ML20129C754
| ML20129C754 | |
| Person / Time | |
|---|---|
| Issue date: | 07/11/1985 |
| From: | Advisory Committee on Reactor Safeguards |
| To: | |
| References | |
| ACRS-T-1428, NUDOCS 8507160319 | |
| Download: ML20129C754 (121) | |
Text
d t A s t-M M
+
UNITED STATES OF AMERICA V(]
NUCLEAR REGULATORY COMMISSION In the matter of:
j ADVISORY COMMITTEE ON REACTOR SAFEGUARDS Subcommittee on Long-Range Plan for the l
Nuclear Regulatory Commission Docket No.
Location: Washington, D. C.
Date: Thursday, July 11, 1985 Pages:222 - 341 ACRSOFRCECDPY x Jo \\lo':1emovgrn4Ql3Alce Court Reporters 1625 I St., N.W.
Suite 921 4
Washington, D.C.
20006 00 (202) 293-3950 bro 711
\\
E 5j7 t 9
a T-1428 l'ille
)
222 1
fj 1
UNITED STATES OF AMERICA v
2 NUCLEAR REGULATORY COMMISSION 3
4 ADVISORY COMMITTEE ON REACTOR SAFEGUARDS 5
SUBCOMMITTEE ON LONG-RANGE PLAN 6
FOR THE NUCLEAR REGULATORY COMMISSION 7
8 OPEN MEETING 9,
10 Room 1167 11 1 7.1 7 H Street, N.W.
.12 Washington, D.
C.
i 13 Thursday, July 11, 1985 14 15 The subcommittee met, pursuant to notice, at 6:35 16 p.m.,
Max Carbon, Chairman of the Subcommittee presiding.
17 18 ACRS MEMBERS PRESENT:
19 Max Carbon, Chairman 20 Dade W.
Moeller 21 Forrest Remick 22 Charles Wylie 23 Harold Lewis
(
24 25 John McKinley, Designated Federal Employee
223 1
PRESENTERS; 2
Floyd Culler 3
John Taylor 4
l 1
5 OTHERS PRESENT; 6
George Sauter 7
Jack Berga 8
9 1
10 11 i
1 '2 13 14 15 16 17 18 19 20 21 22 23 24 25 6
224 1
P ROC EED I NG S 2
MR. CARBON:
The meeting will now reconvene.
3 This is the continuation of an open meeting of the 4
ACES Subcommittee on the Long Range Plan for the NRC that 5
recessed July 10th.
6 I am Max Carbon, Chairman of the Subcommittee.
7 The other ACRS Members present today are Hal Lewis, 8
Dade Moeller, Forrest Remick, and Charlie Wylie.
9 The purpose of this meeting is to continue to gather 10 information from a variety of sources regarding the objectives 11 and bases for a long range plan for the Nuclear Regulatory n-12 Commission.
The topics will be primarily technical issues
(
13 related to the regulation of nuclear power plants with regard 14 to the public health and safety.
It is expected that the plan 15 would cover the period from the present to five to ten years 16 in the future.
17 John McKinley, en my right, is the assigned ACRS 18 staff member for this meeting.
19 A transcript of the meeting is being kept and it is 20 requested that each speaker first identify himself or herself 21 and speak with sufficient clarity end volume so that he or she 22 can be readily heard.
23 We have received no written statements from members 24 of public.
We have received no requests for time to make oral 25 statements from members of the public.
l I
225 O
(,/
1 Do any members of the Subcommittee have questions or 2
comments they wish to make about the proposed agenda?
3 (No response) 4 Gentlemen, as I said before, we are just very 5
pleased and delighted that you could come this evening.
6 Since we have already talked a bit about the format 7
this evening, and since you have already said that you would 8
start out by going through our list of issues and then get 9
into general topics later, let me say to the Subcommittee 10 members unless you hava any points.to bring up at the moment, 11 we will just launch right into it.
[a*)
12 MR.' CULLER:
We are adaptable, too.
If you want to 13 start asking questions, that is perfectly fine with us.
14 MR. CARBON:
Sometimes we start it with general 15 comments and then the issue.
Since you suggested the other 16 way, we are flexible, and you just take your pick.
17 MR. CULLER:
'The way we are going to present this I
18 evening -- if presentation is what you should call it 19 would like to make a few comments on goals.
Then John is 20 going through the list of technical issues, which will bring 21 up lots of discussion.
And, I hope you will interrupt either 22 one of us.
23 When we talk about the technical issues, I am sure i
24 other topics will turn up.
one of 25 At the end, there are a few thoughts
i 226 1
which Hal knows about -- that I would like to introduce in the 2
way of organization and purpose for ACRS, different than it is 3
now that may or may not be a good idea, but I think it 4
deserves some discussion in the present context of changing 5
patterns a little bit.
6 First off, I wish to thank you for the invitation.
7 and your great courtesy of adjusting your schedule to ours.
8 We all appreciate the opportunity of being here and discussing 9
with you thoughts.
You are as well aware of the state of the
.10 industry, of the nuclear industry in the United States as we, 11 probably more so.
We see it more than you do from the utility
/ \\
12 point of view, and I am sure that Warren Owen and many other 13 people have told you the probability of having a new reactor 14 order is essentially zero.
15 We are finding in economic studies at EPRI that 70 16 to 80 percent of the new plants coming on still provide the 17 cheapest power other than hydro or geothermal in limited 18 locations in the United States.
19 The enthusiasm for nuclear power as a potential 20 future economic baseload plant still remains in the industry, 21 and people who have done recent studies show that the cost 22 differential, given proper conditions of management, of 23 regulatory process, speed, that nuclear power would provide
~s 24 the lowest cost power with the exceptions that I have given, 25 in the future maybe by 15 or 20 percent.
227 m(,j}
/
1 Ed Zabroski, who many of you know and Chauncy Star, 2
have been looking at the economic patterns.
3 We have most of the data from France, and much of it 4
from England, and four great bundles of information on total 5
cost from the utilities, not just architect / engineer costs, 6
somewhat on a privileged basis.
7 But, we are finding basically that many of the 8
accusations about the process of regulation and to some extent 9
the regulations themselves in at least 20 to 25 percent of the 10 cases in nuclear power have contributed significantly to 11 capital costs, and they certainly contribute to the delay.
[V}
12 And it is just almost unequivocal, 13 So, your attention to what ACRS can do, and what NRC 14 might do in changing these patterns is important.
15 I think you realize too, and I think this is a and I think 16 double-edged sword, as nearly as we can tell that the 17 the Staff of NRC has determined the same 18 individual utility's liability with a nuclear plant to losses 19 for damage within the plant are of such magnitude, that the 20 level of safety requirements to protect that investment is 21 probably significantly higher than that required to protect 22 the public health and safety.
23 Now, you can make that argument it will come around
~.)
24 and bite the industry as well as biting the NRC, and we 25 realize it, but there is a basis for some of the little
228 g
1 arguments that I will make in that thought, and I wish to
~
2 state it.
3 MR. REMICK:
Could I just interrupt a second.
4 MR. CULLER:
Yes.
5 MR. REMICK:
Do you have a feeling that the 6
utilities in general, all 55 nuclear utilities appreciate that 7
point?
8 MR. CULLER:
I would daresay that they do not.
- And, 9
of course, in safety it has to be all or nothing at all.
And 10 that is why I say in part this argument can come around and 11 bite us.
It isn't evident from your knowledge, and certainly m
1 '2 from ours, that all the utilities act the same way in any 13 particular issue or proceed with the same dispatch or control 3
14 and understanding.
15 They all have the same intentions.
But there are 16 great levels of understanding that differ from one group to 17 another, and I am sure you are well aware of that.
18 The second observation that I am sure John will 19 expand on, is an important one. We are also certain, because since we are at the development interface, and 20 of utility 21 our development is not lab, it is engineering scale looks, and 22 as you may have heard, and I think John presented to you, the
/ '
23 advanced light water program which we have initiated.
One of 24 the great inhibitions for getting approval for that program 25 and we started three years ago to get our board, our advisory
.~
229 O%/
1 structure to improve -- is that they did not wish to look at 2
improvements to LWRs, because the improvements might be used 3
as items for required retrofit.
4 And I guess the sum of this set of diffuse forces, 5
very real, very real, is that the regulatory processes, and to 6
some extent the actual content of the regulations, inhibit 7
improvements in operation, and certainly do not always allow a 8
progression from one safety level to another because of the 9
concern of having costly backfits and such required.
l 10 And, there are other tyrannies of the regulatory 1.
11 process. We see it in little things and big things all along,
(
12 and no more necessarily in nuclear than is now appearing in 13 fossil with acid rain and a lot of other things.
There is a 14 tyranny of regulatory processes that essentially squash real 15 advance.
16 We are very pleased, however, by the endorsement 17 that the Nuclear Regulatory Commission and I think the 18 interest that the ACRS has expressed in the advanced light 19 water program.
There is an effective setup to pursue the 20 safety aspects and safety reviews as John's groups begin the 21 studies to improve operability, to improve the inherent safety 22 where possible, and to cut costs and increase lifetime, and 23 move certainly to go in the direction of standardization.
24 Now, with those introductory comments, we looked at 25 the four general suggestions on goals, primary goals, in your
230 1
letter.
And we have had quite a bit of discussion.
2 Obviously the first goal, that of protecting the 3
health and safety from accident and operating -- the first 4
two, one is protecting from accidents, and the others is 5
protecting from operating error and such, protecting health 6
and safety, we think they are obviously the important and 7
prime ones.
8 But we, I suppose, would word them a little 9
differently.
I will not read what is proposed, but I will 10 make them available to you, the primary goals.
11 Certainly, if the public health and safety must be
(
12 protected in your words and ours to an acceptable level, the 13 definition of acceptability becomes a primary issue.
Safety 14 goals, you are discussing today and many other things.
15 It is a society value that is decided probably by 16 Congress or an elected body, but certainly ACRS and to a 17 limited extent, NRC, must do the analyses upon which adequacy 18 is based.
And I think that maybe one of the more important 19 things now is to decide on what will be used at least for some 20 period of time to define what adequate and acceptable levels 21 of safety might be.
It provides an anchor for a while that 22 will help stabilise the licensing and regulatory process.
23 So, although I put it second, since it is such an 24 important element of the first goal, I think that it deserves 25 real attention.
231
(
1 MR. LEWIS:
If I could just interrupt you for a 2
second, you know,.of course, that Congress is never going to 3
tell the NRC what level of safety is required.
4 MR. CULLER:
Yes.
5 MR. LEWIS:
And so the only way in which this can 6
happen is if NRC from its own limited perspective, doing its 7
bungling best, proclaims what is meant by adequacy, and sort 8
of challenges the world to object to it.
9 And in principle, that is what is going on with the 10 safety goals exercise.
11 MR. CULLER:
We realize that-r 12 MR. LEWIS:
I don't know whether it will ever 13 terminate.
f 14-MR. CULLER:
If it doesn't, the necessary stability 15 that would revitalize nuclear power in the United States may 16 not occur, because there is a real stability that.has to move 17 into place before, I think, a single CEO in the United States 18 would be willing'to commit to a new reactor.
19 MR. LEWIS:
But the stability, just to pursue the 20 point for one moment, depends not only on the enunciation of 21 a safety goal policy or something resembling it, but it also 22 involves a commitment.of the NRC to use it in some way.
23 Because there have been conversations the last ecuple of days 24 about a safety goal policy being something that you polish and 25
-put on the shelf, like we put our trophies on the nantlepiece,
.. ~.
232 fN
\\s-1 and then go on about normal regulatory practice.
2 It is going to be extremely difficult to get the 3
degree of stability that you are talking for which will 4
involve both the enunciation of a policy and a commitment of 5
the staff to implement that policy and a commitment of the 6
Commission to somehow do that in the face of the inevitable 7
uncertainties.
8 And I don't want to be pessimistic, but 9
MR. CULLER:
I think you are realistic.
It will be 10 difficult.
But I think somebody at some point in time now 11 must stand up and take the gab.
And if not, then I think that
)
12 the United States basically will not use the nuclear opt' ion
\\
(d 13 for quite a long time, until a whole generation of people die 14 and we have lost most of the trained personnel who understand 15 how to put the 16 MR. LEWIS:
There is no guarantee the next l
17 generation will be better than this generation.
18 MR. CULLER:
Well, there is the principle of the 19 hiding hand.
If you remember in development economics, the 20 principle of the hiding hand says that you can overpromise and 21 overse11, and run over in costs.
And the only thing that 22 saves the project is that the people who come after you are 1
fs 23 smarter than you are.
\\
24 And in many respects I think it is true in j
i 25 technological advance.
l
233 4
(m
(_,
1 I think you are right, Hal, and George Sauter and I 2
-- and I think most of you know, George is with us this 1
3 evening -- and John discussed this in an indirect way last 1
4 night.
What George suggested is that we may face a period in 5
which there is a double track, where we sort of poter out in 6
the old practices and try to change a little bit, but adhering 7
as best possible to a set of safety goals and different 8
procedures for handling as best as possible.
And that the 9
changeover may take a long time.
10 So, the idea of double tracking is one that I had 11 not thought of, but in thinking about it today it makes a
[V) l'2 little bit of very pragmatic sense to me, and I don't know how 13 to implement that.
14 One other comment, now, on goals.
I don't know 15 exactly what the wording of the Atomic Energy Act of 1954 and 16 the Revisions in the Safety Act were at the time NRC was 17 created, but my memory says there is a second element to 18 protecting against -- there is a second element of 19 responsibility within the NRC and ACRS.
20 There is also not a promotional activity for nuclear 21 power, but to see that operable systems can operate, and that 22 says that as safety is pursued, we think that there has to be
("N 23 a little bit more balance and that some kind of benefit-risk, 24 including comparisons to risks in other generation modes 25 because nuclear power should not be compared to zero risks.
1 234 e
f\\
1 Every other generation method has its risk, too.
2 What we are saying is there is a balancing off, a balance of 3
some kind that does not go as low as practical or all the way 4
down without considering the cost or the increased cost of 5
electricity.
6 My last comment is that the importance of the cost 7
of electricity becomes more and more apparent to us.
Sam 8
Schurr and our economists were studying basically the 9
catalytic effect of availability of power since about 1900 10 until now.
It is quite evident the work horse in any 11 industrial society is electricity.
(
12 There were two periods in the past where the GNP has 13 grown almost directly with the growth of electricty in those 14 periods, but there were different slopes, first the 15 introduction, and second the period of wide application-cf 16 motors and dispersal of the industry throughout the United 17 States, and we may be at the beginning of a third because of l
l 18 the availability of better measurements, of controls, of being l
~
19 able to deliver energy in five or six different forms at the i
20 right place in the right quantity for exactly the right period 21 of time.
[
22 There are whole new processes emerging, and we are
("'g 23 doing quite a bit of research there.
The industrial potential b
24 and the jobs in the United States will depend upon our being 25 able to deliver competitively-priced electricity to move these
235 b
i 1
new techniques into production.
So our concentration is 2
soming more and more on costs, and as a consequence, we think
'3 that in the overall objectives -- let's call them primary that are NRC should have some reflection of the 4
goals 5
importance of balance of safety versus risk, of safety versus 6
economic costs.
It has to be made and it should be a goal to 7
look at it.
8 Those are the primary goals.
I listed another set 9
that are more temporal.
I call them secondary goals but they 10 could be objectives or five-year planning objectives.
One is 11 to stabilize regulation, and that may take 10 or 15 years, but 12 unless it is stabilized, I think the industry won't take off,
)
13 and as part of this, one must cut the time for construction.
14 And it is even possible that the NRC should schedule 15 a rigorous schedule -- three years, two years to license 16 and figure out what is needed, and that the schedule become a a
17 controlling element as'during the period of the environmental 18 impact statements when the AEC became responsible for 121 19 rewrites.
They would schedule one every three months -- I'm 20 sorry, one a week, and completion within six months.
21 The schedules were roughly met, but the importance 22 of the pressure to complete the rewrites was there.
I
(~%
23 therefore suggest that the whole process of scheduling be
(
24 taken as a primary function and adhering to schedule within l
l 25 NRC.
Obviously, the establishment of safety goals is l
l
236 O
1 important, and with those, once established, and they will 2
probably have some numerical value, the use of hierarchical 3
ranking of importance of safety measures or of incidents that 4
occur in operation of what is important in construction 5
-inspection can be illuminated by probabilistic risk analysis 6
-- failure mode and probabilistic risk'so that we do not waste 7.
a lot of time and a lot of money on issues that are of no 8
consequence.
9 With the numerical safety goal, I think this 10 hierarchical ranking system by PRA offers a little bit of 11 absolute advantage, b u't its importance can be exercised even 12 if a numerical -- its practice can be exercised even if a 13 numerical safety goal is not established.
The relative 14 position of a failure sequence versus another.
If a 15 particulan set of valves fail in the secondary system and its 16 probability of interference or carrying to, say, a core 17 damaging condition is 10 to the minus 3,
then immediate action I
j 18-should be taken, but if the analysis says the probability of l
l 19 engaging the core is 10 to the minus 5,
then nothing should be l_
20 done for-a while and the process s h'o u l d not be interrupted, j
21-the completion.
l l
22 As you know, the French now use this system in l
l 23 construction control as well as operation control They rank i
if the probability is 10 to the minus 4 for having a 24 25 serious accident, they let the construction go on.
If it's 10 l
i
237
]
,j/
1 to the minus 3,
they fix it right away.
Under no 2
circumstances do they stop construction, however. They just 3
reverse schedules.
4 MR. LEWIS:
Could I just explore that for a moment?
5 In one of the NRC Staff documents, there is a similar 6
statement that at 10 to the minus 3,
we fix; 10 to the minus 7
4, we think about it; 10 to the minus 5,
we let it go.
That 8
sort of thing.
But you know, if you start applying these 9
things to individual plants, there is a range of uncertainty, 10 part of which is the differences among the plants, and I 11 wonder whether I should worry as much if one plant out of 100
[V h
12 has a 10 to the minus 3 probability of core melt and the other 13 99 have 10 to the minus 5.
Then I'm still talkir.g about 1000 14 years before the first core melt in the industry.
Why should 15 I worry about that?
16 I'm only asking this as an instructive question for 17 me.
18 MR. CULLER:
I have not thought about this integrative effect, and I will do so, and we will talk about 19 20 it sometime.
My own feeling is that eventually each of the 21 plants, even during design, will use PRA as a method of I
22 deciding what levels of safety to build.
l 23 MR. LEWIS:
That's right.
f D l
24 MR. CULLER:
With the individual plant PRAs, then f
25 the measure of what an incident of a delay or a bad QA might
r i
\\
o
/
238
/~T L'
'mean in involvement in serious accident would be very real,~
s-5'-
2 and I would;ddge it without thinking abaut it on that plant
/
3 'l.alone./
s 4
MR. TAYLOR:
If I were the chi ~f executive officer e
5 responsible for that plant with 10 to the minus 3,
I would 6
want td 'do something about it.
)
7 MR. CULLER:
That's the way I would react.
\\
V 8
MR. LEWIS:
I know that's the way everyone comes
\\
9 down.
i
-i 10 MR. TAYLOR:
Simply because I would say my chances 11 of losing my investment, being brought to the edge of t
( )
1 '2 bankruptcy, appear to be 100 times greater, and my assooaates-
,13 and I cannot afford that difference.
I'm going to do 14 something about it.
s 15 MR. CULLER:
Or. involving a suit.
a 16 MR. LEWIS:
I'm trying to distinguish probabilsties N
3 17
-- this'is the wrong forum for this debate, but trying to 18 distinguish probabilities which we fix in our heads by s
19 automatically multiplying by 100 for the whole population, and 20 probabilities that we apply to a single plant.'
If the g;
I'y
\\
21 objective is.to not have a core melt for 1000 years or 22 somethang like that, then you can tolerate a plant or two in s
x
)
23 10 to the minus 3.
{V 24 MR. TAYLOR:
But the person who is responsible z
25 for that investment can't tolerate it.
239 1
MR. LEWIS:
Even the chief executive, if you know 2
it's not going to happen for 1000 years, then it's somebody 3
.else's watch.
I'm not arguing for outliers, but I'm also 4
being told my time is up.
5 MR. CULLER:
I think it's a legitimate question, and 6
I think thinking of it on a more global basis is something I
'7 have really never really talked to anybody about.
I know what 8
my reaction would be.
I would fix it.
9 MR. TAYLOR:
One other comment.
Getting that 10 signal, 10 to the minus 3,
the next question you have to ask 11
-is what caused that 10 to the minus 3,
and inevitably you will 12 find something you are going to take care of.
(
13 MR. LEWIS:
When-you say 10 to the minus 4 plus or 14 minus a factor of 10, you are thinking some of those are 10 to 15 the minus 3.
For peace, this is the wrong place.
16 MR. CARBON:
In concern for time 17 MR. CULLER:
I just have two other quick 18 suggestions.
I already mentioned scheduling as an issue.
I 19 think it.is possible within NRC and under the direction, 1
20 perhaps, of some instructions and questions from ACRS, to 21 increase the level of coordination and oversight if the 4
22 Licensing Board, Appeals Board and the local site reps who 23 rule; and since-I am on advisory committees in almost every 24
.one of the involved agencies here, I find that the 25 coordination between agencies, and certainly within NRC,
240
{}
1
(_/
1 leaves a lot to be desired.
2 The lessens from the licensing boards are not 3
properly taken care of, and I suggest this.
No other 4
important secondary goals.
I think now if you are willing to 5
let John start off on the technical issues.
j 6
MR. CARBON:
Fine.
Any other questions before we do 7
so?
l i
8
[No response.]
I 9
Fine.
Go right ahead if you will, John.
10 MR. TAYLOR:
My assignment is to go over the 11 technical issues which you have proposed and comment on
()
12 them. The first one is simplification.
Simplication is a goal 13 which we have set in our own advanced light-water reactor 14 program as, if anything, a primary goal, so per se, I
15 certainly would support ways in which the NRC could aid and 16 abet and encourage that movement.
17 I'm somewhat' concerned at the way the statement is 18 presented, however, in that it seems to emphasize the NRC in 19 some way engaging in the design process to achieve 20 simplification, or in a very direct sense through regulatory 21 actions affect the design process.
22 One element of simplification is the regulations s
23 themselves', and I would think that that would be a foremost 24 consideration in a simplification program for NRC, that they 25 would evolve ways in which the regulatory requirements were
241 O
(s,/
1 simplified.
2 I think we all feel -- and I think this is shared by 3
NRC today, for example -- that the technical specifications 4
which the operators abide by are far too complex to achieve 5
effective operational safety, so I would think that would he s 6
primary emphasis.
7 MR. CARBON:
Excuse me, John.
I think the thrust 8
here really is to simplify the regulations, the tech specs, 9
the requirements which NRC imposes, with the thought that that 10 then would allow the designer to do things.
11 MR. CULLER:
Going more in the direction of criteria 12 and standards to be met by the industry?
13 MR. TAYLOR:
My judgment is the designer has to take 14 some steps himself a l m o s t.
independent of the regulations to 15 achieve significant simplification.
One very important 16 element in our own efforts is to achieve design features which 17 put substantially less burden on the operator, independent of 18 how the tech specs are written, less burden on him.
19 MR. CARBON:
But this is aimed at fewer 20 requirements, simpler requirements, simpler regulations.
21 MR. TAYLOR:
That's good.
Then another point of 22 comment is the goal is stated to improve the safety of 23 existing as well as future plants. Ford has alluded to this in gs 24 his earlier discussion.
We have found our ability to move 25 ahead for future plants has been substantially improved by
242
(\\,,
1 separating the work we are doing for future plants from the 2
work we are doing on present plants.
3 My own judgment is that the present plants are 4
adequately safe.
I will make that with a qualification that 5
continued in-depth PRA avaluation will be appropriate to be 6
sure there are not some of those outliers that you mentioned, 7
Hal In my judgment, if they are found, they should be taken 8
care of.
But to set a goal for continued improvement of the 9
existing systems is going to militate against what I think is 10 an essential first step, and that is achieving stability, 11 which Floyd mentioned to you, distinct from moving ahead and
(
12 saying there are things that can be done to improve the' safety w
13 for systems that are built in the future.
14 MR. WYLIE:
I agree with what you have said.
The 15 original thrust of this as far as existing plants go was that 16 we felt the NRC, and certainly the industry, with the 17 industry's help, should look at what has been imposed on i
18 existing plants with the idea of relaxing the requirements l
19 where they have been overly complicated, and simplifying the l
20 operation of the plants.
That was what that was intended to 21 do.
22 MR. TAYLOR:
I think we also feel there are elements 23 discussed later in your issues of conservatism which was 24 decided upon years ago before full knowledge of the phenomenon l
25 had been identified, which could be eased, and out of that i
l
1 243 O )
i easement a better balance could be achieved.
- Again, s,
2 operability of the systems, some conservatisms could be 3
achieved on these present plants to be benefit of giving the 4
operator a less complex task to perform and still have full 5
assurance of safety.
6 That doesn't improve the safety of the system per 7
se, and what I am suggesting is if in a broad sense the 8
industry is in a continued process of changing in the interest 9
of improving safety and so on, we will n o 't get stability, as 10 distinct from letting the present regulations sit in a stable 11 form, being flexible in terms of allowing an individual
[ h 12 utility to present a case in which he has the data and the d
13 justification to make a change which he believes will satisfy 14 the issue, much like, in fact, we thought the process was 15
-evolving in earlier days in setting the General Design 16 Criteria, the various appendices, to provide guidance to the l
17 applicant as to what would be satisfactory to meet,the 18 regulatory requirements, but also allowing for a case to be l
19 presented differently approached, which would also be 20 satisfactory.
21 We seem to have lost that additional case to be 22 presented and are geared in only on the formally specified i
23 requirements.
Appendix K,
for example, today is being l
l 24 addressed on a case-by-case basis with some plants.
My 1
25 thought is let's not talk about changing Appendix K for the i
~r, 244 (D
\\s'/
1 present plants, but let us be willing to listen to 2
case-by-case justifications of improvement, for whatever 3
reason, and worry about changing Appendix K and other things 4
like that for future plants.
5 MR. CARBON:
I have a kind of basic question.
Much 6
of.the thrust of this really was in the first clause in the 7
second paragraph, that a likely thrust would be for the Staff 8
to initiate basic system study to review which safety 9
requirements are really important and ease off or eliminate 10 regulations for those that are not, with the thought that in 11 fact'perhaps to question whether a lot of unnece.ssary 12 regulations have been imposed.
13 Would it be quite helpful and beneficial to existing 14 plants if the Staff did carry out what would be basic systems 15 study to see which ones are really important and try and get 16 rid of a lot of them that are not?
That actually reflects up 17 to the last sentence in the first paragraph.
It was the 18 belief that perhaps that would lead to added safety, whether 19 that was your goal or not.
20 But I guess you are saying that you are not 21 particularly interested in seeing that done.
22 MR. TAYLOR:
For the present systems.
And the 23 reason for that is that what is badly needed in this industry s-24 is the stabilization of the regulatory process --
25 MR. CARBON:
Even if it means fewer -- even it it
245 1
meant removal of some regulations?
s-2 MR. TAYLOR:
I suggested a case-by-case evaluation 3
could be made because certain utilities can achieve benefits 4
from some simplifications and easements, but if you do it on a 5
hroad basis, that means your requirements are changing.
Every 6
utility will have to move.
7 Now, if those requirements are primarily of the kind 8
that make it easier for the operator to operate the plant, I 9
am sure it is fundamentally acceptable, but many of them will 10 reflect themselves in the need to change equipment, to change 11 many procedures, and change itself has its element of cost, fI 12 confusion, inadequate training to keep up with it, and so on.
O 13 My own judgment is it is best to see if we cannot 14 stabilise with the present systems on the judgment that they 15 are adequately safe, still scrutinising to look for those 16 outliers and taking action there and concentrate on the 17 improvements for the future systems.
18 MR. WYLIE:
One effort has already been started, and 19 that is to simplify the tech spec.
20 MR. TAYLOR:
That one certainly has great merit 21 because it bears principally on operator failure. If you can 22 define it with that distinction, I find it a lot easier to
/~%
23 agree.
24 MR. LEWIS:
For the others, there is some position l
25 you don't want to change the laws, you just want to get more l
l i
246 s
1 lenient judges?
q, 2
MR. TAYLOR:
Don't want to change the laws.
Would 3
like to see the flexibility that we believe was built into the 4
system originally.
You have got a prescription.
If it is met 5
to the letter, it is satisfactory.
If it is not met to the 6
letter and you can justify technically a different approach to 7
achieve the basic safety, then allowance for that is provided, 8
not that the safety be reduced.
No way am I suggesting that.
9 Now, what we have learned heavily in the industrial 10 side I mentioned earlier, that the risk of the loss of the 11 plant investment puts a higher demand on keeping away from 12 severe accidents, which is a fundamental safety issue on the g
13 utilities and the public risk itself.
14 MR.
LEWIS:
You think all utilities are aware of 15 that at this point?
16 MR. TAYLOR:
I will just echo what Floyd says.
In 17 an industry with as many people, variegated background, 18 experience, there are very perceptions of any given 19 generalization.
20 MR. REMICK:
I guess I am still a little bit 21 confused, John, on where you draw the line.
You~ favor the 22 tech spec change.
I certainly do.
You don't Appendix K.
Yet 23 to me, Appendix K is very specific, very proscriptive, and it 24 seems to me if you consider it on a case-by-case basis, it is 25 going to be pretty tough and you are sure subject to people
247 Ok,)
1 litigating it because it is fairly proscriptive and they can 2
say you are not meeting it.
3 MR. TAYLOR:
What you are gaining, as I see it, is a 4
~ stable statement of regulation.
5 MR. REMICK:
That I can understand.
You want to 6
stabilize rather than add confusion.
That I can appreciate. I 1
7 don't know how we pick out things. That is okay.
~
8 Source term.
I assume you would favor relaxfng 9
source term, and tech specs you would.
10 MR. TAYLOR:
Certainly.
11 MR. REMICK:
There are certain things you would like
[
12 to see reviewed, but others not touched, and I'm not sure how V) 13 we know where 14 MR. TAYLOR:
The source term work we are doing is 15 being done -- at least from our own viewpoint -- with two 16 primary objectives.
One is long term. It is the true 17 understanding of the public risk associated with severe 18 accident. If it could be accomplished and was shown to involve 19 a lower level of risk, that would be a tremendous step 20 forward, in our judgment, in this industry.
21 The second is short term, and it is to make more 22 reasonable the emergency requirements.
Would not affect the
-w 23 plant.
Some people may wish to use the source term to make
%J 24 changes on a plant.
My own judgment is that I would not 25 pursue that.
Simply use that information to get a more
248 1
balanced and reasonable emergency planning regime for nuclear 2
plants.
3 MR. REMICX:
I still don't know what you are 4
suggesting of what ones that we look at to see if there should 5
he changes and what ones we should best leave alone because of 6
stability.
It seems like it.is not clear.
7 MR. TAYLOR:
Think of
$100 billion power generator 8
the utility is trying to run, and I am suggesting that we 9
stabilize requirements which would cause him to have to change 10 that plant.
11 MR. REMICK:
I see.
It might be changing existing 1'2 plants.
13 MR. CARBON:
Your top priority is stability above 14 all else.
15 MR. TAYLOR:
That's right.
16 MR. CARBON:
Then simplification is great if it 17 doesn't affect stability.
I 18 MR. TAYLOR:
One thing that has influenced me in 19 this judgment comes from our recent experience in this 20 advanced light-water reactor program.
When we proposed it, we 21 were rejected by our utility advisory committee who must pass 22 on all our programs.
They rejected it for several reasons, 23 but a dominant reason was we have no idea what regulations we 24 have to build a new plant to, so why should we waste our money i
25 trying to design one?
2
l 249
(N
\\ss 1
In response to that, we undertook a program of 2
working with NRC to develop a methodology of closure of open 3
safety issues, taking a few specific cases to go into in depth 4
in terms of closure, to have a methodology and then specific 5
cases where the issues could be or would be closed.
That 6
program has been amazingly successful, from my viewpoint.
We 7
did develop a methodology. We found some 650 issues that were 8
open, many of them trivial, but nevertheless open, and we got 9
agreement from-the Staff that using this methodology, we could 10 close over 500.
11 Now, it is 550 of those issues for future systems.
(
12 When we got three-quarters of the way through that program, 13 that rigid attitude on the part of our advisories began to 14 melt visibly.
They said: you know, maybe there is a chance 15 that we can know what we are going to have to design and 16 license a plant in the future to, and subsequently we got 17 approval of that.
That's the dominant thing in their minds.
18 MR. SAUTER:
Can I take a shot at responding to your 19 question, Forrest?
I think you have to think in terms of at 20 least two kinds of stability.
For new plants, I think you 21 think in terms of stability of the requirments, stability of 22 the regulatomy process.
Perhaps simplifying the regulations 23 and the regulatory procedures would be a way to approach that.
24 Also, I think in terms of the present plants, you 25 have to think about the stability on somewhat of an individual
250 1
plant basis.
If you design and are operating a plant 2
according to one set of requirements, now somebody comes along 3
and, in an attempt to stabilize and simplify the requirements, 4
presents you with another set, which means you now have to do 5
things differently than you have done before, that's an 6
unstable situation used in that plant.
7 MR. LEWIS:
You are right, that's a different 8
issue.
That's the backfit issue.
9 MR. SAUTER:
I think in terms of stability, we are 10 talking about both of these issues.
11 MR. LEWIS:
I think we shouldn't.
I think we should
(
12 keep them separate.
I think that's what you are saying.
13 MR. SAUTER:
I think what I hear John saying, and I 14 agree with it if you are attempting to set up a different set 15 of requirements, a simpler, more stable set of requirements 16 for new plants, which in turn then get applied to some of the 17 present plants and it means they have to do things 18 differently, that's an unstable situation for them.
19 So I think if you are going to do that, you at least 20 have.to offer the present plants some kind of grandfather.
21 Either you can choose to go with these new ones or you can 22 choose to stay l
23 MR. REMICK:
Isn't there some risk of that in the
(
24 tech speo thing that out of that might come some plants are 25 going to have to do something slightly different?
I don't l
l l
251 i's,)
1 know.
2 MR. TAYLOR:
Yes.
3 MR. REMICK:
And toch specs, hopefully everybody 4
will benefit, but out of it I can see some plants just might 5
have tech specs that don't meet the new thinking.
They are 6
going to kind of have to have backfit procedures reporting or 7
something like that 8
MR. SAUTER:
I don't think there is any system that 9
can guarantee somebody isn't going to have to do some 10 changes.
The name of the game is to try to minimize putting 11 somebody at jeopardy: that I have gone out of my way to comply a
12 with this set of requirements, and now you are laying 13 something else on me.
14 MR. REMICK:
1.was trying to say what do you people 15 think are good things that should be done and other things 16 that should not be done.
l 17 MR. CULLER:
You have raised a good question, and I r
l 18 think, again, it is one of those that deserves some closer l
19 attention.
Certainly the simplification of operating l
20 requirements and the back away of some of the margins and all f
21 would make it easier to operate.
22 MR. TAYLOR:
Let's take another example, seismic I
23 requirements.
We have identified, and others have, too, a
(
24 substantial number of what we judge to be overconservatisms in 25 piping design to meet seismic adequacy.
I am not proposing a l
252.
~ s_)
1 change in the basic approach that regulates the seismic issue 2
at this juncture.
On the other hand, people have approached 3
the Commission and said in this instance, with this new 4
knowledge, I can meet the safety requirements and I won't have 5
to put the snubber or this pipe support in, and that, to some 6
extent, has been accepted.
I see nothing wrong with that 7
approach as distinct from saying, hey, we are going to have a 8
whole new set of seismic requirements for the present systems, 9
most of which.have already been. built, and in order to make a 10 change to " improve" that, would entail substantial cost and 11 change.
[VI 12 MR. LEWIS:
You know, John, in the case I don't 13 want to dwell on the seismic one.
The argument, as I 14 understand it and even believe in it, is that removing the 15 snubbers will make the plant safer.
It is not just a matter 16 of relaxing a regulatory requirement; it is that the 17 overstiffening has produced a degree of vulnerability that was 18 absolutely unnecessary.
It was just a mistake.
19 MR. TAYLOR:
I don't think that we have reached the 1
20 stage that we can truly quantify that we are making it safer i
21 or not.
There may be some 22 MR. LEWIS:
1 think you can.
This debate we should 2
23 have in another forum, but I think that fact can be quantified f-w b
24 in a number of ways, but let me just turn to another question.
l l
i l
25 MR. TAYLOR:
I have heard the arguments, namely, if l
l
253 oO k) 1 you don't have as many obstructions in the pathways for m
C 2
equipment maintenance and surveillance and so on, you are 3
going to make the plant safer by some elimination.
That is 4
qualitative judgment.
It may well be true, and with time, it 5
could reflect itself in a future set of design requirements.
6 MR. LEWIS:
Again, I do think that the evidence is i
7 quite clear that flexibility of piping makes it more resistant i
~
8 to damage than rigidity.
9 MR. TAYLOR:
Rigidity is coming, as you well know, 10 from pipe whip considerations, from blowdown considerations, 11 and there is a delicate balance that has to be --
()
1'2 MR. LEWIS:
Let's conduct this technical debate at 13 another time.
But another thing that troubles me a little bit and I think it is clearing up a little bit now, but there 14 15 has been some confusion about stability of regulation of 16 existing plants and stability of the regulatory process.
I I really want to be sure that I 17 think I heard you say, John that you would like to see future plants licensed 18 did 19 according to existing regulatory practice, even though we know 20 more, just because that is something that we know now.
21 MR. TAYLOR:
I didn't say that.
I said just the 22 opposite.
I said we should separate this issue of changing, 23 even improving regulation with concentrating those O
24 improvements on the future plants, just as we are trying to do 25 that in our own work, and we should be very leery about
(
254 1
changing for present plants.
2 MR. LEWIS:
Let me then just reemphasize my 3
inability to understand because, in terms of what the utility 4
4 executives are talking about, nobody wants to, I am told, 4
5 order a plant now because he has no idea what the regulatory 6
requirements will be on new plants.
The only way he can know there are two ways he can know.
One is that NRC can 7
8 formulate absolutely out of the whole cloth a set of 9
regulatory requirements for future plants based on the fact 10 that we really do know a lot more now than we did when the 11 existing set was developed.
(
12 The other possibility is they can say, fellows, you 13 know, even though we know more, we are not going to take it 14 into account, and this is the set you are going to have to 15 design to.
That would give you stability but at the price of 16 using --
17 MR. TAYLOR:
That I have not asked for, Ha1.
In 18 fact, in our work with NRC on our advanced plant, they have 19 agreed to some changes in Appendix K for " future plants."
For 1
20 future plants.
And we have elicited that agreement.
We 21 wanted it.
And that is just one small example of what we hope i
22 will occur as we go.
23 MR. LEWIS:
The concern that came up here was that s
24 if you do establish new requirements for future plants, 25 somebody will want to make them retroactive to existing
255 (s) x_/
1 plants.
2 MR. SAUTER:
Let me give an example.
Suppose out of 3
whole cloth for new future plants, we could construct a set of 4
regulatory criteria which were performance oriented some way 5
or another as opposed to the present proscriptive one.
That 6
might be quite satisfactory and, I think, perhaps a better way 7
to go for future plants.
But if in doing that they were then 8
en masse applied to the present plants, I could see where that 9
would result in some major backfits.
10 On the other hand, perhaps the way to go in order to 11 get the price of getting that better system for future plants
()
12 would be, as Floyd mentioned, to run temporarily a two-track v
13 system of proscriptive set of regulations for present plants 14 and a performance-oriented set for new plants.
?
15 MR. LEWIS:
There is plenty of regulatory experience 16 with grandfathering things that are licensed under 17 regulations.
18 MR. CULLER:
I think that is what we are basically 19 saying.
20 MR. SAUTER:
We need a grandfathering.
21 MR. CULLER:
Not exactly grandfathering, but a very 22 careful preservation of a stability for the existing plants.
23 If there are real challenges to safety, then changes have to s
(
24 occur, or if there are great simplifications in operations or 25 such, maybe that should occur.
But the new plant criteria
256 b
\\m,/
1 certainly should not be retrofitted onto the old, and that is 2
basically what I think John was saying.
3 Forrest, you asked a question, and we will talk some 4
about it.
If we have a better, finer differentiation with an 5
existing plant study, we will try to write it up and send it 6
to you because it is a good question.
7 MR. CARBON:
Let me interrupt and urge moving on.
8 MR. TAYLOR:
My own judgment, we have a substantial 9
defense-in-depth.
We will certainly in the future systems 10 examine an issue.
I don't.think it needs extensive 11 reexamination for present plants.
Transition from I h 12 deterministic to probabilistic regulation, here again we are U
13 faced with the same dichotomy.
14 I think in the. future there would be great merit in 15 moving toward more probabilistic regulation approach.
I don't 16 see giving up deterministic regulation fully incidentally in 17 that process, but more of a balance, more of a utilisation of 18 probabilistic assessment in the regulatory arena for the 19 future.
20 However, again, for the present systems, I suggest 21 that the work that is done now to use PRA as an evaluative 22 tool separate from the regulatory process itself is valid, and 23 it has already shown itself to be valuable, particularly in 24 finding those outlyers.
Beyond that I don't think I would go.
25 Any comments?
257 (O'
x_/
1 (No response.)
2 That's covering Number 2 a lot faster than Number 1.
3 MR. CULLER:
That's known as " front-end interest."
4 MR. TAYLOR:
Quality assurance.
I think that is a 5
job the industry has got to take care of.
In many respects, 6
the quality assurance specific plans and standards which have 7
been established have been established by the industry in one 8
form or another, and the difficulties that have occurred have 9
been the inability to in very real detail follow through on 10 what was established.
So it's up to the industry to square 11 that away, as I see that, and I don't see how a regulator can.
)
12 Where it seems to be, of course, of greater 13 difficulty is in the area of construction, and I think a very 14 major revision is going to be required.
It presumably will 15 primarily have a benefit for the future because it will 16 require major revision.
17 MR. LEWIS:
It's not only in construction, John.
We 18 just came up from having heard of a trip on a plant that was l
19 generated by somebody using a voltmeter on the ampmeter scale, 20 and this is the same plant that did the same thing a couple of f
21 months ago by somebody using a voltmeter on an ampmeter 22 scale.
Why does it happen twice in a row?
Quality control on 23 people is just as important as on 24 MR. TAYLOR:
Oh, yes.
I mentioned construotion 25 because that was a big visible problem for the industry
1 258
,~
1 entailing enormous costs.
But you're right, it exists in s,
2 every element of work for these operating plants.
Maintenance 3'
is a very important one.
4 MR. LEWIS:
I can't understand how a plant 5
management could let a major event like the first cne occur somebody pumped a 6
and then not take steps to make sure 7
different system.this time, but by the same process.
It seems 8
to have been overlooked.
I don't understand it.
And that's 9
an issue of quality control.
10 MR. TAYLOR:
But who is going to fix that?
It's up 11 to the industry.
And when I say " industry," I am not saying
(
12 just the individual utility.
We have a very big program today 13 in maintenance technology where we are trying to describe the 14 maintenance processes effectively.
We've set up a maintenance 15 equipment application c e n (e r in Charlotte to bring people in 16 to train them on systematic preventive maintenance processes, 17 inspection and so on.
18 INPO is very active, of course, in establishing and 19 following standards and maintenance.
NUMARC has become 20 extremely active.
Those activities are going to be the ones 1
21 that solve this problem if it's going to be solved.
I can't i
22 see how a regulator can, in fact, solve the problem.
23 MR. LEWIS:
I can't either, and that's why I think O
l 24 the responsibility, I agree with you, is on the industry.
25 I was very dismayed by this particular experience in i
i a
--,-,-~n-,,
259
\\m/
1 which apparently nothing happened between the two repetitions 2
of exactly the same state.
That responsibility was at the 3
plant.
We will hear more about it.
I don't want to play that 4
one here.
But that's really not the sort of thing that is 5
resolved by a large program to quantify the best way to do 6
maintenance.
It's resolved by attention to very simple 7
things.
8 MR. TAYLOR:
I am not suggesting that our program is 9
a total solution.
I am saying the solution takes the 10 individual utility and centralized activities of this nature 11 to be fully effective.
12 MR. CARBON:
Is it your belief that on the 13 construction side, where we have had these bits of QA problems 14 showing up way late in the game, that industry will be able to 15 handle the problem, that they will effectively take care of 16 it?
17 MR. TAYLOR:
My judgment is that we can essentially 1
and a little bit of this what I have suggested 18 go back 19 work has started now; it's going to take so long it's
- 20 certainly not going to bear very much on the present plants 21 a project which I call " TRACE," to look at the QA and 22 construction difficulties that exist and trace them back to 23 the setting of the standards in that particular train and 24 reexamine those standards establishing their technical 25 justification, make them unambiguous so an inspector who
260 O-1 cannot be expected to have the full knowledge isn't put in a s
2 position of judgment as to what is correct, and make sure 3
they're practical in terms of construction and installation 4
capability.
5 You know, there is a myriad of such standards in the 6
wwhole setup which will have to be changed.
So my thought is 7
that that process slowly, carefully, systematically has to be 8
reexamined and reestablished for the future.
9 Now, having done that, the thought which has been
~ ~
10 very much promulgated of having the regulator in some sense 11 sign off at various stages in construction rather than waiting 12 until you're already done, will have to be reintroduced into 13 the picture, for obvious reasons.
14 If there is a problem, it should be identified early 15 in the game.
As you well know, the industry is strongly in 16 the terms of future simplified regulation for one-step 17 license.
Such an arrangement is only practical if a very 18 thorough checklist is established which, in effect, assures 19 the regulator that what he was promised would be built has i
20 been built, and that can be used to apprise the regulator as 21 you go that you're building the plant the way you said you 22 would.
23 MR. CULLER:
My answer to your problem, and a little 24 bit to Hal's, is I suspect if the industry doesn't do it, it i
25 won't get done.
261
(' %
\\ss 1
MR. CARBON:
I would agree with that.
2 MR. CULLER:
As a consequence, the enforcement of 3
the responsibility I think the industry takes to heart.
4 Now, exactly how it gets done and how rapidly is always a 5
question with 55 groups involved.
Nonetheless, I know that 6
it's taken seriously, and they are looking at ways of keeping 7
real-time track of all the inspections, so at the end of the 8
day they can say, "This was inspected, and here it is on the 9
scope.
I know that it was done."
I believe there is real 10 effort being made among the leaders, and the leaders will, one 11 way or another, through INPO and other things force a more O
(
j 12 uniform approach.
%d 13 There are other techniques, some that the Japanese 14 use.
They set up one team to do the same thing around all of 15 their circuits, and it's conceivable that these collective 16 opportunities for the industry to act together, set up a 17 quality control system and group that does work for every one 18 of the utilities may be one of the solutions.
19 I don't believe it will be long in coming, but there 20 will always be mistakes like you had, and some of the answers i
21 to that kind of a mistake is you make such a technique of 22 jamming in two electrodes into an instrument, a nonnecessary 23 inspection.
s 24 MR. LEWIS:
You give a guy voltmeters instead of 25 multimeters is one simple solution.
l l
l
262 O
'N-1 MR. CULLER:
There are simple solutions to keep 2
repetitive mistakes from occurring.
3 MR. LEWIS:
There is a problem with even a team 4
going around doing the same thing, because if that team makes 5
a mistake, it can make a lot of mistakes.
6 MR. CULLER:
I am sorry.
You know about party i
7 circles and the business of audit.
There has to be an audit.
8 The teams can't run unaudited.
I think a hierarchical 9
auditing from an industry group or whatever is being done is 10 necessary in an audit by the NRC on what the industry is 11 maintaining is a necessary part of it to check on the quality 12 all the time.
(
13 But going in and doing it as part of a government 14 agency is not going to provide a solution any more than the 15 solution which the industry itself can provide.
16 MR. TAYLOR:
If our goal was to end up with perfect 17 performance, we wouldn't need defense-in-depth.
I think we i
18 need defense-in-depth so we can tolerate imperfect performance 19 at times, no matter how marvelous our system will be, mistakes 20 will be made.
21 And I don't condone repetitive mistakes, 22 particularly of the nature you described, Hal 23 MR. LEWIS:
I agree with most of what you have said, O
24 John.
My concern is really at a much more elementary level.
25 It really is true that if in our laboratories once a year at
263 p) t
\\s,
1 the beginning of the year some student will measure a voltage 2
with an ampmeter and pop it, and then it won't happen again.
3 It really won't happen again.
And I am disturbed about it major events at the same plant.
And, in 4
happening twice 5
fact, the last guy did it to both sides of the system.
He ran l
6 from one to the other and did them both.
I 7
MR. CULLER:
Was it the same man who did it thet 8
second time?
9 MR. LEWIS:
No.
The first event was some time ago.
10 This last event, the guy did it twice within the space of a 11 couple of minutes.
12 MR. CULLER:
Maybe he's still learning his first i
~
13 lesson.
14 MR. MOELLER:
You mentioned QA being decided by the 15 industry.
To what degree does the NRC paper trail help or 16 hinder the QA of industry?
j 17 MR. TAYLOR:
I wish I oculd say that we don't need a l
18 paper trail, because true quality doesn't come from the 19 paper.
True quality comes from the people who do their job, 20 who set the requirements in the first place.
21 My biggest experience in the past has been with QA 22 systems that have passed gloriously all the audits, and the 23 requirements haven't been right.
So you know they are doing N
24 their thing to the wrong requirements and we don't have the 4
25 quality we want, and they have to dig below the paper to go
1 E
l 264 i
/(J 1
after real quality.
s 2
But I can't conceive of a regulatory process that 3
does not involve paper.
We are stuck with it.
4 MR. CULLER:
It has to have tracking.
5 MR. TAYLOR:
Therefore, that paper should all out 6
requirements which are unambiguous and practical to achieve, 7
because the regulator must check against those requirements 8
rigorously.
You can't give excerptions just routinely.
9 MR. REMICK:
I don't disagree with anything you've 10 said about if we're going to have an effective improvement in 11 QA, industry has to be involved in making changes.
But aren't 12 there also some regulatory changes that have to be 13 considered?
For example, you talked about, I think, the need 14 for hold points or certification points.
Isn't there a need 15 to go along with some changes on the regulatory side?
16 MR, TAYLOR:
Very much so.
17 MR. CULLER:
'I think we failed to say something.
18 What we think needs to be encouraged in the 19 industry-regulatory interface is that the growing cooperative needs to be supp;amentary, complementary 20 effort 21 encouraged, with mor. and more being pushed toward the 22 industry because they s a *i e the hands-on control.
23 But there are certain criteria that have to be met
\\
24 by the industry, and there must be a strong audit function, 25 paper trails and other things like this.
There must be a way
265
)
.\\_/
1 of checking because you could not, or NRC could not, discharge 2
its responsibility without having some way of checking that 3
the performance is adequate.
4 But I do think that there are rather strong moves 5
being made slowly within the industry that I think should be 6
encouraged.
They are loosening up a little bit and making 7
INFO records available for looking at and things like this.
8 And as a consequence, with the ice beginning to break, I think 9
it's asking the industry to increase its responsibility to 10 some of the routine checks and monitoring, and actual 11 execution, particularly in quality control, has a very good
(
12 chance of working.
13 I don't think it should be prescribed exactly, but 14 to continue to urge that.the industry pick up more and more 15 and then to back the regulator out of the actual execution and 16
- to allow the audit function and check function to come more i
17 and more into play.
18 John, we might better go on because all the p
19 questions are fun to talk about, but I realise you have had 12 i
20 hours2.314815e-4 days <br />0.00556 hours <br />3.306878e-5 weeks <br />7.61e-6 months <br /> already.
21 MR. TAYLOR:
Number 4 is unquantified conservatism.
22 And we discussed that to some extent already.
Strong feeling 23 that these conservatisms should be quantified.
I think 24 through the NRC safety research program, our program, 25 suppliers programs, have generated a tremendous amount of
t t i
t 266 s
(\\s /
1 quantitative data to permit this to be'done in a logical, 2
effective way, and apply it in the regulation for future s
3 plants.
i I repeat myself again.
For pr es entp! ant s, my own
.4 5
thought would be:
use that~ knowledge in terms of case-by-case 6
circumstances as distinct from making major c h arig e,s in the t
7 regulatory prescriptions.
i
)
8 Any questions on that?
s 9
(No response.)
', 6 '
10 MR. TAYLOR:
5 is appropriate level is s a'l ty.
And 11 now we are into that important issue of safety goal which 1,
1 12 Floyd has mentioned.
I, too, would support his comment that 1,mportant part of the stabilisation process, in spite of 13 an s\\
\\
14 the hard difficulties of establishing it, is the safety goal t,
s 15 that is measurable and can be used to say, along with\\the t
h 16 appropriate cost-benefit evaluations, that there is no need to 1
i 17 make this improvement on this plant at this time.
18 MR. CULLER:
I have a comment here. I suppose,'maybe s
19 two.
One is ah observation and one is a comment.
i t
20 First off, the whole issue of informing the public i
+
21 of logical levels of safety needs to have a champion, and I i
I
's.
22 don't know whether it is the NRC who is supposed to develop 23 intelligible and knowledgeable statements about what levels of C
24 safety are required or not, but somewhere within the matrix of 25 thinking, somebody has to be a public interpreter, and,it 4
6 L..-
n 267 1
can't be done s
2 MR. CARBON:
Excuse me.
Let me just interject.
If 3
isn't NRC, could you suggest who it might be?
4 MR. CULLER:
We have talked a lot about this, and and that's the comment Iam going to make, I 5
one element 6
suppose, a thought -- Chauncey Star, as you know, sort of 7
gurued risk in a formal sense, and he's still doing it.
We 8
talked for three years about the desirability of having a data 9
analysis function where there are risks to radiation and risks 10 to toxics and other things like that.
That takes this i
11 horrible mess of data, somewhat as we did 30 years ago with
)
12 the cross-section data, and cleaned up the background.
13 And we've been talking with both Stanford and 14 Harvard and their medical and engineering groups about setting 15 up a comparative risk center, basically where the data are 16 analysed with care, and then the risks due to radiation as 17 compared to the risks due to smoking, sort of what Lord Walter 18 Marshall has proposed, the index, an understandable index, of l
a puff on Sunday is worth about a 19 "a
puff on Sunday" 20 hundredth of a millirem as far as causing cancer deaths I
L 21 and to begin to develop a feeling in each regulation that it l
l 22 has passed what the relative hasard that this regulation l
i 23 specifies in terms of risks that are understood.
s l
24 Now, such a comparative risk center has to be l
25 absolutely clean as a hound's tooth.
It can have no position i
l l
268 o
i 1
on policy or anything else.
2 And we have been discussing w i [h the Stanford group 3
for almost two years the possibility of doing this.
And there 4
is real interest.
And Kennedy himself is now involved.
5 One element of increasing public understanding is to 6
understand risk in a coaparative sense in real terms.
People 7
cannot understand 10 to the
-5 or 10 to the
-6.
I have talked 8
to my wife about this all of her life, and she says, "It 9
doesn't mean anything more to me now 40 years later than it 10 did when you first started talking."
And that's the truth.
11 As a consequence, I think we have to improve the
()
12 means of communication.
Nonetheless, I think maybe the NRC is 13 the only group in position to take on this educational 14 program, either that or the ACRS itself.
I was going to 15 comment on that at the end.
It's conceivable that your role 16 may be able to interpret what the NRC is saying to one form or 17 another, and that's not so easy to do, but nonetheless, it may 18 be helpful.
19 MR. REMICK:
Somebody has to be able to understand l
l 20 our letters.
21 (Laughter.)
22 MR. CULLER:
You have to be able to understand us.
23 MR. REMICK:
We can't write a letter that people 24 understand.
25 MR. CULLER:
It's true.
It's difficult in this
269
,_s 1
arcane language that we have.
2 Now, go ahead, John.
3 MR. CARBON:
Excuse me.
Could I ask a question back 4
here?
We keep using the words " safety goals," but we really 5
don't mean that, do we?
What I mean is this afternoon someone 6
commented, for example, that FAA, I guess, has safety goals 7
which are two or three orders of magnitude better than they 8
can achieve.
9 MR. LEWIS:
I don't think that was true.
He did say 10 it.
I don't think it's true.
11 MR. CULLER:
I don't think that's true.
We looked p
12 at that.
13 MR. CARBON:
What we really mean, isn't it how safe 14 is safe enough rather than safety goals?
15 MR. CULLER:
Yes.
16 MR. TAYLOR:
In fact, what was very impressive to me 17 about reading of the aviation safety goal is that they 18 originally formulated it and had it, in fact, in an amendment 19 in their regulatory framework, and realized as they attempted 20 to put it in practice, they were not able to measure 21 achievement.
And they changed it so they not only could 22 achieve the goal, but they can measure their ability to 23 achieve it.
And that's what we must have.
(
24 MR. LEWIS:
But the measure of their success is that 25 most airplane accidents nowadays are caused by pilot error,
270 O
i and that means they've done a good job on the mechanics.
2 MR. TAYLOR:
- Yes, And they still have more to do 3
with the pilot.
4 MR. LEWIS:
But there will always be --
5 MR. TAYLOR:
Automation of nuclear power plant 6
operation.
Our own trust here is not to seek out increased 7
automation, either for present systems or for future, but to 8
attempt to utilize modern computer technology, graphics, 9
ability to give substantially greater aid to the operator, 10 diagnostic aid, quickly provided.
11 Let him make the decision, let him be in charge of
()
12 the plant, but give him substantially more effective 13 information with which he can act.
14 MR. CULLER:
Let me expand a bit on that.
One of 15 the areas that we decided to get into with both feet three 16 years ago was artificial intelligence.
And we've had some of 17 the best guys in the nuclear division on it for a while.
18 We had our first application seminar for a lot of 19 things about three weeks ago, and I swear to God I was 20 impressed with what has come out.
There are devices being 21 proposed that will do such things that will assure you that 22 the reactor is shut down when it says it is shut down, and 23 there is no device now available that will do that.
24 MR. LEWIS:
Don't count on those either.
25 MR. CULLER:
Well, I don't either, but the chances
_-_ ~ - =-
l 271 1
of our providing information enough to the operator so that he 2
knows what to look for under circumstances, I truly think that 3
what John is saying, this business of providing the operator 4
with quick information through artificial intelligence is 5
going to be a great assist.
6 You may expand, but I was truly impressed with what 7
has been done in two years.
8 MR. LEWIS:
We must talk about that at some other 9
time.
10 MR. CULLER:
Why don't you come up --
11 MR. TAYLOR:
Floyd is talking about a lot of things 12 that are in the laboratory and not the plants.
13 MR. LEWIS:
My experience is the promises are 14 magnificent.
15 MR. TAYLOR:
The issue with utilizing these 16 techniques in the plants that are advanced, of artificial 17 intelligence, is that they're not familiar and they're not i
18 really trustd by the operator and he is very leery, therefore, l
l 19 of utilizing.
t 20 So one of our frontrunner programs for initial 21 application is a shutdown analyser.
It doesn't operate when
(
i 22 the plant is running; it is utilisable after he has had an 23 inadvertent scram to give him a more in-depth diagnosis and, 24 hopefully, faster, of what caused that scram.
25 We've gotten explicit desire from the utilities for
J 272 1
this because of the Salem affair, where the operator wasn't 2
quick enough to understand that he had lost redundancy in the I
3 few earlier versions of that particular incident.
The utility
)
4 ends up with a tremendous fine and terribly bad publicity.
5 They're still worried, does the operator have that extra 6
subtle capability to see through some of these seemingly 7
simple events?
We think an AI system, an expert system, would 8
be a substantial aid to him.
9 MR. LEWIS:
We will talk about
't h a t, but not on 10 Max's time.
11 MR. CULLER:
Come up and talk with us about that.
()
l'2 MR. TAYLOR:
If we can get the operator to use that 13 in a situation where, if he takes action or not it's not going 14 to hurt the plant, then get confidence in it, then we can talk 15 about putting something to help when he's running the plant.
16 MR. CULLER:
I have a personal opinion, too, about i
17 automation.
I favor the general policy of not forcing the
(
18 operator to make a quick decision under stressful conditions.
19 And I like the British and French automation, automatic l
E 20 control for half an hour.
It gives the operator time to 21 collect data and think about what happens before he has to 22 start reacting.
He can override the system.
The reactor has 23 an automatic control for approximately half an hour.
24 MR. TAYLOR:
My own version of that is we have in 25 the present system that type of automation, and in future
- - ~.
4 273 O
ksI 1
systems we would make it that that type of actuation is not 4
2 needed as fast itself, let alone operator action on top of it, 4
3 that the operators can, in fact, be more leisurely, along the 4
lines that Floyd has mentioned.
But this again would be for i
5 future systems.
6 MR. LEWIS:
This is a very important issue, of 7
course.
s 8
MR. CULLER:
We understand that.
I think John and I 9
agree pretty much, and we have both said about the same 10 thing.
I don't think we would automate a reactor for i
11 operation.
Incidentally, the high-flux isotopes reactor
)
12 during shutdown on the high flux and startup because it goes i
13 through a period of very limited temperature override, has an 14 automated takeup and shutdown.
The power density of the 1
15 high-flux isotope reactor is very high and the period is very i
16 short.
So when we built it in 1963, we automated.
We run it 17 on the computer.
We can override it in an instant, but the 18 computer takes it up and brings it down.
4 19 MR. REMICK:
It's not quite what we're talking 20 about.
But something I remember from a recent trip to Japan.
1 21 At the Fukusaama site, their Unit 4 at the Daini site is going I
22 to have a new control room, somewhat like nuclear but 23 different.
The Japanese version.
They have voice recognition 24 so the operator can stand there and ask reactor pressure, 25 reactor level and so forth, a nice digitized female Japanese j
274 O
1 voice comes back and tells him what it is.
He can ask for 2
equivalent of SPDS.
He can just ask the page he wents.
He 3
doesn't have to punch any numbers if he wants to see another 4
page.
5 MR. CULLER:
This is the kind of thing John is 6
talking about, operator aid.
7 MR. TAYLOR:
The control center on the mirror 8
machine in Livermore, if you've seen it, has nothing but touch 9
types which is there is not a switch or a lever in the whole 10 damn control room.
4 11 MR. REMICK:
You can just point to CDT --
12 MR. TAYLOR:
They have to touch it.
But that's 13 all It's beautifully laid out in color.
14 MR. REMICK:
There is a whole series of infrared 15 and the person just points.
You don't have to touch it, just 16 put your finger close.
17 MR. LEWIS:
That's not automation.
18 MR. TAYLOR:
No.
That's my point.
I don't know 19 whether we want to go above where we are in automation per se.
20 Safety research, I strongly feel that NRC should 21 maintain a major role in safety research.
Being in the 22 research side of the business, you can take that as you wish, 23 too.
I 24 But we feel that that provides NRC with the 25 independence of technical judgment that is essential to make
- ~ _, _.,, _,. - -
275 1
it a credible agency, and I would strongly support it.
I am 2
actually somewhat concerned that every time a budget cut is 1
j 3
taken at NRC today it tends to be taken out of the research 4
program.
1 l
=
5 MR. CULLER:
Hal, we went back and reviewed the 6
UGAVE(?) Report to a limited extent.
7 I think the assignment of roles in those two tables 8
that we put into that report are still valid.
There are 9
certain things the industry should do.
We assigned a role to 10 the Department of Energy, which I do not think it will pick up i
11 necessarily, and that's doing what we called at the time
(
12
" creative research," looking for new ways of bringing on 13 better safety.
14 NRC, because of its interest, maybe should pick up 15 part of that cudgel.
I am uncertain about DOE right now.
16 It's an important issue.
Most of yours is confirmatory, but 17 not exploratory.
I am sorry -- in NRC.
But I think that role 18 becomss'important.
19 It's also smportant that the industry do that, 20 particularly in the areas that John is going into in looking 21 at advanced reactors, The safety systems have to evolve a new 22 form there, but I think NRC needs that, and I recommend -- and 23 Jack brought along copies of the 1983 ERAB report that was 24 done both for NRC and DOE by a subcommittee of relatively good 25 guys, of which Hal wa a member.
We went over the whole safety
. -~. - - - -. - -..
1 276
/
1 program internationally and everywhere else, the safety 2
research program.
And the report is almost valid today; not 3
-much has changed.
So I recommend it to you at least for 4
superficial reading.
5 MR. TAYLOR:
,On e other comment.
I don't know that I 6
would word it, "should NRC do research to permit relaxation."
7 That is, they should be doing research in areas of uncertainty 8
with the objective position of reducing those uncertainties
(
9 and gaining greater understanding.
If they find in the 10 process of this, relaxation, fine.
If they find in the 11 process there is going to be additional restriction, okay, 12 that has to be applied, too.
So I would keep that as a pretty 13 neutral type of approach.
14 MR. CULLER:
I have another quick comment.
As we 15 clearly enumerated in this ERAB report, one of the difficult 16 things is to take advantage of research already done within 17 NRC, within EPRI.
There is a lot more known from ten or 15 18 years' worth of safety research than has been incorporated 19 back into the regulatory process.
And that'particular 20 mechanism within NRC needs' attention, and that's one of our 21 principal conclusions in 1983, and I repeat it now.
Taking 22 advantage of what research has been done is not evident within 23 the regulatory process.
24 MR. WYLIE:
Let me ask a question.
We just got 25 through going through the research budget for this year,
- ~
277
,~
f\\
1 and during that process this came up time and time again, 2
"Well, industry ought to be doing this research, maybe 3
industry should be doing this" -- you mentioned that a moment "that NRC should be doing certain things, and industry 4
ago 5
should be doing certain things."
6 Is there any active coordination going on between 7
industry and NRC to identify these areas where industry is 8
going to be doing it --
9 MR. CULLER:
We almost jointly plan.
10 MR. TAYLOR:
We meet with Bob Minogue, myself and my 11 senior staff and Bob and his senior staff, roughly every four O) t 1 '2 to six months, and we each lay out our existing programs, and 13 we try to avoid any place where there is duplication.
14 And very often we wil come up with joint projects.
15 We have quite a few joint, projects.
Each of us put money into 16 the contractor on an agreed-upon joint plan, set up a 17 management committee to run the project, and some of our best 18 projects are done that way.
19 MR. CULLER:
Yes.
20 MR. TAYLOR:
Because the guy who puts his money in 21 wants to use the results, so NRC wants to use them and we want 22 to use them.
Some of our best projects.
23 MR. WYLIE:
They are in, I call it, " aging it."
24 That's not exactly the right term, but it has to do with how 25 long we're going to run these plants with the equipment we
278 b
1 have.
2 MR. CULLER:
Longevity.
3 MR. WYLIE:
Yes.
That's one area the NRC was 4
spending quite a bit of money on, and they got cut back quite 5
a bit in this budget out.
6 MR. TAYLOR:
That's too bad, because we, too, have a 7
program there and, in fact, have been eliciting support from 8
DOE, in light of the reductions and so on we see on the NRC 9
side, because to be done properly it's a very expensive 10 in-depth program.
But we believe it's extremely important 11 from the utility viewpoint.
The ability to generate power out
[a) 12 of a system without having to put new investment is the most 13 important economics that we can take.
14 MR. WYLIE:
That has to be a joint effort between 15 industry --
16 MR. CULLER:
We recognize that.
We have a very 17 stable program and a relatively good-size one.
18 There is another reason that NRC should do it, I
19 think, and that is, sometime beginning around the year 2000 20 you have to decide on whether you're going to relicense 21 reactors whose license are up, and it's going to take ten 22 years to figure out what the status of the plant is.
23 So we have started researching measurement 24 techniques of trying to decide how much lifetime is left as 25 well as trying to decide what kind of maintenance to extend
279
/%
ms 1
to critical items within the plant that will be necessary as a 2
routine to keep those things that could deteriorate because of i
3 another element's failure from deteriorating.
1 l
4 So it's a broad program, and it certainly is going 5
to be important in relicensing and ultimately in the decision 6
to shut down, and you most certainly are going to be involved 7
in that process by 1990 because you're going to have to begin 8
to think about what you're going to do with the licensing, the 9
relicensing of reactors, or the shutdown of same.
10 MR. LEWIS:
I am not very well informed on this, but 11 I have always had the feeling that it will never be economic
' ("( )j 12 to shut down a major reactor.
13 MR. CULLER:
That's correct.
14 MR. LEWIS:
It will always be economic to fix what 15 needs to be fixed, to keep it going.
16 MR. CARBON:
Except maybe the pressure vessel.
17 MR. CULLER:
'We are looking at that, incidentally.
18 I don't know whether you want to discuss this or not, but our 19 own view is the lifetime of the reactors may be 80 or 100 20 years.
21 MR. LEWIS:
I would anneal the pressure vessel or 22 replace it.
An aircraft has an infinite license.
It really I
23 does.
l 24 MR. TAYLOR:
We discussed that with the aircraft 25 industry.
_... _ _ = -, - -
280
)
( s/
1 MR. LEWIS:
You just replace the parts that wear 2
out.
It's like George Washington's ax.
3 MR. CULLER:
One of the things that John is going to 4
do in looking at advanced light-water reactors is to consider 5
the complete replacement of every component within the plant 6
so that the balance-of-plant and all the concrete and other 7
things like that can go on and on and on.
So we're going to 8
try to lay things out to see what the cost effects and safety 9
effects will be so we can get at everything for easy 10 replacement and maintenance and begin to consider reactors 11 like dams at least or maybe reactors like dams-and-a-half, p) 12 even.
(
13 But the full data on the pressure vessels and all 14 indicate there is relaxation in the embrittlement and all.
j 15 Certainly, there is a 60-year life.
We're certain of 40 and 16 50, but certainly 60 or 70.
And the limiting is the 17 embrittlement in the pressure vessel.
18 MR. LEWIS:
You can replace it.
19 MR. TAYLOR:
We've put a lot of urgency on the 20 program at this juncture because we realize there are many 21 actions a utility can take to reduce the burden on his plant 22 to give him greater opportunity for extended life without 23 replacement, et cetera.
The whole episode of cyclic burden, g
t 24 more care is taken to minimize than that than you're going to 25 move out toward the end of those fatigue limits a lot slower.
l
281
.(\\/
1 MR. CULLER:
Reducing the number of challenges to 2
the safety system is another important element.
So it's an i
3 important program, and we were very concerned at the loss in 4
the NRC budget because we had a reasonable basis for beginning 5
this effort with Minogue.
6 MR. TAYLOR:
There is tremendous cost effectiveness 7
in joint efforts.
Let me repeat that.
And this effort, in 8
view of the essential involvement of NRC in making judgments 9
technically on life extension, NRC should have a vigorous 10 program, just work together so that we're not duplicating, 11 we're sharing resources, 12 MR. CULLER:
In many of these areas we recognize we j
13 can't do all of this work ourselves because there is the 14 regulatory view and consideration that has to go along 15 parallel with the development of technique.
Otherwise, we're 16 out of phase, we really don't know where to go.
17 The whole question of this research is an important 18 ene, and I hope you can raise it in the consciousness of 19 budgeteers.
20 MR. TAYLOR:
Standardization --
21 MR. CARBON:
Excuse me.
22 MR. TAYLOR:
Standardization, that's an extremely 23 important question, what do we really mean.
I don't think we 24 made an adequate definition of standardization from the 25 industrial side or from the regulatory side.
We are trying to
282 1
help to answer that question.
Our adavnced LWR program has as 2
a subsidiary goal a contribution to standardization.
We the deliverables in this program is not 3
believe in setting 4
a design, but an in-depth set of utility requirements which 5
will elicit from the design the goals of simplification, 6
reduced probability of core-melt, et cetera, et cetera.
And 7
we feel that setting those requirements will be a significant 8
step towards a standardized system.
9 We are evaluating as to the degree to which we can 10 enlarge that objective to stipulating functional systems 11 requirements, formfit function type of standardisation, which
(
12 would be, in my judgment, the ultimate standardization you 13 would like to get to, a formfit function, set of 14 specifications where, with that definition, the uttlity or its co,mpetitive market and solicit 15 agent could go out in a i
16 suppliers to provide equipment and systems to fit the formfit 17 function specs, 18 It's very difficult, and we are hoping to help 19 define that.
We certainly don't mean duplication.
20 MR. CULLER:
And not involve duplication.
21 MR. TAYLOR:
That is not in the interests of an 22 effective technology, and it certainly doesn't fit the United 23 States industrial competitive structure, multisupplier, many 24 utilities, and so on.
25 MR. CARBON:
I wasn't sure that I understand your
m I
l i
283 1
word " duplication."
You don't want duplication?
I am not 2
sure that I understood what you meant.
3 MR. CULLER:
Replication.
4 MR. TAYLOR:
Replication.
5 MR. CULLER:
I think we would duplicate all the j
6 requirements and such.
In the evolution of our thinking about 7
this, we have said it's obviously crazy to have a whole 8
standard plant which you can't change or improve some elements 9
of, So we would break the plant down into functional groups.
10 You may standardize the package around the steam generator and 11 its controls with a set of requirements and criteria, and when 12 things look like we have a major chance of improving those, we i
13 will change that whole package, so that there is a constant 14 evolutionary pattern that is available, but the standards 15 ought to be recognized and changed only when there is a formal 16 move to change. the standards and requirements criteria.
17 MR. CARBON:
But you're not going to build 727s and 18 747s and so on.
You're going to. slide from the 727 to a 747, 19 is that right?
20 MR. CULLER:
Well, we may slide from 727A to 727Z.
21 MR. CARBON:
That's a better example.
22 MR. CULLER:
I think we will try to preserve the 23 concepts and the requirements, but there may be internal 24 elements within the package that change from time to time to 25 improve either operability, use better materials or something
l 284 s
1 like that.
2 MR. TAYLOR:
Significant progress was made on the 3
NSSS by individual NSSS suppliers and standardization.
The 4
Westinghouse system, if you wanted a 600-megawatt unit, you 5
got two pumps, two steam generators identically the same, and 6
if you wanted a 900-megawat.t unit, you got three of them.
If 7
you wanted a 1200-megawatt unit, you got four of them.
8 Changes were made as field experience showed you 9
had to make changs, and they were applied across the board.
10 That is a formfit function-type of specification.
11 Westinghouse in that scheme went out to some 12 suppliers for charging pumps, heat exchangers and various 13 other kinds of equipment, giving geometric specs, envelope 14 requirements and flow, temperature, functional performance 15 specifications, and accepted the technology which that 16 individual supplier provided him in meeting the specs, and was 17 able to make choices on a competitive bid basis.
18 That's what we have to try to extend to the entire 19 plant, and then the more difficult issue is can you extend 20 that. concept across a group of NSS suppliers as well and have 21 it make sense?
22 MR. CULLER:
We ultimately believe that our
-w 23 standardization of requirements and criteria for performance g
V 24 by the utilitis will evolve a standard design amongst the 25 suppliers.
I mean you will see the same system you will see
r 285 r
(
1 time and again.
If it doesn't, then we will take another look 2
at it.
But we don't intend to build, as the French do, five 3
identical units in a row, although the utility could elect to 4
buy five at a time and have them built that way.
5 MR. CARBON:
Do the Commissioners use 6
standardization that same way?
7 MR. TAYLOR:
No.
8 MR. CULLER:
No.
9 MR. TAYLOR:
My impression is that Commissiner Zech 10 looks upon it as replication.
I have not heard from any of 11 the other Commissioners really a very clear statement as to O
12 what they mean.
( j 13 MR. CULLER:
I think Joe views it a little bit more I talked to Tom Roberts once about 14 like John does than any 15 a year ago.
His idea was, between replication and some 16 evolutionary changes allowed.
17 MR. TAYLOR:
So there is a perception, which is not 18 quite accurate, on the part of some people that the French 19 standardization program is a duplication or replication 20 program and that the naval standardization program is a l
21 replication program.
It's not true.
They have consciously l
22 standardized and even replicated many elements of the program; I
23 for example, site requir:sents have required them to have 24 different cooling systems and other affairs of that sort.
25 Now, seismic is an interesting one.
We are going to l
l I
L
I 286 m
1 try to emulate that.
In the seismic case they chose an 2
envelope seismic requirement, and when the site required 3
higher requirements they made up the difference with seismic 4
buffers and they built the plant upon pillars of laminations 5
of neoprene and steel so that that little bit above the 6
envelope requirements was taken up in the buffers.
7 MR. LEWIS:
How many French plants are built?
8 MR. CULLLl:
Two plants in southern France and two 9
plants in southern South Africa have those buffers.
That's 10 an intriguing way to handle it.
11 MR. LEWIS:
I didn't know any French ~ plants actually
[N 12 had it.
13 MR. REMICK:
There is an important point here.
I 14 hope the Commission is aware of what industry is thinking, 15 because they way they were proposing it, you would have a 16 grand rulemaking for a standardized plant that you basically 17 wouldn't make changes.
You want a rulemaking every time you 18 want to change one of these modules.
19 MR. CULLER:
We will most certainly preserve the 20 element.
21 MR. TAYLOR:
If you are talking the kind of 22 requirements we intend to set and you extend that to formfit 23 function, there is a hope that that could match a set of 24 regulations which covered the waterfront.
2 25 MR. CULLER:
That's why we have the agreement with
287
/"\\s,)
1 NRC to work along as we evolve these.
2 MR. TAYLOR:
It gets back to the issue that the 3
industry would like to see happen.
Don't tell us how to 4
achieve safety; just set us the goals that you want to -
5 achieve.
6 MR. CULLER:
Let me make sure that we don't give you 7
the impression that we think for each plant some element of 8
this package will change.
We don't.
But it would be a little 9
bit foolish to say this is the kind of p l'a n t we're going to 10 build for 20 years and find three years later or four years 11 later that there-is an elegant, simple way to operate the l'2 steam generators without collecting crud and such, that kind 13 of thing.
If that kind of change occurs, we will go back and 14 we will restandardize with NRC that element, and we will make 15 sure by the other standardization of teh package around it 16 that it will fit.
17 MR. REMICK:
Am I correct, in their thinking there 18 was something like five years' certification subject to some 19 kind of renewal for three or five years, something like that?
20 MR. TAYLOR:
Yes.
21 MR. REMICK:
Is this consistent with your thinking' 22 MR. CULLER:
Yes.
I think so.
All we're saying is 23 we don't believe a replicate is necessary to satisfy that 24 condition.
25 MR. TAYLOR:
Nor that rigor mortis is necessary.
288 O
1 MR. CULLER:
Rigor mortis cannot be set in, 2
otherwise we're not going to survive very long.
3 MR. TAYLOR:
Again, you don't see that in the French 4
program or the naval reactor program.
5 MR. LEWIS:
I never heard the term " rigor mortis" 6
used in connection with birds.
7 MR. TAYLOR:
He shouldn't have used it.
8 (Laughter.)
9 MR. CARBON:
For whatever it's worth, some of us 10 were looking at transcripts of the Commissioners meeting, 11 yesterday, from just this past February, and at a quick look
(
12 it seemed to me they were thinking duplicates or replicates.
13 MR. TAYLOR:
I know Commissioner Zech has expressed 14 that because in a presentation we made of our advanced LWR 15 program, that subject came up, and it was very clear to me 16 that's what he had in mind.
And we said that was not what we 17 had in mind and evident disapointment on his part.
18 MR. CULLER:
I think what we may do is sponsor one 19 of our classic workshops where we get everybody together and 20 let them argue about what they mean.
21 MR. TAYLOR:
I don't think we're ready to do an 22 intelligent job yet.
23 MR. CULLER:
We will be.
24 MR. TAYLOR:
We have in the effort to define this, 25 have a sheet, a two-dimensional array which starts at the top
289 O.
1 broad requirements, works itself down to the detailed 2
requirements, standards, systems design, component design, 3
maintenance, operation, and construction, and across the other 4
way, the level of standardization which we think we an 5
achieve.
6 And we're going to go with each one of these boxes 7
and see if we can determine are we going to be able to 8
standardize this part, are we going to be able to standardize 9
this, and how?
Some will be duplication or replication in 10 the sense of the requirements. Others will be formfit function 11 standardization, and probably when you get down to a certain 12 level, there will be no standardization.
g 13 MR. CULLER:
Incidentally, there is another 14 element.
I think maybe the Japanese may be approaching it 15 this way.
I am not absolutely sure, but they've done 16 something else that we haven't really gotten into place, and 17 I doubt whether the utilities will support.
The Japanese are i
18 setting up big proof-test centers.
They have appropriated 19 some $500 million to set up a testing center tc test the gear 20 that.goes in for failure rate.
21 Now, this goes through all kinds of analyses and 22 we have discussed this for years -- about having some center
~%
23 to test gear to failure.
We did that on the valve work, as
\\.
i 24 you all probably remember.
We did get some failure rate data 25 on the big blow valve.
l 290 tO' 1
But, the total information package, the proofing out 2
of equipment, are you going to buy on performance 3
requirements, and wire these for manufacturers so they 4
have to test to failure?
Or how will you guarantee 5
performance without replication of some kind?
Have you 6
decided that?
7 MR. TAYLOR:
We haven't gotten into that detail.
8 Certainly, we're going to establish the performance testing 9
requirements, that's right.
10 MR. CULLER:
There is another element that always 11 comes out of this that no one ever talks about.
You really
(
12 have to have some idea of how the gear is going to perform if 13 you do good PRA, and we do not exactly buy to performance 14 requirements now, test performance requirements, on individual 15 pieces of equiraent always, but it's an element that has to 16 emerge as I think we go through this exercise right now, and I 17 don't know how NRC proposes to handle it, or anything else.
18 Now, enough on that one.
19 MR. WYLIE:
Let me ask one question.
You say you're 20 not ready to put it on your workshop and talk about it now.
21 When will you be?
22 MR. TAYLOR:
I would say a year from now.
23 MR. CULLER:
Yes.
If we had gotten started two l
l 24 years ago, we would have had it already.
We're just getting 25 well launched, l
291
. (~
\\
1 MR. REMICK:
I would suggest if you do that, you 2
should have a part for the industry lawyers to get hold of the 3
NRC lawyers for well over a year.
You know, the idea of 4
having a rulemaking for standardization is great.
I think 5
it's great.
What do you mean?
I haven't found anybody yet 6
who knows what it means.
7 MR. CULLER:
It will be a year after we have the 8
workshop before we involve any of the lawyers.
9 MR. REMICK:
If they're going to have a rulemaking, 10 they better decide what they mean.
11 MR. CULLER:
The r u l ema k f. ag has to come a lot after O) 12 we have the concept.
(
13 MR. LEWIS:
Do they have to have a rulemaking?
14 MR. CULLER:
It's the proposed way to handle it 15 rather than the normal licensing process.
You would have a 16 rulemaking.
Whether that's an ECCS type of rulemaking or it's 17 a rulemaking that we routinely go through.
Nobody has really 18 thought out what they mean by it.
It sounds great to i
19 everybody.
20 MR. TAYLOR:
It's a marvelous buzz word.
l l
21 MR. REMICK:
I haven't found anybody who knows what l
l 22 it means.
Sort of we tend to have rulemaking for each 23 standard design and a rulemaking for sites, then you get to l
24 match those designs for the site.
It's the concept, but what l
25 that rulemaking is, nobody knows.
I 1
l w
w.-+w-
-ww-ww-w
--t m.
m
--m+
ww w
www.
-e w-m
292 1
MR. LEWIS:
But they don't have to have one, or do 2
they?
e 3
MR. SAUTER:
At the Severe Accident policy statement 4
it would require a rulemaking for each standardised 5
MR. REMICK:
That's what it says in there, yes.
But 6
what it means --
7 MR. TAYLOR:
It looks like the lawyers got there 8
first.
9 High-level waste, Floyd is very' active in this on 10 the Scientific Advisory Board to EPA, and the subject of 11 standard-setting.
()
12 My own plea with you and the NRC is to get deeply 13 involved in what used to be exclusively NRC's jurisdiction in 14 setting standards.
EPA has run away with this, and if the 15 present approach is a c c e p t,e d, forget it, we won't have a 16 repository, in my judgment.
What is it, ten orders of 17 magnitude?
l 18 MR. CULLER:
Not quite.
It's 10 to the minus 13 19 now.
It may be minus 10 to the minus 12.
t 20 MR. TAYLOR:
This is in terms of cancer incidence?
21 MR. CULLER:
Yes.
22 MR. TAYLOR:
What is the normal cancer incidence, 23 about 10 to the minus 37 1
f two in 10 to the minus 24 MR. CULLER:
Yes.
One in 25 4,
four in 10 to the minus 4.
That's not right.
It's 20,000 1
293
't 1
in 100,000.
I am sorry.
I back off.
My mind is not working.
2 We are meeting with Dan Egan tomorrow because there 3
is a new writeup on the requirements.
4 First off, the special committee looking at 5
standards pointed out the way it was being set, essentially 6
due to ALARA for salt mine and other repositories, and took 7
the level at which waste can be contained or a modeling effort 8
to show, and used it as a standard which essentially said that 10 to the
-9 probability of 9
there was about one chance 10 having a thousand cancer deaths over 10,000 years.
And that 11 was -t h e dose criteria.
)
12 We said that this set radiation levels so that the 13 radiation equivalent would be like jumping four feet in the 14 air for a year's dose.
And this was so far lower than any 15 other radiation standard that we were concerned about being 16 able to prove that we could license such a plant by proving 17 that it would meet these requirements; and secondly, that the 18 unnecessary conservatism we thought would eventually translate 19 itself into a measure of hazard and the question ultimately 20 would be asked, "Why are occupational exposures and exposures 21 to background radiation not set lower than this?"
22 The regulations, as set up, have been tightened just
's 23 a little bit.
The procedure for estimating the probabilities 24 of whether or not the repository can meet the required release 25 rates and several other elements of conservatism now appear to
s 4
294 m
1 us in a rewrite that just came out two weeks ago to make it --
to make it impossible to prove that you can meet 2
possibly i
3 the requirements in the licensing process.
4 So we are. going to talk;,with Dan tomorrow and see 5
whether or'not there is any way to relieve these.
It is N
6 important, however, that NRC fully.' recognize the part of the 7
proof and the management of the waste farm which they are 8
going to administer has prescribed regulations on how they are 9
going to decide whether the system is safe or'not and how ii 10 they calculate the probabilisties.
In addition to that, there is a role that NRC most 11
+
)
12 certainly should undertake -- and I am taking this advice from I
13 Bob Catlin, whom many of you know -- Bob suggests for NRC that t
14 as the repositories are built, presuming that they are, that 15 NRC should set up a way of validating the model by which the 16 licenses are granted, there is a rate at which mateerials 17 might move, and all.
The collection of data over a long 18 period of time should be checked back in evaluation of models.
19 Now, we recognize that no one knows how to do this 20 at these low levels, so there are several phases.
The 21 feasibility of checking the licensing model should first be i
22 studied.
And as Bob's contribution and my contribution to 23 this discussion with you, that's the thing that he suggested, 24 plus the general awareness that the levels being set are 25 exceedingly low and that it may be very difficult to execute a
,,_._...m m
295
/
1 rational licensing procedure that will guarantee that we can 2
meet the low levels.
4 r
i 3
MR. LEWIS:
When you quoted this 10 to the
-9, were i
4 you inventing those numbers?
i 5
MR. CULLER:
No.
{
6 MR. LEWIS:
Those are real numbers?
i 7
MR. CULLER:
Yes.
I will send you Bob's comparative 8
risk chart as to where the long-term waste repository falls s
f i
9 in the protective l~
I 10 MR. TAYLOR:
It's like 10 to the-13.
I 11 MR. LEWIS:
I would be grateful.
I assumed you were
()
12 joking.
13 MR. CULLER:
No, no.
It's beyond death by i
14 meteorite.
15 MR. LEWIS:
Which we don't know very well.
4 16 (Laughter.)
3 17 MR. CULLER:
We don't know.
Take two orders of 3
1 18 magnitude here.
1-j 19 MR. TAYLOR:
The present judgment is that the I
I 20 repository. design, in fact, is down in that vicinity, but the 21 problem is to prove it.
There's no way we can prove it.
22 MR. CULLER:
The truth of the matter is Dan did a 23 very good job of modeling this, and salt and the other things 1
24 will come out to 10 to the
-9 or something like that, just 25 naturally.
And that's where the level was set.
296
(
t\\
1 MR. LEWIS:
I wasn't aware it was that far off.
2 MR. CULLER:
Let me get Bob to send you some stuff, 3
and you will understand it very quickly because he has 4
calculated it out.
5 MR. TAYLOR:
Are you going to be here tomorrow?
I 6
can give you that chart tomorrow.
7 MR. LEWIS:
Yes, I will be here.
8 MR. TAYLOR:
I will give it to you tomorrow, drop it 9
off.
10 MR. CULLER:
Bob's in town tomorrow.
11 MR. TAYLOR:
I have a copy of it.
()
12 MR. REMICK:
Is it something that can be copied?
13 MR. TAYLOR:
Yes.
14 MR. REMICK:
It would probably be better to give it
-15 to John so we can all get copies.
16 MR. CULLER:
we will do it.
I think Bob may be 17 coming in to talk to the Waste Disposal hearings tomorrow or 18 something like that, I am not sure.
19 MR. LEWIS:
It's really peripheral to this 20 committee's job.
21 MR. CULLER:
The only reason brought it up was just 22 the general awareness of what's happening.
/~%
23 MR. TAYLOR:
Besides the industry, what it means to U
24 NRC, you will have a regulatory task that's impossible.
25 MR. CULLER:
Absolutely.
Perhaps.
297 Oks 1
MR. MOELLER:
There are a lot of aspects to this, as m
2 I am sure you know much better than we, because you've been 3
involved in the Science Advisory Board.
But the full 4
committee is going to consider the subject tomorrow, and we 5
have had a rather interesting time in dealing with EPA, NRC, 6
and DOE on this matter.
7 For example, the EPA people have called a few days 8
ago and told us they will be unable to come and talk to the 9
full committee tomorrow on this subject.
And further 10 complicating the situation is that both the DOE staff and the 11 NRC staff, not the Commissioners but the working staff, the
)
l'2
' Dan Egans of NRC and DOE have signed off on EPA's proposed 13 standards.
14 MR. CULLER:
But the standards have been rewritten 15 since they saw the last draft.
It's very different.
I just 16 got the revised copy last week, two weeks ago, and we've just 17 been through it.
That's why we asked Dan to come talk with us 18 tomorrow morning at 0:00 o' clock because I don't understand.
19 What's more, the whole process was essentially everyone agreed 20 that it was awfully safe and why not let it stay safe?
21 The technical advisory review committee, however, 22 reacted -- everybody on it -- as being too high, and the "S
23 recommendations basically were, for probably good reasons, not g
N.)
24 taken up by the full committee and then acted on.
We 25 recommended at least an order-of-magnitude reduction in the
298 1
severity of the limits and perhaps more but at least a factor 2
of ten was desirable.
3 Okay.
I will get you the information.
4 MR. LEWIS:
I always give popular speeches saying, 5
"There's no problem I know in which the difference between the 6
perceived risk and the actual risk is greater than in that 7
question."
8 MR. CULLER:
That's very true.
Essentially, I am 9
not being totally critical of EPA.
They did a pretty good job 10 in figuring out how to approach it, but they set the limits 11 too low.
' \\
12 MR. TAYLOR:
Decommisssioning.
From the viewpoint 13 of the industry, a stable, well-designed set of regulations 14 governing decommissioning is needed, and my understanding is 15 NRC is, in fact, on the track of establishing those.
So 16 I don't know that in any new initiatives are needed.
But 17 certainly, to complete the present effort in this respect is a 18 very important framework from which the uti.ities can then 19 begin to do their own long-range planning and decommissioning 20 and clear up issues of uncertainty on costs and so on, which 21 will continue to exist in a strong way if the regulatory 22 framework is not established.
23 MR. CULLER:
Is that our last one?
f-s i
D 24 MR. TAYLOR:
Sabotage.
My own judgment is that the j
25 best way we can effect the sabotage threat for future plants
299
[
e-1 is by the approach we are taking of making the plants easier
-(
2 to operate, therefore, in my judgment, harder to mess up, and 3
provide these highly effective diagnostic aids which, in 4
effect, make much more visible something that is inherently 5
wrong in the system, maybe in an in-depth sense, and don't put 6
any specific effort on sabotage per se, because it leads one 7
into almost a nightmare of intellectual what-iffing that I 8
don't know how one could contend with.
9 If we are successful in the goal of having the 10 ability for the operator to flash up tremendous amount of 11 in-depth information onto a CRT with a push of a button, we 12 have a very fine chance of him being able to spot some subtle 13 saboteur effort, an insider-type effort.
14 MR. CULLER:
The only thought I had here is that, as 15 you know, the international agency has produced, with the aid 16 of almost everyone in the world, a set of guides, which are 17 just not publicized, on taking care of sabotage and direct 18 intervention.
Not everybody follows them.
But before taking 19 off on any major programs, I think talking with the IAEA 20 in-depth to find out what experience they have had and what i
21 the reactions would be is a logical argument.
l l
22 I have reviewed them.
I have been on the scientific 23 advisory committee there for about 12 or 15 years.
I know 24 what most of them are.
They are not widely distributed 25 because it would defeat the purpose of the guides.
And they l
300 C
k 1
are never discussed in the literature, as a consequence.
But 2
there has been a lot of work on it.
3 MR. CARBON:
One spot that some have wondered about 4
with the effort to come up with standardized designs, 5
whatever, the GESARs and so on, we are coming out with 6
reactors that probably will be in operation for 75 years or 7
something,.and as far as I know, there isn't much effort or 8
thought given at the initial design stage to things that you 9
might be able to do quite easily and cheaply at that point 10 that would be effective against potential insider sabotage or 11 some such thing.
12 Part of this is aimed at saying, "Is there merit in i
l 13 giving more or special or some attention to what you can do in t
i 14 the way of design at this point in time, when we are starting l
15 to build these reactors?
Maybe they will be ordered five or l
l 16 ten years from now, but with us for the next 50 to 75 years.
17 Is there merit in paying perhaps more attention to the design 18 of them from the dis-sabotage aspect than we do?"
19 MR. TAYLOR:
My problem is to convert it from what 20 we are trying to do, which is to essentially make them 21 fail-safe with the approprite redundancies without reducing 22 the necessity for immediate action in the event of something 23 going wrong with equipment or an operator making an error, 24 combined with the ability to have a very effective modern 25 diagnostic capability, including the ability to flash up on a
301 (3
(,,/
1 screen in-depth information on the plant with indications of 2
the state of that piece of equipment or system as against what 3
it should be.
4 That is the primary defense against insider 5
sabotage.
If you open up the design process to answer the l
6 question, "What can someone do beyond what is covered by those 7
kinds of capabilities, you get into a nightmare of 'what 8
ifs.'"
9 MR. CULLER:
Incidentally, there is another body of 10 information, and I think many of you remember when Eugene 11 Wigner was concerned with civil defense after the Cuban
[
)
12 missile crisis.
He spent nine years at Oak Ridge, three
%s/
13 months a year, running a civil defense study.
And sabotage 14 and such to the nuclear systems were analyzed in pretty 15 interesting depth by a relatively small group of people.
16 Conrad Chester is still there and did the work.
He 17 had Schlesinger involved and a lot of the Army defense of the 18 United States and all, in looking at the consequences.
19 What I remember of the principal conclusions are 20 these:
Most of the accidents initiators that a saboteur 21 can manage essentially are no worse -- are covered by the 22 automatic shutdown and protection systems that had been 23 designed in for normal safety.
The only things were direct Os 24 penetration of a concrete pressure vessel with shaped charges.
25 So Connie went out and they built a vessel of some
__,-.--..--s-
.~
302 1
kind and shot shaped charges at i t to see what would happen.
2 We've driven telephone poles into containment vessels on 3
rocket sleds and such.
4 So there is a hardness measure from external and 5
analytical work that has been done in the civil defense 6
program at Oak Ridge, and there are still guys around who 7
remember every detail of it because every now and then it 8
becomes an important issue again and they go back and do some 9
more work.
10 As you look at i t, you might want to involve these 11 people in the ex-civil defense group at Oak Ridge.
They can 12 dig up pretty good analytical approaches and maybe some 13 conclusions from that-work.
14 MR. LEWIS:
One thing you just said, I remember the 15 American Physical Society study ten years ago dealt with 16 sabotage in much those terms.
They said that the group 17 couldn.'t imagine an act of sabotage that could produce a worse 18 accident than normal causes could produce, but can only 19 increase the probability of it.
20 MR. CULLER:
That's correct.
That was about the 21 conclusion.
A lot of things were tried.
Connie is the most 22 inventive guy I ever saw anyhow.
He can think of all sorts of 23 weird things.
There was an excellent group of people.
We
)
24 pulled the best physicists and all out of the lab to work with 25 Wigner over a long period of time.
So it was handled
.--e w
e----,--,e p an a e
w w-
~-,,n--,,r,-.,.--,,n,.,--,--mn,---.--r-a-
-e
303 I(j i
responsibly.
It may not answer all of the questions that 2
could be asked now, but nonetheless, it's background that 3
shouldn't be lost.
J 4
MR. REMICK:
I asked the Japanese if they were 5
giving consideration to sabotage in their design, standardized 6
design, and whether they thought they should.
And their 7
answer was, " Absolutely not."
8 MR. CULLER:
I think that's a mistake, and I think 9
the international agency would advise that that's a mistake.
i 10 There is a lot of discussion that has gone on at pretty high 11 levels about that.
12 MR. REMICK:
I guess I wouldn't say it's a mistake.
13 They were saying basically what you were there, that there is 14 no need for additional work for sabotage.
15 MR. TAYLOR:
The distinction between ours and theirs 16 is external sabotage.
We have much tighter security 17 arrangements around our plants than they do.
I would think 18 that that's what the difference is they are eliciting.
i 19 MR. REMICK:
If I recall, I have better figures in 20 my notes.
I think for a four-unit site, they have 40 guards 21 total, all shifts.
That's the total for their guard force, I 22 think, is a total of 40, 23 MR. TAYLOR:
We're not advocating forgetting the 24 sabotage issue in the external sense.
25 MR. REMICK:
I was thinking of design now.
I am not
i 304 fN, 1
thinking terrorist type of thing.
The thing we were talking 2
about, that sabotage 3
MR TAYLOR:
Or the ability for some troublemaker to 4
enter the plant, not part of the staff, has been apryopriately 5
screened and trained.
That's an important protection against' 6
sabotage, the external security.
7 MR. CULLER:
John, do you have any other thoughts 8
now?
9 MR. TAYLOR:
Containment and siting I guess would 10 drive everybody crazy if we go through each of these items.
I 11 am willing to quit.
12 MR. CULLER:
Oh, I am sorry.
13 MR. TAYLOR:
Shall we continue on, or are you just 14 too exhausted?
15 MR. CARBON:
We would like to continue.
16 MR. LEWIS:
How long are we going to continue, Max?
17 MR. WYLIE:
We are scheduled to 9:00.
18 MR. CARBON:
Probably 9:00 o' clock or something.
19 MR. LEWIS:
If they stay, I may stay a little 20 longer, but if you just got back from overseas --
21 MR. TAYLOR:
Containment and siting, certainly we 22 have seen an evaluation of TMI-2 in our source term work that 23 the strength of containment and its ability to hold together s
b 24 for substantial periods of time if a severe accident occurs 25 areextremely important to capability.
305 b'
(,,)
1 I am not suggesting that we need at this juncture to 2
make major changes in the present containment system.
As we 3
evolve future designs, we are going to look at such issues as 4
you have raised, very, very carefully and make sure that, if 5
anything, we improve that barrier in that whole 6
defense-in-depth concept.
It's an essential one, and it's 7
proven to be better, perhaps, than we thought it was as a 8
result of the evaluation of TMI-2.
9 MR. CULLER:
We have also been exploding hydrogen'in 10 Horton spheres to look at shockwave rates of propagation of 11 various waves and heat flashes and such to get a better 12 indication of what might happen and how to protect against 13 hydrogen overpressure.
And I think the vessels we now have 14 are adequate for that.
There was some question on the smaller 15 ones with ice condensors.
Ignitors will take care of that.
4 16 MR. TAYLOR:
We have tested those, and they seem to 17 perform quite satisfactorily.
18 MR. CARBON:
Let me pause for a minute and explore 19 this question of when we break up.
20 (Discussion off the record.)
21 MR. TAYLOR:
Backfitting.
22 MR. CULLER:
Do you want to say anything about 23 siting?
5 24 MR. TAYLOR:
Oh.
And siting.
The whole content of 25 the comment underneath is on containment.
So I just didn't
306 1
prepare any response to the siting.
s 2
MR. CULLER:
Let me give a derivative comment from 3
lots of others, and I think you all know these, and I won't go 4
at all into any expansion.
If we get standardized plants, in 5
whatever definition, and we get sort of a prelicensing of 6
standardized plants, the problem of the site is as much of a 7
problem to the utility in prequalifying sites, and it's under 8
the federal control, it's nominally under federal plus state 9
control with the safety requirements being set by the NRC.
10 But the other conditions are controlled very 11 distinctly by the state itself.
Siting is an important
)
12 element in cost at times in qualifying the site, and it is 13 also likely to be a restraining factor.
I don't know what 14 ACRS can do about this, because of the jurisdiction of the 15 federal site does not exte,nd into state sites exactly.
16 Nonetheless, it's an important element and deserves some 17 consideration.
18 We have suggested, as NRC has suggested, and DOE, 19 that the. utilities probank sites, that they start qualifying 20 sites either for fossil or nuclear stations well in advance.
21 I am not sure that that's going to be uniformly done, and even 22 though we may be able to speed up the process of construction 23 and licensing by standardization and preapproval, the approval 24 of a site is also a major factor.
And other committees have 25 observed this just as I am sure you will
307
/
k 1
But I point it up only because it is not exactly 2
under control in anybody's imagined future plans.
3 MR. TAYLOR:
So then we are moving into the 4
regulatory process, backfitting, establishing a backfitting 5
policy that moves the threat of continuing backfit from the 6
utilities is probably the most important step at stabilisation 7
that can be made.
8 Now, I am not suggesting that backfit can guarantee 9
that there will be no backfit under any circumstances.
.There 10 might well be a circumstance that requires it, carefully 11 identified, reviewed, subject to cost-benefit analysis.
But I
)
12 know that backfitting policy statement is being worked on, and 13 one that owuld give some reasonable assurance of a 14 stabilisation in terms of the incessant change that the plants 15 have been subjected to up to now would be extremely valuable 16 to the utility and would remove the concern that you have 17 expressed here of a reluctance to want to talk about 18 improvements even in future plants.
It would help i
19 substantially that type of concern.
I 20 You also suggest certification points, which remind 21 me of a prevalent industrial practice in the ASME code system 1
22 where changes are made in codes and there is no desire on the f
23 part of the professional community or ASME to quell change.
1\\
l 24 But when a code change is decided upon, is scheduled to be put I
25 into place at a certain time, and equipment that is built i
I i
l 308
\\
1 prior to that does not have to be redone to meet the new 2
requirement.
It meets the old requirement.
New equipment 3
effective on the date specified meet the new requirement.
So 4
you remove the destabilization of what you have got, and you 5
remove the threat against improvement for future equipment.
6 We strongly encourage that.
7 MR. CARBON:
Do you find CRGR to be an effective 8
MR. TAYLOR:
It's an extremely important step that 9
was taken by NRC.
It has been somewhat hamstrung by 10 Commissioner guidance at hte very early stage of its 11 forulation, which to the best of my knowledge it's still rN l
(
)
12 operating under that kind of constraint, and it would be very 13 helpful if it were given more authority than it has today.
14 MR. CULLER:
This is one of the more important 15 technical issues that I think you identified for discussion, 4
16 and I know other industry people surely must have commented to 17 the same effect.
We fe'el it distinctly in the research and 18 development side as well.
19 MR. REMICK:
I guess it's not in there.
Do you 20 think that the cost-benefit analysis should be part of the 21 backfitting?
22 MR. TAYLOR:
Yes.
23 MR. CULLER:
Yes.
We think this hierarchical
(
24 analysis of PRA as it gradually becomes available as a 25 technique, individual PRAs on plants will provide a basis of
309
\\
,/
1 saying the backfit requirement for this system will provide 2
only a marginal increase in safety; therefore, don't do it.
3 And I think one of ~ the proposed wordings of backfitting rules 4
are that there must be a substantial increase in safety and an 5
evaluation of costs that would justify it.
And I think thete 6
are tools that can be brought to bear again that will make the 7
judgment a little bit simpler.
8 I am certain that the industry and the NRC will move 9
and more to hierarchical analyses with PRA, and once they're 10 individually done for each plant, then the determination of 11 the importance of backfit for safety will be a little bit
[ \\
12 easier to make and specific to the plant rather than generic.
V 13 MR. CARBON:
We have been in discussions of this 14 type where the question comes up as to what costs should be 15 included when you do something like this.
16 MR. CULLER:
We're familiar with it.
17 You first.
18 MR. TAYLOR:
All right.
Our judgment is the more 19 standard cost-benefit evaluation of the cost to make the 20 modification against the benefit achieved should be the basis, 21 not looking into the cost avoidance associated with the 22 probability of this action avoiding an accident, et cetera.
23 MR. CULLER:
Or the utilities' cost in a risk of a
(
24 very major loss, economic loss.
25 MR. TAYLOR:
Substitute power and all the potential
310 m
1 consequential loss should not be a basis.
2 MR. CULLER:
Before you came in, we were talking 3
about this, and all of us could think of the system biting 4
back at both the NRC for regulation and not using the lower 5
standards, which they don't, or the industry in saying we can 6
basically -- we're much safer because we need economio 7
protection, safety standards that will protect our economio 8
investment, that are more stringent than the health and 9
safety.
10 Well, the argument in each case can loop back and 11 bite.
As a consequence, I think the nice thing to do is to 12 know that that kind of information is undoubtedly true, but 13 not to use it in an evaluation procedure, and we haven't 14 talked a lot about this, but that was a conclusion after about 15 an hour of fooling around.
16 MR. CULLER:
Is that enough, Max?
17 MR. CARBON:
I think so.
18 MR. CULLER:
If you want, we will explore it some 19 more and write to you, but it's an imperfectly formed thought.
20 MR. CARBON:
That's fine.
21 MR. TAYLOR:
On performance versus prescriptive 22 approach, we talked about that earlier.
I would strongly 23 advocate a movement towards performance type of regulation for 24 the future, and again try to stabilize the present regulations 25 for the present plants.
311
,O
\\ g/
1 And I think the end result of moving to performance 2
will be to achieve, if anything, more effective safety at a 3
substantially less agony and expense in the regulatory 4
process.
Should the NRC encourage smaller plants?
I would 5
say not specifically.
Let the industry determine whether it 6
wants to go ahead with smaller plants or larger plants, and 7
make your judgments on their safety adequacy.
8 We in our own program have a separate effort that is 9
directed to developing conceptual designs'of small plants 10 where our purpose is to see if there is unique benefits 11 achieveable from the small plants that cannot be achieved from
)
12 big plants, such as in high level of modular construction, 13 such as in a substantially higher degree of achievement of low 14 coremelt probability, such as easy achievement of seismic 15 requirements, and see if there is something about the small 16 unit that can't be extended to the large and therefore you 17 are not forgoing the economy of scale which has led us to the 18 larger systems.
19 MR. CARBON:
Part of the thought there was, could 20 you perhaps go to smaller modular units which could be more 21 thoroughly tested for safety?
22 MR. CULLER:
Yes, we saw that i
23 MR. TAYLOR:
It is possible to perform tests less 24 expensively with the smaller units.
My own judgment is that 25 we have learned such an enormous lore that bears on safety
312
)
s/
1 from these prsent systems, which are quite large, that we 2
would not, even though it was easier and cheaper to test the 4
3 small units, we probably would not achieve the same level of 4
understanding of their safety characteristics.
[
5 MR. CULLER:
We explored this 4
6 MR. TAYLOR:
Until they were fully deployed, and 7
we've gotten the same level of experience.
8 MR. CULLER:
We explored this while it was still at 9
Oak Ridge, when safety was still part of the AEC, and the 10 subject basically was how would you test for temperature and 11 pressure simultaneously?
You're going to run the plant under 1,
i 12 nuclear conditions and run a scram or what?
4 13 What we decided is it's almost equally difficult to 14 test a plant of one size or another.
The complexity is very 15 high.
You didn't want to make the plant radioactive before 16 you did the test.
Then the providing power for a large or i
17 quarter-core or tenth-core of electric heat available shaped 18 to the burnoff characteristics of the fuel became an almost 19 impossibly difficult test procedure.
i 20 Again, there is lore on that period of looking at f
21 how and would conduct tests if you want to conduct tests, if i
22 you wanted to de so.
It's a good idea, and it probably i
l 23 deserves some exploration.
There may be merit in it, but I 24 don't think that it would offset part of the economies of 25 scale, and as John said, we think we know more and more about
.i I
w,,,w-
,,ne,,
,,~,._-,n---n_,
,,,_,-,_--aee
,w,,.--..._n.m,,--,-+-eeve-,-w-
313 O(,/
1 how the reactor is operated under safety excursions.
2 MR. TAYLOR:
On the last element, we have 3
experienced a high degree of cooperation and interest by the 4
NRC staff in our advanced light-water program and the safety 5
closure issues.
No inhibitions of any kind that we have been 6
able to see, except for the issue, of course, of their ability 7
to staff and provide the resource allocation while still 8
meeting their other obligations.
9 We have gotten a letter from Chairman Palladino 10 supporting this adanced LWR program and pledging substantial 11 staff support in carrying out the reviews that are involved.
12 Go I see no inhibition there.
I would suggest that ACRS might 13 inject in a more explicit form the issue of how they would see 14 themselves and NRC treating the evaluation of advanced systems 15 and the formulation of regulation in the process.
It's not 16 made too explicit in this writeup that we have here today.
17 Now, our present plans, there is no question of 18 timidity, let me call it, on the part of the industry to be 19 proposing changes of any kind in this present uncertain 20 regulatory atmosphere.
The adversary relationship is 21 counterprodutive.
I don't know that anything can be done 22 about it.
In our own work on the research side, we're 23 gratified that we can sit down on regular intervals and 24 discuss our mutual programs to the best of my knowledge, 25 there is no inhibition in sharing our knowledge and data with
314 l(_)
1 each other on the research side.
2 Because of the adversary arrangement, we don't talk 3
to the NRR Branch of NRC without the presence of duly 4
constituted utility personnel, whose interests are very much 5
involved in any decision or move on the part of NRR in this 6
very adversary relationship.
7 We accept this constraint actually very readily 8
because it means that we can keep back from the adversary 9
f r o n t ', and by being back from the adversary front, have a much 10 better chance of retaining technical objectivity and 11 credibility, which is probably the most important quality that i
12 we can provide for the utilities in times of issues of
(
13 technical question and judgment.
14 MR. CULLER:
The industry actually provides us with 15 a cover.
We are never asked to enter the adversarial 16 hearings.
There are times when we will perform as an expert 17 witness or something like that, but we rarely have to be 18 cross-examined or anything like that in any kind of hearing.
I 19 And if we did it very frequently, more than once every three 20 or four years, we would soon lose our position of technical l
21 objectivity.
22 MR. REMICK:
I don't think we were limiting it, l
23 though, to the hearing process.
It's a whole concept of if 24 it's from industry, you've got to look very, very closely at i
25 the motives.
It's that concept, because certainly they would
315 m
s 1
only be looking at their best interests.
It's a feeling if 2
you can't trust INFO or you can't trust EPRI because the money 3
comes from the utility.
That kind of adversarial relationship 4
that exists that is much broader than the licensing or hearing 5
process.
Or if'the Commission does something, then there are 6
brsnches of Congress that are clobbering them.
we just read 7
Just one recently, it was obvious 8
about it in the papers -- the reason the Commission obviously 9
did it was because of pressure from industry lobbyists, 10 Price-Anderson.
11 MR. TAYLOR:
I wish it could be understood on both 12 sides fully, as fully as possible, that the true demand upon 13 the utility to protect its investment is greater than the 14 demand to provide publio safety.
In other words, if you have 15 a severe accident, you have overnight lost that investment, 16 and it's a catastrophe for the utility.
It's a partial 17 catastrophe for the industry.
The credit rating reductions 18 for the entire nuclear industry have been extremely severe as 19 a result of TMI-2.
If you don't have a severe accident, you 20 don't worry about public safety.
21 So there is a common objective that should reduce 22 this adversarial situation, but I have no hope that in the 23 future, in the foreseeable future, it will change.
24 MR. REMICX:
There is a broader one, national 25 objective of alternate energy source.
Nuclear, that should
1 i
I 316 1
drive us from a national perspetive.
That's what drives the 2
Japanese.
3 MR. CULLER:
I have a suggestion here, and I don't 4
think it will bear fruit.
But I do think, since you are in 5
this state of reexamination, the possibility or removing the 6
quasi-legal procedures should be explored in a formal sense 7
and not by the staff itself, but get some group in, using 8
advice from overseas, the Swedes, the British or the French.
-9 Now, you may even wish to get a group of people who 10 would actively look at this system versus theirs.
I don't 11 believe that anything will come of it, but the idea of having
()
12 to take depositions in a safety incident almost is almost 13 quasi-legal -- well, they are legal because the idea now, 14 because of the presence of the courts as a final arbiter, is 15 you have to have legally gupportable evidence.-
I do think 16 that is desirable before we go further down the road of having 17 safety defined by lawyers or wheth'r the man is honest or e
18 dishonest, and the credibility.of a witness in investigations 19 of safety seems to be an exaggeration of a process that should 20 more frequently discuss substance rather than the procedural 21 aspects.
22 MR. REMICK:
Floyd, that is not an NRC 23 statute.
Congress has written in the Atomic Energy Act that
!- k 24 these decisions should be made following the Administrative i
25 Procedures Act.
So it's a congressional type
.-..,,,-%,->.,-.%--,y-.~---,.,-y,,,-,,,--.m
.y,w,.~ m mw,--
.---,-,--.,--s-.-,.r---
,m-,.<,,r
317
\\_/
1 MR. CULLER:
I understand that.
But unless the NRC 2
challenges it as a procedure that might be suspect, no one 3
else will And that's the only reason for suggesting it.
I 4
don't think anything will come of it.
I think the lawyer 5
groups and all understand the advers. rial process and only 6
that as a basis of making decisions.
And it may be the only 7
one that's available to us, but I would not give it up 8
entirely.
It may be a tilt at a windmill, but other countries 9
manage.
And as a consequence, it is not therefore not 10 impossible.
11 The British use the " white paper" technique.
They 12 bring in a board of inquiry, frequently a judge of outstanding 13 reputation to make a determination, and then they stick with 14 that determination.
15 There are other processes, and I do believe opening 16 the issue to intelligent thought rather than controversy is 17 something that the ACRS might be able to do as an independent 18 look at the procedure.
19 MR. SAUTER:
I would like to make a comment on 20 this.
I thin one of the root causes, not the only one, but 21 one of the root causes of the advancing nature of the 22 adversarial relationship is the fact that as a regulatory and I won't say this is solely the result of the 23 process i
24 NRC; I think it's the whole community that has contributed 25 that we have tended to lose sight of the overall objective,
318 (O
s,/
1 the end product, if you will, of regulation, and have become 2
very focused on the means.
3 We lost sight of the fact, I think, largely that the 4
overall objective here is to have safely operating plants, and 5
we have become very narrowly focused one way of achieving i
6 that; namely, the present proscriptive body of regulations 7
that is involved.
8 So meeting those regulations has itself become the 9
end of the process.
The means have become the end, and in so 10 doing, we have given.over the game largely to the lawyers 11 because now it's a legal question.
Have you satisfied this b
12 requirement or not?
It's not a technical question.
Are we 13 promoting a safe plant?
14 MR. CARBON:
I think we share your views on that.
15 MR. SAUTER:
I think, in terms of long-range focus 16 for the NRC to try to shift away from that focus back to a 17 focus for objectives through, in my view, moving towards 18 performance-oriented regulation rather than proscriptive l
19 regulation, will go a long way to alleviate this problem.
20 MR, TAYLOR:
The subsidiary question here is the 21 role of fines.
We had a brush with this issue last year when 22 a group of CEOs asked us.to undertake a study of the i
23 enforcement system that existed with NRC, And a serious issue 24 in their minds in asking us to do this study was the concern 25 that fines were counterproductive, that fines could be
=-v
319 f
o(
1 absorbed by the utility in view of its many other major 2
financial commitments.
But to the operator and the managing 3
supervisor, they were enormous, and they put them enormously 4
on the defensive, and once they were threatened with a fine, 5
then the whole emphasis was to protect themselves, find the 6
data that would take care of that, diverting them from a 7
balanced view toward operating that plant properly.
had no evidence 8
They also expressed some concern but if this process continued, perhaps some of their 9
for it 10 people would hold back information, which would be the worst 11 thing for everybody.
12 So we began to lay out a plan for that study, and
)
13 our first move was to begin to elicit the opinions of 14 behavioral experts.
We got a very strong refrain that the 15 effective way to make improvements is to emphasize the 16 positive, identify the places where someone had done their jot 17 well, to those who aren't doing it well, have some reward 18 system for those who did it well.
19 And they cited examples of such things as Equal 20 Employment Opportunity enforcement practices, where the 21 negative did not get any improvement and the positives did.
22 Our board decided that was not a proper role for 23 EPRI to play unless NRC was deeply involved, and we have k
24 informed NRC, in fact, with a letter to Chairman Palladino, of 25 our planning. And then we, of course, wrote back to say the
.e, 320
% s/
1 board did not approve our proceeding with this program unless 2
NRC, in fact, engaged us to do so.
3 NRC since, as you know, has set up a small panel 4
consisting of Dennis Wilkinson, one of our people, Jack 5
Parris, who would have, in fact, been deeply involved in this 6
work, and I can't remember the other person.
A much less in-depth study than we would have carried out if we were to 8
go.
9 So I would suggest as one of the long-range planning 10 elements the introduction of behavioral experts to look at the 11 issue of motivation associated with enforcement actions like 12 fines, find out whether or no this concern that we have heard 13 from senior utility executives that these can be 14 counterproductive, is right, and what ameliorating actions
~i 15 might be taken.
16 MR. CULLER:
Fines give the image of criminal 17 negligence, and it certainly shuts off any kind of progressive 18 thought at the time an investigation starts, of' fines are a l
l 19 likely outcome.
That's basically the advice we got at our l
20 preliminary study.
t 1
21 MR. CARBON:
Why don't you just skip 4 and 5 in the i
i 27 interest of time.
l l
23 MR. REMICK:
Let me just ask one question there.
24 had a thought about this myself.
Would you say that all fines l
l 25 are bad?
l
321 0
1 MR. TAYLOR:
No.
2 MR. CULLER:
No.
We wouldn't say that at all.
3 MR. TAYLOR:
We should study these.
4 MR. CULLER:
If there is a real case where you can 5
see criminal negligence or something like that, man, I think 6
they ought to really be hooked, and I wouldn't charge them a 7
little bit of money, I would charge them a lot.
But in the 8
routine investigation of trips and incidents and things that 9
have even safety significance, I think we're much better off 10 in achieving the end result of safer operation or safer design 11 or whatever by a cooperative-like investigation rather than 12 trying to hold the fine as a way of beating people into an 13 awareness of safety.
14 Now, that could be wrong, but most of the got,certainly went that way.
15 preliminary advice we 16 MR. WYLIE:
Would you say the FAA approach to 17 regulation and a separate investigative body like NTSB has 18 merit in the nuclear industry?
19 MR. CULLER:
We think so.
I am sure John does, 20 too.
It has its limits.
l l
21 MR. TAYLOR:
There is a big difference in the 21 situation if there is a plane crash, that's it, the plane is 23 no longer operating, et cetera.
So the scheme of everything
\\
24 is frozen.
Investigators come in and take care of it.
It
[
25 makes a lot of sense.
I i
-._____,,~__s.--
322 b
\\s /
1 If you apply that literally to the nuclear power I don't 2
plant, I am not sure it's correct.
I understand that to some extent that approach was 3
know the details 4
applied to Davis Besse.
But here you have a plant that is 5
still alive, and plants in a shutdown condition are not 6
automatically safe no matter what happens, and the thought of 7
freezing is not exactly clear to me is the safest approach to 8
take.
and I may not have the right 9
Turthermore, I hear t.h a t lawy+rs were brought in immediately along 10 information 11 with.the investigation.
As a result, the plant management was 12 saying, "We have to have our lawyers," and you and I know what
(
i 13 happens in terms of open communication if that is what 14 happens.
15 MR. REMICK:
You mean in the NRC response team we
\\
16 have lawyers?
\\
\\
~
1 17 MR. TAYLOR:
I was told that is hearsay.
I would e.
18 ask you to check it'teiore you believe it.
l 19 MR. CULLER:
Incidentally, my reaction is almost 20 identical to John's.
~
r 21 MR. SAUTER:
I have heard the thing 22 MR. TAYLOR:
It's only h arsay.
23 MR. WYLIE:
Maybe another comment.
Of course, you 24 know this type of(t,hing has been proposed for the NRC.
25 MR. CULLER:
I think you have to approach it with
323
-s
)s 1
care, as John said, because there has no harm done to anyone, 2
except the implication that harm could have resulted, whereas 3
in the case of the FAA they are investigating an accident 4
that killed somebody.
5 MR. TAYLOR:
It's over with, you know-and you don't 6
have a 7
MR. CULLER:
Say there are cracks in the wing or i
8 something like that that appear under test.
I forget the number 9
MR. WYLIE:
They investigate like 3500 a year.
But it's of all levels.
They generally 10 11 look at about six or seven crashes a year.
i 12 MR. TAYLOR:
I think what we're saying is the 13 process could be applied, but it has to be applied with 14 consideration of the real differences in the two fields.
15 MR. WYLIE:
One thing that has gone forth, though, 16 is tha use of designated representatives in the party systems 17 in their investigation where it's not all the regulatory 18 bodies.
In this case of Davis-Besse, this is 100 percent l
l 19 NRC people.
20 MR TAYLOR:
Yes.
That's an issue.
That designated 21 representative idea has a lot of merit.
In fact, it directs l
l 22 itself in the next area as a way to enlarge the up-to-date 23 experience and knowledge of the regulatory staff.
But the O
l 24 question is, is it tolerable in a highly adversarial i
25 arrangement in a nuclear program?
It probably is unacceptsble l
1
324 O\\
1 in our present atmosphere.
2 MR. CULLER:
I wonder at times what would happen if 3
you asked the utilities to investigate.
You would probably 4
get a better report.
5 MR. WYLIE:
Sure you would.
6 MR. CARBON:
Why don't you pick out the most 7
important parts in Numbers 6 and 7,
and maybe we can try to 8
move ahead.
9 MR. TAYLOR:
Indicators of performance.
There is 10 real movement on the part of INPO to establish those in a more 11 complete, highly measureable, visible form.
I would suggest
(
12 that that's the best way for that to happen.
And then NRC is 13 in a position to monitor progress and overview and audit that 14 type of thing as distinct from NRC setting it up.
15 An elite to operate the plants, well-trained people 16 are certainly required.
I don't know that I would want to 17 describe that as "an elite."
18 And, of course, strong judgment is that the industry 19 can, should provide effective high-quality training and this 20 will not be aided by a contral government training capability.
21 Personnel concerns.
One inference in part of those 22 questions was, we have got to find another role for the s
23 Staff.
And you know, in a business you never do that.
You 24 find a real function that is needed, and you look for the 25 people to do that.
And if you don't have that function, then
325 O
1 you make arrangements for hopefully, humane separation.
2 And I would strongly urge that to be your basis for 3
planning.
Don't ask what you are going to do with the people, 4
just ask what you really should do and then find out whether 5
you need people and what kind and so on to do it.
6 MR. CARBON:
Well, part of this one has that flavor, 7
and that is certainly true.
And the question could be raised, 8
can the government do that?
9 Part of it also though, has the flavor, the agency 10 has been heavily design-oriented.
And the future maybe its 11 problems will be more on the operational side.
And will it
(
12 have the expertise, or does it have the expertise for 13 high-level waste when it has been working in the design 14 reactors?
-15 So, it is more than than this.
16 MR. TAYLOR:
Yes.
Well, we earlier commented 17 strongly, a greater involvement by NRC in the high-level 18 wastes regulatory standardsetting area is very important.
19 MR. CULLER:
And in being able to follow the waste 20 21 MR. TAYLOR:
My own judgment 22 MR. CARBON:
I think the point was that as the NRC 23 areas of responsibility change in the future, will it have the gN 24 kind of people it needs?
25 MR. CULLER:
The only way that it will is by
I 326 w
1 anticipation.
And I think the idea of a longer-range plan 2
must identify what future responsibilities will be placed upon 3
the NRC, and an approach to providing the necessary knowledge.
4 My own recommendation in many respects is to push 5
as rapidly as industry can and will respond, to push more and 6
more of the primary responsibility for audit, for control, to 7
the industry, and put the NRC a little bit more and more into 8
auditing what is going on to make sure that the criterion 9
standards for operation are met.
10 Now the industry is going to move a little bit 11 slowly this way.
But they are moving in that direction.
And
(
12 John has cited several areas where this is moving.
4 13 The cooperative part of that I think is the way the 14 major burdens for having a large number of people in NRC do 15 the operating control can,be shifted to industry.
And that is 16 the direction to go because they will have a better 17 responsibility base than the inspector coming in from the 18 outside.
19 MR. TAYLOR:
On the specific question of changing 20 the capability of people to meet a different type of work, my 21 own experience with this has been very favorable when we talk 22 about professional people, where it is near impractical -- in 23 fact has ended up in my own experience with literally having 24 to lay off hundreds and hundreds of people while you are 25 hiring hundreds otherwise -- is between such things as factory
327 g3 1
personnel versus professional personnel 2
But, I have seen just specifically in that in 3
Westinghouse moving people who have been spending a good part 4
of their career in reactor design to various operations in 5
radwaste work which has been strongly emphasized, and making 6
the transition beautifully.
7 So, I think by dint of some training and recognition 8
that you have got capable people on the staff that can adapt 9
in the technology environment, and that can be accomplished.
10 MR. CULLER:
There are other sources of training and 11 knowledge and the national labs are heavily layered with (O) 12 them.
And they have been in these areas a long time.
They 13 can send people out, get them acquainted with the basic 14 science behind it, and then bring them into the regulatory 15
- role, 16 MR. CARBON:
Let me suggest, unless the Members have 17 questions, that we jump to your broad --
18 MR. REMICK:
Just one thing.
I would be curious if 19 you had a ccament on the $1000 a manrem?
20 MR. TAYLOR:
We are satisfied with that.
21 MR. CULLER:
You mean is it enough?
22 MR. REMICK:
Well, it was one of the issues there, 23 the last item there.
I wanted to know if you had any views 24 based on that?
25 MR. CULLER:
My response in a semi-offhand way is
328 O
\\,)
1 that $1000 a manrem may add up to a $10 million cost.
And 2
that is pretty expensive.
There are no insurance judgments 3
that have yet made it there.
4 MR. REMICK:
Does it give you a problem?
5 MR. TAYLOR:
I think it has been accepted.
It is an 6
operating tool.
7 MR. MOELLER:
It is a surrogate for economic and 8
other losses.
9 MR. CULLER:
Of course.
10 We have no real opinion or objection, I suppose.
11 MR. SAUTER:
As long as it is made clear that it is 12 a surrogate.
Some people think you ought to add the other 13 economic costs on top of it.
14 MR. WYLIE:
That is a big problem, this business of 15 economic on-site costs being added on top of that.
16 MR. SAUTER:
I think it is acceptable if it is made 17 clear it is as a surrogate not only for actual public health, 18 but for everything else.
1 19 MR. CULLER:
The whole issue of economic on-site 20 loss.really deserves some attention, I think, because it is 21 likely to stick around for a while.
And I don't know, it the 22 seems inappropriate to us to use that plus other j
23 surrogate of $10 million, let's say.
But, I don't think our g
24 opinions are going to be enough to resolve that issue with 25 staff positions and other things like that that move in that
4 329 1
direction.
2 Let me preface this remark basically by three or 3
four years' worth of thinking summed up in a moment.
4 The regulatory process that we have been involved 5
in, technical regulation in the United States, has the unique 6
characteristic of taking a broad law in Congress, or a set of 7
laws and setting up essentially the regulatory criteria, the 8
procedures by which they will be enforced.
They also set up 9
the appeals procedure and are judge and jury as well as the 10 investigation.
4 11 The process is so powerful that it has all sorts of 12 internal loops that deserve attention, and it becomes a fourth 13 branch of government basically, which no one will touch to see 14 exactly how it performs because it is politically a difficult 15 task.
And no one in Congress really wants to get back into 16 the morass.
17 The Administration, although responsible, has an 18 indirect responsibility, because the independent regulatory 19 commissions are creatures of the Congress and the l
.20 administrative appointments as such are the only way that real 21 checks and balances can occur.
22 It seemed to me that prudence in any system requires 23 some sort of an overall audit and evaluation of performance.
f-24 And years ago I started thinking, in the first big wave after 25 Three Mile Island, what do we do in the way of regulatory l.
330 1
change?
2 I suggested the role of the ACRS should be clearly 3
defined as an auditor of performance of the NRC, and that a 4
formal audit of performance should be established.
5 This says you should not be a creature of funding by 6
the NRC.
That you should have a degree of independence, 7
perhaps, to perform this function.
8 If this is done, it is not something that will be 9
done without staff and without a reasonable amount of time.
10 In order to provide for this activity, auditing, and 11 another important role that I think ACRS already has, I 12 suggest that it might be desirable for ACRS not to review the 13 licensing actions as routine.
To remove ACRS from the routine 14 work pattern of NRC as a "me too" or separate identification; 15 and that the emphasis within ACRS moves to the responsibility 16 that you already have to explore new safety issues or possible 17 improvements of safety in principle; to become the long-range 18 thinker; to pick out major issues of safety and in difficult 19 cases to adjudicate in licensing or operating problems; at the 20 request of NRC to perform special reviews.
21 Basically, that is the whole of the proposal there.
22 Now, it sounds simple.
I know it would be very 23 difficult.
Hal and I talked about it while we were doing this 24 safety review for ERAB.
He said, you know, it is a good 25 idea.
I have talked to some of the Conmissioners about it in
)
w
331 C
\\
1 the process, over the last few years.
2 I think it is a very good idea.
And I think in the 3
long run it will improve safety in the United States, 4
particularly if NRC has a long-range plan to have someone come 5
in and evaluate their performance and to report to the 6
Congress, to the President, and to the utilities 7
independently.
8 Then I have two other suggestions, that as a result 9
of being on multiple committees, EPA and DOE here, and others, 10 there is a woeful lack of coordination among the agencies that 11 are responsible for the generation of safety and regulatory 12 information in the United States.
And I have sat through 13 hours1.50463e-4 days <br />0.00361 hours <br />2.149471e-5 weeks <br />4.9465e-6 months <br /> of essentially confusion that results from one group 14 picking the other.
15 There is a formal council that is supposed to meet, 16 but I swear to goodness I don't think that it is going to
~
17 occur within the government agencies themselves.
18 At one point, something that did not get printed in 19 the safety report -- I suggested to DOE that a super 20 coordinating group in which there was industry representation 21 to make sure that there was coordination, might be desirable.
22 Many things are lost.
23 It is clear to me that NRC doesn't quite know what 24 it is going to be getting from EPA and I know that DOE has 25 already shucked it off as too complex an issue for them.
l
)
1 332 Os 1
They, too, have a responsibility, and Ben Rusche is going to 2
have trouble administering this new activity.
But they have 3
not yet responded forcefully to what the criteria setting with 4
the EPA are.
5 Now, all of those issues in intergovernment level 6
should be resolved to mutual agreement before these major 7
policies hit the Congressional Record.
It places the industry 8
in the position of having to deal with the standard setter, 9
the guy who thinks he is going to execute, and others.
10 So, I think attention paid to formal cooperation 11 between the NRC, the EPA, the ACRS, DOE and certain other the NCRP and the ACRP, who are responsible for the 12 agencies 13 fundamental standard setting is important to safety in the 14 United States, and you all might want to look at it as part of 15 this review.
16 Those are the two issues that I wished to bring up.
17 I have stated them almost to the extent that I have thought 18 about them, but not quite.
19 There is another one which I will bring up, but I 20 don't have any hopes for this at all.
One of the great 21 internal coordinating responsibilities within NRC is to 22 coordinate the actions of the various review boards and panels 23 with the NRC itself.
There are other outside coordinating l
24 functions.
l 25 At one time when we were discussing the importance
333
(
1 of a five-man Commission, I suggested again when the 2
reorganisation acts were proposed, that each Commissioner be 3
given a portfolio.
4 The first portfolio go to the coordination of the r
5 various agencies; the appeals boards, the hearing boards, and 6
the NRC Staff functions and the NRC itself.
One Commissioner 7
has that portfolio.
8 There are potent relationships that must exist 9
between states and the federal regulatory commissions in 10 safety, and another Commissioner would have the responsibility 11 of coordinating with 50 States.
12 The third Commissioner would have the responsibility 13 of coordinating with the industry.
14 And the fourth Commissioner would coordinate among 15 the agencies in Washington who are responsible for safety.
16 Now, it is different than the existing set.
I still 17 think that these functi'ons are important, because on the 18 outside we work with all parts of this operation.
The 19 disconnected nature of what goes on is more apparent to those 20 who work within the agencies themselves.
21 I make those three suggestions now in the hopes that 22 there is some possibility of improving the exchange of 23 information and the resolution of significant problems in all A
24 of these three coordinating functions.
25 Now that is it.
I am finished.
334 1
MR. CARBON:
These strike'me as very interesting 2
ideas.
I guess I don't personally have much comment at the 3
moment.
4 MR. CULLER:
I don't expect you to.
5 Now, I think we have finished, have we not?
6 MR. TAYLOR:
We're going to be thrown out of hare if 7
we don't leave.
8 MR. CARBON:
I have at least a couple of questions.
9 Do you want to raise any?
I will raise mine right away.
We 10 didn't have down here, and you have not mentioned, but in the 11 long term is breeding of importance?
,/
12 MR. CULLER:
Sure.
13 MR. CARBON:
How long a term?
let me say 14 MR. CULLER:
I don't think that we 15 this.
I think the breeder was probably the more economical 16 form of reactor, and it maybe safer the way the system works.
17 As far as the classic argument of when do you have to have a 18 breeder to make sure we don't run out of uranium, I don't 19 think that that is anywhere near.
I think it is imprudent not 20 to have a breeder program to take advantage of one of the i
21 better technologies that is evolving.
I think you will find 22 the Russians are thinking of putting only breeders up starting 23 in 1990 in Western Russia.
n' l
24 MR. CARBON:
They are?
~
l 25 MR. CULLER:
Yes.
That comes out of the scientific
335 O)
(,
1 advisory committee meetings.
I think we will find that 2
running at no pressure and the high-heat-transfer 3
characteristics, we can design a breeder that will coast 4
itself down almost.
The British have operated their little 5
reactor up at Dune Ray at full power, shut the pumps off, and a
6 let it coast itself down by natural convection.
7 EBR-2 has done it.
8 In our work at Como, which is sort of coordinating 9
the big designs with the Department of Energy, we have looked 10 at the breeders, and we are certain we can design two or three 11 shutdown systems that will carry the reactor through by 12 natural heat removal.
fires, probably 13 There are problems with breeders 14 aodium fires being as important as any.
15 We think that it is a good reactor, and with a 16 little bit of care and thought, that it may be a reactor 17 system that should be introduced.
It's certain that other 18 countries are going to do so.
I think that it would be 19 imprudent to not continue in some breeder effort in the United 4
20 States.
21 We have worked out a long-and short-range plan 22 at the request of the Department of Energy and coordinated 1
23 with them basically what to do in the post-CRBR period.
We Cj g
4 24 have always worked on the long-range and never the CRBR.
So 25 these plans, if you want to review sometime and the thoughts
336 s
1 that exist, we would be happy to do so.
4 2
We think that the breeder is important and that 3
breeding is important.
4 MR. CARBON:
Can you see the government or EPRI or 5
groups of utilities putting up money to finance building 6
breeders soon?
7 MR. CULLER:
There is no possibility, no 8
possibility, that any reactor development will be funded by 9
the utilities, nad any new reactor, gas-cooled, that requires 10 demonstration, a breeder, whatever, that has to be built will 11 have to come from the government.
There is not that kind of f
12 risk money, nor do the utilities have the ability.to raise 13 money, borrow it, put it into the rate base to cover any new 14 nuclear system.
15 So if reactors are to advance, if breeding or 16 high-temperature reactors are brought in for chemical purposes 17 and high-temperature heat, it will be done with government 18 money.
19 MR. CARBON:
This surprises me a little bit, such a t
l 20 clear-cut statement, because I keep reading where four or five 21 utilities are supposedly serious about building a 22 high-temperature gas cooler.
23 MR. CULLER:
They're serious if the government pays g
24 for the prototype.
25 MR. CARBON:
The whole works?
337
'N 1
MR. CULLER:
Yes.
And the government is paying.
2 The utilities are putting in, what, $5-6 million?
3 MR. TAYLOR:
More like three.
4 MR. CULLER:
Three.
And the government is putting 5
in 50.
6 Now, I am not being -- I don't mean to be flip about 7
it, but the utilities do not have the latitude, because they 8
are regulated, of raising that kind of investment money.
The 9
biggest sum that was ever raised 'as for the CRBR, and it was w
10
$256 million.
And that could never be recreated.
11 MR. CARBON:
My last question.
Some people I think
(
12 believe that we may very well change the organizational 13 structure and have energy-producing companies or something 14 like that, selling to utilities who, in turn, will 15 distribute.
Could you give us your views on the likelihood 16 that anything like that will come about?
17 MR. TAYLOR:
I don't think I heard --
18 MR. CULLER:
Energy suply companies, people who 19 supply 2nergy and the utilities distribute it, say, a utility 2d who generates selling to utilities who sell.
21 MR. TAYLOR:
I think there i a probability for 22 that.
Right now it's happening in little, small ways.
Big 23 industrial companies are building some of their own 24 generators, in part, to use for their own energy's need but 25 also selling substantially in the PURPA legislation picture
,._,-._.y
338
/~s k,)
1 where they get marginal prices for that energy, two to three 2
times the average price which the utility costs, which the 3
utility has to sustain in generating power in its system.
4 I can see in the nuclear program, with a 5
stabilization of the regulatory situation and an evident need, 6
a group of investors picking up one of these plants that has 7
been canceled, on which billions have been spent at a very 8
attractive price, finishing it, and furnishing electricity to 9
the utilities.
Those things are beginning to happen in small 10 ways and can happen in big ways in such circumstances as I 11 have described.
12 MR. CULLER:
I believe there is, of all the energy 13 resources that can be used to generate power, nuclear may have 14 as great a need for a coordinated supply, maybe not one 15 company, but a company who would be.willing to design, build, 16 and operate.
And it has been considered by the utilities.
We 17 have suggested maybe regional supplies.
Of course, you know, 18 there are many people who use this as a major theme for 19 reorganization.
20 I think that the utilities will move in that i
21 direction to some extent.
There are other collective ways of 22 getting coordination, the simplest of which is to form an 23 engineering company which provides continuity in the O
24 engineering for the utilities, where as one company builds a 25 reactor once every ten years, the engineering responsibility
1 339
\\
1 moving from one plant to another provides for a learning curve 2
and a continuity of experience.
And all of these ideas have 3
been discussed.
4 Right after Three Mile Island I proposed a customers 5
engineering group to accumulate this kind of experience rath'r e
6 than having each utility.
It was not exactly rejected, it 7
just wasn't thought about.
But as time goes on and 8
competitive pressures and cogeneration emerge, the utilities 9
are thinking more and more about different ways of configuring 10 the industry.
And within EEI now there is a very high-level 11 committee looking at competitive forces and how they may 12 affect in the future the structure of the industry itself.
13 It's all set up, and some of the thoughts that 14 you're asking about will be explored in this policy analysis strategy and policy analysis group.
15 16 MR. REMICK:
John, did I understand you to say -- I 17 might have misunderstood what you said.
I thought you said 18 in the cases where the utilities are being forced to buy the 19 cogenerated electricity, it is at two to three -- what would 20 it be, the utilitiy costs?
21 Is that what you said?
22 MR. TAYLOR:
The average cost of the system.
i I
23 They pay the avoided cost, which is the cost that O.
24 would be associated with the newest facility they could build.
25 MR. REMICK:
That is in the range of two to three --
340 0
1 MR. CULLER:
Idaho Power has about 5 mil power, a 2
little bit less than that.
Becaust they have a lot of hydro, 3
I think it's about a 4 mil power.
In a small unit that's 4
coming on line.
They have to pay 9.5 mils or 10 mils for the 5
avoided cost of the maximum cost because they can't add more 6
hydro.
As a consequence, they have taken it to court.
7 MR. REMICK:
Earlier, Floyd, you said something 8
about hydro being very inexpensive.
9 In Japan I found it was the highest cost electricity 10 of their four modes.
They list that as the highest.
th'nk so, Forrest.
11 MR. TAYLOR:
I don't i
\\
(J 12 MR. REMICK:
The reason, they said, is the capital 13 costs.
They're doing seismic considerations of dams.
14 MR. TAYLOR:
Maybe they're talking about a new plant 15 because their present plants --
16 MR. REMICK:
I have to check my notes.
I believe 17 this was the basis of existing plants in Japan.
Hydro was the 18 highest.
I have the figures somewhere, amps per kilowatt 19 hour2.199074e-4 days <br />0.00528 hours <br />3.141534e-5 weeks <br />7.2295e-6 months <br />.
Nuclear lost.
It was just slightly lower than coal, 20 It was surprising.
We asked questions about it.
21 Don Rubio, by the way, I think was on that same 22 MR. TAYLOR:
I think he was on that same trip.
23 MR. REMICK:
Where that discussion was held.
I am t
24 trying to think if he was there.
He was there certain days.
j 1
25 MR. CULLER:
I wanted to tell you something else
341 O
\\s, 1
about the cogeneration.
2 There are several locations in the United States 3
where essentially a kind of abuse is occurring.
There are 1
4 many cogenerators who will put in a steam turbine and sell 5
in some states it's only ten percent of the total heat to 6
generate steam; and the rest of it, 90 percent of the energy 7
goes into generating electricity.
And Don Jordan in Houston 8
Power and Light has about 5,000 megawatts proposed this way on 9
marginal cost replacement.
10 So that a fully well-intentioned program is being 11 abused in locations around the United States now, and some l'2 relook at this whole operation has to occur.
13 MR. CARBON:
Some legislatures are already relooking 14 at it.
15 Well, considering the time, let me thank you both 16 very much for an excellent series.
17 MR. TAYLOR:
-I am honored you would sptnd this much 18 time with us.
19 MR. CARBON:
I also thank the subcommittee members.
20 (Whereupon, at 9:40 p.m.,
the meeting was adjourned.)
21 22 23 24 25 1
~.
%s/
1 CERTIFICATE OF OFFICIAL PEPORTER 2
3 4
+
5 This is to certify that the attached proceedings 6
before the United States Nuclear Regulatory Commission in the 7
matter of:
ADVISORY COMMITTEE ON REACTOR SAFEGUARDS e
9 Name of proceeding: Subcommittee on Long-Range Plan for the Nuclear Regulatory Commission 10 11 Docket No.
()
1e place:
Washington, D. C.
i ta cate:
Thursday, July 11, 1985 14 15 were held as herein appears and that this is the original 16 transcript thereof for the file of the United States Nuclear 17 Regulatory Commission.
13 (Signature) g g g (Typed. Name of Rep 6rter)
Mimie Mefizer
~20 21 22 23 Ann Riley & Associates. Ltd.
24 25
-