adequate protection, but 24 that could lead to a high risk, i f~n l
( ,) 25 But they don't do that. So I don't know that l l
NEAL R. GROSS COURT REPORTERS AND TRANSCRIBERS 1323 RHODE ISLAND AVE., N W.
(202) 234-4453 WASHINGTON, D C. 20005-3701 (202) 234-4433
571 1 you could ever --
,r 3 2 MEMBER KRESS: Figure out --
t !
'~
3 MR. SHERRY: -- quantify what 4 MEMBER APOSTOLAKIS: is.
5 MEMBER KRESS: -- the level associated with 6 what they're allowed to do.
7 MR. SHERRY: And they no doubt operate the 8 plant --
9 MEMBER KRESS: It has a very precise 10 mathematical meaning.
11 MEMBER APOSTOLAKIS: Yes. What does it mean?
12 MEMBER KRESS: It means it is --
13 MEMBER APOSTOLAKIS: The smallest upper bound?
/ h
' w/
14 VICE CHAIRMAN POWERS: Right.
15 MEMBER KRESS: Yes.
16 MEMBER APOSTOLAKIS: That's not what it is.
17 MF: >ER KRESS: The smallest one you can place 18 in --
19 MEMBER APOSTOLAKIS: It's not even an upper 20 bound because of the things that are missing. That's what 21 I'm saying, that --
22 MEMBER SHACK: Well, it's at least upper bound 23 on the quantified --
24 MEMBER APOSTOLAKIS: On the quantified, yes, f
( ,) 25 MEMBER SHACK: And we're setting everything NEAL R. GROSS COURT REPORTERS AND TRANSCRIBERS 1323 RHODE ISLAND AVE., N.W.
(202) 234-4433 WASHINGTON, D C. 20005-3701 (202) 2344433
572 1 else aside.
,- 2 MEMBER APOSTOLAKIS: No. We have --
(
r i
~
3 VICE CHAIRMAN POWERS: Put that out of your 4 mind. It's gone.
5 MEMBER APOSTOLAKIS: I think we should repeat 6 it, "at least upper bound on the quantified level of risk 7 to be associated with the adequate protection."
8 MEMBER KRESS: That's all right.
9 MEMBER APOSTOLAKIS: I mean, these are the 10 commissioners. Three of them --
11 MEMBER SHACK: We're replacing precision with 12 clarity here. You know, it's more precision than the 13 concept is worth, p
s 14 MEMBER APOSTOLAKIS: For heaven's sake. It's 15 upper bound when you know that it's not an upper bound?
16 MEMBER SHACK: It's three decimal points on a 17 PRA number, George.
18 MEMBER APOSTOLAKIS: I think Dana is right.
19 Just because we're not dropping anything, you decided to 20 object. I think when we come to words, you should be on 21 my side.
22 CHAIRMAN SEALE: All right. Next paragraph.
23 MEMBER APOSTOLAKIS: Wait a minute. Wait a 24 minute. What did you do with the least upper bounds?
,n i j) s 25 MEMBER KRESS: It's left in there.
NEAL R. GROSS COURT REPORTERS AND TRANSCRIBERS 1323 RHODE ISLAND AVE., N.W.
(202) 234-4433 WASHINGTON, D.C. 20005-3701 (202; 234-4433
573 1 MEMBER APOSTOLAKIS: The author agreed to
,, 2 change it.
(
\
i l
3 MEMBER KRESS: No. I agreed to do what you 4 said.
5 MEMBER APO!srOLAKIS: Yes.
6 MEMBER KRESS: That leaves upper bound in 7 there.
8 MEMBER APOSTOLAKIS: Yes.
9 MEMBER KRESS: "Least upper bound on the 10 quantified level of risk."
11 MEMBER APOSTOLAKIS: What is it? Okay. I'm 12 sorry. I'm sorry.
13 VICE CHAIRMAN POWERS: George, as the official I l
U 14 junkyard dog, it is imperative that you pay attention.
15 MEMBER APOSTOLAKIS: No. I'm talking without 16 paying attention. j l
17 VICE CHAIRMAN POWERS: I've noticed that. And )
18 you're being told by the junkyard dog to knock it off.
19 MEMBER APOSTOLAKIS: So where are we, "Within i
20 the spectrum"? I 21 CHAIRMAN SEALE: Yes.
22 VICE CHAIRMAN POWERS: By the way, Tom, that 23 sentence is eloquent. I like it.
t 24 MEMBER KRESS: Thank you.
/^N
!, j 25 MEMBER FONTANA: Is that a true statement?
NEAL R. GROSS COURT REPORTERS AND TRANSCRIBERS 1323 RHODE ISLAND AVE., N W.
(202) 234-4433 WASHINGTON, D C. 20005 3701 (202) 234-4433
574 1 Even if you're including only the quantifiable risk, is l I
7._s 2 that still true? l 6
~
3 MEMBER KRESS: Yes, probably still true.
4 Nobody knows if it is true because nobody has done an 5 evaluation to see where the risk level of plants is with 6 respect to the safety goals. It's all speculation, but 7 it's speculation based on knowledge and IPEs and l l
8 NUREG-1150 and going down and quantifying low-power and l
9 shutdown risk and other things, i l
10 So nobody knows if this is the case. That's 11 why the words "it is likely" are in there.
l l
12 VICE CHAIRMAN POWERS: I like this paragraph. l l
13 I think it -- l p-14 CHAIRMAN SEALE: Yes.
15 VICE CHAIRMAN POWERS: I like it.
16 CHAIRMAN SEALE: It really summarizes the 17 whole --
18 MEMBER KRESS: It takes care of everything.
19 VICE CHAIRMAN POWERS: There's a whole bunch 20 of stuff in there that has a --
21 CHAIRMAN SEALE: Gets pulled together.
22 VICE CHAIRMAN POWERS: Yes, yes. It tells you 23 where you stand in space.
24 CHAIRMAN SEALE: And time.
n k_j 25 VICE CHAIRMAN POWERS: Well, I don't know NEAL R. GROSS COURT REPORTERS AND TRANSCRIBERS 1323 RHODE ISLAND AVE., N W.
(202) 234-4433 WASHINGTON, D C. 20005-3701 (202) 234-4433
575 1 about time. Space, f3 2 MEMBER APOSTOLAKIS: I t.' i nk the paragraph is
! \
3 fine, but if I were a commissioner and read the last 4 sentence, I would say, "What the hell are these guys 1
5 telling me?" The problem that is difficult is difficult.
6 VICE CHAIRMAN POWERS: No.
1 7 CHAIRMAN SEALE: No, no. We're telling them l 8 that Gary Holahan's way of treating the plants that are 9 currently above 10 is a workable way of doing it. l 10 VICE CHAIRMAN POWERS: But it has to be done !
I 11 within the confines of the backfit rule.
l l
12 CHAIRMAN SEALE: Exactly.
13 VICE CHAIRMAN POWERS: I mean, it is taking a
~.
l
'/ 14 risk base and segmenting it and telling you what you can 15 do within those regions of the risk base. I think it's 16 almost picturesque.
l 17 MEMBER APOSTOLAKIS: All right. l 18 MEMBER FONTANA: I'll bet if you go to a QHO, l l
19 there's not --
20 MEMBER KRESS: What?
21 MEMBER FONTANA: Some plants above the safety 22 goal.
23 TECHNICAL SECRETARY SUMMERS: Mario, you've 24 got to talk into the microphones.
/~N
(,) 25 VICE CHAIRMAN POWERS: Mario, I will guarantee MEAL R. GROSS COUPT REPORTERS AND TRANSCRIBERS 323 RHODE ISLAND AVE., N.W.
(202) 234-4433 iVASHINGTON, D.C. 20005-3701 (202) 234-4433
576 l 1 you --
1
,- s 2 MEMBER KRESS: Did you hear that, Dana?
) I 3 VICE CHAIRMAN POWERS: I will absolutely l l
4 guarantee you that almost no plant satisfies the QHOs. l l
l 5 MEMBER KRESS: So there you are. l 6 MEMBER APOSTOLAKIS: The quantified --
7 MEMBER FONTANA: Only talk about the 8 quantified --
9 VICE CHAIRMAN POWERS: The QHOs are very 10 quantitative. l 11 MEMBER FONTANA: Yes, but they're calculating 12 them only with the things you can calculate.
13 MEMBER KRESS: Yes.
/
t i
' 's- 14 VICE CHAIRMAN POWERS: Yes.
15 MEMBER FONTANA: I don't know. You're going 16 to have to convince me.
17 VICE CHAIRMAN POWERS: It's an enormously l
18 simple calculation to do.
l l 19 MEMBER FONTANA: The ones I've done are so far 20 below it it is just --
21 VICE CHAIRMAN POWERS: But the ones you have f 22 seen have only counted risk at power. Now take that --
23 MEMBER APOSTOLAKIS: But we have eliminated 24 that. We have --
f~%
() 25 VICE CHAIRMAN POWERS: No, no, no.
NEAL R. GROSS COURT REPORTERS AND TRANSCRIBERS 1323 RHODE ISLAND AVE., N.W.
(202) 234-4433 WASHINGTON, D.C. 20005-3701 (202) 234-4433
1 577 .
d 1 Unquantifiable. These other things are quantifiable. And l g 2 you just take all those numbers that you got for risk at
( "!
3 power. You multiply them by three. And then you take 4 about 1.4 times twice the result of your product that you 5 got.
6 MEMBER FONTANA: Square root of 2 is what you i
7 said.
8 VICE CHAIRMAN POWERS: And you will then add j 9 that into it, and you will find that it will violate the 10 safety goal, every single one of them.
11 MEMBER FONTANA: I'11 talk to you later.
12 CHAIRMAN SEALE: Or you can just multiply it 13 by five.
~s i 1 Y~ / 14 MEMBER APOSTOLAKIS: By the way, should we --
15 VICE CHAIRMAN POWERS: It's actually --
16 MEMBER APOSTOLAKIS: I mean, we do have to >
17 treat everything --
18 VICE CHAIRMAN POWERS: -- five plus one over 19 alpha.
20 MEMBER APOSTOLAKIS: -- that the CDF goal is 21 violated. And we keep referring to the QHos. Should we 22 just say that someplace?
23 MEMBER KRESS: We might want to somewhere, but 24 --
n
) 25 MEMBER APOSTOLAKIS: Before you reach the NEAL R. GROSS COURT REPORTERS AND TRANSCR!BERS 1323 RHODE ISLAND AVE., N.W.
(202) 2344433 WASHINGTON, D.C. 20005-3701 (202) 234-4433
578 1 point where you say " elevate a CDF ancillary goal to the
~s 2 status of a fundamental." This is a major result of this
/ i
\ l
'^
3 IPE exercise for the last eight years.
4 If you look at this, there are two figures in 5 the executive summary. One is on core damage frequency, 6 clearly violated. And one is on containment, conditional 7 containment, probability, which is ridiculous.
8 VICE CHAIRMAN POWERS: Which just devastates 9 it.
10 MEMBER FONTANA: They both have no meaning.
11 MEMBER APOSTOLAKIS: They both have no 12 meaning?
13 MEMBER FONTANA: They don't. The core damage (3
(_) 14 frequency, if you don't relate that to f4ssion product 15 released, it gets out of the containment, and gets some at 16 the site boundary is a subsidiary thing that in itself 17 doesn't have a meaning.
18 VICE CHAIRMAN POWERS: Not according to this 19 letter, it's not.
20 MEMBER APOSTOLAKIS: Not according to this 21 (;'atter. Yes, 10".
I 22 MEMBFR KRESS: We'll change that.
23 VICE C:9.l.RMAN POWERS: 10" according to this 24 letter is about to go up to be a God-given carved in stone
/-
t w/
/ 25 --
NEAL R. GROSS COURT HEPORTERS AND TRANSCRIBERS 1323 RHODE ISLAND AVE., N W (202) 234-4433 WASHINGTON, D C. 20005-3701 (202) 234-4433
- - - - . . . - - . . - . - . . - . _ . . - . - . _ - . ~ - . - , ~.
} 579 :
j- 1- MEMBER KRESS: Sacred calf.
2 MEMBER FONTANA: But you put it in with LERF. I
- O 3 MEMBER KRESS
- No, no. You --
I 4 MEMBER FONTANA: It's the LERF that's going to 5 --
?
j' 6 MEMBER APOSTOLAKIS: You're supposed to do j 7 toth. t i- i 3 8 VICE CHAIRMAN POWERS: Hey,'if you look at~his l
- 9 numbers there, on the containment failure, he's got ,
i-1 10 bypass, he's'got early containment failure, and he's got j i
). -\
l- 11 late. Throughout the late, just take the early and the 12 bypass. It still blows the CCFP out of the water.
13 MEMBER APCPTOLAKIS: And we said earlier, b
V 14 remember, our emphasis, all CDF and CCFP --
15 CHAIRMAN SEALE: I would urge that we go ahead 16 and finish this thing and then go back and-decide where to l
17 put that in.
18 MEMBER APOSTOLAKIS: Yes I think, yes.
19 CHAIRMAN SEALE: Okay. So next paragraph.
20 VICE CHAIRMAN POWERS: Tom, I espouse and 21 support this " Regulatory Transparency" paragraph even 22 though I know it's false as sin. Okay? It ain't going to 23 help a bit.
24 MEMBER KRESS: It's made up of whole cloth; 25 right?-
NEAL R. GROSS COURT REPORTERS AND TRANSCRIBERS 1323 RHODE ISLAND AVE,, N.W.
(202) 234 4433 WASHINGTON, D.C. 20005-3701 (202) 234 4433 ll
580 1 MEMBER SHACK: We'll go the whole nine yards.
,- 2 MEMBER KRESS: Go the whole nine yards.
\
(
- s. 'J
~
3 CHAIRMAN SEALE: But that ain't a first down.
4 VICE CHAIRMAN POWERS: This is not going to l
5 make it clear to the public. Actually, all it's going to 6 do is infuriate the public.
7 MEMBER KRESS: think it would help. ,
1 8 VICE CHAIRMAN POWERS: No, I think it will 9 not help at all. It will help us.
10 MEMBER APOSTOLAKIS: Yes. I don't think the 11 public has anything to do with this.
12 MEMBER KRESS: Well, the NRC commissioners are 13 very sensitive to the public. That's why I put it in
- i
\/ 14 there.
15 MEMB~iR APOSTOLAKIS: Yes. But, I mean --
16 VICE CHAIRMAN POWERS: However, should the
'7 commissioners ask us just exactly how this is helping with 18 tran m y, I guarantee you I'm going to go --
19 MEMBER KRESS: I could explain the level of 20 risk if I have a value there and say, "I make all the 21 plants conform to this level of risk." But if I tell 22 them, "I make them conform to all the regulations," they 23 say, "So what? What does that mean?"
24 And then I say, "Well, it means they're safe."
(-
(,) 25 I think it would be --
NEAL R. GROSS COUV CEPORTERS AND TRANSCRIBERS
% i RHODE ISLAND AVE., N W.
(202) 234-4433 WASHINGTON, D C. 20005-3701 (202) 234 4433
581 l' -VICE CHAIRMAN POWERS: 'I think you do, Tom.
g 2 MEMBER KRESS: I think it'would clarify it.
\~) '
-3 VICE CHAIRMAN POWERS: I think I could 1
- 4. probably do a dance of seven veils for them because I 5 said: Look, rules and regulations are made by bureaucrats ;
6 in Washington that you don't know, that don't know your .
1 1
7 plant,.that don't know your society.
8 MEMBER KRESS: That's the idea, yes.
9 VICE CHAIRMAN POWERS: Whereas, this risk, 10 this is an absolute thing that I compare against your life 11 and the risks that you face day to day.
12 MEMBER KRESS: Right. That's it. And that's 13 the reason they did the. safety goals, one reason.
-s 14 VICE CHAIRMAN POWERS: I think they did the 15 safety goals to define how safe is safe enough, but-I 16 agree with you. You can do the dance of the seven veils 17 here.
18 CHAIRMAN SEALE: Are we going to do the dance 19 of the seven veils?
20 VICE CHAIRMAN POWERS: I think we just let the 21 paragraph go as it is.
22' MEMBER KRESS: I second the motion.
23 CHAIRMAN SEALE: Okay. Next paragraph.
24 MEMBER APOSTOLAKIS: How about -- wait. Do l%,g 25 you really need to say "and if enforced"? Drop "if NEAL R. GROSS COURT REPORTERS AND TRANSCRIBERS 1323 RHODE ISLAND AVE. N.W.
(202) 234-4433 WASHINGTON, D C. 20005 3701 (202) 234-4433
582 1 enforced."
2 MEMBER KRESS: Oh, you want to drop that?
I \
3 MEMBER APOSTOLAKIS: "If implemented"?
4 MEMBER KRESS: I want " enforced." This goes 5 back to the concept of whether you should enforce the 6 safety goals as acceptance criteria for individual plants.
7 VICE CHAIRMAN POWERS: I think you don't need 8 the " enforced," Tom.
9 MEMBER KRESS: Okay. I'll cede. What do you 10 need there?
11 VICE CHAIRMAN POWERS: You don't need 12 anything.
13 MEMBER KRESS: Okay.
,r'~S k) 14 MEMBER APOSTOLAKIS: Yes.
15 CHAIRMAN SEALE: Okay. Next paragraph.
16 MEMBER KRESS: Will the "if enforced" be 17 understood?
18 MR. MARKLEY: Tom, could you define regulatory 19 transparency for me?
20 MEMBER APOSTOLAKIS: Well, if it's not 21 enforced, who cares?
22 VICE CHAIRMAN POWERS: No. You don't want to 23 ask that question.
24 MEMBER KRESS: It means --
yy (j 25 VICE CHAIRMAN POWERS: Sit down. You don't NEAL R. GROSS COURT REPORTERS AND TRANSCRIBERS 1323 ftHODE ISLAND AVE., N W.
(202) 234-4433 WASHINGTON, D C. 20005-3701 (202) 234-4433
- . - . . - . . - .- _. . . - . . . . . - , - . _ ~ _ . . . . . . . , . - . . . . . .
t 583 1c want to ask'that. question.
s
.. 2 MR. MARKLEY: I want it to be opaque so I can
?I 3 see it.
4 MEMBER KRESS: It's a buzzword. Okay.
j 5 CHAIRMAN SEALE: Okay. That first paragraph
~ 6' is a lead-in to these.
- 7 VICE CHAIRMAN POWERS
- Can I write additional 8 comments that say that early fatality QHOs generally do 9 not control the risk if we look at shutdown and external 10 event accidents?
11 MEMBER KRESS: I would rather work'that in 12 here.
13 MEMBER SHACK: I mean, why do we have to '
14 restrict it to the early fatalities?
15 MEMBER KRESS: We. don't. It's just that my 16 Attachment 1 only dealt with early fatalities. But you 17 could use either one of them if you wanted to. I mean --
18 VICE CHAIRMAN POWERS: Yes. Don't you say 19 that in your attachment? .l 20 MEMBER KRESS: I don't think I really say it 21 very clearly in the attachment.
22 VICE CHAIRMAN POWERS: Don't you say that, 23 "Here's a way to derive these things" --
24 MEMBER KRESS: Yes, and --
25 VICE CHAIRMAN POWERS: -- "using a measure of NEAL R. GROSS 4 COURT REPORTERS AND TRANSCRIBERS I 1323 RHODE ISLAND AVE., N W.
(202) 234-4433 WASHINGTON, D.C, 20005 3701 (202) 234 4433
584 1 consequences"?
, 2 MEMBER KRESS: Yes.
3 VICE CHAIRMAN POWERS: And you say, "We're 4 going to use the prompts"?
5 MEMBER KRESS: Yes, I do that.
6 VICE CHAIRMAN POWERS: I think you try to 7 cover --
8 MEMBER KRESS: You may just very well have put j i
l 9 in the other --
l 10 VICE CHAIRMAN POWERS: I think you've got it l l
l 11 Covered there. l I
12 CHAIRMAN SEALE: Yes. And you're 13 over-qualifying it in a sense in terms of the n l
( ) l L/ 14 applicability cf the idea. Why don't you just say "QHO," l 15 period? l 16 MEMBER KRESS: Okay. Then you have to put an 17 "s."
18 CHAIRMAN SEALE: Okay.
19 MEMBER KRESS: Okay?
20 CHAIRMAN SEALE: Safety goals early 21 expression.
22 MEMBER KRESS: Next paragraph. That's almost 23 a non-sequitur there.
24 CHAIRMAN SEALE: Yes.
,i~y j 25 MEMBER APOSTO;%KIS: So you're in Number 2 NEAL R. GROSS COURT REPORTERS AND TRANSCRIBERS 1323 RHODE ISLAND AVE., N W.
(202) 234-4433 WASHINGTON. D C. 20005-3701 (202) 234-4433
585 1 now?
,s 2 MEMBER SHACK: No. I think that's an
/
i i
- 3 interesting piece of information.
4 MEMBER APOSTOLAKIS: Except under very special 5 circumstances, does it need any elaboration?
6 VICE CHAIRMAN POWERS: Well, I think it's that 7 part that gives me the biggest heartburn.
8 MEMBER SHACK: I thought Dana was going to 9 replace that with a more specific statement.
10 VICE CHAIRMAN POWERS: The trouble is I don't 11 know. I mean, I honestly don't know what controls under 12 shutdown conditions. I think it's the latent.
13 MEMBER KRESS: I would guess it's the early
\- 14 shutdown also.
15 VICE CHAIRMAN POWERS: It very much could be 16 except it's a low-intensity source term. You don't get 17 this burst of a very rich radioactivity. You get a 18 continuing stream of radioactivity out.
19 MEMBER KRESS: I'm not sure.
20 CHAIRMAN SEALE: Well, actually, you know, 21 this is the place where you could really make the argument 22 that the present setup is incomplete in the sense that the 23 real risk here is to plant --
24 VICE CHAIRMAN POWERS: Could we say something
,/ 3
(_,) 25 like this, Tom? And I am very much thinking out loud NEAL R. GROSS COURT REPORTERS AND TRANSCRIBERS 1323 RHODE ISLAND AVE., N W.
(202) 234-4433 WASHINGTON, D C. 20005-3701 (202) 234-4433
586 1 here.
f3 2 MEMBER KRESS: Okay.
i 4 3 VICE CHAIRMAN POWERS: "The early fatality QHO 4 generally controls the risks trom reactor operations" --
5 MEMBER KRESS: Right.
6 VICE CHAIRMAN POWERS: -- or " reactor power 7 operations," let's say. "Our level of understanding" or 8 it is likely that the QHOs on latent effects are also met.
9 Our understanding on shutdown modes of operation and 10 accidents initiated by external events isn't sufficient 11 yet to make definitive statements on the controlling 12 risks."
13 MEMBER KRESS: Okay.
! i V 14 TECHNICAL SECRETARY SUMMERS: I didn't get l
15 that all, but --
l l
16 MR. DUDLEY: Another suggestion that captured l 17 Dana's thought is to move that thought to the end of this 18 section, after Bullet 3, and then at that point make a 19 statement, " Risk assessments to verify compliance with the 20 criteria will need to include the risk associated with 21 shutdown and low-power operations, fire, and external 22 events."
23 MEMBER KRESS: Okay. But that doesn't capture 24 the --
., ~ .
(,,) 25 MR. DUDLEY: The concern that way.
NEAL R. GROSS COURT REPORTERS AND TRANSCRIBERS 1323 RHODE ISLAND AVE., N W.
(202) 234-4433 WASHINGTON, D C. 20005-3701 (202) 234-4433
587 1 MEMBER KRESS: -- concern that it could be one
,_ 2 or the other. We were distinguishing between them. It
+
\') 3 might be good to have that in there, too, though.
4 She didn't get all of what you said there.
5 VICE CHAIRMAN POWERS: Yes. But I --
6 MEMBER KRESS: You're going to read it.
7 VICE CHAIRMAN POWERS: I absolutely cannot 8 compose looking at that screen.
9 MEMBER KRESS: Well, you compose and then tell 10 him the way. You're like Hal Lewis. j I
11 CHAIRMAN SEALE: That's about the nicest thing i O
4 12 you've said about him lately. !
)
13 MEMBER APOSTOLAKIS: Should there be a --
()
(_)
14 VICE CHAIRMAN POWERS: What did I do? Do I l
15 owe you money or something like that?
16 MEMBER APOSTOLAKIS: I think -- !
i 17 MEMBER KRESS: Hal Lewis used to take umbrage i 18 at you guys messing with his words.
19 MEMBER APOSTOLAKIS: A lot of the stuff under 20 this section we've said before in other reports.
21 Shouldn't there at least --
22 MEMBER SHACK: We've said most of this letter 23 before in our reports.
1 24 MEMBER APOSTOLAKIS: But, I mean --
,r x (j 25 MEMBER KRESS: Do you think I just made this NEAL R. GROSS COURT REPORTERS 'AND TRANSCRIBERS 1323 RHODE ISLAND AVE., N W.
(202) 234-4433 WASHINGTON D.C. 20005-3701 (202) 2344433
- . - - . . . - . - - , . - - - ~..__. - - - . . . . - - - . . . _ ,. . . - . . - . . . - . . - . . -
i 588 -!
1
-1 stuff up out of whole cloth?
2 MEMBER APOSTOLAKIS: "We believe the ancillary l
3 CDF goal should be elevated to the status of a fundamental
, 4 goal." And it's left at-that. "As stated in our report.
l 5 of".such and such.
.6 MEMBER KRESS: Oh, gosh.
7 MEMBER APOSTOLAKIS: I mean, this is not new.
8 MEMBER KRESS: I'know it, but that's just such 9 a-- l 1
10 MEMBER APOSTOLAYIS: We had more reasons' I l
11 stated there, I remember. We cited Rick. We said why we l'2 didn' t want the 10-2 I mean, here is a report a few l 13 months later that says less. What's the idea?
14 Can we get that letter? I mean, what's the i
15 idea of the November-- .oh, you have it? Okay. Well, 16 actually, it was deleted. All of that was deleted in the 17 final version. It was in some draft.
18 MEMBER KRESS: And for a good reason it was 19 deleted.
20 MEMBER APOSTOLAKIS: Why, to be repeated now?
l 21 (Laughter.)
22 MEMBER APOSTOLAKIS: Is that the reason?
23 CHAIRMAN SEALE: Okay. I think you've got 24 something you're typing there. Okay?
i 25 TECHNICAL SECRETARY SUMMERS: It doesn't have NEAL R. GROSS I COURT REPORTERS AND TRANSCRIBERS 1323 RHODE ISLAND AVE., N W.
(202) 234-4433 WASHINGTON, D.C. 20005-3701 (202) 234-4433
4 . . . . . . . _ . . . . ..- . _ _ . . . _ . . _ - . . _ . _ . _ . . . _ . . . _ _ . __.
589 i ,
1 a concern to'--
f 2 CHAIRMAN SEALE: Do'you want to make it a 3?
] 3 VICE CHAIRMAN. POWERS: No. It goes under 4 Point 2.
u
! 5 MEMBER KRESS: It goes.under 2, yes.
l 6 TECHNICAL SECRETARY SUMMERS: Two.
Two.
i 7 MEMBER APOSTOLAKIS: I just don't know about
- i. !
8 the -- I'm not sure.
9 CHAIRMAN SEALE:- Definitive conclusions about a
10 what?
l - 11 MEMBER SHACK: Whether-it's the early or late l
12 --
i
[ 13 VICE CHAIRMAN POWERS: You could say, 1 o
O -14 " Definitive conclusions concerning the limited QHO."
l
[
15 CHAIRMAN SEALE: All right.
i 16 MEMBER APOSTOLAKIS: I'm not sure, though, L 17 that by putting low-power and shutdown operations and~
i 18 external events on the same level you really mean that. I l-19 think we understand external events much better than 20 low-power and shutdown operations, which we haven't done, 21 in fact.
22 VICE CHAIRMAN POWERS: George, I don't think 23 you'll find that out because I think you run into problems 24 that you don't know what the ancillary damage around the 25 plant is that interferes in emergency operations very NEAL R. GROSS COURT REPORTERS AND TRANSCRIBERS 1323 RHODE ISLAND AVE., N W.
(202) 2344433 WASHINGTON, D.C. 20005-3701 _
(202) 2344433
590 1 well.
, 2 MEMBER APOSTOLAKIS: But we were forbidden F
(m)
\ /
3 from including that in our calculations. We've tried to 4 do that.
5 CHAIRMAN SEALE: That's what he says. We 6 don't know it.
7 VICE CHAIRMAN POWERS: 'You have to --
8 MEMBER APOSTOLAKIS: Yes, but --
9 VICE CHAIRMAN POWERS: -- understand that in 10 order to handle the evacuation problem. Evacuation is 11 what controls QHO.
12 MEMBER APOSTOLAKIS: That's correct. But if 13 they tell me I should not consider the impact of the
\~/ 14 earthquake on the surrounding community --
15 VICE CHAIRMAN POVEFS: Who is this, Moses?
16 MEMBER APOSTOLAKIS: The NRC. We did it for 17 Indian Point and Zion. We showed that there will be no 1
18 Chicago when the earthquake that will demolish Zion l 1
1 19 occurs.
20 VICE CHAIRMAN POWERS: Oh, I was thinking no 21 Chicago when the earthquake struck Indian Point.
r,va 22 MEMBER APOSTOLAKIS: It's none of youk 23 business what Chicago does.
l 24 VICE CHAIRMAN POWERS: Well, I mean --
( 't ,
) 25 MEMBER APOSTOLAKIS: I mean, it's another NEAL R. GROSS COURT REPORTERS AND TRANSCRIBERS 1323 RHODE ISLAND AVE., N W.
l (202) 234-4433 WASHINGTON, D.C. 20005-3701 (202) 234-4433 i
591 1 trick of the nukes to show, you know, blah blah blah blah.
73 2 So we said, "Okay. Chicago exists."
t ;
V 3 VICE CHAIRMAN POWERS: George, you're talking 4 about a long, long time ago.
5 CHAIRMAN SEALE: Yes.
6 MEMBER APOSTOLAKIS: Really?
7 CHAIRMAN SEALE: Yes. It was.
8 VICE CHAIRMAN POWERS: Zion and Indian Point.
9 You know, you're almost a decade old.
10 MEMBER APOSTOLAKIS: Do you really want to put 11 low-power and shutdown operations on the same level as 12 external events or fire?
13 VICE CHAIRMAN POWERS: Absolutely. I think 7
)
'V 14 low-power and shutdown --
15 MEMBER MILLER: The least.
16 VICE CHAIRMAN POWERS: The least.
17 MEMBER KRESS: Okay. Can we go back up --
18 MEMBER APOSTOLAKIS: What you're saying is --
19 I mean, this can be a little misleading because people may 20 think we really don't know how to handle fire and how it 21 leads to core damage.
22 And you mean something else. You really mean 23 the impact of these events on the surrounding community on 24 evacuation and --
/>
- \
q) 25 VICE CHAIRMAN POWERS: George, right now I am HEAL R. GROSS COURT REPORTERS AND TRANSCRIBERS 1323 RHODE ISLAND AVE., N W.
(202) 234-4433 WASHINGTON, D C. 20005-3701 (202) 234-4433
. .. . . . . ~ , . - . - . . . - . ~ - _ . - . . _ . . - . . - . - - . - . - - . . . - . - . - . . - -
l 592
[ sympathetic to your point of view, and I encourage you to ,
J
^
2 find some words to refine what's written up there and
..t
(--
3 bring them'to us once you have them. '
I 4 MEMBER KRESS: In the meantime --
{ 5 CHAIRMAN SEALE: Junkyard dog.
j 6 MEMBER KRESS: In the meantime, could we go j 4 l 7 back up to the introductory paragraph in front of this?
- 8- CHAIRMAN SEALE
- All right. Right there, f
j 9 MEMBER KRESS: 'I think we've done some damage f 10 here and we need to think about it. "In our opinion, the-i
! 11 appropriate choice," which is one thing, "for an i
12 acceptable risk criterion," which is one, "the i-l :13 plant-specific application is the safety goal QHOs."
i'(./n 3
14 There's some damage there. We need to think about how to 1
15 fix that.
16- MEMBER APOSTOLAKIS: Simply is the QHOs. You 17 don't need " safety goal."
18 VICE CHAIRMAN POWERS: Why don't you say that, 19 "In our opinion, the QHOs are the appropriate choice for 20 acceptable risk criteria"?
21 MEMBER KRESS: Okay. Make it all plural?
22 VICE CHAIRMAN POWERS: Yes.
23 MEMBER KRESS: Okay.
24 CHAIRMAN SEALE: Okay.
25 MEMBER KRESS: No, no, not yet. We need to NEAL R. GROSS COURT REPORTERS AND TRANSCRIBERS 1323 RHODE ISLAND AVE., N.W. l (202) 234-4433 WASHINGTON, D C. 20005-3701 (202) 234-4433 1
593 1 get rid of the "and," and we need to get the " ion" off of 2 the --
3 MEMBER APOSTOLAKIS: What?
4 MEMBER KRESS: It's plural now, "are the l
5 appropriate choices for acceptable risk criteria."
6 MEMBER Al'OSTOLAKIS : Yes, yes.
7 MEMBER KRESS: And we need a "that."
8 MEMBER APOSTOLAKIS: "For acceptable risk 9 criteria." Yes.
10 MEMBER KRESS: And we need a "that" in front 11 of the "the" on the first line.
12 VICE CHAIRMAN POWERS: No, you don't.
13 MEMBER KRESS: Yes, you do.
V 14 VICE CHAIRMAN POWERS: No, you don't.
15 MEMBER KRESS: Yes, you do.
16 MEMBER APOSTOLAKIS: Yes, that's correct.
17 VICE CHAIRMAN POWERS: No, no.
18 MEMBER KRESS: Yes, you do.
19 TECHNICAL SECRETARY SUMMERS: Well, we could 20 have said --
21 CHAIRMAN SEALE: We've got an editor. We pay 22 to do that kind of crap.
23 TECHNICAL SECRETARY SUMMERS: We say, "In our 24 opinion" --
(J w
25 MEMBER APOSTOLAKIS: " Number one, the safety NEAL R. GROSS COURT REPORTERS AND TRANSCRIBERS 1323 RHODE ISLAND AVE., N.W.
(202) 234-4433 WASHINGTON, D.C. 20005-3701 (202) 234-4433
. . . - . . . - . - . . . . . - . - . . - . _ . ~ . _ - . . . ~ . - . ~ - ~ ~ . . - . . - . . . . . . . . .
- . 594 a
[. 1 goals are the expression," not ".were," -
"are."'
p
{b i.
2 3 criteria.
. MEMBER KRESS:
Oh, you down on 1?
There are no acceptable risk
- .4 MEMBER APOSTOLAKIS
- Down there, yes. ,
5 MEMBER KRESS: Okay. Okay. I like that.
6 Now, 2 is a bit of a disconnect from the lead-in sentence 7 because now we switch to early fatality QHO controls the 8 risk and --
9 MEMBER APOSTOLAKIS: What does it mean it is 10 likely? That's a disconnect. You're right. )
11 MEMBER KRESS: There's some sort of disconnect 12 now.
i 13 MEMBER SHACK: Well, do you need the second 14 sentence if the first one is true?
l l
15 MEMBER KRESS: Right. You don't need the 16 second sentence, but even without the second sentence, 17 there still is a disconnect. We could zap that one.
18 MEMBER SHACK: Zap?
19 MEMBER KRESS: That.
1
- 20. CHAIRMAN SEALE: Zonk.
21 MEMBER KRESS: Now the question is: Why are 22 we saying that first sentence? See, the reason I had it 23 in there before is because in the introductory statement, 24 I said it was the early fatality QHO that would be the 25 right criteria. Now we've said we ought to choose between NEAL R. GROSS COURT REPORTERS AND TRANSCRIBERS 1323 RHODE ISLAND AVE.. N.W.
.(202) 2344433 WASHINGTON, D.C. 20005-3701 (202) 2344433
595 l 1- the QHOs --
2 VICE CHAIRMAN POWERS: Well, right now it's a O 3 point of information on -- we've said we've got a bunch of 4 QHOs. And then we're going to'say, "Here." It's the 5 early fatality one that's going to control under reactor !
6 power operations. We can't decide on the other operating 7 modes.
8 MEMBER KRESS: This no longer is a reason for j 9 choosing the QHO as a risk acceptance criteria.
l
! 10 CHAIRMAN SEALE: Okay. What about taking that j 11 and moving it after the next bullet and.just making it --
i 1
12 MEMBER KRESS: A paragraph?
13 CHAIRMAN SEALE: -- a paragraph?
14 VICE CHAIRMAN POWERS: Then he's going to have 15 to say we have two reasons for this belief.
16 CHAIRMAN SEALE: Okay. Fine.
17 TECHNICAL SECRETARY SUMMERS: After 3?
18 CHAIRMAN SEALE: Yes.
19 TECHNICAL SECRETARY SUMMERS: As another 20 bullet?
21 CHAIRMAN SEALE: No.
22 VICE CHAIRMAN POWERS: No.
23 MEMBER KRESS: Just as a paragraph.
24 CHAIRMAN SEALE: As a paragraph. Now, the 25 whole thing, Dana's sentence, too --
NEAL R. GROSS COURT REPORTERS AND TRANSCRIBERS j 1323 RHODE ISLAND AVE., N.W. y (202) 234-4433 WASHINGTON, D C. 20005 3701 (202) 234-4433 j
?
596 1 MEMBER SHACK: All of 2 goes in.
,m 2 CHAIRMAN SEALE: Yes.
3 MEMBER KRESS: All of that, right.
4 TECHNICAL SECRETARY SUMMERS: Goes where?
5 CHAIRMAN SEALE: Down there where you just 6 moved that other stuff.
7 TECHNICAL SECRETARY SUMMERS: After this?
8 CHAIRMAN SEALE: Yes.
9 MEMBER KRESS: Yes.
10 CHAIRMAN SEALE: Here you go.
11 MEMBER KRESS: And now the 2 gets -- 3 becomes 12 a 2. Right.
13 CHAIRMAN SEALE: There you go.
l )
'd 14 MEMBER KRESS: And now we need to go up and 15 see what we said --
16 CHAIRMAN SEALE: Introductory.
17 MEMBER KRESS: Introductory. We have two 18 reasons.
19 CHAIRMAN SEALE: We have two reasons. Okay.
20 Would that be okay, Tom?
21 MEMBER KRESS: Yes. It suits me.
22 CHAIRMAN SEALE: Fine. Now, lower-tier 23 acceptance criteria.
24 MEMBER APOSTOLAKIS: Why do you separate out f'%
(, 25 fire out of external events? It's part of external NEAL R. GROSS COURT REPORTERS AND TRANSCRIBERS 1323 RHODE ISLAND AVE., N W.
(202) 234 4433 WASHINGTON. D.C. 20005-3701 (202) 234-4433
597 1 events.
,- 2 VICE CHAIRMAN POWERS: I chose to do it, first
)
S# 3 of all, to give me a parallelness to the construction and, 4 second of all, because the Commission is going to get a 5 letter shortly from the staff saying that they need to 6 think about what they're going to do on creating a 7 performance-based fire regulation. And so fire has a 8 singularity now that needs to be expressed explicitly, I 9 think.
10 MEMBER APOSTOLAKIS: Okay. But I still find 11 this sentence a little misleading because most people when 12 they read this, they will say, "Oh, okay. We really don't 13 know how to" --
,S
, . _ . 14 VICE CHAIRMAN POWERS: And I'm very 15 sympathetic to your point of view and encourage you to 16 find ways to augment and improve the sentence.
17 MEMBER KRESS: I second that motion. l l
i 18 MEMBER SHACK: Okay. Since I have some words l l
l 19 ready, can I make a suggestion?
20 CHAIRMAN SEALE: Yes.
21 VICE CHAIRMAN POWERS: We haven't invited him l
22 to do this. l l
23 MEMBER KRESS: Who asked you? l 24 CHAIRMAN SEALE: Would you please give us your
,')
' l
( ,/ 25 words?
NEAL R. GROSS COURT REPORTERS AND TRANSCRIBERS 1323 RHODE ISLAND AVE., N.W.
(202) 234-4433 WASHINGTON, D C. 20005-3701 (202) 234-4433
l 598 1 MEMBER SHACK: In Bullets 1 and 2, if we l
4
,_ 2 scroll back a little bit -- and what I'm suggesting is we 3 take Bullet 1 and we just move it after the first 4 sentence. If you can just scroll back up so we can sort l 1
5 of see what that looks like? I 6 So there's " acceptable for plant-specific 7 application." Then you would say, "The safety goals, 8 then, in addition, we believe should be elevated to this i
9 fundamental." And we can get rid of that "We have two 10 reasons for these beliefs." l l
11 MEMBER KRESS: Okay. It sounds like a good '
12 idea to me --
13 VICE CHAIRMAN POWERS: Great.
!n\
\_/ 14 MEMBER KRESS: -- or a good suggestion, Mr.
15 Shack.
16 MEMBER SHACK: Whole sentence gone. Oh, boy.
17 Two bullets.
18 MEMBER KRESS: Good.
19 VICE CHAIRMAN POWERS: He'll be in the bar 20 tonight, won't he?
21 MEMBER KRESS: Now what do we do?
22 MEMBER SHACK: That " elevating" just now 23 follows right after the -- no, no. Yes. You just take 24 that and put it right after there, yes.
A j 25 MEMBER KRESS: Right.
NEAL R. GROSS COURT REPORTERS AND TRANSCRIBERS 1323 RHODE ISLAND AVE., N.W.
(202) 234-4433 WASHINGTON D C. 20005 3701 (202) 2344 433
599 1 CHAIRMAN SEALE: Very good.
,.. 2 MEMBER KRESS: Now you just leave that right
!) 3 there.
4 MEMBER SHACK: Right, until George comes up 5 with the words.
G MEMBER KRESS: Right. That helps.
7 CHAIRMAN SEALE: Now do we want to try to 8 figure out a way to integrate acknowledgement of the IPE 9 report into this thing earlier? ;
10 MEMBER APOSTOLAKIS: Are we done with the 11 letter?
12 VICE CHAIRMAN POWERS: No. We're anxiously l l
\
13 awaiting your -- l sh\
t l
l C/ 14 MEMBER KRESS: You've got to have some words.
15 VICE CHAIRMAN POWERS: -- defining words. And l
l 16 we have a question on whether we should acknowledge the i 17 IPEs.
18 MEMBER APOSTOLAKIS: Yes. And I also proposed 19 that the last paragraph, " Lower-Tier Acceptance Criteria,"
20 be dropped. We've already said that. We've already said 21 that attached to this letter are two approaches as 22 examples. Why repeat it? Go back to --
23 MEMBER SHACK: But I thought we worked 24 something for the IPEs in already.
/1
) 25 MEMBER APOSTOLAKIC. No. The IPE is a NEAL R. GROSS i COURT REPORTERS AND TRANSCRIBERS 1323 RHODE ISLAND AVE.. N W.
(202) 234-4433 WASHINGTON, D C. 20005-3701 (202) 234 4433
(.
I' 4-600 1 separate. issue.
. 2 MEMBER KRESS: We did, didn't we?
- 3 MEMBER APOSTOLAKIS
- Yes. Well, we didn't say 4 anything regarding'the subsidiary goals.
5 MR. MARKLEY: I would reference those since 6 those are still' draft documents and not in the final form.
7 MEMBER SHACK: No, no. That's a final --
8 MEMBER APOSTOLAKIS: This is final. It's a 9 draft report. You're right.
10 MR. MARKLEY: They're still draft.
11 MEMBER APOSTOLAKIS: Yes, but this is a major j l
12 effort on the part of-the staff. I mean, we're talking !
13 about --
14 MEMBER KRESS: Can we reference that?
15 MR. MARKLEY: That's why I say you can A 16 reference the draft, but it is a draft.
17 MEMBER APOSTOLAKIS: Yes.
18 MR. MARKLEY: Recognize that.
19 MEMBER APOSTOLAKIS: But I raised another 20 issue. I said we have already said in our letter up front 21 that there are two example approaches attached to this 22 letter. There is no reason to have the last paragraph 23 that says --
24 VICE CHAIRMAN POWERS: I'm very sympathetic to 25 George's point of view on this. The only value of this is NEAL R. GROSS COURT REPORTERS AND TRANSCRIBERS 1323 RHODE ISLAND AVE., N.W.
(202) 234-4433 WASHINGTON, D.C. 20005-3701 (202) 234 4433
601 1 to ascribe authorship, which is clearly indicated on the 3 2 attachments.
('~') 3 MEMBER APOSTOLAKIS: Yes, it is.
1 4 CHAIRMAN SEALE: And more than that. I think 5 it diminishes the paunchiness of the conclusion.
6 MEMBER APOSTOLAKIS: Yes. I mean, the authors 7 are cited, I hope, in the attachments themselves.
8 CHAIRMAN SEALE: Yes.
9 MEMBER APOSTOLAKIS: So there is no problem 10 with that.
11 MEMBER KRESS: Show me the part where it says 12 we've got the attachments, though.
13 MEMBER APOSTOLAKIS: Yes. Let's go back.
()
Y2 14 MEMBER KRESS.: Be sure that we really said 15 that.
16 MR. DUDLEY: Second paragraph.
17 MEMBER APOSTOLAKIS: Second paragraph, "The 18 attachments to this report provide examples of approaches 19 that can be used to quantify lower-tier acceptance 20 criteria."
21 MEMBER SHACK: We can even acknowledge the 22 authors here if you'd like.
23 MEMBER KRESS: No, I don't. That's not 24 important. Okay. Is that the last? No. Leave that and
,m
(,) 25 go to the last paragraph, please.
NEAL R. GROSS COURT REPORTERS AND TRANSCRIBERS 1323 RHODE ISLAND AVE., N W.
(202) 234-4433 WASHINGTON, D C. 20005-3701 (202) 234 4433
602 1 MEMBER APOSTOLAKIS: The last paragraph.
2 CHAIRMAN SEALE: Zonk it.
/ .
+ -
3 MEMBER APOSTOLAKIS: " Lower-Tier" all the way 4 to "LERF."
5 MEMBER KRESS: Now we'll have a signature on a 6 separate page.
7 TECHNICAL SECRETARY SUMMERS: No, no.
8 CHAIRMAN SEALE: That will never happen.
9 MEMBER APOSTOLAKIS: I'm learning as fast as I 10 can because you're leaving.
11 VICE CHAIRMAN POWERS: George, you realize you 12 outdid him. He got a sentence, and you got a whole 13 paragraph. Ooh, boy.
/ 14 MEMBER APOSTOLAKIS: If our Chairman had 15 agreed to have a five-minute break an hour ago, I would 16 have been much more cooperative.
17 CHAIRMAN SEALE: Yes. All right.
18 VICE CHAIRMAN POWERS: Let's see. We still 19 have on the board some clarifying words from George.
20 CHAIRMAN SEALE: Yes.
21 MEMBER APOSTOLAKIS: You really want me to 22 come up with words? The real issue, Dana --
23 MEMBER KRESS: Unless Bill Shack were to 24 volunteer. )
,~
j 25 MEMBER APOSTOLAKIS: The real issue is that we NEAL R. GROSS )
COURT REPORTERS AND TRANSCRIBERS j 1323 RHODE ISLAND AVE., N.W.
(202) 234-4433 WASHINGTON, D C. 20005-3701 (202) 234-4433 1
603 1 don't estimate the impact on the evacuation plans; right,
,s 2 of these events? Is that really what you're saying?
l
\
' '/
3 VICE CHAIRMAN POWERS: I'm sure that's one of 4 them. !
5 MEMBER APOSTOLAKIS: So when you say, "Our i 6 understanding of risks," that's what you really mean?
7 VICE CHAIRMAN POWERS: Well, that's what I l 8 think I mean, now. My understanding of the external 9 events is very poor because I've seen very few PRAs in 10 that area. l 11 MEMBER SHACK: What I suggest is that we sort 12 of print this one as it is so we can take it home tonight l 13 and take a look.
- p. ,
/
)
_./ 14 VICE CHAIRMAN POWERS: No. We have to finish 15 deliberations tonight and vote it final with just the 16 allowance for a quality assurance check on Saturday.
17 CHAIRMAN SEALE: That's the dog.
18 VICE CHAIRMAN POWERS: Our deliberations have 19 to be done tonight.
20 CHAIRMAN SEALE: That's the dog.
21 MEMBER APOSTOLAKIS: That's the rule.
22 MEMBER SHACK: Is this going to be true for 23 all letters?
24 VICE CRAIRMAN POWERS: No.
,a
( ,
) 25 MEMBER KRESS: No. Just this one because it NEAL R. GROSS COURT REPORTERS AND TRANSCRIBERS 1323 RHODE ISLAND AVE., N W.
(202) 2344433 WASHINGTON, D C. 20005-3701 (202) 234-4433
604 1 has -- we didn't announce it in the public --
2 VICE CHAIRMAN POWERS: Similarly, George, I
/
1 fm'i/ i 3 don't know that I'm not in a position to say for the 4 external event PRAs what ends up being the controlling 5 risk.
6 MEMBER KRESS: Could we say something like --
7 CHAIRMAN SEALE: Let's say "or accidents where 8 emergency response" --
9 MEMBER APOSTOLAKIS: Yes.
10 CHAIRMAN SEALE: -- " measures are impeded by" 11 --
12 MEMBER SHACK: Just stop right there.
13 CHAIRMAN SEALE: Okay.
O
. 'w./ 14 MEMBER SHACK: " Initiated by external events i 15 where" --
16 VICE CHAIRMAN POWERS: Fair enough.
17 MEMBER APOSTOLAKIS: Say it again.
18 VICE CHAIRMAN POWERS: " Accidents initiated by 19 external events" --
20 CHAIRMAN SEALE: "Where emergency response is 21 impeded."
22 MEMBER APOSTOLAKIS: Oh, okay. Yes. I'm much 23 happier with this.
24 MEMBER SHACK: And we could take out the "and
(~
() 25 fire."
NEAL R. GROSS COURT REPORTERS AND TRANSCRIBERS 1323 RHODE ISLAND AVE., N W.
, (202) 234-4433 WASHINGTON, D C. 20005-3701 (202) 234-4433
605 1 VICE CHAIRMAN POWERS: And then you just
,w 2 delete "and fire."
/ \
e CHAIRMAN SEALE: Well, we need to finish the 4 sentence, though. You lost too much of it.
5 MEMBER SHACK: Undo, undo.
6 VICE CHAIRMAN POWERS: I am going to follow 7 that --
8 MEMBER APOSTOLAKIS: It's not sufficient --
9 VICE CHAIRMAN POWERS: -- Americans With 10 Disabilities action here for color-blindness or something.
11 MEMBER SHACK: It's the "and fire," though, is 12 what we want to get rid of?
13 CHAIRMAN SEALE: Yes, yes.
t Y ') 14 MEMBFR SHACK: "Is impeded " and just kill the 15 "and fire."
16 TECHNICAL SECRETARY SUMMERS: Not all of this, l 17 though.
18 MEMBER SHACK: No, no, no, no, no. Undo, 19 undo. Okay. Now just the "and."
20 CHAIRMAN SEALE: Yes. Okay.
21 MEMBER KRESS: Vote final. !
i 22 VICE CHAIRMAN POWERS: I move final. j I
23 MEMBER SHACK: Second. !
i 24 TECHNICAL SECRETARY SUMMERS: What about IPE?
g i
() 25 Did we --
NEAL R. GROSS COURT REPORTERS AND TRANSCRIBERS 1323 RHODE ISLAND AVE., N.W.
(202) 234-4433 WASHINGTON, D.C. 20005-3701 (202) 234-4433
606 1 MEMBER APOSTOLAKIS: Well, the way it is
,s 2 written now, it will be very hard unless you do major i 1
\-)
3 surgery. My point was that there are two figures here 4 that show clearly that the subsidiary goals are violated 5 by a large number of plants.
6 CHAIRMAN SEALE: Well, even though we don't 7 specifically refer to them, it might be appropriate to 8 include that in a list of references because they 9 certainly make the point.
10 MEMBER APOSTOLAKIS: Sure.
11 MEMBER SHACK: Well, let's just go back to 12 that sentence where we do the --
, 13 MEMBER APOSTOLAKIS: Wa keep talking about T
\- 14 risk. The whole body of the letter refers to risk.
15 CHAIRMAN SEALE: Yes, yes.
16 MEMBER APOSTOLAKIS: And these are referring 1
17 to the ancillary, although maybe in the last paragraph, l l
18 where we talk about the CDF being elevated to a 19 fundamental goal, maybe there we can add a sentence that 20 says, " Guys, it's violated." l 21 VICE CHAIRiIAN POWERS: Well, it seems to me --
22 MEMBER APOSTOLAKIS: I don't know what purpose 23 that would serve.
l l
24 CHAIRMAN SEALE: Why don't we just include r
[~h l
(_) 25 that in a list of references and be done with it?
NEAL R. GROSS COURT REPORTERS AND TRANSCRIBERS 1323 RHODE ISLAND AVE., N.W.
(202) 234-4433 WASHINGTON, D.C. 20005 3701 (202) 234-4433
607 1 MEMBER APOSTOLAKIS: Okay.
. 2 CHAIRMAN SEALE: Can we do that, Noel?
( )
x '/
~
3 MR. DUDLEY: I'll put it.
4 CHAIRMAN SEALE: All right.
5 VICE CHAIRMAN POWERS: Call for discussion?
6 CHAIRMAN SEALE: Yes. Any discussion?
7 VICE CHAIRMAN POWERS: And this discussion has 8 to recognize that we are --
9 MEMBER APOSTOLAKIS: Well, it seems to me --
10 VICE CHAIRMAN POWERS: -- we have to conclude 11 our deliberations today --
12 CHAIRMAN SEALE: Yes.
13 VICE CHAIRMAN POWERS: -- but that tomorrow
,e y t i V 14 you're going to be allowed only for a QA check.
15 MEMBER APOSTOLAKIS: It seems to me that 16 before we vote though, we should see a hard copy of the 17 revised letter. I mean, it's unrecognizable now.
18 MEMB2R FONTANA: Yes. We can move final.
19 MEMBER APOSTOLAKIS: We can take a break for l
20 five minutes and --
21 VICE CHAIRMAN POWERS: Mr. Chairman, that's a 22 point of order. So it does take precedence over the vote.
23 CHAIRMAN SEALE: Yes. Okay. Let's make a 24 copy for everybody and give everybody a stretch.
./ M l
(-) 25 MEMBER APOSTOLAKIS: Okay. l l
NEAL R. GROSS COURT REPORTERS AND TRANSCFilBERS 1323 RHODE ISLAND AVE., N.W.
(202) 234-4433 WASHINGTON, D C. 20005-3701 (202) 234-4433
i 608 l
l 1 CHAIRMAN SEALE: Be back at 7:00 after.
i 2 (Whereupon, the foregoing matter went off the 7
,( )
l '# 3 record at 5:53 p.m. and went back on the 4 record at 6:02 p.m.)
5 CHAIRMAN SEALE: Okay, gentlemen. We now have 6 a copy.
7 MEMBER KRESS: Is it buff?
8 CHAIRMAN SEALE: It's Draft 4.
9 VICE CHAIRMAN POWERS: Buff.
10 CHAIRMAN SEALE: Buff.
11 VICE CHAIRMAN POWERS: My Draft 4 still has my 12 markings on it.
13 Tom, on Line 43, you have "A quantitative risk
,s i \
_.) 14 acceptance criterion could be used for enforcement 15 decisions." I think it's only for some enforcement 16 decisions simply because we're taking out the 17 unquantifiable risks.
18 MEMBER KRESS: Yes. I think "some" should go 19 in there.
20 VICE CHAIRMAN POWERS: Or how about "for many 21 enforcement"?
22 MEMBER KRESS: Okay. I like that better.
23 VICE CHAIRMAN POWERS: And it is quite 24 attractive because it would be objective. Could we say, fm t
) 25 "An objective criterion" --
NEAL R. GROSS COURT REPORTERS AND TRANSCRIBERS 1323 RHODE ISLAND AVE., N W.
(202) 234-4433 WASHINGTON, D.C. 20005 3701 (202) 234 4433
l 609 1 MEMBER KRESS: Yes. Where is that?
1
-s 2 VICE CHAIRMAN POWERS: -
"for many" --
'") 3 MEMBER KRESS: " Objective risk acceptance 4 criteria"?
5 VICE CHAIRMAN POWERS: Yes. "An objective."
6 MEMBER KRESS: 'An objective."
7 VICE CHAIRMAN PCWERS: We didn't get rid of 8 that. Line 42, Tom, we have "that basically" is back in l l
l I
9 there.
1 l
10 MEMBER KRESS: Did it end up in there again? l 11 MR. DUDLEY: It was in there twice.
l 12 MEMBER KRESS: I see. Where is it? l 13 CHAIRMAN SEALE: Two lines up on the cursor.
'J 14 VICE CHAIRMAN POWERS: Could you just say, j l
15 "The current, unquantified adequate protection"? l I
16 MEMBER KRESS: Yes. That. would be a good way l 17 to do it.
18 VICE CHAIRMAN POWERS: " Current" just with a 19 comma following it.
20 MEMBER KRESS: You need another comma.
21 CHAIRMAN SEALE: Y e.s . Yes.
22 MEMBER KRESS: Boy, that "As the recent IPE 23 study indicates" really grates on the nerves, doesn't it, 24 when you read it?
( / 25 VICE CHAIRMAN POWERS: I'm not sure where you NEAL R. GROSS COURT REPORTERS AND TRANSCRIBERS 1323 RHODE ISLAND AVE., N.W.
(202) 234-4433 WASHINGTON. D.C. 20005-3701 (202) 234-4433
610 1 are, Tom.
j , -- .s 2 MEMBER KRESS: Line 71.
)
3 MEMBER FONTANA: Seventy-one.
4 MEMBER APOSTOLAKIS: Maybe -- are we making 5 comments now or just reading?
6 MEMBER FONTANA: Let's read it over.
7 MEMBER APOSTOLAKIS: Okay.
8 VICE CHAIRMAN POWERS: Well, for one thing, 9 the IPE study should be spelled out, " Individual plant 10 examinations indicate."
11 MEMBER KRESS: That's going to make it grate 12 even more.
13 VICE CHAIRMAN POWERS: That's going to grate 3
,V 14 even more?
15 CHAIRMAN SEALE: What if you put a period 16 after " operating plants" and say, "This conclusion is 17 supported" --
18 MEMBER KRESS: That would help.
19 CHAIRMAN SEALE: -- "by the recent IPE" --
20 MEMBER SHACK: In Line 67, maybe we can make 21 that "Since there are many ways." And we'11 at least have 22 one less "as" around.
23 MEMBER KRESS: Okay. That's a good idea. It 24 still grates on the nerves, though.
( "As shown by the recent IPEs,"
g/ 25 MEMBER SHACK:
NEAL R. GROSS COURT REPORTERS AND TRANSCRIBERS 1323 RHODE ISLAND AVE., N.W.
(202) 234-4433 WASHINGTON, D.C. 20005-3701 (202) 234-4433
611 1 rather than " indicate."
n 2 VICE CHAIRMAN POWERS: I have --
(
3 MEMBER KRESS: No. But you --
4 VICE CHAIRMAN POWERS: I have an " Operating 5 Plants. This conclusion is supported by the results of 6 the recent IPE studies."
7 MEMBER KRESS: That's better.
8 MEMBER SHACK: Could we take the " unnecessary" 9 out of the " unnecessary part of this margin"?
10 MEMBER KRESS: No. We went through that 11 argument once before. Dana wanted to take it out, and so 12 did somebody else.
13 MEMBER SHACK: That's at least three votes.
n t 1
'k M
~
14 MEMBER KRESS: I argued vehemently to leave it 15 in, and I think I beat everybody to death on it. And ,
16 everybody told me to take it out, and I put it back in 17 anyway, I think.
18 MEMBER FONTANA: Is this it?
19 CHAIRMAN SEALE: Where is this?
20 MEMBER FONTANA: It starts at Line 1?
21 VICE CHAIRMAN POWERS: No, no. The 22 " unnecessary portion"? You want to leave that in?
23 MEMBER KRESS: Yes.
24 MEMBER FONTANA: Does this stop at Line 111?
A
- b
(,,/ 25 MEMBER APOSTOLAKIS: Yes.
NEAL R. GROSS COURT REPORTERS AND TRANSCRIBERS 1323 RHODE ISLAND AVE., N.W.
(202) 234-4433 WASHINGTON, D.C. 20005-3701 (202) 234-4433
612 1 MEMBER FONTANA: It kind of leaves you hanging
,s,_ 2 off the end of the cliff, doesn't it?
! i
'~'
3 MEMBER KRESS: Yes. That's why I had that 4 other paragraph on there.
5 MEMBER APOSTOLAKIS: Can I add something on 1 4
6 Line 78? The way it reads is "Within the spectrum of I t
7 risk, it is likely that there are plants with risk levels 8 above the safety goals and other plants with risk with j 9 levels below. (We note that the IPE studies have shown )
l 10 explicitly that the subsidiary goals of CDF and CCFP are 11 violated by a number of plants . ) "
12 MEMBER KRESS: Then you run the if this is 13 indeed the case.
'\ l
(~~/
's-- 14 MEMBER APOSTOLAKIS: It is. Excuse me. What 35 was the --
l 16 MEMBER KRESS: You run the if this is indeed 17 the case because you've interposed a whole line in between 18 and what the case was we were talking about.
19 MEMBER APOSTOLAKIS: Yes because this refers 20 to the safety goals themselves.
21 MEMBER KRESS: No. This refers to some lower 22 and some above. That's the case that I'm indeed referring 23 to.
24 MEMBER APOSTOLAKIS: No. It refers to the im (s_.-) 25 risk levels. And there is question abeat that.
NEAL R. GROSS COURT REPORTERS AND TRANSCRIBEAS 1323 RHODE ISLAND AVE., N.W.
(202) 234-4433 WASHINGTON, D.C. 20005-3701 (202) 234-4433
613 1 MEMBER KRESS: It refers to the sentence just
,._, 2 prior to it.
k' ~ '
)
3 MEMBER APOSTOLAKIS: Yes.
1 4 MEMBER KRESS: And if you interpose another j 5 sentence in between, then you've lost the connection.
i 6 MEMBER MILLER: Even with the parentheses. I 7 VICE CHAIRMAN POWERS: Even with the 8 parentheses, he's still going to destroy the connection to 9 -he employed pronoun.
10 MEMBER MILLER: For what purpose are you going 11 to put that in there anyhow?
12 MEMBER APOSTOLAKIS: Because that's a major 13 result of the study.
)
(_/ 14 MEMBER MILLER: But you're not reporting on 15 that study. You're reporting -- that's not the purpose of 16 the --
17 CHAIRMAN SEALE: George, we had tested that in 18 the previous paragraph.
19 MEMBER APOSTOLAKIS: Yes.
20 CHAIRMAN SEALE: Instead of saying, "As the 21 recent IPE studies indicate," that we, rather, put a 22 period after " Operating plants" --
23 MEMBER APOSTOLAKIS: Yes.
24 CHAIRMAN SEALE: -- and then say, "This r-(xv) 25 conclusion is supported by the results of the recent IPE NEAL R. GROSS COURT REPORTERS AND TRANSCRIBERS 1323 RHODE ISLAND AVE., N.W.
(202) 234-4433 WASHINGTON, D C. 20005-3701 (202) 234-4433
614 1 studies."
,, 2 MEMBER APOSTOLAKIS: Can we continue that and l )
3 say, "Where it is explicitly shown that the subsidiary 4 goals in CDF and CCFP are, in fact, violated by a number 5 of plants"?
6 MEMBER FONTANA: The reason I don't like that 7 --
8 MEMBER APOSTOLAKIS: That's really what they l
9 do.
10 MEMBER FONTANA: Yes. The reason I don't like l
11 this is CDF in itself is not a risk. '
12 MEMBER APOSTOLAKIS: No. That's why we have 13 difficulty putting it someplace.
m
/ \ l
() 14 MEMBER FONTANA: Yes. By itself it is not a 3 15 ' risk. Conditional -- :
l 16 MEMBER BARTON: Containment.
17 MEMBER FONTANA: -- containment failure 18 probability is hokey.
19 MEMBER KRESS: Not very hokey. Real.
20 MEMBER FONTANA: How do you get it?
21 MEMBER KRESS: Calculate it with the --
22 MEMBER FONTANA: How?
23 MEMBER KRESS: Use the models of thermal 24 hydraulics in severe accident and calculate it.
,O
,) 25 MEMBER FONTANA: You've got to do it for every NEAL R. GROSS COURT REPORTERS AND TRANSCRIBERS 1323 RHODE ISLAND AVE., N.W.
(202) 234-4433 WASHINGTON, D C. 20005-3701 (202) 234-4433
l 615 l l
l 1 plant. I
,y 2 MEMBER KRESS: Yes.
(\ ')
l l
3 VICE CHAIRMAN POWERS: You've got a secondary I
4 notion in the work here.
J 5 MEMBER FONTANA: If you're going to -- l l
6 CHAIRMAN SEALE: Gentlemen.
7 MEMBER KRESS: Yes, you do it for every plant.
8 CHAIRMAN SEALE: We've been here before. l 9 MEMBER FONTANA: Well, I'm going to --
10 VICE CHAIRMAN POWERS: George, make a motion.
11 MEMBER APOSTOLAKIS: Look at Line 20, Mario.
12 Line 20 says that if you satisfy the lower-tier acceptance 13 criteria, you ensure that the early fatality QHOs are met.
/
!\ -) 14 MEMBER FONTANA: I agree with that.
15 MEMBER APOSTOLAKIS: And here's a result that 16 says you don't.
17 MEMBER FONTANA: Okay. And what I'm saying is 18 that you shouldn't tar and feather a plant because they're 19 not meeting these lower-level criteria. If a plant 20 doesn't meet the screen, then you do a little true PRA.
21 And I think you're going to find when you do 22 that that in these quantifiable things, I don't think 23 you're going to find many that are going to go above 24 safety goal as defined by the QHO.
()
(, 25 MEMBER APOSTOLAKIS: So this is not a new NEAL R. GROSS COURT REPORTERS AND TRANSCRIBERS 1323 RHODE ISLAND AVE., N.W.
(202) 23+d 133 WASHINGTON, D C. 20005-3701 (202) 234-4433
616 1 thought that we have never really addressed, namely that
, ~3 2 if you do not meet the subsidiary criteria, you may still Y 3 be okay.
4 MEMBER FONTANA: That's right.
5 MEMBER APOSTOLAKIS: We have never said that.
6 MEMBER FONTANA: I just said it.
7 VICE CHAIRMAN POWERS: I think we are at the 8 point that, George, you need to formulate your suggestion 9 as a motion.
10 MEMBER APOSTOLAKIS: Okay.
11 VICE CHAIRMAN POWERS: Just formulate a motion 12 that says that we augment the sentence to include a 13 specific reference to the violation of the CDF and the
.e3
( /
n/ 14 CCFP --
15 MEMBER FONTANA: From the IPE.
16 VICE CHAIRMAN POWERS: -- from the IPE 17 summary. And we'll worry about the words later.
18 CHAIRMAN SEALE: This is down on Line 71.
19 MEMBER APOSTOLAKIS: Yes.
20 MEMBER KRESS: Actually, it's not. He wants 21 to put it on Line 78.
22 CHAIRMAN SEALE: No. I think we talked about 23 that.
24 VICE CHAIRMAN POWERS: I think we talked him
,,m
/ \
() 25 into moving it up to 71.
NEAL R. GROSS COURT REPORTERS AND TRANSCRIBERS 1323 RHODE ISLAND AVE., N W.
(202) 234-4433 WASHINGTON, D C. 20005-3701 (202) 234-4433
617 1 TECHNICAL SECRETARY SUMMERS: So what do I f3 2 put?
(' )
~~
3 MEMBER SHACK: Can we go back to Line 82 and 4 --
5 CHAIRMAN SEALE: Let's take care of this 6 first.
7 VICE CHAIRMAN POWERS: We've got this first.
8 George is ready to formulate --
9 MEMBER APOSTOLAKIS: "This conclusion is 10 supported by the results of the recent independent plant 11 examinations, in which it is" --
12 VICE CHAIRMAN POWERS: Again, George, 13 formulate your motion as an augmentation.
p 14 MEMBER APOSTOLAKIS: Yes. I move --
15 VICE CHAIRMAN POWERS: And we'11 worry about 16 the words later.
l 17 MEMBER APOSTOLAKIS: Yes. I move that this 18 sentence be adopted, "in which it is shown that a number 19 of plants" -- and the words can change -- "a number of I
20 plants violate the subsidiary goals on CDF and CCFP." And j 21 we can wordsmith it for later, but the thought belongs ,
I.
22 there.
23 VICE CHAIRMAN POWER'.+ I don't want to usurp.
24 You need to call for a second.
/3 (a) 25 CHAIRMAN SEALE: Well, do we have a second of NEAL R. GROSS i I
COURT REPORTERS AND TRANSCRIBERS 1323 RHODE ISLAND AVE., N.W.
(202) 234-4433 WASHINGTON, D C. 20005-3701 (202) 234-4433 !
._ . - , . . . _ ~ . . _ _ _ _ . . . _ . _ . _ _ _ _ _ _ . _ . _ _ . _ . . . . _ . _ _ _ _ _ . _ _ - _ . . - _ . . _ . _ .
618 1
.1 that?
I 2 MEMBER APOSTOLAKIS: Probably not.
3- CHAIRMAN SEALE: Do we have a second for the
- 4- part that is in red? j 1
- 5. MEMBER SHACK: I just don't-think we need it.
6 VICE CHAIRMAN POWERS: We don't really need 1
- '7 it. .
1 8 ~ CHAIRMAN SEALE: Okay. Apparently it's going i 9 to die of want of a second.
I- 10 MEFBER APOSTOLAKIS: Fine.
t t
11 C1LIRMAN SEALE: Okay. Kill'it. Then I think
', 12 IPEs ought rc be -- or the Individual Plant Examination t
i i 13 ought to be capitalized. It's individual, isn't it?
1
- l. ( 14 VICE CHAIRMAN POWERS: Yes.
I
'15 MEMBER SHACK: How about, "This conclusion is 16 consistent with the results of the recent IPEs,"?
17 MEMBER KRESS: I would prefer " consistent.
18 CHAIRMAN SEALE: "With"? Fine. All right.
19 Bill, you had a --
20 MEMBER SHACK: Yes. I don't like " unnecessary 4 21 portion." I mean, that --
i 22 MEMBER APOSTOLAKIS: Line?
23 MEMBER SHACK: Eighty-two.
24 MEMBER KRESS: Well, if you 1 + the i
f.
4 25 " unnecessary" in, that makes it all-inclusive in the sense i NEAL R. GROSS COURT REPORTERS AND TRANSCRIBERS 1
1323 RHODE ISLAND AVE., N.W.
i (202) 234-4433 WASHINGTON, D.C. 20005 4701 (202) 234-4433 4
619 1 that you can use any part of that margin as margin that 7s 2 can be bartered with by the process. If you don't put
- 1
\
~
3 " unnecessary" in, that implies that you think the whole 4 margin can be dealt with. So it is a qualifying word when 5 you put it in there.
6 MEMBER SHACK: See, I don't read it that way.
7 I read it as though anything that's left is unnecessary.
8 MEMBER KRESS: No. What the staff says is 9 that when they've got margin between their acceptance 10 criteria in the actual risk status of the plant, that 11 they're going to grant some relief, but they're not going 12 to let you go all the way up. They're going to only let 13 you use part of that margin.
k- 14 And I'm saying that that's the part that's i i
15 allowable and the rest of it is -- I mean, that's the ;
l 16 unnecessary portion of the margin.
17 MEMBER APOSTOLAKIS: But that's not quite what i
18 they said, though. They're not going to let you go to the 19 goal in one shot.
20 VICE CHAIRMAN POWERS: Tom, could I suggest 21 that you change it to say "The portion that is 22 unnecessary" --
23 CHAIRMAN SEALE: Look, you've already 24 characterized that as a difference. Okay? In the rm e h
(_/ 25 previous sentence, you say, "For these plants with risks NEAL R. GROSS COURT REPORTERS AND TRANSCRIBERS 1323 RHODE ISLAND AVE., N W.
(202) 234-4433 WASHINGTCN, D.C. 20005-3701 (202) 234-4433
620 l 1 below the safety goals, the difference in risk can be l
m\ 2 viewed as margin."
l \" ) \
1 l 3 MEMBER BARTON: Right.
l 4 CHAIRMAN SEALE: Just say, "This margin 5 provides" -- "This margin is the region from which l 6 plant-specific regulatory relief can be granted."
l l
7 MEMBER APOSTOLAKIS: I agree with Bill. "It 8 is from this margin that plant-specific regulatory relief 1
9 can be granted." l l
10 MEMBER SHACK: I was going to change it to how l l
11 about -- because I agree with Tom. I don't want to give l l
l 12 them the whole God damned thing. l 13 MEMBER KRESS: That's right.
.Y/ 14 MEMBER SHACK: I would say, "From some portion 15 of this margin, regulatory relief can be granted."
16 MEMBER KRESS: Okay.
17 CHAIRMAN SEALE: Okay. Fine.
18 MEMBER KRESS: I would take that.
19 CHAIRMAN SEALE: Get rid of the " unnecessary."
20 MEMBER FONTANA: On Line 76, do you want to 21 say "likely" or do you want to say "possible"?
22 MEMBER KRESS: "Likely." "Likely" is what I 23 really meant.
24 MEMBER APOSTOLAKIS: It is likely.
(\
l 25 MEMBER FONTANA: You do mean "likely."
NEAL R. GROSS COURT REPORTERS AND TRANSCRIBERS 1323 RHODE ISLAND AVE., N.W.
(202) 234-4433 WASHINGTON, D C. 20005-3701 (202) 234 4 433
621 1 CHAIRMAN SEALE: Okay. Okay.
~- 2 VICE CHAIRMAN POWERS: Call the question.
Nj 3 MEMBER FONTANA: The safety goals that we're 4 talking about on Line 77, for example, are those QHOs --
5 MEMBER APOSTOLAKIS: Yes.
6 MEMBER FONTANA: -- or are these the other 7 ones that we did to take the place of the QHOs?
8 MEMBER KRESS: Which line are you on?
9 MEMBER FONTANA: Seventy-seven.
10 MEMBER KRESS: Seventy-seven. What's your 11 question again?
12 MEMBER FONTANA: Where it says that, "There 13 are plants with risk levels above the safety goal." So
-' 14 are you using the actual QHOs or do you use these things 15 that take the place of them?
16 MEMBER KRESS: We're using the safety goals.
17 MEMBER FONTANA: Themselves?
18 MEMBER KRESS: Yes. That's what it says.
19 MEMBER FONTANA: I think --
20 CHAIRMAN SEALE: QHOs, CDFs, LERFs.
21 MEMBER KRESS: Whatever you want to put in --
22 MEMBER FONTANA: What I'm talking about --
23 CHAIRMAN SEALE: He's asking about QHOs at the 24 exclusion of everything else.
f)
(_) 25 MEMBER FONTANA: That's what I meant.
NEAL R. GROSS COURT REPORTERS AND TRANSCRIBERS 1323 RHODE ISLAND AVE., N W.
(202) 2344433 WASHINGTON, D C. 20005-3701 (202) 234-4433
622 1 CHAIRMAN SEALE: And the answer is no. He
- c. 2 means everything else.
i }
3 MEMBER KRESS: No. I mean the safety goals.
4 VICE CHAIRMAN POWERS: This is substantive.
5 CHAIRMAN SEALE: That's what I mean.
6 MEMBER KRESS: There are two safety goals.
7 There are qualitative and quantitative. Those are the 8 ones I mean, the two safety goals. One of them is on l 9 early fatalities, and one of them is on latent and --
10 VICE CHAIRMAN POWERS: We need to move the 11 question so we can --
12 CHAIRMAN SEALE: Yes, yes.
13 VICE CHAIRMAN POWERS: Call the question.
A l f i V 14 CHAIRMAN SEALE: Okay.
15 MEMBER FONTANA: What was the question? )
16 CHAIRMAN SEALE: Final.
17 VICE CHAIRMAN POWERS: Final.
18 CHAIRMAN SEALE: Do I have a second?
19 VICE CHAIRMAN POWERS: You have a second on j 20 the voting final. 1 i
21 CHAIRMAN SEALE: I do? f i
22 VICE CHAIRMAN POWERS: Yes.
23 CHAIRMAN SEALE: I didn't know that. Okay.
24 MEMBER APOSTOLAKIS: Oh , you moved --
.A l 25 MEMBER KRESS: You can't call the question i.v) i NEAL R. GROSS COURT REPORTERS AND TRANSCRIBERS 1323 RHODE ISLAND AVE., N.W.
(202) 234-4433 WASHINGTON, D.C. 20005-3701 (202) 234-4433
623 1 unless you have a second.
t.s8 2 CHAIRMAN SEALE: I'm sorry. Okay. All right.
( )
3 Those in favor of final?
4 (Whereupon, there was a show of hands.)
5 CHAIRMAN SEALE: Those again.
6 (Whereupon, there was a show of hands.)
7 CHAIRMAN SEALE: Final.
8 VICE CHAIRMAN POWERS: Okay. From this point 9 on, our deliberations are confined to quality checks and 10 minor editorial adjustments.
11 CHAIRMAN SEALE: On this letter.
12 VICE CHAIRMAN POWERS: On this letter. l 13 MEMBER FONTANA: Now, I have a different c
A ,
14 opsonin. I base , couple of different opinions.
l 15 MEMBER APCS'10f_Ai;IS : Yes. You have --
16 VICE CHAIRMAN POWERS: Go ahead, yes.
17 (Whereupon, the foregoing matter was concluded 18 at. 6:22 p.m.)
19 20 27 22 23 24
(%_,- ) 25 NEAL R. GROSS COURT REPORTEGG AND TRANSCRIBERS 1323 ROC E N AND AVE., N.W.
(202) 234-4433 WASHINrJTON, J C. 20006-3701 (202) 234-4433
_ - . _ m .. . . . . . _ . _ _ _ _ . . _ . - _ _ .
i l
I O c'E R T I'F.I.C A T E This is to certify.that the attached proceedings before'the United States Nuclear-Regulatory Commission in the matter of: l Name of Proceeding: 439" ACRS Docket Number: N/A Place of Proceeding: ROCKVILLE, MARYLAND were held as herein appears, and that this is the original transcript thereof for the file of the United States Nuclear Regulatory Commission taken'by me and, thereafter reduced to
'O 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.
FZ ~
41A i
" CORBETT RINER Official Reporter Neal R. Gross and Co., Inc.
5 O
NEAL R. GROSS COURT REPORTERS AND TRANSCRIBERS 1323 RIIODEISLAND AVENUE,NW (202)234-4433 WAS!!!NOTON, D.C. 20005 (202)234-4433
.a . m _ . m 2 m. Arm:n or rHE Aas csA;axAs 439TH ACRS MEETING, MARCH 6-8, 1997 L ,,/
THE MEETING WILL NOW COME TO ORDER. THIS IS THE SECOND DAY OF THE 439TH MEETING OF THE ADVISORY COMMITTEE ON REACTOR SAFEGUARDS.
DURING TODAY'S MEETING, THE COMMITTEE WILL CONSIDER THE FOLLOWING:
(1)
INDEPENDENT SAFETY ASSESSMENT OF THE MAIN YANKEE ATOMIC
, POWER STATION (2)
NATIONAL ACADEMY OF SCIENCES / NATIONAL RESEARCH COUNCIL PHASE 2 STUDY REPORT (3)
DEPARTMENT OF ENERGY PROPOSAL FOR TRITIUM PRODUCTION (4) FUTURE ACRS ACTIVITIES (5)
RECONCILIATION OF ACRS COMMENTS AND RECOMMENDATIONS (6) PROPOSED ACRS REPORTS THIS MEETING IS BEING CONDUCTED IN ACCORDANCE WITH THE PROVISIONS OF THE FEDERAL ADVISORY COMMITTEE ACT.
,~.
\
_/ MR.
SAM DURAISWAMY IS THE DESIGNATED FEDERAL OFFICIAL FOR THE INITIAL PORTION OF THE METITING.
WE HAVE RECEIVED NO WRITTEN COMMENTS OR REQUESTS FOR TIME TO MAKE ORAL STATEMENTS FROM MEMBERS OF THE PUBLIC REGARDING TODAY'S SESSIONS.
A TRANSCRIPT OF PORTIONS OF THE MEETING IS BEING KEPT, AND IT IS REQUESTED THAT THE SPEAKERS USE ONE OF THE MICROPHONES, IDENTIFY THEMSELVES AND SPEAK WITH SUFFICIENT CLARITY AND VOLUME SO THAT THEY CAN BE READILY HEARD.
g i )
w/'
_. ..._ ..._ _ _ .... _ _.._ __ _ ...- _ _....._ _ __.. _ ._ _.._ _ _ _.-_._.._ _ .. _ . _ _. _ ._. _ ._ _ _. 7 . , ,
O O O ,
gaREGy c
[ & i k,
r xA3 L 8
g
- j c ;
o[
%{++$ ***
THE INDEPENDENT SAFETY ASSESSMENT OF MAINE YANKEE ATOMIC POWER STATION Edward L. Jordan Ellis W. Merschoff Jack E. Rosenthal ACRS Briefing March 7,1997
_..._______ _ _ _--__ __ _ _________._____ _ _.__._ _ __7_ .
O O O L PRESENTATION OBJECTIVES
- Describe the Independent Safety Assessment process used to evaluate Maine Yankee
- Discuss the findings and conclusions of the Maine Yankee Independent Safety Assessment
- Discuss Regulatory Lessons Learned i
2
O O O SELECTION OF MAINE YANKEE
- Allegations Regarding RELAP/5YA e Office of Inspector General Inquiry i
i
- State of Maine Concerns l.
e 3
NRC ACTIONS
^
REGARDING MAINE YANKEE ISSUES 1
e December 1995 Allegation from UCS e January 3,1996, Order i
Restricts Power to 2440 MWt (90% of license limit) i Provides basis for operation at 2440 MWt until reanalyses i
performed for operation at 2700 MWt 1 i
Concludes operation as permitted under the Order poses no unduei risk to public health and safety
- Licensee required to perform revised ECCS and containment ;
pressure analyses for operation at 2700 MWt i'
e Staff initiated lessons learned efforty ^ !j
O O O NRC ACTIONS .
REGARDING MAINE ' YANKEE ISSUES (Continued) -
e Independent Safety Assessment inspection -
- Chairman initiative in cooperation with State of Maine '
Immediate findings resolved for August 1996 plant restart Root causes and other long-term findings e On-Going Staff Actions i Plant is shutdown - Restart issues include
- Cable separation l
Generic Letter 96-01 on Circuit Testing
- Offsite power (Suroweic 115Kv Line & TSs)
- Letter dated January 24,1997
- Long term followup of licensee actions-
- Power Uprate review (NPSH, CCW, SBLOCA) j 34
O O O INDEPENDENT SAFETY ASSESSMENT l
- Large Experienced Team
- Independent of NRR and Region l
- Participation by State of Maine
- Technical Team
- Process Team
- Citizens Group
- Modified Diagnostic Evaluation Technique t
4 !
INDEPENDENT SAFETY ASSESSMENT MISSION e Provide an independent assessment of the conformance of Maine Yankee Atomic Power Station to its design and licensing bases including appropriate reviews at the site and corporate offices.
e Provide an independent assessment of operational safety performance providing risk perspectives, where appropriate.
e Evaluate the effectiveness of licensee self-assessments, corrective actions, and improvement plans.
- e Determine the root cause(s) of safety significant
- findings and draw conclusions on overall i
performance.
l 5
o o o
~
MAINE YANKEE INDEPENDENT SAFETY ASSESSMENT TEAM Edward L Jordan Team Manager AEOD Ellis W. Merschoff Team Leader Region 11 Uldis Vanags .
Patrick Dostle David Decrow A n As tant State Representatives Region it i i I I l l Kriss M. Kennedy Ronald Lloyd Thomas O. Martin Alan L Madison Jack E. Rosenthal OPS / Training Maint> Testing Eng. Design / Tech. Management & Orgarnation Analytic Code Support Region IV AEOD RES AEOD AEOD Larry Bell Russell Bywater John Boardman Harold Christensen G. Norman Lauben TTC Region ill AEOD Region 11 RES John Kauffman Peter Prescott George Hausman Brian Haagensen Leonard Ward AEOD AEOD Region III Contractor Contractor Robert Christie Cyril Crane Contractor Contractor George Cha Contractor g
t Michael Shlyamberg Contractor l i
f 6
e TECHNICAL TEAM - DAY-TO-DAY PARTICIPATION IN EACH OF THE FIVE FUNCTIONAL AREAS BEING ASSESSED.
e ULDIS VANAGS ~
. e PATRICK DOSTIE ~
e DAVID DECROW '
i e PROCESS TEAM - OBSERVE THE PROCESS AT IGY MILESTONES TO ASSURE ISAT IS FAIR, BALANCED, AND OBJECTIVE.
e PETER WILEY e DR. FORREST REMICK e CITIZEN'S GROUP - PERIODIC BRIEFINGS FOR GOVERNOR AND i
CITIZEN'S GROUP TO KEEP THEM INFORMED OF PROGRESS.
e DR. DON ZILLMAN e MR. ROGER HEWSON e MS. ELIZABETH ARMSTRONG e DR. EDWARD LAVERTY e MR. THOMAS BROUSSARD Q
LICENSEE SUPPORT ORGANIZATION :
[
i
- Senior level counterparts
- Staff
- Technical
- Administrative
- Development of extensive response library
- Effective link to line organization
- Thorough extent of condition reviews j i
i i
7
\
O O O SAFETY ASSESSMENT SCHEDULE
- June 17-July 12 Team preparation e July 15 Public entrance meeting
\
- July 15-26 First onsite evaluation period i
- August 12-23 Second onsite evaluation period
- October 7 issue report !
- October 10 Public exit meeting i
8
~
O O O t
! SAFETY ASSESSMENT PROCESS e Extensive preparation e Horizontal assessment across functional areas ;
\
- Vertical slice reviews for selected systems e Analytic code review e Interviews
- Safety assessment / root cause evaluation i
[
9
SAFETY ASSESSMENT ACTIVITIES 1
e WALKDOWN SYSTEMS i e EXTENDED CONTROL ROOM OBSERVATIONS .
- VERTICAL SLICE REVIEW i SERVICE WATER HIGH PRESSURE SAFETY INJECTION ;
AUXILIARY FEED WATER (PARTIAL)
EMERGENCY DIESEL GENERATORS (PARTIAL) l t
e PROGRAM / PROCESS / PROCEDURE REVIEW o ANALYTIC CODE REVIEW i HORIZONTAL REVIEW TO SER VERTICAL SLICE REVIEW STEAM LINE BREAK VERTICAL SLICE REVIEW DROPPED ROD e INTERVIEWS 4 -
h j
ANALYTIC CODES e SCOPE
> REVIEWED 13 CODES FOR DISPOSITION OF CONDITIONS
> REVIEWED APPLICATION OF CODES FOR ROD DROP
> REVIEW IN PROGRESS FOR STEAM LINE BREAK e RESULTS
> ALL CONDITIONS IN CODE WERE ADDRESSED
> ROUTINE RELOAD ANALYSES WERE GENERALLY HANDLED WELL
> MAIN STEAM RUrrORE (RETRAN) HAD ERRORS BUT RESULTS WERE NOT AFFECTED
- PROCEDURES
- DISPOSITION OF CONDITIONS
/0
] Q C/ (Q./
~.
MAINE YANKEE ANALYTIC CODES REVd' WED CODE NAME FUNCTION NO. Of SER CONDITIONS"'
CASM0-3G Physics 1 SIMULATE-3 Physics 2 l STAR Physics - Space / Time 2 FROSSTEY-2 Fuel Performance 4 COBRA-IIIC Generic Core and Fuel Hydraulic 11" COBRA-IIIC Plant Core and Fues Hydraulic 0 Specific YAEC-1 CHF Critical Heat Flux Correlation 2 SCU Statistical Uncertainties Applied 0 to Thermal Margin Setpoints RETRAN 02 Mod 2 System Thermal Hydraulic 39 Generic RETRAN 02 Mod 2 System Thermal Hydraulic 0 Plant Specific BIRP Reactivity Balance O_
CHIC-KIN Integrated Singla Channel Fuel, O T/H, Physics GEMINI-II System Thermal Hydraulic 5
"' SER conditions are limitations, application restrictions, and verification and validation issues within which the code and application are judged acceptable to the staff.
" Conditions for COBRA-IIIC Generic are imposed by the author and are described in the Topical Report BNWL-1695.
/A ---
~ .
CODES USED FOR KEY TRANSIENTS TRANSIENT'" FROSSTEY-2 , GEMINI- COBRA- RPS YAEC-1 SCU HAND RETRAN CHICKEN /
II IIIC SETPOINT CHF CALC 02 STAR
. M00 2 CEA DROP X X X X X X CEA WITHDRAWAL X X X X X X BORON DILUTION X LOSS OF FLOW X X X X X X SEIZED ROTOR X X X X X X EXCESS LOAD X X X X X X LOSS OF LOAD X X X X X X LOSS OF X X X X X X FEEDWATER MAIN STEAM X LINE RUPTURE CEA EJECTION X STEAM GENERATOR X TUBE RUPTURE
'" CASMO-3G and SIMULATE-3 are used to prepare input for all the transient calculations b
/0A
SAFETY ASSESSMENT STANDARDS e Regulations - Measure conformance e Assessment - Measure margin of safety
- Superior
- Good
- Acceptable j e Probabilistic risk assessment - Provide l perspective I
// i i
o o o SAFETY ASSESSMENT RESULTS e Overall Performance Adequate for Operation
- Design / Licensing Basis - Generally in conformance
- Operations - Very good
- Maintenance - Good
- Testing - Acceptable i
- Engineering - Good
- Self Assessment / Corrective Actions - Acceptable j
[
i i
12
O O . O 4
LICENSING AND DESIGN BASIS e Licensing Basis
- Generally in conformance
- Lacks specificity
- Contains inconsistencies
- Not well maintained e Design Basis j
- Quality good
- Availability good
- Design deficiencies identified l
- Supports operation at 2440 MWt 13
1 LICENSING / DESIGN BASIS i OPERABILITY ISSUES i
- Component cooling water i
- Reactor water storage tank
- Equipment qualification
- Ventilation
- Logic circuitry i
i
- Containment spray pump
- Check valve testing ,
14
O O O .
i SAFETY ASSESSMENT OPERATIO'N S ,
i
- e Strengths i
- Operator performance
- Use of risk information j t
- - Management involvement
- Shift turnovers
- Pre evolution briefs
- Weaknesses :
- Workarounds and compensatory measures
- Post trip reviews j
- Log keeping i
15
^
, O O O l SAFETY ASSESSMENT MAINTENANCE l
- Strengths
- Knowledge /use of risk
- Motivated / dedicated work force
-- Control of temporary repairs
- Quality of maintenance e Weaknesses
- Declining material condition
- Inconsistent equipment reliability 16
~
O O O SAFETY ASSESSMENT TESTING I
e Strengths
- Steam generator tube testing
- Inservice testing
- Containment leak rate testing l
- Weaknesses !
- Inadequate scope Weak rigor i
- Weak evaluations I 17 1
i
O O O i
SAFETY ASSESSMENT ENGINEERING ,
- Strengths
- Quality of engineering work
- Qualified capable staff 1
- Electrical design work
- Support provided by Yankee Atomic
- Weaknesses j
- Inconsistent problem identification
- Inconsistent problem resolution
- Limited ownership of programs 18
SELF ASSESSMENT /
CORRECTIVE ACTIONS
- Self Assessment i
- Internal / external effectiveness mixed i
- Fragmented problem identification process
- e Corrective Actions
- Weak implementation i
- Fragmented
- Occasionally ineffective
- Improvement Plans
- Limited effectiveness j
- Many individual plans
- Results mixed 19
!tlt ! ! !!! :lsl! ;!l!l1i !!!!
.O E
- S yot
- U gt ns ,
s A r sa as e el h eem
- n cp
- C er ent rt cha r u
T O
t some osoms udg o eo sr cesee up O r gr ad
- R l w
ba daece ol e n nl abed
- T aa l sMiae l ave N eina o ahn E bvis.v oat c e ase M t daddr em ee eraeuo
_O S rt u i v gi t zss
_ S s m ti pi r s f e o E sl e
i cu et opn r t S r s r i
r 1
e par npl a e s .
S s irc hoecmlyc ime n
_ A u a meseen ni o Y C o c svva ueot i
ipa n otc T
E t
o o ndr odpet f o Epaiebpa r cf t uon sd R crdmf F
A e S
-O
_ ;' ! : 3
i SAFETY ASSESSMENT ROOT CAUSE
- Root Cause 2 i
There is a lack of a questioning culture which has resulted in the failure to identify or promptly correct significant problems in areas perceived to be of low safety significance. Management appears complacent with the current level of j safety performance and there does not appear t
to be a clear incentive for improvement.
21 !
i
o o O SAFETY ASSESSMENT PUBLIC MEETING
- With Licensee
- NRC presented results
- Licensee senior management responded i
- With Public i
- Overallimpressions
- - Key concerns 22
O O O REGULATORY LESSONS LEARNED t
I e Analytic code validation
- Compliance with safety evaluation reports i
>. i
- Licensing reviews for power uprates
- Net positive suction head requirements
- Inspection program issues l
l t
23
DigitalInstrumentation and Control Systems in Nuclear Power Plants i
i Safety and ReliabilityIssues i
Briefing on FinalReport March 7,1997 I
[
[
~
O O O' Committee on Application ofDigitalInstrumentation and Control Systems to Nuclear Plant Operations and Safety 3 Douglas M. Chapin (chair) MPR Associates,Inc.
Joanne Bechta Dugan University of Virginia Donald A. Brand Pacific Gas and Electric Co. (retired) 3E James R. Curtiss Winston and Strawn
- D. Larry Damon Bechtel Research and Development I Michael DeWait Federal Aviation Administration f4 John D. Gannon i
University of Maryland Robert L. Goble Clark University
- David J. Hill Argonne National Laboratory Peter E. Katz Calvert Cliffs Nuclear Power Plant en Nancy G. Leveson University of Washington j Christine M. Mitchell Georgia Institute of Technology
- Carmelo Rodriguez General Atomics Company
-3 James D. White Oak Ridge National Laboratory i
, ~
O O O~ ,
Committee's ChargeforPhase 1 Define the important safety and reliability issues (concerning hardware, software, and human-machine interfaces) that arise from the introduction of Digital Instrumentation and Control Technology in Nuclear Power Plant Operations, including operations under normal, transient and accident conditions.
1
O O O IssuesIdentified in Phase 1 Technical Issues Systems aspects of Digital I&C Technology Software quality assurance Common-mode software failure potential Safety and reliability assessment methods-Human factors and human-machine interfaces Dedication of commercial off-the-shelf hardware and software Strategic Issues Case-by-case licensing process Adequacy of the technicalinfrastructure
O O O Committee's ChargeforPhase 2 Identify criteria for review and acceptance of Digital I&C Technology in both retrofitted reactors and new reactors of advanced design Characterize and evaluate alternative approaches to the i T
certification or licensing of this technology Where sufficient scientific basis exists, recommend guidelines on the basis of which the USNRC can regulate and certify (or license) Digital I&C Technology, including means for identifying and addressing new issues that may result from future development of this technology
Committee's ChargeforPhase 2 (Cont'd) t Where insufficient scientific basis exists to make recommendations, suggest ways in which the USNRC could acquire the required information l,
t
Two Major Themes ;
Two major intertwined themes associated with the use of Digital Instrumentation and Control in Nuclear Power Plants are: .
Dealing with the specific characteristics of Digital I&C Technology as applied to Nuclear Power Plants Dealing with a technology that is more advanced than that widely used in Nuclear Power Plants '
advancing at rate and in directions largely uncontrolled by the nuclear industry likely to have a significant impact on the operation and regulation of Nuclear Power Plants i
b i
O O O~ ~l Major Communications Baniers .
i Ways to address the major communications barriers among I participants in key technical communities and individuals are:
Provide better, clearer, crisper statements of regulatory concern and the appropriate acceptance criteria that are valid at any point in time Both the nuclear industry and the USNRC need to be more -
proactive in participating in the relevant technical communities Both the nuclear industry and the USNRC need to strengthen infrastructure in digital systems Address communication problems systematically, e.g., pay increased attention to definition of terms and context, encourage early communication among interested parties ,
Tune up regulatory mechanism to better address situations where advanced technology, like Digital I&C, has out paced regulations !
Committee's Response to Phase 2 Charge 9
In carrying out its charge the Committee recognized:
Only a limited number of issues could be dealt with in the relatively brief duration of the study ,
General, high level criteria would not be particularly useful i
Criteria are legally the USNRC's responsibility Licensing criteria should be forged in a detailed interaction among the regulators, the industry and the public The Committee is not a surrogate for this interaction among the stakeholders. !
1 Committee can best serve by clearly delineating and defining issues ;
and providing guidance to resolve these rather than by developing specific licensing criteria.
~ ~
O O O Committee's Conclusions and Recommendations Committee's Conclusions and Recommendations fall into four (4) broad categories:
Current practice that is essentially satisfactory or requires some fine tuning Points of weakness in the USNRC's approach !
Issues that merit further inquiry and research before satisfactory ,
criteria can be developed ;
Criteria and guidelines that are unreasonable to expect in~ the near future ,
The Committee expects the Nuclear Industry and the USNRC to extend the work of specific criteria development beyond where this report stops.
High-LevelIssues i
Regarding high-level issues related to digital technology, the Committee emphasizes: '
l Deterministic assessment methods, including design basis accident I analysis, hazard analysis, and other formal analysis procedures are applicable to digital systems, as long as they are applied with care
~
~
Software failure probability can be used for the purposes of performing PRA to determine the relative influence of digital system failure on the overall system. Including software failures in PRA is preferable to the alternative ofignoring software failures i e
Hardware and software must be treated together as a system; focusing solely on one or the other should be done with great caution L
O O O High-LevelIssues (Cont'd) 1 Most practical I&C systems cannot be exhaustively tested and thereby shown to be error free. However, adequate approaches exist and can be applied within practical resource constraints to support using digital systems in safety-critical applications in Nuclear Power Plants.
[
N b
t
}
O e Summary Conclusion Regarding DigitalI&C O :
l in NuclearPowerPlants Digital I&C Systems offer powerful capabilities that can affect Nuclear Power Plant safety Digital applications should be treated carefully, particularly in safety-critical applications I
USNRC and the nuclear industry are moving forward with procedures, processes and technical infrastructure; the Committee has several i suggested improvements i
- The Committee considers the use of Digital I&C Systems in new l nuclear power plants and in modification and upgrades of existing plants to be appropriate and desirable
- For existing plants, replacement by digital equipment of older systems I and equipment for which vendor support is no longer readily available is particularly appropriate and desirable 1
o o o ~$
l Commercial Light Water Reactor Tritium Production Project OVERVIEW i
.N ~ ~
w \
3 !
+ !
March 7,1997 Stephen M. Sohinki Director, CLWR Project Office Defense Programs U.S. Department of Energy i
?
~
O O O
~x.
E Requirement for New Tritium e All warheads in the U.S. nuclear weapons stockpile depend upon tritium to function as designed.
e Tritium decays at a rate of 5.5% per year (12.3 year halflife).
e Tritium is rare on earth. It must be man-made with reactors or accelerators. ,
e Tritium production stopped in 1988 when the last of the Savannah River reactors shut down.
e Current stockpile being supported with tritium recycled from dismantled units.
e The 1996 Nuclear Weapons Stockpile Plan and accompanying Presidential .
Decision Directive require DOE to establish a new tritium supply by 2005, if commercial reactors are to be used, and/or by 2007, ifit is to be a new accelerator.
l Q O DOE's Dual-Path Strategy
- Based on Secretary of Energy's December 1995 Record of Decision for the 7'ritium Supply and Recycling Programmatic Environmental hnpact Statement.
Design, build, and test critical components of an Accelerator for Production of Tritium (APT)
Initiate purchase of a Commercial Light Water Reactor (CLWR) or irradiation services
- By late 1998 the Secretary will select CLWR or APT as the primary tritium supply. The other alternative, if feasible, will be developed as a backup.
- In any case, the CLWR option will be used to establish a contingency capability that can deliver new tritium on short notice.
Complete qualification activities with the Nuclear Regulatory Commission Fabricate initial inventory of tritium-producing materials Construct Tritium Extraction Facility (TEF)
Establish options contracts with utilities
O O sDual-Path Strategy for Tritium Production FY 1996 l 1997 l 1998 l 1999 l 2000 l 2001 f 2002 l 2003 l 2004 l 2005 l 2006 l 2007 i 200812009 l As of Dec 31,1996 T tium-Producing Burnable Abso er Rod Teshng & Certif cation
@ CLWR Extraction Process Developrnent & design ,' ,'
Commercial Light Water Reactor l l Construct SRS Extraction Facihty
. . . . l Tntium Project ^
6 Procurement' Process'& Conditional PIrchase of Reactor or Irradiation Services l l l l l l Fabric'ate First' Core Lo'ad of Ab' sort >er R'ods I e
g y .. ,
PLANNED l l ,, .
ACTIVITIES l l l l Maintain options contracts for reactor and/or irradiation services purchase , g APT NEPA Review. Permitting. and Safety Appfoval l 1 l
^cce o p APT Conceptual Design of Trit:um APT Engineering Development Project APT Engineenng Desen I 8 8 8 I I I I g g g g
{ g i i i i I I i i i I i i DECISION l l Secretarial Selection of Primary Techr$ ology l l l l l l I I I I I I I I I I I I I APT Site Preparation g l l l l l and Construction , , .
- g g g g g -
l l Equipment Mfg ACCELERATOR l l g , ,
SELECTED [ "d '"S*[ " " lg. l l l l-APT Commissioning and Startup l l l l l l : i i i i :lI i I l l l l l l Continuing Tntium Production Operations l /
I I 1 1 i
CLWR l
I l
I I lI . . .
I I I Transfer NRC License Finalize CLWR Purchase g 1 l PURCHASE I I Irradiate First Batch of Abso er Assemblies j First Extraction of' Tritium Gas I SELECTED l l l l 1 I I I I I I I g Continuing Production Operations l l , -/
1 I i i I I I I I I l l l ll l
CLWR l l Amend NRC Licenses. Execute options contracts.
IRRADIATION . . . . l . - - - -
inadiate First Batch of Absorber Assemblies l First Extraction of Tritium Gas l SERVICES l , ,
SELECTED Continuing Tritium Production Operations '
l l
/
r . . . .
O O O
- 1 Sy' stem for Producing Tritium N" rs in Commercial Reactors Tritium-Producing Burnable Absorber Rod !,
Manufacturer Tritium-producing rods are irradiated in one or more reactors for 1-2 years
?k V3 i s After irradiation, rods i are shipped to DOE's Tritium-Producing
/
- L 7 Savannah River Site rods are installed * ,
in reactor fuel y 3Ter& tt y _'_'____
assemblies ------N---------
Savannah River Site New tritium Tritium Extraction is sent to the Facility removes tritium recycle facility i
from the rods to support the ruet ASSEMBLY stockpile
t Light Water Reactor Tritium Target Qualification Project t
Tritium Producing j Burnable Absorber Rod Technical Discussion ,
Dr. Jerry L. Ethridge Pacific Northwest National Laboratory i
March 7,1997 l
l l
,.. m _ ..,~.,_m.e_,,
q
~
O O .
O~
Light Water Reactor Tritium Target Qualification Project Production of Tritium from Lithium in a Reactor Helium-4 Neutron ,
1h < SBC Lithium-6 Lithium-7 tium (H )
Half Life (tm) of Tritium = 12.3 years Helium-3 Paabc Northwest National Laboraftwy >
O O O~
Light Water Reactor Tritium Target Qualification Project Tritium-Producing Burnable Absorber Rod for.
Commercial Light Water Reactors (TPBAR) h ic ,
+ c- Alummate ,
Pellets : <,
Nickel . 14 / 5
'j 7 ;
Plate -
Zircaloy : *g F;: j e ",y:x i t : s; ;
- 5. g Aluminide Coating on
..; &g s
} Rea h h 't s inner Surface a Stainless Steel Cladding
!/
e.:
j!'
i
---f gm $s
!S if$ -
02 ~t I n
3.g 3 x & ,
- .j.
Not to Scale Pactic Northwest Nabonal Laboratory I
_______________.__________.-____________.-.-_________-___..______..____________.______-._______-___._.-___________m. _ . _ _ _
O O O' Light Water Reactor Tritium Target Qualification Project Functions of Tritium Producing Burnable Absorber Rod Components Stainless Steel Cladding - Similar to reactor burnable poison elements.
Contains all components.
Aluminide Coating - Prevents diffusion of tritium through the steel cladding into the reactor coolant. Also prevents hydrogen in the coolant from entering the rod.
Zircalloy (zirconium alloy) Tritium Getter - Absorbs free tritium gas.
- Nickel Plate - Prevents oxidation of the tritium getter material.
Lithium Aluminate Pellets - High-temperature ceramic material containing Lithium-6, the material that transmutes to tritium when a neutron is absorbed.
Zircalloy Liner - Removes free oxygen to improve tritium gettering.
During and after irradiation, tritium is tightly held in the ceramic, the tritium getter, and the liner until released by the extraction process. !
l }
l t
! I i
l Pac 6c Northwest Na00nal Laboratory l
+
~
O O O~ :
f Light Water Reactor Tritium Target Qualification Project i Basic Concepts I
Production: 8Li + 1 n o
> 4He + 3H + 4.8 MeV Retention: 2(aH2 O) + Zr > ZrO2+ 2(aH2) (Liner) 3H2 + Zr > Zr3H 2 (Getter)
PRF* ~
1 04 (Barrier) i Distribution: 3H in pellets 3H in Zr liner 3H in NPZ getter
- PRF (permeation reduction factor) = ratio of tritium permeability in bare stainless steel to that of a coated tube; i.e., high PRF means low tritium release. -
Pacdc Northwest Nabonal Laboratory
Light Water Reactor Tritium Target Qualification Project E
Lead Test Assembly Cycle 5 Irradiated LTAs are retumed to PNNL 1 for post-irradiation examinations Parts are purchased and :
assembled by PNNL !
i
\
. .c t
4
= LTAs to be
- irradiated for i
\lg one operating ;
cycle in theq !
- Watts Bar w 2:
Reactor . 43
- Ei 7 j N @hN Lead Test Assembly (LTA)
Tritium Producing Bumable !
Absorber Rods (TPBAR) are j sent to Westinghouse, Columbia, SC -
to be inserted in four fuel assemblies 3
na Fuel assemblies with LTAs ;
-Y Paate Nor1thwest Nabonal Laboratory are sent to Watts Bar !
)
f Light Water Reactor Tritium Target Qualification Project 9l :
Tritium Producing !
Burnable Absorber Assembly l 1
- g. -
8E 8 y e ' '
? w-- hu l
i:
stl i:,
l i -
i , '
/l -l ' '
it'67e i
> i 4
ll Ql i
, t, '
! ! '. l i
l /
l ll <1
, E ,
s
/'
j ji s .
i, W
ll .'
c' * ,
1 1 i i h O'
l 0l ; ,
4 tk
, l
- i. ;
4 I' i f d ! EIGHT (
ql d< (!j
'i TRITIUM i i i . J. !. . ! ! PRODUCING i i
!, a. idi j
- , i,.:
l:
ii j
. BURNABLE ABSORBER l
f j 2 RODS 12 FT LONG l 3/8" DIAM i, l : K l-(TPBAR) r i
l
. 1 J
l l 4 a
- a s :
i e
V Y
[
J U
b V ej Pacific Nomwest Nabonal Laboratory r
I
w O O O ~!
Burnable Absorber Rod (TPBAR) 1
- The TPBARs perform the same function as the burnable absorber rods normally found in a commercial reactor.
The TPBAR has essentially the same external dimensions as i burnable absorber rods.
\
The TPBAR substitutes Lithium-6 as the neutron-absorbing j material in place of the usual Boron-10. i
- The Lithium-6 is in a ceramic form (lithium aluminate).
- When Lithium-6 absorbs a neutron, it transmutes into tritium.
The TPBAR's neutron-absorbing characteristics are very similar to those of rods containing Boron-10.
4 -
TPBARs contain no fissile material, i.e., no uranium or plutonium.
Pacstic Northwest Na00nal Laboratory l
Light Water Reactor Tritium Target Qualification Project i
Confirmatory Information ,
Supporting LTA i
l
\
t f
i e
I Paa6c Noithwest Nadonal Laboratory
[
t
Light Water Reactor Tritium Target Qualification Project -
i l
! Two Types of Reactor Tests
- l i
Three Capsule Test Rods (S-1, B-1, WC-1)
- Permeation measured by periodic capsule Flush / Sampling / Analysis Loop-1 Test (8 rods in a bundle) :
- Permeation measured by daily sampling and analysis l
t I
l Pacinc Writm'est Nadonal Laboratory f
~
O O O~
Pacific North west National Laboratory TRITIUM PERMEATION DATA
~
i 2000 1800 ,
,f EstirmtedLimitingPerfmmnx:
. 10 Ciper550 days ,
.! 18 .
l u
g }- 1200 . .___
ge ,
Trendline estimates fir N 1000
.E5 '
LTXliii~edmATRTests /
.s
= Y 800
. /
/
b= 8 1 f ;
- -; y /
4 _
o i 0 :n ,
f j= p ,
0 50 100 150 200 250 300 350 400 450 500 550 Tine (days) ;
c l oop-l x WG1 : S-1 : B-1 Trend . . Linit i
77QP
Light Water Reactor Tritium Target Qualification Project Free Gas Analysis of WC-1 Gas Distribution Volume i (Percent of total) Predicted Measured 4He 90 to 99% 99.1 %
3H 10-s <104 i 3H2 O 10-2 <104 Tritium Activity in Gas, mci 60 30 i
NOTE: Results confirm 3H was produced during irradiation, and that the getter and lithium aluminate retained most of the tritium produced.
1-Pacdc Normwest Na00nal Laboratory
l Light Water Reactor Tritium Target Qualification Project
, f
- LTA Report Overview ,
i Describes the design and fabrication requirements for TPBAR LTAs. :
Evaluates the impacts of four LTAs on the operation '
of a commercial light water reactor.
Addresses applicable areas of review from NUREG-0800, the " Standard Review ~ Plan" Paanc Northwest Nabonal Laboratory
O O O ~!
Light Water Reactor Tritium Target Qualification Project Report on the Evaluation of the Tritium Producing Burnable Absorber Rod Lead Test Assembly (LTA Report) -
CONTENT (Standard Review Plan Section):
TPBAR Design (4.2)
Nuclear Design (4.3)
Thermal Hydraulic Design (4.4)
Materials (4.5.2, 6.1.1)
Impacts of Irradiation (Various)
Quality Assurance (17.1)
Safeguards and Security (n/a)
Regulatory Issues (Various)
Paanc Northwest Naconal Laboratory
O O O ,
, Light Water Reactor Tritium Target Qualification Project
[
LTA-TPBAR Mechanical Design Evaluation l
i t
Components Evaluation Summary l l
i Cladding /End Plugs -
Structural integrity maintained under all conditions (except LBLOCA) using most conservative assumptions (see below)
Acceptable stress / strain (including creep l and bow)
No cladding collapse No failure predicted due to fatigue / fretting l Inr.ignificant corrosion / erosion Pellets -
Nonreactive in H2O Stability demonstrated to design GVR t Pacshc Normwest Nanonal Laboratory
o o o Light Water Reactor Tritium Target Qualification Project LTA-TPBAR Mechanical Design Evaluation (Continued)
Components Evaluation Summarv !
Getters / Liners -
Nonreactive (Zircalloy 4) ;
Design accommodates swellir:j due to hydriding Springs Conventional fuel design used (302SS) i i
I
[
PactMc Northwest Nabonal Laboratory b
O O O :
Light Water Reactor Tritium Target Qualification Project t
6 LTA-TPBAR Nuclear Design Evaluation Attribute =
Summary of Effect
- Reactor Nuclear Design -
Minimal impact limited number in any one assembly (8 total / assembly) limited number in core (32 total) placed in non-limiting positions 6Li loading set to achieve maximum tritium production within functional requirements (s1.2g/ rod)
NRC-approved core design methodology will be used -
i i
Pack Normwest Nabonal Laboratory
O O O Light Water Reactor Tritium Target Qualification Project ]
LTA-TPBAR Nuclear Design Evaluation (Continued)
Attribute Summary of Effect Power Distribution -
Minimal impact gaps, fabrication tolerances, and local power distribution were
- evaluated Control Requirements No. impact Will be performed by Westinghouse using Reload Safety Analysis modified (SLi depletion, 3H decay)
PHOENIX-L (will be V&V'd prior to use)
Pacdc Northwest Nabonal Laboratory
O O O Light Water Reactor Tritium Target Qualification Project Comparison of TPBAR with Burnable Absorber Rod .
Design Parameters Parameter WABA Rod BPRA Rod TPB A R I 17x17 FA 17x17 FA 17x17 FA t
Overall Length, in. 150^ 152 152 Total Weight, lbs 1.9 1.8 2.26 Absorber Length, in. ,
134^ 142 142 .
Absorber O.D., in. [ ](*) [ ](*) 0.303 !
Thick ness, in. [ ](*) [](*) 0.040 Absorber Material Al 2O3 -B 4C SiO2-B 2 03 LiAIO 2 Outer Cladding O.D., in. 0.381 . 0.381 0.381 i Thickness, in. [ ](*) I [](4 0.0225 Inner Cladding O.D., in. [](4 None None ,
Thickness, in. [](4 Cladding Material Zircaloy-4 304SS 316SS Fuel Assembly Guide [](4 [] (4 [](4 Thimble I.D., in.
Fuel Assembly Dashpot [](4 [ ](*) [](4 i 1.D.,in.
i 1(4 Denotes Westinghouse proprietary data.
Paatic Northnost Nabonal Laboratory
O O O ;
Light Water Reactor Tritium Target Qualification Project Comparison of irradiation Conditions for the Tritium !
Test Rods and the TPBAR i
Paraweter WC-1 Test S-1 Test B-1 Test Loop-1 Test TP8Mt ,
Total Exposure. EFPD 281 143 143 217 450 to 500 550 Max :
Axial Peak Thermal 4 4 6 3 -1 !
Neutron Flux. l 1 055 n/cW/s i Axial Peak Thermal 0.98 0.5 0.75 0.56 -0 . 5 I Neutron Fluence. l 10n n/cd i I
Arial Peak-to-Average 1.3 1.3 1.3 1,46 1.2 Thermal Neutron Flux Ratio l i
Cladding Fast Neutrun 0.7 0.7 0.7 14 10 F l ux . 1058 n/cW/s (E>l MeV)
Fast Neutron Fluerte . 0.17 0.09 0.09 2.60 -5 10" n/cW (E>l MeV)
Axial Peak n-a Power 0.60 0.60 0.9 0.45 (1.5) BOL Level. kW/ f t (kW/m) ( 2. 0) (2. 0) (3. 0) 0.60 (2.0) l EOL O.40 t
( 1. 3)
PacMc Nor%est Natmenal Laboratory l
i
O O O :
Light Water Reactor Tritium Target Qualification Project Comparison of irradiation Conditions for the Tritium Test Rods and the TPBAR (cont'd) ,
Parameter WC-1 Test S- 1 Tes t B- 1 Tes t Loop-1 Test TPBAR Axially Averaged n-a 0.43 0.43 0.63 0.32 ( 1.05) BOL Powr Level . kW/ft ( 1. 4) ( 1. 4) ( 2.1) 0.50 (kW/m) ( 1.64)
EOL i 0.35 (1.15)
Rod Average GVR 116 53 80 68 174
. Axial Peak GVR 159 73 110 99 209 Rod Average ti 15.7 7.2 10.8 8.4 65 Burnup. % .
ti Enrictment, at.% 57 57 57 59.2 -22 Tritium Produced per 2700 1240 1860 1559 3900 4 ft (1.22 m) Rod. Ci (0.28) (0.13) (0.19) (0.16) (0.40)
(9) pH of Water 5. 5 to 10 5.5 to 5.5 to 6.65 i0.2 6. 5 to 7.0 10 10 Paahc Nortneest National Laboratory
O O O' l Light Water Reactor Tritium Target QualiHcation Project i
LTA - Thermal Hydraulic Design Consideration Evaluation Summarv
- T-H Design -
TPBAR cladding temperature criteria met Tech specs prevent locating LTA in limiting position Negligible thermal effects on operational i transients i
- Core T-H Design i No. impact (assumes proper location) l PacMc Northwest Nabonal Laboratory I
O O O~
Light Water Reactor Tritium Target Qualification Project ;
Thermal and Hydraulic Comparison Table i
Design Parameters Re f. Plant With LTAs w/o LTA ,
Reactor core heat output, M W. 3,411 No c han ge Reactor core heat output, !O' Btu /hr 11,641.7 No c hange llent generated in Fuel, % 97.4 No change Sy stem Pressure, nominal, psia 2,250 No c han ge l System pressure, minimum steady state, psia 2,204 No chan ge Cnnlant Fin w e
Total therm at flow rate,10* !! ./h r 144.7 No chan ge i i
Effective flow rate for heat transfer,10'Ib./hr 131.7 No c hange !
Effective core flow, gpm 372.400 No change I Average flow p er asse mbly , gp m [ ]N No change Effective flow area for heat transfer, ft" 51.3 No c han ge Average velocity along fuel rods, ft/sec 15.4 N o c han ge I Average mass yclocity,10* lb./hr.ft 2 2.57 No change Onnlant Tem n erni nre Nom inal in tet.*F 559.0 No change Average rise in vessel,*F 58.4 No change Average rise in ccre,*F 63.5 No change A verage in core,*F 592.5 No change ,
Average temperature in vessel,*F 588.2 No chan ge
}
Pactic Northwest National Laboratory i
o
~
O '
Light Water Reactor Tritium Target Qualification Project Thermal and Hydraulic Comparison Table (cont'd)
Des 1
21.I No change Power dens 1ty. kW per 11t.er of core 104.5 No change Specific power. kW per kg uraniten 38.4 No change funi rnntral Tmmrat iro Peak at peak linear power for prevention of fuel 4.500 tb charuje centerline snelt.*F TPRAR Peak Trennra t e rn Peak IPBAR absorber tanperature. *F N/A 750 Prmenen nrrn Across core. pst 24.4 12.4 No charwje Across vessel . Including ncIlle. psi 44.5 16.7 tb change Paafw Northwest National Laborasary t ~ O~
- O O Light Water Reactor Tritium Target Qualification Project .
LTA - Materials , i Component
- Material Evaluation Summary '
- Cladding /End Plug 316SS (20% CW) - Compatibility Demonstrated negligible corrosion chemistry control prevents SCC will go through thorough weld qualification Highly stable under all conditions Pellets LiAIO 2 to design bumup .f Getter / Liner Zircaloy-4 . Standard reactor material
- Ultrasonic, eddy current, radiography and He-leak testing will be p~e rformed under strict QA requirements for all components PactRc Northwest Nabonal Laboratory i
~ O O O~ Light Water Reactor Tritium Target QualiRcation Project LTA Operational impacts - Normal Operations - tritium releases represent an insignificant incremental increase in the total off-site exposures to the maximum exposed in'dividual - Refueling Operations no new equipment / tools necessary for handling TPBARs ~ - shipping performed by DOE in approved cask occupational exposure less than a few millirem . Off-Normal No TPBAR failures are predicted during Conditions I,11, Ill and IV except LBLOCA Relocation of absorber materials will not occur except during LBLOCA (loss of ~1 pencil) Cladding Defects minimal impact on core reactivity no impact on coolant chemistry off-site doses from cladding breach are insignificant Pacnfc Northwest Nabonal Laboratory O O O~ Light Water Reactor Tritium Target QualiRcation Project LTA Operation impacts (cont'd)
- LBLOCA represents most limiting accident .
- like the fuel, TPBAR failure may occur during design basis i LOCA i - adequate core cooling is maintained - no increase in calculated radiological consequences of the i l accident to off-site individuals - negligible contribution to combustible gas inventory (less from equivalent number of zirconium clad WABA rods) l r [ l ! i i l i l l l PacMc Normwest Nabonal Laboratory ~ O O O 7 Light Water Reactor Tritium Target QualiHcation Project LTA Report Conclusions Off-site doses due to release of tritium from TPBARs over the operating cycle are insignificant - nearly all tritium produced is captured in a solid matrix and can be released only under extremely high temperatures. a TPBARs will not fail during analyzed accidents, with the exception of a Large Break Loss-of-Coolant Accident (LBLOCA). - Off-site dose consequences of non-LBLOCA accidents are inconsequential. i Pacdc Northwest Naeonal Laboratory - . - . - _ . . . . . _ _ - - _ _ _ . - - . _ . _ - - . - . . _ . - . - _ _ - - . . _ - . - - - - - . - . - . . . . - - . - ~ _ - . - . . _ . . ~ O O O~ Light Water Reactor Tritium Target QualiRcation Project t LTA Report Conclusions.(cont'd) I LTAs are compatible with existing fuel assemblies and conventional handling tools, equipment, and procedures. = The LTAs use materials with known and predictable performance characteristics and are compatible with PWR coolant. TPBARs operate within established thermal-hydraulic criteria and will not fail during normal operation or anticipated operational , occurrences. - The LTAs do not adversely affect reactor neutronic or thermal-hydraulic performance - perform essentially the same as standard burnable absorber rods. ; I Pac 6c Northwest Nabonal Laboratory i Light Water Reactor Tritium Target QualiHcation Project t i LTA Report Conclusions (cont'd) - The LTAs will not cause a difference in the analyzed behavior of the reactor during a LOCA. - Off-site dose consequences of LBLOCA show no increase. - No impact on host facility Technical Specifications. - Evaluation of test irradiation at a commercial light water reactor is within the criteria of 10 CFR 10.59. l Pacdc Northwest Neeional Laboratory + _ - - _-_ -___ - - _ - _. _ _ - _ _ _ - - _ _ _ - - _ _ - - _ _ _ - _ _ _ - _ _ - _ - _ - - - - - _ - - _ _ - - - _ _ _ - - - _ _ _