ML19247B862
| ML19247B862 | |
| Person / Time | |
|---|---|
| Issue date: | 08/08/1979 |
| From: | Bender M, Etherington H, Seiss C Advisory Committee on Reactor Safeguards |
| To: | |
| References | |
| ACRS-T, NUDOCS 7908150053 | |
| Download: ML19247B862 (56) | |
Text
,n._-,,,--..
^ * ~ " ~ - -
u-
((M
^
Y' 3
lA
%e.
Y NUCLE AR DEGUL ATO RY COMMISSION r.
}g
's
^
ADVISORY COMMITTEE ON REACTOR SAFEGUARDS
- =
rh k'v-T T,
we
.-A IN THE MATTER OF:
)
SUBCOMMITTEE MEETING on 5
REGULATORY ACTIVITIES b
'g-
.v+
-. < g,.
r a
(;<
Jh Place.
Washington, D.
C.
Date..
Wednesday, 8 August 19 79 Pages 1 - 56 F.1 n
/??
,1 (
O l (,
- 'T J n :: e' (20:} 34-nco Telechone:
x.'.
[~
8
',1'k
~
. s
~
E f
4 t
j ~ : r,,
-a?
ACE -FEDERAL KE'OllTERS,1NC.
>b Official Rep <>ncrs
%h Ac 444 North Capitoi Street ci Wcshington, D.C. 20001 -
- 7
' ML 7908150 p [.3
- NATIOMWIDE COVERAGE. DAILY
[:, ; ~ :
.x,,
1 I
PUBLIC NOTICE BY THE CR6342 h
2 UNITED STATES NUCLEAR REGULATORY CO:G!ISSION' S 3
ADVISORY CO?e1ITTEE ON REACTOR SAFEGUARDS 4
,Te dne s day, 8 August 1979
'S 6
The contents of this stenographic transcript of the 7
proceedings of the United States Nuclear Regulatory 8
Commission's Advisory Con'Jaittee on. Reactor Safeguards (ACRS),
9 as reported herein, is.an uncorrected record of the discussions 10 recorded at the meeting held on the above date.
11 No member of the ACRS Staff and no participant at this 12 meeting accepts any responsibility for errors or inaccuracies 13 of statement or data contained in this transcript.
14 15 16 17 18 19 20 21 22 23 24 ce-Federal Reporters, Inc.
25
2 CR6 342 1
UNITED STATES OF AMERICA 2
NUCLEAR REGULATORY COMMISSION 3
ADVISORY COMMITTEE ON REACTOR SAFEGUARDS 4
SUBCOMMITTEE MEETING 5
on 6
REGULATORY ACTIVITIES 7
8 Room 1046 9
1717 H Stree t, N.
W.
Washington, D.
C.
10 Wednesday, 8 August 1979 j
I The ACRS Subcommittee on Regulatory Activities met, pursuant 11 to notice, at 8:45 a.m.,
Dr. Chester P.
Siess, chairman of the 12 sdacommittee, presiding.
13 BEFORE :
14 DR. CHESTER P.
SIESS, CI: airman of the S ubcommittee 15 MR. MYER BENDER, Member 16 MR. HAROLD ETHERINGTON, Merrbe r 17 PROF. WILLI AM KERR, Member 18 19 20 21 22 23 k
24 Ace Federal Reporters, Inc.
25 ul ? o,'
- A9
- ts
3 6342 01 01 l hhp i
DR. SIESS:
The mee ting will come to order.
This 2
is a meeting of the ACRS Subcommittee on Regulatory 3
Activities.
I'm Che ster Sie ss, Subcommittee chairman.
The 4
o ther ACRS members present today are Harold Etherington and 5
William Kerr.
The. purpose of the meeting this morning is to 6
discuss.two ma ttersi one is a proposed Regulatory Guide, or 7
Provision 2 to Regulatory Guide 1.136, Ma teri al,
8 Constructicn and Testing of Conrete Containment.
The o ther 9
is a proposed limi ted revision, it says here, of Appendix J 10 to 10 CFR 50 on leak rate testing as it pertains chiefly to 11 airlocks and pene trations.
I believe that's Type C, is 12 that?
13 MR. ARNDT:
B.
14 DR. SIESS:
Type A is the integrated leak rate.
15 Type B is penetration in airlocks.
What's type C?
16 MR. ARNDT:
Type C is the small pane tra tion s.
17 DR. SIESS:
Okay.
This meeting is being conduc ted 18 in a ccordance with the provisions of the Federal Advisory 19 Commi ttee Ac t and the Government in the Sunshine Ac t.
My 20 Gary Qui ttsch eiber, seated at my right, is the designated 21 federal employee.
The rules for participation in today's 22 meeting have been announced as part of the notice tha t wa s 23 published in the Federal Register on July 24th.
A kh 24 transcript of the mee ting is being kept, and will be made 25 available as stated in the Federal Register notice.
'de have
_;9
D '
r4 o
v.i
'342 01 02 kap i
received no written comments or requests for time to make ggg 2
oral statements by members of the publi c.
3 We did receive, dated Monday, just for the members O
4 of the Subcommi ttee, f rom Sam Duraiswamy, some revisions to 5
Part 50, Appendix J.
Did everybody ge t those?
6 PROF. KERR:
Yes.
7 DR. SIESS:
I'll try to factor those in.
I 8
glanced at them.
I didn't try to integrate them.
What 9
order would you like to take things up, Bill?
10 MR. MORRISON:
We'd just as soon take up Appendix 11 J first, but it's not crucial.
Whatever your pref erence is.
12 DR. SIESS:
Okay.
Appendix J, as I understand it 13 this is essentially an a ttempt by the staff to write in to 14 the regulations what they have been doing for a numbor of 15 years, in o ther words, codify the exceptions that you've 16 been making.
That's my recollection, because I've been 17 seeing reports back and f orth, or le tters back and forth, 18 between the staff and licensees on testing airlocks.
Am I 19 right that the things tha t every time you open an airlock 20 you nave to test it to f ull pre ssure?
21 MR. ARNDT:
That's correct.
22 DR. SIESS:
To do tha t, you have to put a strong 23 back on the inside door, which means you have to be inside.
24 MR. ARNDT:
I don't know if you had to be inside, 25 but you did have to put a strong back on.
One of the O
5 342 01 03 p
i problems was tha t when the current Appendix was i ssued, it 2
really didn't have provisions for existing designs, on what 3
are now the older airlock designs, and on a number of plants 4
this created a problem because it was a pretty hard and f ast 5
position as stated in the regulation.
And we found it 6
nece ssary, because of the design of the airlocks, the older 7
ones, to grant some ex em ption s.
8 DR. SIESS:
How did the airlock design diff er?
9 Wha t is there about a current design airlock that allows you 10 to do a full pre ssure test?
11 MR. ARNDT:
I think there are several 12 differences.
One is on the design pressure tha t the airlock 13 could be tested a t.
Another is whether they have double 14 seals on the doors or not.
15 DR. SIESS:
Now, if they've got double
.c tls, you 16 let them test the seals.
17 MR. ARNDT:
Right.
I believe that on some of the 18 older ones they didn' t have that possi bi lit,,
therefore, 19 you had to test in be tween the doors, you had to test the 20 c hamb e r.
21 DR. SIESS:
W ha t I thought you were saying is that 22 Appendix J as written could be complied with for the newer 23 designs.
24 MR. ARNDT:
Yes.
At the time it was written, it 25 was written looking ahead to what was anticipated in the way G
I
6 6342 01 04 p
I of designs coming out.
2 DR. SIESS:
Bu t A ppendix J, as wri tten, still 3
didn't let you just test between the seals, did it?
4 MR. ARNDT:
No.
5 DR. SIESS:
So that would have meant you still had 6
to test the airlock itself, internally.
7 MR. ARNDT:
Right.
8 DR. SIESS:
Does tha t mean that the new designs 9
have a stronger fitting on that inner door or something?
An 10 inner door opens inside, so you depressurize the lock and 11 you try to open tha t door.
12 MR. ARNDT:
Correct.
13 DR. SIESS:
And the new designs permit you to 14 pressurize that thing up to f ull pressure?
15 MR. ARNDT I believe they do, yes.
16 MR. ETHERINGTON:
Pressurizing the lock requires a 17 seal in the wrong direction and it doesn't test the seal in 18 the right direction, of the inner door, isn't that right?
19 MR. ARNDT Tha t's right.
20 MR. ETHERINGTON:
So it i sn' t a good test in that 21 respect.
22 MR. ARNDT:
I t's wha t we had.
23 DR. SIESS:
You could argue, Harold, that if the 24 door would hold against the pressure trying to open it, it 2S woula be even tighter against the pressure trying to close g,6
- Dc
7 6342 01 CS llhp I
i t, but I'm not sure tha t's nece ssarily true.
2 MR. ETHERINGTON:
No, I don't think it is, I mean, jgg 3
the seal is on the wrong side.
4 DR. SIESS:
I mean if it was just the door that 5
came up against the gasket like this, it would obviously 6
be --
7 MR. ETHERINGTON:
Yes, if that's all it were, yes.
8 MR. ARNDT:
Now, as f ar as the changes going in 9
here, we pointed out the major change, what we're doing is 10 really in the Appendix, revising one paragraph at this time.
11 DR. SIESS:
Now, let's see, the new sheets we got 12 essentially replace the old ones.
13 MR. ARNDT:
Correct.
There was one sheet, 14 replacement sheet, on the te; of the paragraph.
15 DR. SIESS:
The first page was the old Appendix.
16 MR. ARNDT:
Correct.
17 DR. SIESS:
of course, that's still valid and 18 p.ses two, five, six -- okay.
19 MR. ARNDT:
page two, enclosure one, which was the 20 first page of the pro po se d new te x t, was revised.
We sent a 21 replacement page for that, where we added the phrase "or 22 within 24 hours2.777778e-4 days <br />0.00667 hours <br />3.968254e-5 weeks <br />9.132e-6 months <br /> of opening, when the reactor is in an 23 operating mode requiring contain aent integrity."
The other 24 three replacement pages were in the a ttachment to the value 25 impac t statement which addressed the changes -- excuse me,
8 342 01 06 p
I I lef t out one other, there was an editorial change on that 2
text on page two, enclosure one.
At the bottom of the pa ge,
3 we inserted the words no t le ss than."
4 So for these two changes in the text that were 5
sent down, we had to make corresponding changes to the value 6
impact statement.
7 DR. SIESS:
So the underlined material is the 8
change to the change.
9 MR. ARNDT:
Correct.
The underlined material is 10 the result of some comment received af ter the package was 11 sent down to the ACR5.
Now, the fundamental change is to 12 eliminate the requirement on every plant to test the airlock 13 af ter every single opening and to use a position which is 14 consistent wi th what we have been using in individual 15 exemptions that have been granted, which is on a three-day 16 maximum period.
We have tightened up the wording a li ttle 17 bit also, with respect to the penetra tion which might have 18 been tested but then reopened before you go back to 19 operation or maybe af ter you've gone back to operation.
20 DR. SIESS:
Let me interrupt you a minute.
When 21 you talk abou t opening the penetration, this i be maybe 22 electrical penetration tha t has a plate over it..ith a 23 gaske t?
24 MR. ARNDT:
What we started out considering was a 25 situation which I understood had occurred, where a type A
.sa n,o
6342 01 07 9
llbp I
test had been run following a type A test on equipment ha tch 2
had been used.
It had been opened, but before going into 3
opera tion.
4 DR. SIESS:
I understand about hatches.
5 MR. ARNDT:
Tha t came under the type B test.
6 DR. SIESS:
But I'm trying to make a distinction 7
between hatches and pene tra tions.
8 MR. ARNDT Right.
9 DR. SIESS:
And I'm trying to visualize opening a 10 penetration.
Now, there are all kinds of penetrations.
A 11 pipe penetration is a built-in piece of equipment, it's got 12 all sorts of sleeves and stuff.
What kind of penetrations 13 can be open?
14 MR. ARNDT:
Electrical penetrccions, I understand, 15 on the one plant had been opened during operation because of 16 some repair maintenance work.
It had to be done.
17 DR. SIESS These are things that have a bolt at 18 the cover of some thing that you can take off and get in and 19 work on the insulaticn.
Are the connectors inside there, 20 the elec trical penetrations?
21 MR. ARNDT:
I'm not that f amiliar with them.
It's 22 been a long time since I've seen a diagram of the electrical 23 penetration, but I don't bclieve they have connec tors in 24 them.
I think it is a atinuous wire through a penetration s
25 with a potting compound, which surrounds the wire and s,
10 6342 01 08 llhp i
directed in such a way that you wouldn't have any streaming 2
path directly through the penetra tion.
If you had any gap 3
between the potting compound and the wire --
4 DR. SIESS:
This is mainly aimed at airlocks, 5
manlocks, equipment locks, but it doe s also cover 6
7 MR. ARNDT:
Correct.
8 DR. SIESS:
You consider an airlock a penetration?
9 MR. ARNDT:
No, we don't, in the sense of this 10 paragraphs.
Airlocks, we have covered by themselves with 11 the sub-paragraphs here under 2-B.
The pene tra tions would 12 include those tha t are no t airlocks, personnel or escape 13 locks, would include the equipment hatch, which is not a 14 double door unit, it's a single door unit.
And i t would 15 include electrica] mechanical pene tra tion s.
Some of them 16 are individually tested.
Some are tested as multiple units, 17 and that's covered by the portion of this paragraph under 18 2-A.
19 DR. SIESS:
Now, what kind of penetrations employ 20 a continuous leakage moni toring system?
These are 21 penetrations that are pre ssurized internally, and with some 22 means of monitoring leakage?
23 MR. ARNDT Right.
I know they exit.
I don't 24 have examples of which ones.
25 DR. SIESS:
What about a containment like
342 01 09 11 llp i
Conn. Yankee, Haddam Neck?
If I'm not mistaken, Haddam Neck 2
keeps pressure on the containment at all times.
They can (gg 3
argue they've got a continuous leak rate test on the 4
containment; is that true, do you know?
5 MR. ARNDT:
I'm not familiar with that particular 6
plan t.
7 DR. SIESS:
Or sub-atmospheric.
8 MR. ARNDT:
The subject of continuous leakage 9
monitoring, we may be giving some f urther thought.
The re is 10 planned anather revision of A ppendix J, which is a general 11 revision, not just this paragraph, but there are a number of 12 clean up items as far as clarifications.
13 A lot tha t we've learned in the years that it's 14 been out, that could be improved.
We plan to make that IS general revision.
As part of that we are considering 16 looking at that continuous leakage monitoring system and 17 wha t kind of credit can be given to it in the leakage 18 program.
We decided this time with this paragraph not to 19 pursue any change to the wording as we have it on that.
20 DR. SIESS:
Why did you make the change to, not 21 less than p sub A.
Did somebody want it to go higher?
22 MR. ARNDT To put it in line with the previous 23 paragraph, which says, not less than, and the fact that if 24 we left it tested at p sub A, it means specifically at a 25 single pressure level, no more than, no le ss than.
And that o~7 c
ai
6342 01 10 12 ll)p I
wasn't critical to us.
2 DR. SIESS:
Who caught that, the lawyers?
gg) 3 (Laughter.)
4 MR. ARNDT No, our tech spec people caught t ha t.
5 DR. SIESS:
P sub A is what?
6 MR. ARNDT:
Design Accident Pressure -- excuse me, 7
i t's the peak calcula ted.
8 DP. SIESS:
Accident pressure which is a little 9
less than design pressure.
Harold, do you have any 10 questions?
1I MR. ETHERINGTON:
No.
12 DR. SIESS:
Bill.
13 PROF. KERR There was one place I was looking 14 for, where it seemed to me one might also have considered 15 inser ting "no t le ss tb -a p sub A. "
I thought I wrote i t in.
16 MR. MORRISON:
The bottom of page two, enclosure 17 one.
18 DR. SIESS:
It's down there.
On page three of 19 enclosure one.
20 M R. /> c.'J DT I t hink I see where you mean.
Airlock 21 door seal testing shall not be substituted for the six-month 22 test of the entire airlock at p sub A.
23 DR. SIESS:
To be consi sten t, I guess i t could khh 24 rca d, not le ss than.
25 PROF. KERR:
There's also, on page ^ight of eEz s
- SS
342 01 11 p
i enclosure two, a similar one which says p sub A.
2 DR. SIESS:
That's in the value impact statement, 3
Bill.
4 PROF. KERR:
Yes.
5 MR. ARNDT Yes.
6 DR. SIESS:
In which one, seven?
7 PROF. KERR Yes.
8 DR. SIESS:
That says pA or reduced pre ssures, but 9
t ha t is not the same kind of --
10 PROF. KERR In the event of the testing, it 11 canno t be at p sub A.
You ought to check.
I ju s t thought 12 it seemed a li ttle inconsistent.
13 MR. ARNDT:
Yes.
This re peats the wording that 14 shows up on page three, enclosure one.
And if I change one, 15 I'll c hange t he o t he r.
16 PROF. KERR:
Ot he r t han t ha t, C he t, I have no 17 comments.
18 DR. SIESS:
Let me ask, in general, a couple of 19 questions.
I've been reading LERs and in a number of 20 instances where they had a leak in the airlock it was 21 because the seal got damaged, something got caught in the 22 seal.
23 MR. ARNDT:
Right.
24 DR. SIESS:
I s there something either in tech 25 specs or just normal operating procedures that says when you
' ~ %,,,
14 342 01 12
~N.,..,.
p 1
open'that lock and you close it, that somebody sort of looks j
2 at the seal'tb be sure there's nothing s ticking,JfA-6 that w',
n 3
they didn't leave a wnid,1,ng cable rt;;;mi$Through there or
%.._ p' 4
something else?
Do you know wha t they do in the plant?
5 MR. ARNDT:
Af ter reviewing that information you 6
referred to, it occurred to us also about maybe putting 7
something in about' a visual in s pe c tion.
"Look at the seals 8
before you close the door."
However, we felt that that was 9
not. really a ppropria te in a regulation.
10 DR. SIESS:
No.
I just wondered what people do.
11 MR. ARNDT This is something that I think we'll 12 be taking up with our tech spec peo pl e.
13 DR. SIESS:
If somebody just goes in and comes 14 out, they're not likely to damage it, but if they had to do 15 something in the pl an t, they've hauled stuff in there --
16 PROF. KERR :
They also might put in something that 17 says, " Don't get your f oo t caught in the door when you close 18 it."
19 DR. SIESS:
Tha t's hard to do, with those doors.
20 MR. ARNDT If there's a reasonable way of 21 emphasizing that to the plant personnel wi thout imposing on 22 them some kind of paper requirement tha t i s going to burden 23 them, where the NRC inspec tor is going to come in and say, 24 Show us the log that you initialed that every time you close 25 this thing, you also swore that you looked at the seal.
h'
_ n)
15 342 01 13 kp i
DR. SIESS:
Well, the way you encourage tha t is to 2
penalize people for not doing it, I guess.
Does that test 3
frequency go up every time they have a f ailure?
4 MR. ARNDT No.
They test until things are the 5
way they should be, and test frequency is not altered.
6 DR. SIESS:
That brings me to another 7
question. Before
...ee Mile Island came along, we had an 8
item for f uture discussion in ACRS on leak testing.
I 9
wanted to get the staff in and talk about leak testing in 10 general, type A and the other types.
And I m'ght as well 11 bring it up here, so you guys can start thinking about it.
12 I think we'll get back to it one of these days.
13 I've been reading integrated leak test reports 14 that come in, and very frequently, when they start making 15 the integra ted leak rate test, they can't get the thing 16 pumped up because somewhe re there's a leak.
They go around 17 and they fix all the leaks and then they finally make the 18 test.
And when they get through wi th the test, it's down at 19 a half or a tenth of one percent of whatever the tech spec 20 figure is, and it just goes along like that for another five 21 years, I gue ss, whatever the period is.
22 I couldn't ge t an awful lot of con f iden ce in the 23 leak rate test when you had to patch up all the leaks before 24 it could pa ss i t, but just before you made the test, the 25 leak rate might be just about anything.
9 6/6
'6i
16 6342 01 14 2p 1
MR. ETHERINGTON:
But they are penalized, aren't 2
t hey, if they have to fix things?
They're per.alized by (g) 3 requiring more frequent testing.
4 DR. SIESS:
No, not on the integrated testing.
5 MR. ETHERINGTON:
On the initial one, you mean?
6 DR. SIESS:
The one they make every five years.
7 MR. ETHERINGTON:
I don't think they should be 8
allowed to go around fixing things without penalties.
9 DR. SIESS:
A ppa ren tly, they can.
Now some of 10 these leaks aren't leaks, in other words they're not 11 necessarily that stuf f would leak to the atmosphere.
It 12 would just go through the first valve into another system or 13 some t hing.
Every penetration doesn't go to the atmosphere, 14 right?
15 Some of them go to closed systems outside 16 containment, in fact, into the PHR, so I've never been able 17 to get a feel for the significance of an integrated leak 18 rate te s t, for example, where there was a greater leakage 17 than was allowed, because what I'm interested in is leakage 20 that provides some pa th to the atmosphere.
Obviously, I 21 think maybe we've learned f rom Three Mile.
We haven't got 22 the whole pic ture ye t, but if stuff gets into the secondary 23 side -- not secondary side, but auxiliary side, it can 24 eventually get to the atmosphere, going out a leaky pump 25 seal through fil ters, e t ce tera.
So something ge ts ou t
_L
342 01 15 17 4
4 tru'p 1
eventually.
All the systems aren't watertight.
2 3
4 5
6 7
8 9
10 11 12 13 14 x
4 I5 d
16 17 18 19 o
20 21 22 23 24 25
n 342.02.1 18 l
But I was wondering if rnybody has made any kind qggsh 2
of a study, I guess it might have to be probabilistic, as 3
to wha t is the probability at any given time that the leak 4
rate is below some value.
5 It's not a simple system where leak rate increases.
6 And at the end of five years, you pull it down back here.
7 It goes back up, you pull It back down to here and you never 8
let it get above a certain value.
9 I know you think of some degradation because that's 10 why the leak rate test has to be below the total allowable 11 leak rate.
12 You're supposed to allow for some degradation 13 between tests.
But there's just no indication that the 14 degradation is any kind of a continuous thing.
It's some 15 valve that's not closing tight.
That's been the fix on those 16 cases, take a valve out.
17 How much confidence do we have that at any given 18 point in time if we pressurize that containment to TA, the 19 leak rate would meet the part 100 limit, for whatever that's 20 worth?
21 MR. ARNDf: I don't know whether we have any 22 probability such as you mention, whether the allowable 23 leakage limit is existing at any given po int in time.
But 24 Appendix J, in setting up the type B and type C tests, type B ggg 25 for the penetrations and type C for the containment isolation
[
/
(3 f
' l
342.02.2 19 ggph valve 3 establishes ditterent frequencies, di f f e rent intervals 2
on these most likely sources of leakage. And in between your 3
type A, integrated leak rate test, you'll be performing these 4
local and individual tes ts on your most likely sources of 5
leakage.
6 And I think that the combination of these has given 7
us, qualitatively, net quantitatively, a feeling that we have, 8
or are maintaining an acceptable leakage level in that 9
structure.
10 DR. S I E SS : I'm inclined to agree, but I'd get a J1 lot more confidence out of type 3 and C and the airlock tests, 12 which are made at frequent in t er va l s, than I do out of the 13 integrated leak rate test.
14 And I guess one of the questions in the back of 15 my nind is what puroose does the integrated leak rate test 16 serve?
17
'Ihere are no sources of leakage other than 18 pe ne tra t io n is ol a t ion valves and airlocks.
The staff always 19 assumes that something gets out somewhere through the concrete 20 or through the liner or something else.
21 I've s een enough concern about minor welds, you know, 22 because when they make an integrated leak rate test and 23 they don't n ee d it the first time, it's invariably a pre tty 24 g ood leak.
You don't get big leaks tnrough penetration, 25 right?
342.02.2 20 sh I
When penetrations fail, a leak rate, a t.se B test, 2
it's not a big amount.
When an isolation valve f ails, a 3
type C test, it's usually a pretty good leakage because the 4
valve didn't seat right.
5 MR. ARIIDT: Right.
6 DR. SIESS: Or an airlock.
7 MR. ARNDT: It may be also that the valve wa sn't 8
closed the way it should have been closed.
That cou.! d provide 9
a leakage path.
10 DR. SIESS: The bi ggest potential for a leak is a 11 valve.
12 MR. ARNDT: Right.
13 DR. SIESS: How when you do a valve test, I gue ss 14 type C isolation valve, what you do is_ pressurize between the 15 two isolation valve le vels.
So if one of them is leaking in, 16 it still shows as a leak.
17 So that's a conservatism, I guess, Harold.
18 MR. ETHERIliGTOI!:
Yes.
19 DR. SIESS: If one of them is leaking in and one of 20 them is leaking out the same amount, you really get twice 21 as much leakage as would be important.
But you count it that 22 way.
23 So there's a conservatism in that.
24 The same way on pene trations, right?
Penetration is 25 ususall y closed off at both ends.
It has to be.
If you want
- f. 'i
? C f) a
~:
342.02.4 21 sh 1
leak testing, yo u internally pressurize the penetration, don't 2
yo ?
u 3
MR. ARNDT: I think there
.re different ways of 4
testing tae penetrations.
They may Le one-sided tests such 5
as a channel around the perimeter, or, for example, a bellows-6 type of penetration between two, say, skins of the bellows, 7
testing in between.that rather than a linear unit.
8 DR. SIESS: Has anybody ever made a study of the 9
type B and C tests to see what the failure rate is, for 10 example?
If we were going to go into a probabilistic analysis il and how much the y f ailed, whether I'm cc rrect in my f eeling 12 that it's type C on isolation valves that usually represents 13 the biggest leakage and not type Bs on penetrations.
14 MR. AHNDT: Fron what I've heard, that is the case.
IS But I don't know of any studies specifically.
I don't know 16 if anybody else has heard of any.
17 DR. SIESS: Any failure on a type B or C test 18 requires an LER, doesn't it?
19 MR. ARUDT: Yes.
And we get a list o f those.
And 20 in preparing this change to this paragraph, we had extracted 21 from the LERS going back some time what kind of failures had 22 been related to airlocks specifically.
And we came up with 23 the inf ormation you mentioned, that a lot of this is really 24 something that can be visually picked up -- dirt on the seals, 25 the seal is out of its groo ve, there's an obstruction which F:6
- oi
342.02.5 22 sh I
lays across the seals and prevents the door f rom closing 2
properly
- readily apparent situations which could lead to 3
leakage.
4 But I don't know that there have been speci fit 5
studies made to take these failures and list them and relate 6
them to the general population and say, we ll, the probability this type of penetration failing is X-times that other, 7
t 8
you know, that general population 9
DR. SIESS: And the f ailure of a type D or type C or e en an airlock test does not nece ssarily mean you would v
10 11 have had leakage in an integrated leak rate test.
12 MR. ARNDT: That's correct.
13 DR. SIESS: The inner door or the airlock could be 14 leakin), not awful, but if you pressurize the.ontainment, 15 the outer door would have held it.
16 MR. ARNDT: Correct.
17 DR. SIESS: And the same could be true on some 18 isolation valves where there are check valves.
You get the 19 pressure on then from the check.
20 But think about the question of what function the 21 integrated leak rate test serves.
It nust serve some function.
22 I t can't be extremely important because you only require it 23 e verY five years.
24 MR. ARNDT: It performs a function that reassures us 25 that the re hasn't been something overlocked, that we haven't
.oc
342.02.6 23
>h I
covered by installed leakage testing capability in some local 2
pe ne tra t io ns, in such tremendous number of penetrations 3
through this structure that it gives us that assurance, at 4
least early on, that everything has been really covered.
5 DR. SIESS: Not just early on.
It's more likely after 6
30 years that somebody's forgotten to do something.
7 MR. ARtJDT Also, it covers the situation where you 8
might hypothesize that the containment liner has deteriorated, 9
for one reason or another, and it has not been picked up in 10 the course of the operation, or through other existing 11 detection systems.
12 For example, some plants have leak chases all over 13 the containment liner welds and they can test those weld seams 14 directly without testing the whole containment.
15 But only some plants have that.
16 So the overall leakage test makes :;ure we haven't 17 missed something and also tells us that on the general surf ace IS of the containment, there has been no undetected deterioration 19 of that surface.
20 Many people consider that very unlikely, but it's 21 still something that has to be checked.
22 So you can feel sure that there are no small leaks 23 involed.
24 DR. SIESS: Le t me postulate something.
Let's take 25 a pre-stressed concrete containment with a steel liner and g
342.02.7 24 sh 1
let's go in and drill a one-1,ch hole in the liner and then 2
run an integrated leak rate test.
klh 3
Do you think you'd find that hole?
4 MR. ;WNDT: With no pressure on the containment?
5 DR. SIESS: An integrated leak rate test.
6 MR. ARNDT: Probably not.
7 DR. SIRSS: That's P/A.
It's a half P/A.
8 MR. ARNDT: I doubt that you would.
9 DR. SIESS: I doubt you would, too.
It would have 10 to go between the liner and the concrete far enough to fi d
.11 an outlet somewhere, wouldn't it?
12 MR. ARNDT: And the pre-stressing would probably 13 keep it uncracked.
14 DR. SIE SS: What about non-pre-stressed?
You might 15 pick it up, right?
16 MR, ARNDT: fhere, you'd again
- hal f of the 17 calculated peak accident pressure, the stress has been very 18 low on otter rebar, and it's unlikely that you would open up 19 anY new or existing crack paths through that concrete at that 20 pressure level.
That structure would still be very lowly 21 stressed, and there shouldn't be any direct paths from the 22 liner out to the outside.
23 DR. SIESS: That sort of shoots down your integrated 24 leak rate test for liner deterioration.
25 MR. ARNDf: It goes, except there is some --
m Ec 342.02.8 25 sh I
DR. SIESS: Except for gross materials.
2 MR. ARNDf Some thought being given and, ac t ua lly,
3 in many cases, being done, not to use the half P/A as the 4
integrated leak rate test pressure
- that's outside the 5
subject of what we're discussing here.
6 DR. SIESS: That was one of the ones I would come to 7
next.
8 MR. ARNDT But going to the full calculated peak 9
accident pre ssure as the type A test pressure, and at this 10 point, f or conventionall y pre-stressed reinf orced concre te J1 containment, you're raising your stresses to a point where 12 maybe you'll see something.
13 DR. SIESS: Pre -s t re ss e d, you still wouldn't crack it.
14 MR. ARNDT: Not in pre-stressed, no.
15 DR. SIESS: Now wha t's the status of that.
When we 16 were discussing surveillance on, what was it, unbo nd3 d 17 containments or something, somebody said that the leak rate 18 people were planning to go to P/A.
Therefore, we'd e ssent ia ll y 19 get full pre ssure test on a containment; you know, like the 20 s truc tural integrity test every five years.
21 Has that been slowed down?
22 MR. ARNDT: What we did initially, that Regulatory 23 Guide 1.90 on pre-stressed inspection, we were trying to 24 dov tail with what we understood was developing in respect to e
25 Appendix J at that point in tine and anticipate what the change bln
342.02.9 26 h
1 might be in the direction it had been going -- as f ar as I 2
know, it's s till going -- is to pref er a full pressure" vest 3
rather than a half pressure test for the integrated leak 4
rate test.
5 DR. SIESS: Do you remember the meeting we had with 6
the German RSK last fall to discuss leak rate testing? I 7
think Charley Hoffmeyer might have been there.
8 MR. ARNDT: I was at one meeting with them. but I 9
don't recall it dealing with integrated leak rate testing.
10 DR. SIESS: You might try and find the report that
~
11 we made on that meeting.
They had some comments about the
[
12 integrated leak rate test at low pre ssures.
And I guess that 13 they were talking -- I can't remember now whether they were 14 talking about high pressures or not -- but they thought that 15 the scatter and the results at low pressure just didn't give 16 them much confidence.
17 MR. ARNDT: I think that there's quite a controversy 18 between advocates f or low pressure and high pressure testing.
19 Both sides feel that their condition will reveal some 20 situations that the other test won't.
21 For example, if you take a penetration with a 22 potting conpound or a compression seal, as you observe a 23 higher pressure on it, that's going to tighten up more, it's 24 going to pack tighter, it will seal tighter.
Therefore, maybe 75 your low pre ssure test will be considered a more valid test
342.02.10 27 llhsh I
than your high pressure test.
2 Howe va r, if you have penetration of a different type 3
where you don't tend to seal the penetration pattern with the 4
higher pre ssure, then your high pres sure test would be the 5
more valid one because in that condition, you're going to 6
release more through a given opening.
7 DR. SIESS: Now to get back to your concept of the 8
reason for the integrated leak rate test, there's a couple 9
of things.
One is that it could disclose gross dc Jradation 10 in some portion of leakage barrier that was not subjected
.11 to type B or C tests, such as the liner or the concrete or 12 wha te ve r it might be, aren't right.
13 The other one, I guess wasn't that clear?
I can put 14 it this way:
If it could detect leakage through a penetration 15 that for some reason had not been subjected to a type B or C 16 test --
17 MR. ARNDT: Right.
18 DR. SIESS: You know, things do get overlooked 19 sometimes.
If that's true, then the fact that orior to making 20 an integrated leak rate test they find that sone penetration 21 of some isolation valve is leaking excessively, they go in 22 and fix it, you know, in order to make the type A test.
23 This doesn't bother you because, presunably, that 24 valve was checked every six months, or whatever the rate was 25 for the type C tests.
So that the interval between tests is 8):
Ji 5
342.02.!!
28 sh I
low enough that the fact that it failed just before the 2
integ ated leak rate test is no different than failing in 3
some other six-months interval. Presumably, your interval, 4
your test interval is set low enough to maintain some level 5
in between tests.
6 MR. ARtJDT: I'd have to say yes to that.
I think 7
the confid.ance level is the re, but it's qualitative, not 8
quantitative.
9 DR. SIESS: It seems to me that since we really hang 10 so darn much on leak rates, I look at Three Mile Island, 11 and, of course, we don't think that the containment leakage 12 there contributed significantly to the o ff-site releases, 13 unle ss you went to call the failure to isolate and get the 14 stuff out to the auxiliary building.
15 But, of course, Three Mile didn't see very much 16 pressure.
A pre ssure spike, a ppa re ntl y.
I don't t'elieve 17 that the re is any radioactivity release associated with that 18 pressure spike. I haven't tried to correlate the radioactive 19 sequence with the other sequence.
But sitting the re at, what, 20 2 or 3 psi, Three Mile Island apparently didn't leak.
It 21 had a few billion curies inside it, but not very much go 22 outside.
23 But it does seem to me that either we're pu tting 24 too much emphasis on the leak rate or we're not enough 25 concerned about just how good it is.
There must be enough data i
342.02.12 29 2h I
f rom type B and type C tests to look at integrated leak rate 2
tests so that somebody coutd do some kind of a reliability-type 3
study to s y with what confidence we know what the leak, a
4 what rate would be in between tests.
5 Darn it, if it isn't as high as we'd like it to be, 6
then we forget about the darn thing or change it.
7 MR. ARNDT I think that this is a good suggestion 8
for us to pay attention to when we go for our general 9
revision on Appendix J.
10 DR. SIE SS: I think it's a possible research subject.
Il MR. ARNDT: On the three different types of subjects.
12 DR. SIE SS: I think that there's a research component 13 in here.
I'm not sure.
If there isn't now, it has to be 14 formulated better.
15 MR. ARNDT: I think we'll look into this. At the very 16 least, we ha ve various reports on test results and LER reports 17 on failures.
And perhaps we could compile the data from 18 that and derive something from that data that would tell us 19 where the most probable failures are and what the reason 20 is.
21 DR. SIESS: There's a frequency magnitude issue.
22 The license frem Three Mile Island here need to be looked at.
23 The leak rate cones into the licensing process and the saf ety 24 review pretty much in terms of LOCA.
25 You assume that we've got P/A, peak calculated
.;a
3342.02.13 30 sh I
accident pre ssures.
We take a source term from Reg Guide 1314 2
and we take a leak rate of the tech specs to which these 3
tests are related.
We take sone 5 percent meteorology anu 4
we compute some doses.
5 And, of course, we compute some doses which are 6
absolutely outrageous in view of Three Mile Island.
M 7
On the other hand, our source tena may be equally
(
/
8 7
outrageous in terms of Three Mile Island. I don't know.
9 What the whole leakage thing fits into that scenario which --
10 I'm not sure it's a bonding scenario any more, or an acceptable
.11 scenario.
12 13 14 15 16 17 18 19 20 21 22 23 24 25
.t.,
- o
_: u.
31 6342 03 01 bp i
Certainly, it's not acceptable to the public in 2
terms of the numbers we come out with in the SER.
The 3
numbers we come out with in the environmental impact 4
s ta teme nt, on a realistic basis for a LOCA, are much 5
smaller.
In fact, for Three Mile Island 2 the environmental 6
mpact statement Class 8 large LOCA was 600 and some-odd 7
person-rems within a 50-mile radius, which is abcut the same 8
order which we got for three-mile.
9 I'm having trouble fittfng the leak ra te thing 10 into the overall picture now.
I just don't have tha t much 11 confidence.
12 MR. ARNDT:
Your comments are good ones, and I 13 think when we go f or our general revision --
14 DR. SIESS:
know how we got where we are.
15 MR. ARNDT:
-- we can f actor the se in, pa rdo n.
16 DR. SIESS:
I know how we got where we are but 17 when we go back and tie the leak rate entirely to the Part 18 100 siting cri teria, you knok, which leads to unacceptable 19 consequences by public standards, I think it need looking 20 a t.
I don't know whether Standards should look at it, or 21 Licensing, or wha t.
22 MR. ANDERSON:
There is something other than the 23 minimum considered in the WASH-1400 re pe c t.
They considered 24 t ha t there was goir.g to be hair leakage based on the 25 probabilities.
- ~
- , i
32 342 03 02 ggp i
DR. SIESS:
WASH-la00 always carried everything to 2
a core melt.
3 MR. ANDERSON:
I think they had studied somewhat 4
the probabili ty.
We'll double-check that.
5 DR. SIESS:
I don't remember whether WASH-1400, 6
you know, looked at reali stic acciden t pre ssures, or whether 7
they just sort of used the LOCA analysis to get the 6
pressures and so forth.
9 MR. ARNDT:
I think also the Le ssons Learned --
10 Threa Mile Island 2 Lessons Learned Task Force had addressed 11 itself in recommendations to looking at Appendix J.
I don't i
12 recal] --
i3 DR. SIESS:
I don't recall that.
14 MR. ARNDT I don't have the re por t wi th me, but I 15 t hought they were considering looking a t tha t al so.
16 DR. SIESS:
Maybe when we get through with Three 17 Mile, we'll get back and talk to the s ta f f about leak rate 18 testing.
Really, the more basic question to me is not the
[
t 19 leak rate testing but where we stand on leak rate, and why.
20 Because I've seen in the past few years the s truc tu ral 21 design of containments change significantly simply because 22 of the leak requirement.
I don't think the meteorology has 23 changed that much, but something has changed to where 24 every thing i s a double containment.
25 We'll just see more and more of those, and it's t
33 6342 03 03 hhkp i
being alctated by a particular calculation, a particular 2
requirement which may be important.
Af ter all, Three Mile (l) 3 Island, if it showed anything, it showed tha t containments 4
are awf ul nice things to have.
5 MR. ARNDT I'm not so sure about the more recent 6
containments, but the meteorology model, I think, was very 7
significant with respec t to the structural design used.
For 8
e xam pl e, in Beaver Valley, the application of an additional 9
barrier. I think was due to meteorological modeling of the 10 site.
Where the original design might have been fine on one 11 site, here it was felt that additional protection for that 12 site was desirable.
13 DR. SIESS:
I think maybe there were some changes 14 in the me teorology.
But I haven't seen any site -- o f 15 course, I ha v e n' t seen very many cps recently, but I'm 16 talking about four or five years, I ha ve n ' t seen any site 17 where the old two-tenths percent leak rate was acceptable, 18 6nd that's what we were using ten years ago.
We've still 19 got some of chem.
The leak rates were two-tenths percent, 20 then they started going to one percont; then they started 21 going to doubie containment, to ge t down to one-tenth 22 percent.
You know, we've done some pre tty extreme things.
23 PROF. KERR:
One of the things that happened was 24 changing a goal Lo half the goal a t the CP stage.
That 25 probably helps some, because I can remember plants coming in
34 342 03 04 p
i with 299.6 rems to the thyroid at the OL stage, and we found 2
that acceptable.
You just don't see that anyrore.
You see 3
at the OL stage, 15 rems to the thyroid, one, two, three, 4
f our, five rem, whole body.
Everybody's got tightened up, 5
oven on Part 100.
I'm not quite sure why.
6 I've got nothing against containments and low 7
leakage, but a double containment at Three Mile wouldn't 8
have changed the releases, I don't think, one bi t.
9 MR. ARNDT:
I agree.
10 DR. SIESS:
And if it's an i solation valve that 11 doesn't close, the fact that you've got a double containment 12 acesn't hel p you one bi t.
Sometimes I think the regulations 13
-- where the reason f or them has sort of go tten lost -- have 14 led to design changes or solutiens that are not necessarily 15 improvements in real sa f e ty.
They're an improvement in a 16 calculation, which may or may not af f ec t the real safety.
17 Okay, this is going out f or comment, right?
18 MR. ARN DT:
Right.
19 DR. SIESS:
Any objec tions?
20 MR. ETHER INGTON :
No.
21 DR. SIESS:
The next one we've got is Revision 2 22 of 1.136.
1.136 is entitled Material, Construction and 23 Testing of Concre te Con tainments.
It's a qualified 24 endorsement of ASME boiler and pre ssure vessel code sectir.
25 3, revision 2, original draf t.
And revision I was the
)
}
35 342 03 05 p
1 original ground.
Revision 1 Just dealt w.' th Article CC 2
2000, which was Materials and the new draf t has
,anded 3
that to deal with essentially all of the parts of CC 1002, 4
4, 5, 6 and 7, and 3 is Design; am I right?
5 MR. MA:
Tha t's right.
6 DR. SIESS:
That's covered in another one, isn't 7
it?
8 MR. ANDERSON:
No.
9 DR. SIESS:
Don't we have a rate guide that has 10 load f actors and stuf f like that?
I thought we had a couple 11 of rate guides tha t addre ssed tha t.
12 MR. ANDERSON:
That was on 349, other concrete 13 structures.
14 DR. SIESS:
Okay.
I t's on 349, and actually, all 15 the staf f has added to cover five more chapters is about two 16 items, right?
17 MR. MA:
That's correc t.
IC DR. SIESS:
I went through positions.
One was not 19 changed; two was not changed; three simply gives a cross 20 reference to Reg Guide 1.107; f our wa s no t c hang ed ;
- five, 21 I think is new; and six is new.
They deal with 5000 and 22 6000.
Rather than go t hrough the se i tem-by-i tem, let's just 23 see what comments we have.
It's fairly clean.
I have a 24 c ou pl e.
25 In Item 3, where you simply change the ref erence m7-
.c.
_oi
36 6342 03 06 llhp i
1107, that's on the chloride limits, which is about as hot a 2
subject as you can get into.
And I think now you're quoting 3
313, and I don't know tha t 318 -- all I can tell you is that 4
318 is not considering the chsage right nou I can tell you 5
tha t since I'm chairman of the 318 Subcommi ttee.
6 MR. ARNDT:
Okay.
7 DR. SIESS:
I just wonder if just for the benefit 8
of the user, whether it wouldn't be just as well to pu t al 9
the words in Position 3 instead of cross ref erencing another 10 guide.
To make a person go look up another guide I
li haven't any strong feelings, but --
12 MR. ARNDTs The re's a particular reason for this.
13 Items one through four are sort of in the commi ttee 14 de velo pmen t proc e ss.
There are being changes made to the 15 code.
Changes have not been made ef f e c ti ve, they're not out 16 in an addendum to the code, and with the exception of number 17 t hr e e, items one through four are going through as shown.
16 Item number three, the Commi ttee is a ttempting to make some 19 progre ss on i t, but there 's still f airly intense discussion 20 between the Commi ttee and us as to what the chloride limits 21 should be.
There are other f actors than just the chloride 22 limits, although this ref ers specifically to t ha t.
23 DR. SIESS:
No it doesn't.
By reference to 1107, 24 it limits other things than chlorides.
25 MR. ARNDT Right, the chemical requirements.
1
+r
37 342 03 07 p
i W ta t I'm saying is between the Code Commi ttee and Reg Guide 2
107, there's also some controversy as to the time at which 3
it should be in place in the structure and tension, without 4
having any environmental protection.
That's outside of this S
. ref e renc e he re --
6 DR. SIESS:
All I was saying is there's not any 7
c hange in substance, but would it be convenient for the user 8
if you just simply re peated in here the requirements f rom 9
107 instead of referencing them?
10 MR. ARNDT:
Not rea ly, because what we're hoping 11 is -- maybe too optimistically -- tha t the Code Committee 12 will decide that the contents of 107 or the discu ssions that 13 we're having af ter the issuance of 107, can come out in an 14 addendum to the code, in which ca se, if they do that, we'd 15 have no need for the regulatory guide any longer.
But 16 what's more likely is tha t they may come out in some 17 modified form, which maybe they and we can agree upon, and 18 it may not fully be exactly the same as 107 now i s.
19 Now, if we come up with something that's a li ttle 20 diff erent f rom the current 107, then we have to go back cnd 21 c hange 107 and thi s guids, both, whereas if we wind up with 22 a compromise posi tion wi th the Committee, different than 23 wha t's currently in 107, and we had the wording in here, 24 we'd have to change both rather than one.
25 DR. SIESS:
So you could only change one.
But o
1
_ ') n'
38 6342 03 08 llkp I
this assumes that it is satisfactory, that if you change 2
107, the change also applies here.
3 MR. ARNDT:
Correct.
4 DR. SIESS:
And t ha t i s tru e.
5 MR. ARNDT:
That is a fact.
6 DR. SIESS:
So t ha t is an advantage I can see.
7 okay.
8 MR. ETHERINGTON:
What happens if the code 9
conforms completely to your Reg Guide requirements?
Do you 10 just withdrae the Guide, or do you leave it and endorse the code in the Guide ?
12 MR. ARNDT:
If they have pu t all the points that 13 we're concerned wi th into the code, then we feel that 14 there's no longer any need for the Guide itself.
I t's a 15 redundant piece of pa per.
And so long as we have no further 16 points of con tention on that subject, then we don' t need the 17 Regulatory Guide.
18 DR. SIESS:
But you do need the ref erence.
19 MR. MORRISON:
What you're saying, I don't 20 understand.
21 DR. SiESS:
Let me try.
If the code is completely 22 a cceptable wi thou t any reservation, the way you would accept 23 i t would no t be by a Reg Guide thing, it's completely 24 a cce ptabl e.
But by referencing it, there's a place in Part 25 50 where you reference acceptable standards.
39 342 03 09 p
1 MR. MORRISON:
We had the codes and standards in 2
Part 55(a).
But I think what Gunter is saying is a little 3
conf using.
If the code was comple tely acceptable to us in 4
our present course, we would still need a Regulatory Guide 5
that would endorse it as an acceptable way.
6 DR. SIESS:
Why not?
I f you pu t i t in 55(a) --
7 MR. MORRISON:
If we think it's mature enough, 8
then we can do this with any standard where we think we'd 9
want to put it in the regulations, under the Regulatory 10 Gui de.
Tha t is another option open to us.
But we'd have to 11 endorse it by one niechanism or another.
12 DR. SIESS:
You can either put out a Reg Guide 13 that endorses it up to a certain addendum, or you can put in 14
.55(a), which is the same thing.
15 MR. ARNDT Right.
Yes, I can cae where I may 16 have confused the i ssue.
17 DR. SIESS:
You're always a year and a half behind 18 in 55(a), which doesn't bother me.
It used to bother 19 Dr. Bush quite a bit.
20 MR. MORRISON:
We're usually that f ar behind on 21 our other standards that we're endorsing on Regulatory 22 Guides, too.
23 MR. ARNDT:
We wouldn't need a special ty guide 24 like 107 if all of t he se provisions were in the code.
We 25 would continue a guide like this guide, which is endorsing
'a
40 6342 03 10
'rdp I
-- we're a ttempting through this guide to endorse the code 2
as a whole.
3 DR. SIESS:
What does 107 cover?
4 MR. ARNDT:
It covers grouting, the in-service 5
inspection of containments -- excuse me, it covers the 6
grouting of pre-stressing tendons.
7 DR. SIESS:
There are codes which are acceptable 8
and these are the ones, Section 3, Division I,
is covered by 9
55(a).
10 MR. MORRISON:
55(a).
11 MR. ANDERSON:
We're planr.ing on revising 50, 12 55(a) to broaden it.
Right now it only covers Class I metal 13 components.
One of our priority tasks is to complete t he 14 rewrite, 50. 55( e ) and to broaden it, including starting to 15 pick up metal c -atainment s.
I don't believe we're going to 16 pick up concrete containments f or some time.
We want to see 17 this code mature somewhat.
18 DR. SIESS:
Well, it looks to me like CC is 19 getting pretty close to wha t you want.
20 MR. ARNDT It is, but it's not at the point where 21 we're in close enough agreement with the Code Corami tt ee t ha t 22 we could go and put it in the regula tion.
23 DR. SIESS:
The se are -- r.aybe they're a li ttle 24 better than nits, but t'.y're not great big i ssues.
25 MR. ARNUT Agreed.
O a
41 34 03 11 kap i
DR. SIESS:
They're relatively minor differences, 2
considering what's in them.
3 MR. ARNDT Our procedure here has been to attempt 4
to pick up as much of Section 3, Division 2 as we can in 5
adjustable bites.
This is the first expansion of the 6
original Guide.
The first issue of this Guide covered the 7
materials only.
This issue of the Guide goes beyond that.
8 DR. SIESS:
You have still got diff erences on 9
3000, I know.
Let's go to page six, where the two new items 10
-- the first one of them is on 52-10.
I have one comment.
11 In line three, i t s hould be " tho se embedments," I believe.
12 "Those emcedments shown on the drawings."
I think it's a 13 little clearer that way.
Anybody have any questions about 14 52-10, Item 5?
15 (No response.)
16 DR. SIESS:
Wha t abou t Item 67 That seems to be 17 some thing the code ought to pick up and fix, because I don't 18 think -- did they argue about that, gentlemen?
19 MR. MA:
No.
20 DR. SIESS:
Because the way i t was worded, if you 21 read English, it didn't require a remedial measure, but 22 there wasn't much point in mentioning i t, unless you 23 intended to require it.
24 MR. ARNDT:
I'm not sure that that was t he 25 intentional wording in the code.
~,
e r !
42 6342 03 12 kp 1
DR. SIESS:
It was just very poor wording.
2 MR. ARNDT:
But once they make the changes to the 3
code 4
DR. SIESS:
They might make it before this goes 5
final.
6 MR. ARNDT:
It's a possibility.
7 DR. SIESS:
Mike, do you have ny questions?
8 MR. BENDER:
No.
9 DR. SIESS:
Anybody else?
Any objec tion to thi s 10 going out for comment?
11 PROF. KERR:
I have no objection.
I have a 12 comment on page three, 1.1363, where the staf f believes 13 these recommended limi ts are more conser vative and can 14 provide better assurance.
I'd f eel be tter if it were IS something like "needed assurance" rat 'er than "be tter 16 assurance."
One can always become more conservative --
17 DR. SIESS:
Where is that, Bill?
Oh, I see it.
18 Item 14.
19 PROF. KERR:
One can always be more conservative 20 and provide be trer assurance.
It seems to me what you want 21 i s -- you don' t thin.
the existing is good enough and you 22 want some needed insurance.
23 MR. ARNDT Correct.
24 MR. BENDER:
Can I ask an extraneous question?
25 PROF. KERR No.
43 342 03 13 p
1 MR. BEND 5R:
I'll ask it anyway.
Having become 2
aware of the rock anchor problem, what's being done about 3
ge tting something to deal with those kinds of adversities?
4 Tha t doe sn't come under thi s.
5 MR. ARN DT I t's no t covered here.
I'm not eu to 6
date, exactly what's happening.
I think that there is some 7
activity currently going on, looking at this.
8 MR. BENDER:
There's water intrusion i".to the 9
rocks.
Tha t much I know.
The problem is wha
- ner, wi th all 10 the attention we're giving to grouted tendo'.s in the 11 concrete, we aren't ignoring the most impf rtant matter, 12 which is whether the integrity of the ri. ck anc hors tha t are 13 used for that purpose have really been properly controlled 14 by a regulation.
I don't want to tr. e the ma tter f urther.
15 It just reminded me of i t, in looking at this.
16 MR. NJDERSON:
On your suggestion, we looked at 17 t he word "be t ter" and I'm not si;re we can provide the needed 16 a ssurance of avoiding unf oresean problems, because we can't 19 foresee all of them.
20 PROF. KERR:
We ll, ~ny point i s yc 1 can always make 21 things more conservative, c'c1 thus presumably provide be tter 22 assurance, but you can takt next week, you can take an even 23 more conservative stance to provide better a ssurance.
It 24 does not seem to me that tha t's a valid argument for doing 25 70me thing.
The fact that you are more conservative and
44 6342 03 14 ip i
provide be tter assurance --
2 MR. ANDERSON:
I'm not sure we're going to provide 3
all the needed assurance of avoiding unf oreseen problems.
4 PROF. KERR:
Well, if you don't provide the needed 5
a ssuranc e, then you shouldn't use that change, because the 6
change presumably is one which you think will provide needed 7
a ssurance.
Otherwise you wouldn't be recommeding i t, would 8
you.
9 10 11 12 13 14 M
15 o'
16 17 18 19 20 21 22 23 24 25 g~s
/I
CR 6342 #4 DAV/PV 45 I
MR. ANDERSON:
When it comes to unforeseen problems.
2 PROF, KERR:
Assurance doesn't mean certainty, to me.
3 MR. MORRISON:
That's right.
I don't have any prob-4 lem with the "needed assurance."
5 PROF. KERR:
It's in it.
6 MR. ETHERINGTON:
Would it be a compromisc just to 7
leave out "more conservative'*?
"These recommended limits pro-8 vide better assurance.
If you can provide better assurance at 9
small cost, you should do so."
I think that's part of the 10 philosophy.
II DR. SIESS:
I am not sure it's at a small cost.
I2 MR. ETHERINGTON:
Well, I don't know, compared with 13 the cost --
Id DR. SIESS:
That's the chloride.
That's not in the 15 position, Bill.
I6 PROF. KERR:
No more comments.
17 DR. SIESS:
Anything else?
18 (No response.)
I9 DR. SIESS:
Okay.
We'll say "Okay" on sending that i
20 one out for comment.
I doubt if you're going to get very many I
7I comments, because that particular position is already in force
{
i 22 on the chloride.
You're covering grouting in another reg guide;,
23 right?
24 MR. ARNDT:
That's right.
This is just for consis-Ace-Federal Reporters, Inc.
25 tency.
. ~.
n/h i
pv2 46 I
DR. SIESS:
What have you got for us next month, if 2
anything, Bill.
MR. MORRISON:
We have nothing for next month.
4 DR. SIESS:
You guys are busy on something else?
MR. MORRISON:
Our people are diverted on working on 6
licensing cases and new projects as a result of the TMI Lessons l 7
Learned.
~
DR. SIESS:
I want to know when Gunter Arndt became 9
an expert on leak testing.
These structural engineers are 10 pretty verratile.
11 (Laughter.)
12 MR. ARNDT:
We're still on the record.
13 MR. MORRISON:
When we lost our previous expert.
14 DR. SIESS:
Who is that?
15 MR. MORRISON:
Joe Melvin.
He went to MUS.
16 MR. ANDERSON:
About two or three years ago.
17 MR. ARNDT:
As a matter of fact, my association with l
18 6
Appendix J goes back a considerable number of years, when I i
19 first backstopped an individual who had the responsibility for i
20 revising the Appendix in its current form, to its current form.
And he left shortly thereafter, and since I b worked with him on it, I continued coordination on it.
And it's recently been i
23 i
that time has elapsed from the issue of that regulation to a l
k 2A Period where people have to implement. The time period involved Ace Federal Reporters, Inc.
l 25 had passed, and we started getting information in Licensing that c:
,a,
+11
()
'iL
pv3 47 1
indicated the Appendix needed changing. And NRR has drawn up, as 2
a result of their experience with the current one, a number of 3
changes they want made, a number of recommendations for change.
4
'I will be coordinating that action.
5 DR. SIESS:
Bill, leaving out the discussion about 6
leak testing, in general, which really didn'trrelate specifi-7 cally to the guide, what did we accomplish here this morning 8
that we couldn't have accomplished perhaps by just some written 9
comments from members of the subcommittee to the staff?
I am 10 still looking at ways to reduce some of our effort in this area, Il particularly on guides that are going out for cc.7 ment.
Once 12 I suggested we might even consider not reviewing them until 13 after they had been out.
In this case,we had relatively minor 14 items -- that is, noncontroversial -- and yet, you know, we've 15 convened three to four members of the committee and brought you 16 guys down here from Bethesda for an hour and a quarter.
17 MR. MORRISON:
I know this is something that we have 18 discussed before.
I9 DR. SIESS:
Did you get anything more out of this than l
20 you would have we had sent you a few comments in writing on 21 these two guides?
22 MR. MORRISON:
In the regulation in th guide that we !
23 discussed this morning, probably not.
And, in fact, where, I
(j) 24 because of previous commitments we have been unable to meet and Ace FMeraf Reporters, Inc.
l 25 maybe we only had one guy down here, rather than defer it, I s
i
pv4 48 I
suggested that perhaps we could get your comments by mail 2
without having a meeting.
3 But I think it depends on the individual guides and 4
regulations as to the extent of discussion that takes place.
I 5
think that particularly on the more controversial guides, we 6
think that the meetings, even before we go out for comment, are 7
worthwhile.
8 DR. SIE7S:
I w.nted to limit it to these two because,,
I 9
for example, supposa we had a procedure whereby I saw these two '
10 a little bit in advar.ce, looked at them, and said I didn't think 11 there was too much controversy and suggested to the staff that they simply circulate them to the members of the committee to 13 see if there were any comments, and we would not consider them Id at a meeting.
That would have been satisfactory?
MR. MORRISON:
That would have been satisfactory, i
16 I think there is some other factor that we should I7 probably take into account.
Tha'- is the extent of public 18 involvement, too.
Suppose that a g. zide is in that category but 19 l
you also received a couple of request,
_~or public comment.
20 DR. SIESS:
That would affect i.y decision.
But, again, 2I on a guide like these two changes, they both go out for public 22 comment, they will be back before they're made final, and there i
23 will be ample time for the public to comment and to appear before h
2#
this committee if they want.
AceTederal Reporters. Inc.
j 25 Now, in this instance.of just these,two following the m,,a
pva i
49:
I procedure, I suggested we could have saved a meeting if we had l
2 had a couple more substantive issues.
Then I wouldn't see much!
3 point in trying to handle something like this by correspondence..
4 We might just as well bring it in and take it up in the same 5
meeting, unless it's going to extend the meeting longer than 6
would seem desirable.
7 But does the subcoramittee see any reason for not 8
trying that if the occasion arises?
It won't happen P-
) often.
9 MR. E-fHE RINGTON :
No.
I think that would be fine.
10 O f course, it does presuppose that we get some kind o f a look II at this enough in advance to decide whether we'll nerd a meet-
{
I2 ing or not.
13 DR. SIESS:
What I thought is that I would look at 1-4 it; and if I thought that we didn't need a meeting, I would ask '
15 the members to comment or to indicate whether they preferred to 16 discuss it in a meeting.
17 MR. ETHERINGTON:
Of course, you can always call a 18 meeting and then cancel it.
I think that's possible; isn't it?
I9 MR. QUITTSCH REIBER :
Yes.
20 DR. SIESS:
I think the number of times where we could 2I avoid a meeting entirely might be few, because we might have anl l
22 effective guide coming in and two for comments, and we might 23 juast as well take them all up.
24 MR. ETHERINGTON:
This is when we clearly could have Ac eral Reporters ine.
25 avoided a meeting.
1 0 ;_
_/J
pv 50 1
DR. SIESS:
This one we could have avoided a meeting, 2
but we could have a case with three or four effective guides.
3 That means an all-day meeting, but we could cut a meeting down 4
to half a day.
There are various things that we can do.
These 5
Wednesdays are getting pretty valuable to the committee for 6
other meetings, and sometimes half a day is enough.
7 But we'll try this and see if it works.
It's no t out 8
of lire with what you said, Bill.
9 Then, next month we have nothing.
10 Will you have anything the month after that?
II MR. MORRISON:
I haven't really looked that far ahead '
12 on our schedules, but I will let you know at the time of the 13 meeting next month.
14 DR. SIEbb:
Okay.
15 Anything else, gentlemen?
16 MR. ETHERINGTON :
Yes.
I would like to ask a question l
17 about integrated leak rate testing.
It's a very difficult test,!
18 of course, but, as I recall, North Anna 1 ran a leak test, and l
19 they came out with a negative leak rate which di dn' t look right l
20 to them, so they rationalized away the negati'e part and came 21 out with an extremely small positive leak rate.
22 This doesn't seem very satisfactory.
Have you any I
23 thongs on that?
i i
24 MR. ARNDT :
I am not familiar with that, sir, I am Am 'Jeral Reporters, Inc.
}
25 afraid.
I
\\
pv 51 I
MR. BENDER:
I remember it, Harold.
2 MR. ARNDT :
I know the containment itself is supposed, 3
to operate.
9 4
MR. ETHERINGTON:
It doesn't give you much confidence 5
with the real figure they get.
0 DR. SIESS:
They're all using the absolute method now, 7
aren't they, rather than the reference document?
8 MR. ARNDT:
I believe they are.
9 DR. SIESS:
They're putting du Pont monitors ahd I'O thermometers around, and so forth.
You get absolutely no 11 confidence in that whatsoever.
What I do get some confidence I
I2 from on the integrated leak rate test is that they have to 13 calibrate after they've run their 24-hour tests that they then Id must put in a known leak and run for, what, four hours or some-15 thing like that.
And they've got to be able to pick up that 16 known leak within plus or minus 25 percent.
Now, that gives me' 17 a little confidence that what they got was right.
Ib MR. ETHERINGTON:
I don't remember on North Anna, per' 19 se.
20 DR. SIESS:
What were they talking about?
Out-gassing 21 from the concrete?
I don't remember.
But this thing is so 22 darned sensitive to temperature, they put about 20 thermometers l
23 around, you know, and then they got all these complicated --
l 24 it takes a computer program to compute this.
Act ederal Reporters, Inc.
PROF, KERR:
They just failed to calculate the
{
n ~I O
'll l
\\1 i
pv8 52 1
relationship between pressure and temperature correctly.
Then 2
you get a negative leak rate.
3 MR. ETHERINGTON:
But they made all the known cor-4 rections, you see.
5 PROF. KERR:
But maybe the corrections weren't 6
properly done.
7 MR. ETHERINGTON:
I am only saying I don' t have much 8
confidence in the number; that's all.
9 DR. SIESS:
It doesn't give you any confidence except 10 when you've seen the calibration test; you get a little confi-II dence if that works, that the thing is working right.
I think 12 I would give a little more confidence to the calibration test 13 if it was made first, rather than afterwards; but that may be 14 just a minor point.
15 MR. ARNDT:
One of the reasons we would like to make 16 a later, more general, change to Appendix J is s; that we can I7 leave the technical aspects of running conducting a test, the 18 mechanics of testing, to the industry standard that has been l
19 recently developed and expanded, and leave the criteria for the 20 leakage rate in the regulation.
21 But for the reasons you're citing as examples, there 1
22 is going to be considerable change anticipated in the way peo-23 ple go about conducting the tests; and having an industry 24 standard which we participate in and follow along with, which Act eral Reporters, Inc.
25 will be sort of the handbook of doing the test, we find it to 3
.i0 t'
1
53 1
be a valuable way of keeping the testing techniques current 2
and keeping them as accurate in doing them and interpreting 3
them as possible.
4 DR. SIESS:
I think, following what Mr. Etherington 5
said -- and I certainly agree with him -- in terms of the 6
assurance that the containment will not leak in the event of an 7
accident, I don't think the integrated leak test, the interval 8
it's made and how it's made, contributes very much to my feel-9 ing of assurance.
10 I think the Type B and C tests, being sure that 11 penetrations are tight and that valves close and are closed, 12 give me a lot more assurance because they're made 'at more fre-13 quent intervals, 14 But from my feeling about the failure rate there, I l
i 15 don't know how much assurance that I get, and I think the level!
1 16 of assurance really depends more -- if you did a probabilistic' 17 analysis, your level of confidence would depend much, much more 18 on your Type B and C test than on a Type A test.
19 MR. ETHERINGTON:
I don't want my comment to be 20 construed as my not being in favor of integreated leak rate 21 tests.
There are a lot of reaction forces underlying, and you I
gg 22 could develop a crack in a weld that wouldn't be found any other i
23 way.
On your one-inch hole, assuming that you have 10/1000 gap t
24 there, or a bad fit or something, once it gets away from the A
eral Reporters, Inc.
25 hole, I think it will find its way out.
And I find that would l
k
pv10 54 I
be about 2/10 of a percent a day on a one-inch hole for the 10- !
e 2
mill gap.
3 DR. SIESS:
It was 40 psi on that liner.
I am not 4
sure it was there, either.
5 MR. ETHERINGTON:
I don't know.
The roughness, the 6
unevenness.
7 DR. SIESS:
The concrete is cast agains t the steel.
8 MR. ETHERINGTON:
It's cast against the steel.
Does 9
it stay there?
10 MR. BENDER:
There is lots of buckling in those lines.,
II DR. SIESS:
The thing is, the intergrated leak rate 12 test, I believe, made at five-year intervals at half the pres-l 13 sure, if you did a probabilistic type of analysis, you would 14 find that the confidence limits you got from that would be 15 pretty poor compared to what you would get from your Type B and 16 C tests.
17 I am not arguing against integrated leak rate tests, i
18 either.
I9 l
MR. ETHERINGTON:
No, that's right.
That's all I 20 wanted to say.
I l
i 21 DR. SIESS:
It's a gross test.
It's likely to show h
22 up something and nake some sense, beca,se that's really what f
23 you're after.
But I am also not sure that failing it is all 24 that significant, ei the r, if it fails because a valve line Act ' ederal Reoorters, Inc.
25 cloces.
n U
pvll 55 I
MR. BENDER:
If I could just add another element of 2
thought to Harold's point.
An intergrated test that tries to j
establish a leak rate of like a quarter of one percent of the 4
system volume per day is just not within the measuring ccpa-5 bility of the systema.
So, I think tests like the North Anna 0
test --
7 DR. SIESS:
You have got darned few containments that 8
are as high as a quarter.
9 MR. BENDER:
-- Show up that reasurement.
It seems 10 to me that just from the standpoint of overall control, that 11 being able to show that the leakage is less than one percent of 12 the system volume per day is the kind of assurance you need in 13 order to be sure that all the penetrations are closed up.
And Id that's the part that I think has the highest uncertainty.
15 It seems to me, when you're thinking about leak test-16 ing, one of the things that ought to be dealt with is that, just l
to be sure that all the mechanisms do operate.
These contain-18 ments that have buf fered connections on them are permitted to I
19 i
leak in all sorts of ways, and a leak rate test that is masked I
20 by those buf fers is not really much of a leak test, anyhow.
l l
2I So, I think that whole aspect of it can stand some 1
22 scrutiny, what leak rate should you establish where you've got 21 a buffered system, just to the extent that the staf f wanted to 24 I
get some reasonable assurance that the closures are being made At eral Reporters, In..
25 well.
PV 56 I
MR. ARNDT:
I would like to add as a footnote -- I 2
mentioned the standard before; this is the consensus national o
0 tandard, ANCI-N-274, containment system leakage testing 4
requirements.
And we would like to, in future, not have 5
Appendix J be technical as far ac the mechanics of the test are 6
concerned.
7 MR. BENDER:
Would you plan to reference the standard?.
8 MR. ARDNT:
Yes.
The current one does, as a matter 9
of fact.
The reference is sort of the pret cessor to the 10 standard.
II DR. SIESS:
Anything else, gentlemen?
12 (No response.)
13 DR. SIESS:
Thank you, 14 The meeting is adjourned.
end#4 15 (Whereupon, at 10:10 a.m.,
the meeting was adjourned.)'
16 17 l l
18 19 20 21 h
22 23 24 l
Ac deral Reporters. Inc.
25 l
2n3
.