ML20136G388
| ML20136G388 | |
| Person / Time | |
|---|---|
| Site: | Zion, 05000000 |
| Issue date: | 10/17/1985 |
| From: | NRC COMMISSION (OCM) |
| To: | |
| Shared Package | |
| ML20136G381 | List: |
| References | |
| NUDOCS 8511220337 | |
| Download: ML20136G388 (193) | |
Text
%
0 HG i(Ai UN11ED STATES NUCLEAR REGULATORY COMMISSION
~w m --
-m e
em_.
w.,-me..
e-
-6
-- ~
IN THE M A i i tR OF:
DOCKET NO:
CONTAINMENT LEAX RATES PUBLIC MEETING I.OCATION:
1 - 135 DATE:
T!iURSDA*l, OCTOBER 17, 1985
' t l
Acs-FEDERAL REPORTERS, INC.
c p ui. w e, 444 Nonh Cicitol Screet 0S11220337 051110 Washington, D.C. :201 PDR ADOCK 0500 5
p) 3.33
. NATIONWIDE COVTJtAC2
7_.
\\
1 UNITED STATES NUCLEAR REGULATORY COMMISSION
~v
~
2 3
IN THE MATTER OF:
Public Meeting with 4
Dr. Zenory Reytblatt O
5 6
o 7
REPORT OF PROCEEDINGS taken at the meeting 8
of the above-entitled cause on Thursday, October 17, 9
1985, at NRC Offices, Region III, 799 Roosevelt Road 10 (Bldg. 4), Glen Ellyn, Illinois at 2:00 o' clock p.m.
11 PRESENT:
12 U.
S.
Nuclear Regulatory Commissions
(~0 MR. ROBERT BERNERO, Division Directort 6.
13 MR., JAMES.SHAPAKER, Section Leaderr MR. JOHN HUANG 4 Engineerr 14 MR. FRED A.
MAURA, Rx Inspectorr MR. CARL PAPERIELLO, Region III 15 Division Directort and
~
MR. LUIS REYES, Branch Chief.
16 DR. ZENOVY REYTBLATTr 17 MR. EDWARD GOGEL, President, C.A.N.P.;
MR. CHARLES NISSIM-SABAT, Northeastern 18 Illinois Universityr MR. RICHARD HOEFLING, Attorney, Nuclear 19 Regulatory Commission MR. TIMOTHY W.
WRIGHT, Attorney, BPIr 20 MR. ABE SKLAR, IIT 21 COMMONWEALTH EDISON COMPANY:
MR. PETER LeBLOND, 22 MR. FRANK LENTINE, MR. TOM HAMMERICH, 23 MR. ROBERT LANE, MR. MIKE TURBAK, f'
24 MR. JIM GLOVER, N
MR.
L.E.
- DAVIS, 25 MR. DENNY'FARRAR e
i 1
_ =.
i E
1 MR. BERNERO:
Good afternoon.
My name is Bob q
2 Bernero.
I'm director of the Division of Systems 3
Integration in the Office of Nuclear Reactor 4
Regulation of the NRC, and we're here today to hold 5
a meeting with Dr. Zenovy Reytblatt on issues 6
concerning the leak testing of reactor containment
.o 7
structures or containment buildings.
8 I*have brought with me today two 9
members of my staff who are sitting on my right:
10 Mr. Shapaker and Mr. Huang, who will be speaking 11 later on the technical issues which are the subject 12 of the meeting.
I have also brought Mr. Hoefling 97
~_
13 from our office of the Executive Legal Director,'who 14 is sitting in the first row, and I would like to 15 take the opportunity to thank our hosts, the NRC's 16 Region III, which graciously provided the quarters.
17 Mr. Carl Paperiello is the host here, and in the 18 staff here, Mr. Fred Maura, also sitting at the e
i 19 table directly opposite me will be speaking later.
20
'He is a specialist in this field.
21 It is a public meeting, and I notice 22 that representatives of Commonwealth Edison Company 23 have come, they are here, and I would like to say 24 just a few words before I start the meeting.
The 25 format I would like to follow is actually inspired i
6 2
~ ~ - - - -
l by the purpose of the meeting.
.q 2
We have a process by which members of 3
the public can make petition for relief of one sort 4
or another with respect to some safety question or
~
5 safety issue.
And in March of this year 6
Dr. Reytblatt filed two letters that were treated 7
together as a petition, and they concern technical 8
issues related to containment leak testing.
9 The process by which such things are 10 handled is spelled out in our regulations, and that 11 process has completed and a response was sent to him 12 by my superior, Mr. Harol'd Denton, Director of m
13 Reactor Regulation, denying the petitions.
14 Dr. Reytblatt wrote in what could be 15 called appeal form, and we have chosen in order to 16 enhance possible communication to hold this meeting.
17 It is not a formal appeal of the process, th e 18 petition process.
There is no structured formal o
19 appeal to that process, but nonetheless since 20 Dr. Reytblatt expressed very strong concerns, we 21 chose to hold this meeting and hold it here in the 22 Chicago area for his convenience in order to discuss 23 the technical issues, because apparently 24 Dr. Reytblatt does not accept or does not understand 25 the technical arguments upon which we denied the 3
r-1 petition.
2 So for the format of the meeting, what 3
I suggest is that first we cover the technical 4
issues, and I'll start with Mr. Shapaker and work 5
through the NRC staff, in presentations of the 6
technical issues, and as al'so as a courtesy to 0
7 Dr. Reytblatt, I believe you received some slides 8
that we sent out to you last week 9
DR. REYTBLATT:
Yes, we did.
in order that you could have 10 MR. BERNERO:
11 some material in advance.
Then having engaged in 12 whatever dialog is appropriate to these issues that
-a 13 you have raised, later in the meeting you, of course,
-~
14 can make whatever remarks you choose to make, and 15 then we've left cn opportunity for Commonwealth 16 Edison to give whatever perspective or comment they 17 choose to make, because much of the debate and 18 certainly the examples are with respect to 19 Commonwealth Edison plants here in the Chicago area, 20 Zion in particular.
a 21 So the format of the meeting
.I would 22 like to follow then is that I will turn the floor 23 over to Mr. Shapaker, and we can proceed through
=
24 there.
25 Dr. Reytblatt?
4
1 DR. REYTBLATT:
Yes.
I emphatically object to 7
2 that format.
The reason being that I am acquainted 3
with the slides of Mr. Shapaker.
What he is doing, 4
he is presenting my point of view, but I am here and 5
I can do this much better.
That's number one.
6 Number two, Mr. Shapaker is known for 0
7 his distortion of facts which he did in his slides 8
again, and I wouldn't trust him to present my point 9
of view.
And number three, I am sort of a plaintiff 10 here and he is sort.o f a defendant, and whoever 11 heard of a defendant presenting the plaintiff's 12 point of view before the plaintiff.
So I suggest 13 that we either have no presentations, but just
.m 14' engage in technical discussion, so that we would 15 take the unanimous decisions.
We do not want to 16 spell out our opinions.
We want to get the 17 technical stuff done.
18 MR. BERNERO:
Doctor --
O 19 DR. REYTBLATT:
If this is not acceptable, then 20 I suggest that I would start presenting my g
21 complaints.
22 MR. BERNERO:
Dr. Roytblatt, let me go back to 23 the structure with the following comments:
=-.
24 First of all, Mr. Shapaker and 25 Mr. Huang will be presenting what are the technical 5
r rq 1
point of view or position of the NRC in the way it 2
regulates nuclear reactors particularly with respect 3
to leakage.
If you choose to coalesce and say that 4
that represents your position, than you are free to 5
do so.
But they are presenting the essential 6
technical basis with respect to issues that you 7
raised.
8 And I do not accept your argument that 9
we have what would be to me unstructured discussion, 10 nor to shift the thing.
You made a representation 11 in writing, you made your points in writing, as a 12 petition to the agency.
You challenged our response 13 as not having responded in technical terms.
And 14 wh'at we have chosen to do is put together in what I 15 hope is an exhaustive treatment the technical terms.
16 Please be patient and sit through them.
You will l
17 have an opportunity to speak.
18 DR. REYTBLATT:
I vehemently again object O
19 presentation by Mr. Shapaker.
It is not clear what 20 role he played in preparation of director's decision o
21 which contains false statements.
Unless l
22 Mr. Shapaker here declares openly that he has no 23 part in preparation of the statement, that I submit w
i 24 it a proprietary document on criticism of practices, 25 and because of that it hasn't boon responded, I 6
1 vehemently object.his presentation of any of my 2
views.
3 MR. BERNERO:
He is not 4
DR. REYTBLATT:
I even object to his presence O
5 here.
6 MR. BERNERO:
Dr. Reytblatt, you are entitled to 7
register your objections.
They are duly noted.
He 8
is here as a responsible member of the NRC staff.
9 He is, as I say, representing the NRC's technical 10 position, and if you choose to object as you have 11 done so in the past, and also made allegations of 12 impropriety and illegal action, I can advise you 13 that we have referred your objections and 14 allegations to our office of Inspector and Auditor 15 for the investigation as is appropriate.
16 But nonetheless, he is a responsible 17 agent in my staff.
My staff is responsible for the 18 licensing criteria of reactor systems analysis Q
19 including containment systems and the testi,ng e
20 thereof, and he is the appropriate spokesman as well 21 as Mr. Huang.'
22 So with that, we will proceed and I 23 would like Mr. Shapaker to start.
=--
24 DR. REYTBLATT:
I again vehemently object to 25 this.
I would like to present my standpoint, and p
1
t S
1 then I will respond, if they have anything.
I am 2
afraid that 3
MR. BERNERO:
Dr. Roytblatt 4
DR. REYTBLATT
-- Mr. Huang's presentation will 5
end after the first line, because the first line is 6
nonsense.
0 7
MR. BERNER0s Dr. Reytblatt, please sit and be 8
patient.
We will start with that presentation which 9
addresses the contentions you raised in your March 10 letters, and then you will have ample opportunity to 11 speak.
12 DR. REYTBLATT:
I am asking for postponement of 13 the meeting.
I am going to call to Mr. palladino 14 right now, and let me speak, and then after that 15 Mr. Shapaker, if he wants, he can speak.
16 This is not the right thing to do.
If 17 I complain, I have to present my complaints myself.
18 I would not object his speaking if he talks about O
19 generalities, that's fine, but presenting my 20 standpoint of view, I object.
I can do that myself.
21 I don't need Mr. Shapaker to do that for me.
22 MR. BERNERO:
We are not asking Mr. Shapaker to 23 represent your views.
m 24 DR. REYTDLATT:
Okay.
Then tell him this.
25 MR. DERNERO:
Mr. Shapaker was instructed by me 8
1 to prepare a technical presentation addressing the l
2 points that you made in writing.
And if you are i
3 unwilling to sit still and listen to that technical 4
presentation, if you would resort once again to 5
allegations of impropriety and dishonesty on the 6
part of anyone who attempts to communicate with you 7
and if you wish a five minute delay of the meeting 8
while you try to call the chairman --
9 DR. REYTBLATT:
Okay.
feel free to do so.
I will l
10 MR. BERNERO:
11 give you five minutes.
12 DR. REYTBLATT:
Sure.
13 MR. BERNERO:
This meeting is as a courtesy to 14 you.
15 DR. REYTBLATT:
Okay.
Thank you.
16 (Short recess taken.)
17 MR. BERNERO:
I would like to reconvene the 18 meeting now.
As I said, the meeting will go in the a
19 format that Mr. Shapaker and Mr. Huang and Mr. Maura 1
20 will present the technical' bases by which we i
21 regulate containment leakage, and then Dr. Reytblatt 22 will have ample opportunity to express his views, 23 whether he differs or objects to what has been said, fL 24 and we can carry from there.
s 25 Jim.
9
1 MR. SHAPAKER:
Okay.
Dy way of introduction and r}
2 summary, I will be presenting what the allegations 3
are and reviewing them and presenting the 4
conclusions that are to be demonstrated.
5 Follow-on discussions will be by John 6
Huang, who will provide a technical discussion of O
7 the ILRT methodology, and by Fred Maura, who will l
8 present a parametric analysis of the ILRT and a l
l 9
discussion of the inspection practices by the 10 regions.
11 MR. DERNERO:
Excuse me, Jim.
There are a l
12 limited number of copies of these slides up here.
m L.
13 If more are needed, we can have them made.
14 MR. SHAPAKER:
As Mr. Dernero stated, the 15 purpose of the meeting is to accommodate the 16 petitioner's request for a technical discussion of 17 the director's decision under 10 CFR 2.206, and this 18 was dated July July 3, 1985.
O 19 The petitioner's request was prompted 20 by the decision to deny the petition as stated in 21 the two letters identified'there, March 6th and 22 March 0, 1905, and wherein he was seeking immediate 23 actions with respect to containment leak rate v
24 testing.
25 The basis for the decision is that the 10
1 allegations made about Zion 1 containment integrated c
2 leak rate testing do not support the concerns upon 3
which the petition statements are based.
4 The March 6, 1985 letter sought 5
immediate postponement of all containment ILRT's.
6 The expressed concern in this letter was over the 7
determination and confirmation of weight 1
8 coefficients used in calculating the containment l
9 atmosphere weighted averago absolute drybulb 10 temperature.
11 The March 8, 1985 letter sought an 12 immediate ban on the use of the computer program to l
13 process raw test data until auch time that it is 14 debugged and revalidated.
The primary concern here i
15 was with the computational validity of the software l
16 to average temperature data.
17 A summary of the allegations about th e, 1B Zion 1 test which formed the basis for the 0
19 petitioner's concerns are shown on this slide.
The 20 first statement is that the Zion 1 software does not 21 properly compute subvolume temperatures.
22 The second allegation is that the Zion 23 1 software allows for the manipulation of weight p-24 coefficients to obtain acceptable results.
The U
25 third allegation is that the equation used to 11
a 1
calculate containment air mass is wrong.
And the 2
fourth allegation is that the current procedure for 3
weight coefficient determination and verification 4
may lead to underestimating abnormally high leak 5
rates to the degree that leak rates appear 6
acceptable.
1.*
I 7
With respect to the first allegation, 8
which is that the Zion i software does not properly 9
compute subvolume temperatures, we looked at a data 10 set from the Zion ILRT referenced by the petitioner, 11 and we examined it for evidence of the software's 12 inability to add and divide correctly as the P
b-13 petitioner stated.
The staff concluded based on the 14 results of a hand calculation, that the software 15 does in fact properly compute subvolume temperatures.
16 This is a slide showing the data set 17 302 of the Zion 1 December 1983 test cited in the 18 petitioner's example for subvolume 2.
And it is o
i 19 noted that the data set actually presented in the L,
20 petitioner's example is not the one that was cited I
21 nevertheless, one can use any data set and show the l
22 software's acceptability.
23 Here the sensor -- the sensors that l
24 were used --
Use your pencil as a pointer.
12
1 MR. SHAPAKER:
Okay.
These are the sensors that 2
were used in subvolume 2.
This is a typo here.
3 This should be OOS for sensor 4 and 14, and that 4
designates that those sensors were out of service.
5 The remaining 7 sensors were used to record the 6
temperature and subvolume 2.
7 This is the equation that was used to 8
calculate the average temperature in'that subvolume.
9 It's merely the summation of the temperature sensor 10 readings divided by the number of sensors.
And for 11 this data the average temperature in subvolume 2 is l
l 12 63.34 degrees Fahrenheit.
Now, this value is the 13 same as that which is reported on the data set sheet 14 from the Zion test being cited.
15 Yop can see here that 63.34 is the l
l 16 temperature in subvolume 2 that was reported, and if 1
17 you use the temperature sensors that were in service 18 during the test, that are identified by the arrows 1.
o 19 here, you will obtain that result.
l l
20 These were used on the basis that l
21 at the bottom of the page it notes which sensors 22 were locked out, meaning which sensors were not used l
23 during the test.
Four and 14 were the sensors in b
24 subvolume the drybulb temperature sensors in s
25 subvolume 2.
)
13
3 1
The designations for the sensdrs in
,,y 2
subvolume 2 are 22 as you see in parentheses here.
the first digit is that it's subvolume 2.I The 3
22 4
second digit is that it's the drybulb temperature sensor 4 and 14, 5
sensor.
Where you see a 20 in 4 6
that is subvolume 2 drybulb temperature sensor which 7
was out of service.
8 DR. REYTBLATT:
Was it?
9 MR. SHAPAKER:
Indeed.
10 MR. HUANG:
Yes.
e 11 MR. SHAPAKER:
The second allegation re$tated is 12 that the Zion 1 software allows for the manipulation 13 of weight coefficients to obtain acceptable results.
14 The staff has concluded that leak 15 testing practices in the industry, coupled with NRC
^
16 regulations and inspection practices, provide 17 adequate assurance that meaningful tests are being 18 conducted.
A subsequent presentation by Fred Mauia* ;
e 1-9 will characterize inspection -- the involvement of 20 the inspectors in monitoring such tests and in 21 performing independent calculations to assure that r
22 meaningful tests are conducted.
23 The third allegation is that the 1
24 equation used to calculate containment air mass is 25 wrong.
As noted in the American National Standard,
', w.
14
~
4 q
1
. entitled "Contain, ment System Leakage Testing i
2 Requirements," ANSI /ANS 56.8-1981, it's the equation L
3 given there, the equation is not wrong as the f
4 petitioner states.
The formulation for the j.
5 containment temperature for that parameter value 6
that should be used is what is at issue.
The ANSI e
7 standard does not prescribe a formulation for T.
8 The two T formulations that are 9
currently in use are shown on this slide.
The first i
10 formulation is one divided by the sum of the weight 11 coefficient divided by the temperature.
And the 12 second correlation is the sum of the product of the r
13 weight coefficients times the temperature.
14 The staff agrees that the first i
15 formulation is mathematically more correct; however, 16 we also conclude that within the range of 17 temperature patterns experienced during ILRT's, the 18 difference in leak rate using the second formulation i
u 19 has no safety significance.
And subsequent 20 presentations by John Huang and Fred Maura will sho'w 21 the impact of the T formulation on the calculated 1
22 leak rate.
i l
23 The fourth allegation is that current 24 procedures for weight c> efficient determination and W.4 25-verification may lead to underestimating abnormally 15 I
O
.OW.
m
& 4 n
-..--. I.
t
mg 1
high leak rates tp the degree *that leak rates appear 2
acceptable.
3 The current industry guidance on the 4
determination and verification of weight 5
coefficients is felt to be adequate.
The need for a 6
precise knowledge of the weight coefficients is 7
diminished by the acknowledged importance of 8
stabilization of containment atmosphere of test 9
conditions, and this stabilization is done to assure 10 the validity of the mass equation.
11 The presentations again by John Huang 12 and Fred Maura will provide an overview of the data
.- s i
13 analysis technique and the requisite test conditions, 14 and they will discuss the significance of the 15 verification test.
These first two bullets will be 16 presented by John Huang, and Fred Maura will present 17 the results of a paremetric analysis using actual 18 test data.
19 So with that, I will turn it over to 20 John to proceed with the ILRT methodology.
s 21 MR. BERNERO:
John?
22 DR. REYTBLATT:
Would you please leave two 23 slides, a slide where you show the data from Zion 7
24 leak rate test.
I'm going to use it.
25 MR. BERNERO:
Yes, they will be available to you.
E 16
r 1
DR. REYTBLATT:
Two slides.
4 2
MR. HUANG:
My name is John Huang.
I'm going to 3
discuss briefly again, you know, how the ILRT's 4
conducted, the formula I use to calculate the leak 5
rate, and what kind of assumption and the 6
requirement to go with those calculations, and the 7
significance of the verification tests.
8 I would like to just show you some 9
formulas in the equations that were used in the 10 process here.
And the whole thing starts off with 11 the equation of state.'
We usually call it p'erfect 12 gas law, Pv.
Here the small v is specific volume, 13 so therefore, it is P,
small v,
and RT.
7-14 And in the process of ILRT, we are 15 measuring the pressure and the temperature-of the 16 containment, so we can determine the specific volume 17 inside a containment, and the specific volume can be 18 then expressed by this formula, which is the total a
19 volume in the containment and then the mass inside 20 the containment.
21 Normally, we assume in the process 22 that the volumes in the containment are constant, 23 and then we can calculate at any time by knowing the 1
24 pressure and the temperature of the containment and
/
25 the mass stayed back inside the containment.
6 "17
1 so for any delta P in the time 2
interval, degree can be defined by calculating the 3
difference in the mass and then divided by the time.
4 And another thing I wanted to point out to you is 5
when we are doing the ILRT, we are not just 6
measuring two points.
We follow the history and the 7
trend of the changes.
You can see the mass 8
' fluctuate along a kind of constant line.
So we are t
9 not just picking these two points and give you a 10 number.
We really follow.
So sometimes we read it 11 24 hours2.777778e-4 days <br />0.00667 hours <br />3.968254e-5 weeks <br />9.132e-6 months <br /> and 36 hours4.166667e-4 days <br />0.01 hours <br />5.952381e-5 weeks <br />1.3698e-5 months <br /> just to get a trend.
12 MR. BERNERO:
Excuse me.
If you're going to 13 speak from the audience please identify yourself 14 clearly for the stenographer.
15 MR. GOGEL:
My name is Edward Gogel, G-o-g-e-l.
16 Isn't that first equation there Pv equals RT 17 strictly speaking only valid for an infinitesimal 18 point in space?
That if you want to actually 19 consider a real macroscopic volume in the massive 20 air cited, you have to writ'e an integral, and could 21 you please do that and do this analysis with the 22 integrals.
23 MR. HUANG:
I will get into that part of it.
a~
24 Theoretically, you are right.
And this equasion 25 represents we present, you know, a temperature at 18
1 a particular given mass.
,j d
2 MR. GOGEL:
Wouldn't it better to be more 3
precise from the start rather than go back and try 4
to o
5 MR. BERNERO:
Hear him out.
We'll get to it.
6 MR. HUANG:
Yes, I will get to that, and I will D
7 tell you the approximation we have to make, what 8
kind of assumption we have to use.
9 MR. NISSIM-SABAT:
Dr. Huang, I'm Charles 10 Nissim-Sabat.
My question is what is the pressure 11 we're talking about?
12 MR. HUANG:
The pressure is the pressure of the
.- J x-13 volume in the. containment.
14 MR. NISSIM-SABAT:
Yeah, but how many 15 atmospheres?
16 MR. HUANG:
It's varied.
We have a different 17 kind of containment.
I think usually varies from 18 for ice conditions for like 15 pounds and up to 60 0
19 pounds.
Sixty pounds, which means 60 psig.
If you 20 use absolute pressure, it will be from 20, 27 up to 21 75.
That's the range of most containment.
22 MR. NISSIM-SABAT:
27 to 75 23 MR. HUANG:
Psia in abso' lutes.
_~
24 MR. NISSIM-SABAT:
In psia, okay, and you've 25 determined that the vander Waals' equation will not
~
19
1 be necessary for that pressure?
2 MR. HUANG:
If you really want to get into the 3
theoretical part of it, as y o'u can see, even this 4
equation is an approximation representing the mass 5
inside the containment.
Vander Waals' equation may 6
be in some cases-even more accurate than this 7
equation.
We are engineers, you know.
This is why 8
I have to tell you what assumptions I have to make 9
in order to use this equation.
10 MR. BERNERO:
Carl, did you want to add 11 something?
12 MR. PAPERIELLO:
Carl Paperiello.
Can somebody 13 state at which pressure this test is actually run?
14 MR. MAURA:
The specific decibal that we run it?
'15 MR. PAPERIELLO:
Yes.
What pressure when it's 16 run?
17 MR. HUANG:
You mean for Zion?
18 MR. PAPERIELLO:
When they run the integrated a
19 leak rate test, I don't believe it is the same as this pressure.
20 the pressure that 21 MR. HUANG:
The Zion?
22 MR. BERNERO:
I think he's referring you are 23 referring to the option of half, l.
24 MR. PAPERIELLO:
Okay.
_u" 25 MR. BERNERO:
Yes, you can rur it full or at 20
l 5
1 half.
2 MR. HUANG:
Yeah, you can run it at a half.
3 MR. BERNERO:
At varying speeds.
4 MR. HUANG:
Yes, I'll show you in the previous 5
slide, and the data we have measuring from this ILRT, 6
you know, shows some kind of scattering movement o
7 along this constant line, and it is pretty hard to 8
just use one number to represent your leak rate.
So currently we are using a linear least square 9
we 10 analysis to analyze this data.
You know, we analyse 11 a whole bunch of data and try to fit into a linear 12 least line.
m 13 Normally we call this technique a mass s-14 point technique, and use this, the statistical 15 analysis,,to come up with the final results, which 16 is represented by these formula here.
And we 17 represent the results in terms of a percent per day 18 of a original containment mass.
And even this o
19 equation because this scattering of the data, we 20 record it from the instrumentation, and we the 21 final result will be corrected with an upper 22 confidence limit.
23 In other words, we really analyse how 24 they fluctuate and try to make sure the results will 25 represent the 95 percent of the data we measure.
21 I
=q 1
This is where I want to discuss now 2
we recognize the formula we use, you know, the 3
equation we use, and is kind of approximation of 4
what is going on inside of a containment during this 5
test.
So in order to use these equation properly, 6
you know, we have to realize'what assumption, you o
7 know, in the requirements we have to stick to in 8
order to make the result acceptable.
9 And that the first assumption which we 10 made in this process is the stabilization.
You know, 11 in-other words, we don't want to see fluctuation, 12 here and there, and changes rapidly throughout a 13 containment, and even we can get a number.
~.
14 Analytically we don't have any confidence on those 15 numbers.
16 So the importance of stabilization is 17 that we want to assure the spacial temperature 18 gradient changes throughout entire test.
It's a
19 relatively small and slow, because if it change too at certain 20 much, then we really don't know 21 confidence level we really don't know what the 22 number ought to be.
23 More specific, I wanted to point out
=-
t 24 to you a lot of literature already specify what we s-25 mean by stabilization.
I can use another way to 22
1 represent it.
The trend of individuals as you since the containment 2
probably know, we have from 3
is so large, we could not put in infinite number of 4
temperature sensor in there, and we can only use 5
limit number amount of sensor, and we try to map in 6
the temperature inside the containment.
- However, 7
when we are mapping that, we also are interested in 8
getting an average mean temperature from that 9
sensors.
10 And a.very important assumption we 11 made in this process is the trend of individual 12 sensors must follow the general trends of the L-13 containment mean temperature.
That containment mean 14 temperature is the T we use in the equation in the 15 perfect gas.
And if I try to represent in the 16 graphic sense, if we have a temperature distribution 17 like that, the individual sensor cannot vary too 18 much away from that line, and if it is too much away O
19 from that line, we know the stability is being 20 broken or something went wrong, or even the sensor 21 sometime we will determine, you know, just did not 22 work.
23 DR. REYTBLATT:
So what do you do to those r-~
24 stubborn sensors?
25 MR. HUANG:
Excuse me?
t 23
1 DR. PEYTBLATT:
What do you do to those stubborn 2
sensors that do not read the way you want them to 3
read?
4 MR. HUANG:
I can give you my personnel 5
understanding of some cases.
If the sensor, 6
temperature sensor, is all of a sudden don't even O
7 give you a number, say for instance, reading in one 8
of the examples on previous slide, if you read 65 9
degree, all of a sudden you read a =ero, you know 10 something went wrong just by common engineer sense.
11 DR. REYTBLATT:
So you throw them out?
12 MR. HUANG:
We'll talk about that later.
We
- i 13 throw two out in that previous case already.
14 DR. REYTBLATT:
Thank you.
Thank you.
15 MR. HUANG:
We're going to get back to that, all 16 right?
Nine sensor threw out the 9 sensor was 17 not concluded in the average temperature calculation 18 for the subvolume.
o 19 DR. REYTBLATT:
That's precisely the point.
20 MR. HUANG:
Anyway, there are other assumptions j
21 you must make to use those formula is here.
Only 22 one set of weight coefficients may be applied to the 23 entire data.
You cannot change your weight e---
24 coefficient during the process of your calculation 25 just to fit whatever the result you want.
You have e
24
=-
1 to use one set of weight coefficient to order datas.
2 And another thing I wanted to mention 3
is the data must be taken at a regular intervals at 4
random.
No manipulation or selection of data is 5
allowed.
Here I'm giving you the assumptions we 6
have assumed for the test, and then we believe the o
7 test will be acceptable.
8 And we also want sufficient data to be 9
taken for a valid statistical analysis.
Leak rate 10 calculations are not just simple two-point 11 calculations.
You can pick any two points you like 12 and give us a result.
-s 13 And the acceptability, the last one, 14 the acceptability of the test results must be 15 verified.
You know, you can claim you had a 16 wonderful measuring system, and I calculate it great, 17 but this result must be verified by checking against 18 a known leak rate.
In other words, we at the very C
19 end of the test, we are going to use a flow meter to 20 actually leak a low amount of leak rate, and we ask 21 the instrumentation to verify for us.
And so we 22 have to check against each other.
It's not just a 23 number you crank out from the addition and the
=-
24 division.
You really have to check this number 25 against a number, a simple formula, accurate formula.
25
1 You have to check within certain limits.
~..
1 2
Now, the very last thing I already 3
mentioned, the verification test.
Because we have a 4
lot of assumption made in this process, we are using 5
a perfect gas law for describing the mass inside 6
containment, you can always argue there are better 7
formulas.
However, we engineers try to measure a 8
number within certain limits.
We are not hit right 9
on the absolute number you want to measure against.
10 So I wanted to emphasize at this point the 13 importance of the verification test.
12 I wanted to now describe the purpose 13 for which we believe the verification test is.
The 14 purpose of the verification test is to verify the 15 entire ILRT test procedure.
This including 16 selection of your instrumentation, location of the 17 instrumentation, accuracy of the instrumentation, i
18 containment weight coefficient network.
You can use a
19 whatever the network you use, but we are going to j
20 have to verify later to see that network is really j
21 work for you.
Containment atmosphere stabilization, 22 and even the computation of the final leak rate.
23 There are a couple of other things we can also
=~~
24 verify, but these are the most important elements.
e 25 Now, we give you exactly how this 26
1 verification is done.
This is the way we calculate
., j 2
a final containment instrumentation.
This were to 3
be the number we calculate using the containment 4
instrumentation, whatever.is located there, and this 5
is what we call Lam.
6 At the end of that calculation, that O
7 test, we impose a one value.
This value actually 8
comrs off from formula in the containment.
So we e
9 know this number.
This number in our judgment are 10 very accurate.
So after that 11 DR. REYTBLATT:
Only in your judgment.
12 MR. HUANG:
To the accuracy of that flow meter.
T 13 DR. REYTBLATT:
No.
To-the accuracy of your 14 wrong calculations.
15 MR. HUANG:
No.
These La comes from the flow 16 meter.
Let me restate that.
17' DR. REYTBLATT:
La 18 MR. BERNERO:
He's talking about Li.
s 19 MR. HUANG:
I'm talking about Li now.
I already 20 passed La.
21 DR. REYTBLATT:
Sorry, I didn't 22 MR. HUANG:
Lam is the number you calculate 23 using the containment instrumentation, but after 24 that we have to complete a containment using a flov 25 meter, so we know exactly how much is going out of 27
1 the containment.
Then we go back to instrumentation 2
again.
We want to measure a composite.
So that's a 3
one value you know come from the containment.
4 Another value is the one still continuing leaking 5
from the containment.
So that's the sum of the 6
when you know accurately the one you just measured 7
and calculate.
8 So then we are'using this formula to 9
check against whether this containment 10 instrumentation really works.
Le is the composite 11 leak rate, which compose both impose and measure it, r,
- if you subtracted them, this number actually 12 and 13 indicates a new measuring value for the imposed leak 14 rates.
So we compare this number with the one we 15 know.
The difference of this two has to be less 16 than certain value.
In this case regulation require 17 them to be within.25 La.
18 DR. REYTBLATT:
Can you give us a reference A
19 where it says that you have to bleed containment to 20 zero and whether it is done real in practice?
21 MR. HUANG:
You mean the containment may bleed 22 Li? ^
23 MR. BERNERO:
No.
5 24 DR. REYTBLATT:
No, to zero.
It never is done.
25 MR. BERNERO:
No, not to zero.
~
28
1 DR. REYTBLATT:
What you said is just not 2
correct.
It is a false statement.
Go ahead.
3 MR. HUANG:
I am only stating *what NRC believe 4
should be done at this point.
All right?
5 DR. REYTBLATT:
Oh, I see.
I'm talking about 6
what is really done.
7 MR. BERNERO:
Dr. Reytblatt, wait a minute.
8 He's saying the containment is bled at a measured 9
leak rate, and you seem to be say.ing it's bled to 10 zero. pressure, I assume.
11 DR. REYTBLATT:
No.
I am saying that it is not 12 correct.
That instead of saying it is bled, he was T
4-13 supposed or he intended to say that it should be 14 bled, but it is not bled.
15 MR. HUANG:
Let me correct my statement again.
16 I reviewed many of the textbooks.
We'do have a 17 specific textbook requirement during this test to 18 bleed a containment.
4 19 DR. REYTBLATT:
What is this requirement?
Give 20 us'the reference.
21 MR. MAURA:
The reference for the verification?
22 DR. REYTBLATT:
Yeah, a reference for what he 23 says.
1 24 MR. MAURA:
It's the CFR Standard, Appendix R-50.
25 It does cite it.
4 29 t
=
=
=
r-:_=r__:_____-_=__-__,-
1 MR. NISSIM-SABAT:
What is the value of the La 2
there?
)
3 MR. HUANG:
Very, very constant from containment 4
to containment.
In that's maximum allowed for 5
leakage from the containment.
6 MR. BERNERO:
Let me interject a little bit here, 7
because some of you may not know where the 8
specification comes from.
Depending on the reactor 9
design, the specification of leakage ranges from 10 about 1/10th of a volume percent per day up to 11 something on the order of 1 volume percent per day, 12 and the specification is actually derived by a very 13 old dose calculation.
14 There is a postulated design basis 15 accident with a postulated release of a large 16 quantity of fission products inside the containment, 17 and then the risk analysis or accident analysis 18 further postulates that the weather is bad, in bad a
19 whether for dose calculations is stagnant weather, 20 and they calculate at certain distances from the 21 plant that the radiation' dose to a member of the 22 public for a given period of time should not exceed 23 doses specified in the regulations.
They happen to c:m b
24 be 25 rem whole body exposure and 300 rem thyroid.
l' 25 The leak rate that is used in that 30 n
i the leak rate that is used in 1
calculation is in 2
that calculation is translated to the licensed 3
conditions for the plant, which we call technical 4
specifications, and it is.a catalog of requirements 5
of settings and things like that, and that leak rate
~
6 is the leak rate that is supposed to be demonstrated 7
by this testing on a periodic basis.
8 If you go into the appendix that was 9
referenced, there is a discount, 25 percent discount 10 put on that for further assurance.
In other words, 11 don.'t measure that leak rate, but measure.75 of,it, 12 and then in the procedure as Mr. Huang said, there r-4 13 is this certainty level, you know, the confidence
_~
14 level further imposed.
It is a rather archaic 15 method, later revi,ews it's an unnecessarily tight.or 16 strict calculation, but that's the subject of 17 another meeting or rulemaking.
18 MR. REYES:
I just want to clarify that in the 4
19 boilers of Elsa Bay (phonetic) we have some reactors 20 where the maximum range could be up to two or three 21 percent.
22 MR. BERNERO:
A couple of boile.s, right.
But it would be from 1/ loth of one f
24 percent to could be up to two or three percent.
d>'
25 MR. BERNERO:
In general boiling water reactors l
31
=r 1
will have a higher specification than the 1
2 pressurized water reactors like Zion.
3 ha. HUANG:
Okay.
With this calculating 4
calculation procedure and assumption, the t
5 requirement I just described about above, I want to 6
get into a little bit the effect of temperature 7
equations on ILRT results.
I included this on this 8
slide just to make discussion a little bit easier.
9 This is the mass equation we use at 10 any time, and the temperature here, Ti is the mean 11 temperature in the containment.
That mean 12 temperature can be calculated either using equation T
13 A and Equation B.
Here can be in response to the q.
14 remark just made.
15 Temperature distribution is not 16 normally uniform inside containment due to large 17 size and the presence of subcompartments.
Here as 18 an approximation the containment is divided into 0
19 volumes, subvolumes, and the finite number of 20 sensors are used.
We cannot use infinite number of 21 sensors inside containmene.
It's not engineering 22 and practical.
Temperature after that, 23 temperaturo is monitored by assigned sensors.
la
=
24 Remember, the reason we cannot use a 25 finite number of sensors, because the stabilization 32
1 criteria.
We don't want to see the temperature 2
change as presented by examples.
3 MR. BERNERO:
Yes, there is'a 4
MR. SKLAR:
My name is Abe Sklar, that's 9
5 S-k-1-a-r.
I am a mathematician from IIT, and I 6
wonder in discussing the mean temperature, why do 7
you use a particular set of two possible means?
8 There are infinitely many possible ways in which you 9
could determine a mean, and you've chosen two 10 particular ones, and incidentally, the first one 11 will always give you a lower value than the second 12 one.
wr s
13 MR. HUANG:
You mean, equation A will give you a 14 lowe'r value than equation B7 15 MR. SKLAR:
Yes.
16 MR. HUANG:
It beats me.
17 MR. SKLAR:
And I wonder what is the basis on 18 which you chose these particular ones, and I 19 remember a remark that was made earlier saying that that equation A is in some way more correct 20 the 21 than equation B,
and I wonder on what basis this 22 statement was made?
23 MR. HUANG:
Okay.
The equation is' derived from 24 assuming the volume of a subcompartment constant, 25 and I can shov you a way of how to calculate it, but 33
1 we did not prepare a slide for it.
If you plot 2
DR. REYTBLATT:
Be prepared.
3 MR. HUANG:
Maybe another speaker will show you.
4 The first equation A is actually a direct derivation 5
of using the PV equal RT here.
All right?
Because 6
if you measure a temperature T,
then the T will be D
7 equal to PV over RT here, and then in there you 8
wanted to assuming a constant volume, you can 9
eventually derive using an integral maybe, since we 10 are using finite number of sensors, that's why you 11 use an equation, from that approach you come up with 12 this equation where we try to present it here.
All W
13 right?
Since I heard someone is going to present 14 how we're going to get here.
15 But the reason I'm saying, presenting 16 this slide here, these are the two currently used by 17 the industry to calculate the temperature.
And the 18 first equation, the equation A, are supposed to be d
19 used claimed by the petitioner; however, some 20 utility licensing are using equation B.
This is why 21 I wanted to make some comment on the difference of 22 these two equations.
We can always talk about how 23 to derive them by making certain assumptions.
W l
24 We have looked at the equations and we ws 25 have showed you some of our conclusions in the 34
from previous speaker.
Here is what I
=q l
previous 2
found.
You know, I went back to the literature and 3
what I am able to find out is th a t this issue 4
actually concluded a long. time.
And this is what I 5
found.
6 Eighteen years ago Mr. Zapp and even 7
more recent, I three or four years ago, Mr. Frank 8
concluded that equation A and equation B yield 9
comparable results in the leakage rate calculation.
10 But, of course, if the stabilization holds, if the 11 spacial temperature distribution is insignificant.
12 The s t a'f f data recently reconfirmed this conclusion wm Ls 13 by performing our own leak rate calculation using 14 actual test data and we found the same thin 9 15 However, I want to make another 16 conclusion here.
Despite the minor impact on the 17 calculated leak rate from using thess two different 18 equations, the industry will endorse the equation --
d 19 the use of equation A, since it is more 20 mathematically correct.
The reason we say this, 21 because before in old time when we use hand 22 calculation, the second formula, equation B,
do 23 provide a lot of, you know, timesaving and nowadays
],
24 we are using computer.
So there is no point in 25 continuing arguing which equation should be used.
35
1 We also encourage people now to use the equation A,
S 2
and I wanted to show you 3
MR. GOGEL:
ks that going to be put in the form l
l 4
of law any time soon?
5 MR. HUANG:
No, the regulation I don't believe i
6 we should go into that kind of detail.
However, if 7
you ask me for personal opinion, because I do 8
believe these two equations in our range of 9
temperature measurements, the conditions inside the 10 containment do give you almost identical results.
11 There is some results I can show you 12 from the literature.
You can see now from the l
7, 13 temperature here, the difference from engineer point 14 of view is insignificant.
One is 367 and another is 15 363.
Usually we pick the temperature to come up 16 with a result.
You can look at this other one 17 calculation here, you know, the 296 and 299.
So we
[
i j
18 do conclude that the significance of these j
19 temperature is very small.
And we even go out into i
20 using -- you know, this is what I mean.
We did our v
l 21 own calculation here.
All right.
i 22 We're using the Zion data in this case, 23 and we go out and try to crank out a number ourself.
Y 24 We found that the temperature difference only by a s
25 1/10th of percent in degree, I mean, here.
So we i
l 36
=---A e=-=
e=
= --
r-
- n v-a =
- ~ - - - - - - ~ -
~ ~ -
1 really don't see any difference.
2 DR. REYTBLATT:
Can you leave this slide, please?
3 I will have to refer to it someplace.
4 MR. HUANG:
All right.
No problem.
5 MR. BERNERO:
Set it aside for him for later.
6 DR. REYTBLATT:
Set it aside.
0 7
MR. HUANG:
All right.
We also go in and take a 8
look at the effect of weight coefficients on the I repeat this formula 9
ILRT results.
The weight 10 here because for convenience.
This again, the mass 11 equation, and here I only use the mean temperature 12 used in there, so it is a more corrected formula, 13 and the alpha j is the weight coefficient of this.
14 And I wanted to give you a definition of' weight 15 coefficient here.
16 The weight coefficient defines a 17 fraction of the containment volume within which the 18 temperature is assumed essentially uniform during 19 the test.
20 The second point I want to make is the 21 concept using a single set of weight coefficients 22 assumes that the spatial temperature distribution 23 does not change dramatically during the test.
Here N
24 l(>-
~
I again take you back to the stabilization.
25 So from the assumption and the 9
37 l
1 computation formula we intended to use we conclude 1
4 2
we conclude that the concept using a single set of 3
weight coefficients relies on a stable
- spacial 4
temperature distribution.
For this reason the 5
regulation as well as industry sponsored guidance 6
emphasize stabili=ation.
And appendix here requires 9
7 stabilization before you even go into it, start your 8
leak rate calculation.
9 And to even more substantial point, we 10 perform a parametric analysis, and demonstrate that 11 with stabilization, even you vary the weight 12 coefficients to certain extreme, and.the effect 9
13 would be very small on your leak rate calculation.
14 And again, for that we also now took a quick 15 calculation using actually data here.
16 We verified those weight coefficients 17 as you can see.
Case one we're using the other 18 equation.
We get this leak rate, and we now all a 19 sudden change from five volume to one volume.
We 20 give every sensor a leak weight.
We treat it like a 21 one volume.
We get this leak rate, then we use the 22 other formula which is five formula.
23 You can see.
This shows you two c
24 things here.
One, these two equations give you
~.
25 almost identical result, and the case two, yes, 38
1 there is some changes, but in this particular case
.,y 2
lower the leak rate calculated now.
And to help you 3
at least 4
DR. REYTBLATT:
In my petition a presented this 5
low cale on the effect of weight coefficients.
Did 6
you review it?
7 MR. HUANG:
Yes, we did.
8 DR. REYTBLATT:
Show it, please.
9 MR. HUANG:
We're going to get it into that when 10 we discuss it.
11 DR. REYTBLATT:
Because you simply' ignore the 12 facts.
This is not a descent way to discuss the m
13 things.
s 14 MR. HUANG:
We're going tc do that.
I only 15 present the ILRT methodology here we believe should 16 be followed.
We are going to go into your example 17 maybe later when the time comes.
18 Here is the reference we used to make e
19 this presentation, and will help you to understand 20 what a leak rate methodology is.
21 Thank you very much.
22 MR. BERNERO:
I would like to add a few remarks.
23 Some members of the audience have asked questions 24 about the approximate range of pressure to get a c-25 physical feel for the dimensions involved.
P 39
1 When one speaks of a containment leak
-s. ;
2 rate test for a containment like the Zion Nuclear 3
Plant up there on Lake Michigan north of Chicago, 4
the containment volume is approximately 2 million 5
cubic feet.
And the specification is on the order 6
of 1/10th of a volume percent per day.
I believe 7
Zion is exactly
.1 volume percent per day.
So that 8
this gives you a sense of.the gas flow involved. *It 9
is an extremely low leakage rate.
10 And the experimental problem of 11 measuring such a volume is formidable because of the 12 sheer size of it.
It is mostly a free volume, but 13 it includes systems down low and in concrete 14 compartments that are open at the top for individual 15 steam generators and things like that. - So that you 16 do have free gas flow between spaces, but some parts 17 of that 2 million cubic feet are more constrained 18 than others and have metal surfaces and concrete 19 surfaces around them.
20 Now, Fred, do you want to talk about i
21 that scale?
22 MR. MAURA:
Okay.
What we did is we ran a set 23 of studies based on actual test data and I'll open e'.:::,.
t 24 by just showing you the first set was Zion 1983 test.
e 25 And this gives you an idea of the containment, where 40
mr 1
the sensors were placed.
This is the, you call it 2
the upper half of the containment, the open area, 3
and this is the next floor below the refueling floor, 4
and then way down at the basement.
These are just 5
to give you a quick idea of where the sensors were 6
placed.
7 okay.
What we did is we made three 8
cases.
We ' studied three cases, and in case number 1,
9 we used an actual weighting factors determined from 10 the location of the sensors.
In the case of number 11 2,
what I did is I assigned every sensor the same 12 weighting factor.
I said, forget about weighting 13 these coefficients.
We'll just take one big average.
~~
14 The third case was where I arbitrarily went down the 15 line with no rhyme or reason, just started writing 16 numbers, making sure that at the end I would come 17 back to a total of one.
I have to always end up 18 with one, but the numbers meant nothing.
19 The only thing that should I should 20 point out is, for example, on the sensor number 1,
I 21 ended up in case number 3 with a weighting factor 22 almost 2 times its true weighting factor, number yeah, where we say 1/10th here.
23 five, I guess
~
24 This is about 2.
In some cases like down -- this av 25 one down here at point 7,
and the one with point 08, 41
l those are about 4 times their original weight.
eq 2
Based on that, when we ran the results 3
I will put the slide in, although it'doesn't add 4
anything, it is the same thing presented'a different 5
way -- the only difference, it gives youfyour 6
subvolumes instead of just listing all the s'insors.
7 These slides are available, and I don't think you did you get a copy of them?
8 have 9
DR. REYTBLATT:
No, I don't have a copy of them.
10 I would like to see them.
I may use them.
Just 11 keep them.
12 MR. MAURA:
Okay.
We'll take it one at a time.
~.
13 When we analyzed the data, we -- by the way, we-14 placed in the computer every data set.
We did not 15 leave any data sets missing or anything.
every data 16 set for that 12-hour test was used, and we 17 calculated -- our program can do it either two ways, 18 using either equation A or B that John presented.
19 And the results show that for the case 20 1 with the actual containment weighting factors.
21 applied, the results were the 95 percent, point 015.
22 You can see the differences in mass run about 138 23 pounds, and the temperature difference about half a P~
24 degree.
/
~
25 When we went to case numb e r.- 2, it gave
'42 w.
mg 1
us very similar results, but a little over a half a 2
degree difference in the mass.
The difference in 3
mass changed.
It went from 138 pounds to 100 pounds.
4 We ended up with leakage rate of about 5
022.
Case number 3,
in which were arbitrary numbers 6
used, gave the same identical result as in case 7
number 1.
So it indicated that the fact that I 8
weighted four times heavier than others made no 9
difference.
10 When we used a so-called more exact
.11 formula for temperature calculations, for all 12 purposes we come up with the same results.
The only m
13 differences are a pound or two in the change of the 14 mass in the containment at the beginning or at the 15 end.
By the way, the first number it is telling you 16 start up test, and the last is end of test.
So we 17 started with a mass here at the beginning of the 18 test, which would be say 319 versus 181, forgetting 19 about the thousands.
20 So then we since there was another 21 test that was challenged which was the LaSalle Unit 22 183, we also ran that test, and this gives you an 23 idea of what the containment subvolumes are and
[>
24 within the subvolumes th e distribution of sensors, m
25 I don't have a slide for it, but we have the data 43
1 available.
s 2
DR. REYTBLATT:
I have the LaSalle slide.
I'll 3
give it to you, if you want it.
4 MR. MAURA:
Well, that's all right, I don't 5
really need it for the case.
6 DR. REYTBLATT:
I'll be talking about LaSalle.
7 MR. MAURA:
In this case we again try the same 8
three cases, the three different.
Number 1 similar 9
as in Zion, it's the actual weighting factors as 10 calculated based on the volume that they represented.
11 Number 2,
again an arbitrary not 12 arbitrary, just calling it 1 volume instead of 8 7
s 13 subvolumes.
Case number. 3 was not as arbitrary.
14 This time on purpose I decided to shift the weight 15 from sensors that were relativel,y -
.to sensors that 16 were relatively stable, and the reason I did that is 17 because in the Zion case, after I ran the case, I 18 realized that it did not prove anything in the sense 19 that I got the same results.
20 DR. REYTBLATT:
How many sensors there was at 21 that time, the 29 or 307 22 MR. MAURA:
In this case I use 30 sensors.
23 There are 30 sensors.
[
24 DR. REYTBLATT:
Why in the report there are only 25 297 44
l MR. MAURA:
Which report?
,.q 2
DR. REYTBLATT:
LaSalle.
3 MR. MAURA:
Because there was one that was taken 4
out of service, and what I did is I left it in on 5
purpose to show you that it.makes no difference.
6 DR. REYTBLATT:
Then why was it taken out?
7 HR. MAURA:
Because the sensor was drifting out 8
of service.
.If you have a sensor that is varying 9
throughout a test, then you know that something with 10 the instrumentation, not with the air.
The air 11 temperature is not varying at that rate.
If I have relatively next to each 12 two sensors, one next 13 other and one is not varying relative to my average,
~m 14 and I have one that is varying, which one is right?
15 DR. REYTBLATT:
We'll talk about.th a t.
We'll 16 talk about this.
17 MR. MAURA:
So in any case, this time I shifted 18 the weight.
And the only reason I decided on case 19
, number 2 to shift the weight to sensors that were 20 relatively stable, was that case number 2 did the 21 opposite.
So it also gave a good comparison of the 22 two effects.
23 At this time I want to also point out
[_
i 24 that the shift in the weighting was more significant 25 than in the Zion case.
For example, for subvolume 6,
45
1 I made a shift equivalent on case number 2 of about 2
7 times.
I heavier weighted that subvolume 7 times 3
larger than originally, versus in the case of number 4
3, I should point out that subvolume number 8 5
represents about 40 percent of 'th e total volume of 6
that containment, so we went from 40 percent and 7
changed it to make it represent roughly 60 percent 8
of the total volume.
9 When he we ran those cases, what we 10 came up was that as expected, because we had seen 11 this at other plants take place, when you weight the 12 sensors towa.rd-those that are varying, increasing in
~n
~_
13 this particular case, the number that you get which 14 is case number 2, will be a conservative number, is 15 higher than the actual leakage.
If you weight in 16 the other direction, in case where they you put 17 all the weight on the sensors that are stable, than 18 you have a somewhat nonconservative, it is a smaller O
19 leakage, than casa number 1 which is what we state 20 is the true leakage.
21 MR. BERNERO:
Hold on, Fred, while he changes 22 his tape.
23 MR. MAURA:
Yes.
l e=:--
f.
24 Again, this chart indicates that it W
l l
25 for all purposes makes no difference which method of I
~
46
q l
temperature computation we use.
Because in either you're talking in the third 2
case the results are 3
decimal place a change of about 7 thousands of a 4
percent.
~
5 MR..GOGEL:
That's which method of calculating a 6
mean temperature for the entire containment at every 7
point in time you're saying.
8 MR. MAURA:
Right, it makes no difference 9
whether you use the ANSI formula or whether you use 10 the so-called more correct formula.
11 MR. REYES:
For the temperature changes involved.
12 MR. MAURA:
Yes, for the temperature changes m
13 that you see during that test, because that is a 14 critic'al point, because if we start talking about 15 temperature changes of 50, 100, 200 degrees, we have 16 got another ballgame, but we have never in my 27 17 years of doing leak rate testing ever seen changes 18 more in -- I would say the LaSalle is a typical case 19 that there were changes greater than at Zion.
This 20 specific, the '82 LaSalle versus th.e Zion
'83, and I 21 would.say this are probably about as severe as I 22 will see.
'Now, there may have occurred someshere 23 aise, but I haven't seen those.
24 Okay.
So that really summarizes the 25 studies that we made for the two tests.
And then 47
kind of summarizes 1
the final slide I have is the s
2 what the inspection program.
What is the inspector 3
supposed to do or does when he's at a site to insure 4
that some of these areas that John talked about and
~
5 we have talked about today are enforced.
6 And for example, an inspector, the 7
first thing he typically do is review the procedure.
8 He may do that here at the region or he may do it at 9
the site, but in any case, he will review the conforms 10 procedure to make sure that it includes 11 with the license, with the regulation, with the 12 standards that are being used, with their own m
~.-
13 textbooks.
14 Then he will go to the site.
He will 15 make a request from the licensee for drawings, get 16 from them the procedure, he will get a flow path 17 that they plan to implement, what valves will be he will select 18 closed, open, and he will look 19 penetrations, and preferably penetrations that will 20 have an affect on the system, on the test.
21 Penetrations that could effect entry of mass into or that have had a 22 the containment, or could 23 history of problems, and he will look in there and 24 make sure that the line up is correct.
25 He will later perform an independent 48
1 verification that those valves, those volumes have 2
been properly vented, and also of the sensored 3
location.
So here again is the critical sensor
-4 location decision.
He will not just accept what the 5
licensee told him where the sensors are.
He will 6
verify them.
he will obtain 7
He will perform 8-at this point in time the licensee by now is pumping 9
the containment, and during that time he will 10 request a copy of a computer printout out of the 11 computer, what are the fact. ors that you have placed 12 in this computer.
He will get the weighting factors
' -n 13 that the licensee has placed.
He will get any 14 correction factors for pressures, anything that came
,15 out of the calibration of the instrumentation that 16 was done a week or two, whenever prior to the test.
17 He will then use those factors.
- First, 18 he will verify them to make sure they are correct, 19 because there have been times when we have found 20 that mistakes have been made, and maybe the factors 21 don't add up to one, or are different from the 22 temperature surveys that were performed earlier.
So 23 he will verify that.
P' 24 And then he will perform an w--
25 independent calculation of at least a minimum of one 49
z 1
data set.
The main reason for that is up to now, up 2
to prior to the development of the NRC computer code, 3
our capability of calculations were based on a Ti or 4
a HP hand-cale probable calculator, but is limited
~
5 to the number of steps that you can have.
So we 6
could not input to say 30, 40, 50 sensor data in a
7 that small calculator and come up with all the stuff.
8 But if you make one or two to verify 9
that the computer that the licensee is using is 10 calculating masses correctly, there are no errors 11 anywhere, you don't have to do any more than that.
12 You have already verified the accuracy.
After that, U
13 you will witness and verify the temperature 14 stabilization period.
Typically, wh'at I do because 15 you are limited in manpower, is that I schedule 16 myself to be there during that period through the 17 temperature stabilization, and maybe catch the first 18 two or three hours of the actual test.
Then I leave 19 to get some sleep, and will return toward the tail 20 end of the test to try to catch maybe the last four 21 to six hours of the test and try to catch the 22 verification test.
I consider that is the most 23 critical periods, the beginning and the end.
What
-T 24 goes on in the 12 hours1.388889e-4 days <br />0.00333 hours <br />1.984127e-5 weeks <br />4.566e-6 months <br /> in the middle is -- you can 25 always find out if they had a problem when you 50
1 return to the site.
2 So you look at the stabilization.
3 once you're satisfied that the containment is the 4
containment will never be perfectly stable, you know.
yeah?
5 So this is something that 6
DR. REYTBLATT:
What do you say about Zion s
7 weight coefficients?
Do you say they are correct?
8 MR. MAURA:
Yes.
9 DR. REYTBLATT:
Second question, what is your 10 procedure when the verification test fails?
11 MR. MAURA:
You do another verification test.
12 You have to look at the whole thing, and you have to
--- s 13 determine on the basis of what you see.
You know, 14 there is a certain amount of engineering judgment in 15 this thing, that has to be applied.
This is not a 16 black and white art.
17 DR. REYTBLATT:
All right.
This is not black 18 and white.
All right.
19 MR. MAURA:
But 20 DR. REYTBLATT:
All right.
21 MR. MAURA:
But with regards to the first 22 question you had, are they perfect, the weighting 23 coefficients, well, we don't know whether they are 24 perfect or not, but we demonstrated through these 25 slides that even if they are off through say 10 51
=q l
percent, 20 percent, it has no effect.
2 DR. REYTBLATT:
We'll show that this is not so.
3 MR. GOGEL:
In the 20 years you say you've been 4
doing this, how many times have you ordered the P
5 primary test repeated because the verification test 6
has failed?
Not just repeat-the verification test e
7 again, but repeat the primary test.
+8 MR. MAURA:
We have failed tests, if that's wha *.
9 you mean.
10 MR. GOGEL:
Have you actually said repeat the 11-primary test?
I'm not talking about the leak that 12 was higher than the value.
I'm saying you run a e
i.
13 primary test, you run a verification test, it 14 doesn't work in your judgment, and then you go back 15 and say, run,the primary test again.
Have you ever 16 done that?
17 MR. MAURA:
The regulation does not require that 18 the primary test be r.e p e a t e d.
Regulation only 19 requires that the verification test be repeated, 20 corrections to be taken, if there are any 21 corrections to be taken.
You know, you have to 22 explain why the first one failed.
It doesn't mean 23 you have to repeat the entire test.
m
}'
24 DR. REYTBLATT Do you have to review your set 25 of weight coefficients if your verification tests M
l 52
7-1 fails or not?
2 MR. MAURA:
You don't have to.
3 DR. REYTBLATT:
You don't have to.
Thank you.
4 MR. REYES:
It should have gone on already.
~
5 MR. GOGEL:
Have you actually ordered a primary 6
test repeated because of a failure of a verification 7
test in your personal knowledge?
8 MR. MAURA:
Not or the primary test.
9 MR. BERNERO:
Carl, do you want to add something?
there was a 10 MR. PAPERIELLO:
We recently 11 certain, at least part of if, not an entire 12 integrated leak rate test, that was rerun at Cook.
b 13 MR. REYES:
But that was a primary test.
You 14 see, what happens in real life, when y'ou pressurize 15 a vessel of 2 million cubic feet to 20, 3 0,: 40, 60 16 pounds, if it is leaking excessively, you find that, 17 stop the test, go back and find out why, do all the 18 activities you have to do and repeat it.
But his 19 question is, once you do the first step, the first 20 test, and you pass, you get a satisfactory result, 21 then you induce a known leakage and*go through the 22 process of conducting a verification test, and has 23 Fred specifically been in a case where the second 24 test failed and we have ordered them to repeat the v
25 first test.
-w 53
=q l
MR. GOGEL:
If you run the first test, and as t
2 you said, and the first test shows a high leak rate, 3
why don't you just order the reactor shutdown rather i
4 than 5
MR. REYES:
The reactor is down already.
6 MR. GOGEL:
You just said you run a test and you 7
find a leak rate, a high leak rate, a real high leak 8
rate, why do you assume that those same conditions 9
won't be repeated during a real accident situation?
10 MR. PAPERIELLO:
The answer of the question is 11 why.
It's clearly if you have a whole blown in 12 containment, you've got a different problem than if
-r,
a 13 you find that there is a component that's been left 14 in the wrong position.
15 MR. GOGEL:
But if that happens, couldn't that 16 happen during normal operations?
17 MR. PAPERIELLO:
If you pass a test, it can 18 still be put in the wrong position.
19 MR. BERNERO:
I would like to interject here, if 20 I could, with a little explanation.
This is 21 something I think Dr. Reytblatt knows about from 22 conversations with Mr. Stallo at the NRC.
23 You remember I said that the basis for 24 leak rate requirements and the testing requirements 25 goes back more than two decades to the dose rate 54
=
= =r=-
=
= x---
~
=p, 1
calculations based on postulated accidents and 2
fission products in the containment, and it is a 3
very good point.
It says our leak test procedure 4
puts a degree of integrity.or leak tightness into 5
the containment, verifies that the containment can 6
hold that leak tightness under those conditions, and o
7 then the plant proceeds to operate, and you come 8
back one refueling cycle or two refueling cycles 9
later to repeat that.
10 Now, there is also a regime of 11 individual valve tests that are required by the 12 these same regulations that add confidence, but T'
13 nonetheless, it has been established from experience 14 that there needs to be concern about the possibility, 15 in a grotesque example, someone opens the air lock 16 that goes into the containment and leaves the door 17 open.
Now, of course, there is an alarm on it, but 18 let's just assume for the moment that the door is b
19 left open.
That would have a predisposed leak path, 20 and it would be far more significant than subtle 21 differences in containment leakage.
22 We have in our regulations for 23 containment leak tightness, and in a current 24 proposed regulation, a system based on leak 25 tightness integrity.
There are many studies, and 55
~
rr 1
this current proposed regulation speaks to it, there 2
are many studies that suggest a shift, a substantial 3
shift away-from leak tightness regime as the 4
principal emphasis to pressure management or 5
predisposed opening surveillance as a more important
~
6 way to' provide assurance of containment integrity.
7 And there are in the literature now an abundance of risk of nuclear release to ' people 8
studies of risk 9
off site, and how sensitive the risk of a member of 10 the public is to the leak tightness as against some 11 other factor of the containment performance.
12 And I don't want to discount the m
13 question raised.
There is a substantive safety 14 issue.
What means are provided to assure that the are not left open.
Actually, that 15 air-lock d,o o r s 16 example is pretty well guarded against by alarms and, 17 of course,- it's a double door, but there have been 18 examples, and we have just published a report not s
19 long ago on the experience of predisposed opening 20 times at one time or another where a valve was 21 opened and discovered while the plant was in 22 operation.
And the valve had been open for a good 23 period of time.
It is a not a trivial issue and it's really 24 we're not ignoring that, but that is a
25 not directly affecting what we're talking about here t
56
q l
as to the validity of the leak rate testing 2
procedures.
3 Yes, sir you have a 4
MR. NISSIM-SABAT:
I'm so-t of asking my 5
questions out of ignorance, but it seems to me to 6
determine the mass in the container, you need to a
7 know the temperature in separate volumes, but you 8
also need to know the pressure.
And I see almost no 9
mention made of the pressure sensors.
I would think 10 they are much more delicate, much more subject to 11 errors, and I wish I would have seen an analysis of 12 how they work.
m 13 MR. MAURA:
The containment although is divided 14 in subvolumes, 'some of the subvolumes-- I should say 15 all of the subv,olumes are imaginary subvolumes.
16 It's not really a tight subvolume.
They all 17 communicate.
There is one pressure.
18 MR. NISSIM-SABAT:
But what I haven't seen and a
19 again, I have said 20 DR. REYTBLATT:
Can I answer this?
21 MR. NISSIM-SABAT:
Please.
22 DR. REYTBLATT:
In some instances up to six 23 pressure gauges is placed.
Unfortunately, there P-*
I 24 were cases when the change in pressure of all six of v
25 those pressure gauges was significant.
So in other 57
=r 1
words, it was not uniform.
It is a common practice; 2
for example, LaSalle test 1982, that there are two 3
pressure gauges, they are on the same line, and one 4
reading is thrown out.
It is thrown out because of a
5 fluctuations that are wild.
Also the one is the least 6
selected that gives the leak pressure g
7 leak rate.
8 I can attest to this, because I had and I 9
available data from two tests, and I was 10 raised this question, and they just told me, well, 11 we don't want.that pressure gauge, it doesn't work 12 right.
c 13 MR. MAURA:
The only comment I will have,is th a t 14 the 1982 test was witnessed by myself, and I don't 15 remember s,eeing you there.
16 DR. REYTBLATT:
This is what Mr. Reyes told me 17 on January 5,
1984.
18 MR. REYES:
I would like to find out what you a
19 consider significant.
You know, if the pressure 20 gauges are displayed through the containment, which 21 Fred had a slide there in terms of height, you're 22 going to have some difference in the location.
23 DR. REYTBLATT:
Are you talking about head 24 pressure?
No, we are talking about the changes.
25 MR. REYES:
But that vessel is sealed, okay.
58
1 DR. REYTBLATT:
So what?
2 MR. REYES:
Let me finish.
And if you look at 3
the readings between the instruments, there may be 4
some significance, but they may be instrument error, S
and I would like to find out what you consider 6
significant, because as you notice, we have some 7
accuracy.
We are not saying this is a perfect 8
measurement.
There are some tolerances in terms of 9
the readings.
10 DR. REYTBLATT:
I will speak about this.
11 MR. MAURA:
- Anyway, just to finish what an 12 inspector would do, he would perform independent 13 calculations then of the type A test.
He will 14 observe the individual sensored data for trends.
He 15 will review the log, what is going on, any procedure, 16 changes that are performed, and this would guarantee 17 that, for example, weight coefficients are not 18 changed arbitrarily throughout the test.
e 19 When the time for the superimposed 20 leakage comes, he will actually verify that the 21 actual readings the actual readings of the flow 22 meter.
He actually will go to the flow meter and 23 take his own independent readings.
He will not 24 depend on the licensee's reading, and he will then j
25 also perform independent calculations of that test 59
eq l
to make sure that it comes within let's say the 2
point 25.
3 At that point he will, if everything 1
4 went right, you know, complete his inspection and
!l" 5
ret. urn home.
If things went wrong, there would be 6
more discussions than with the licensee.
7 I don't have anything else to add at l
8' this time unless you have some questions.
9 MR. BERNERO:
Well, Fred, what I would like to 10 do -- thank you.
I would like to just recapitulate 11 for the NRC that when you wrote in your two petition 12 letters, Dr. Reytblatt, we took what you said, your 13 written petitions, did analyze them, did analyze 14 them technically, and based'on what you've just l
15 heard, these are the reasons why the NRC prepared, 16 the NRC staff prepared and presented to Mr. Denton 17 the denial'of your petition that was tendered to you 18 earlier this year.
19 Now, we are satisfied within the l
20 reservations I expressed earlier about' predisposed 21 openings and other safety issues, we are satisfied l
l 22 that the present leak test methods provide I
23 sufficient reproduceability and sensitivity to 24 measure the objectives, the 1/10th volume percent or 25 5/10the volume percent, whatever it might be for l
l 60
_ _ _ _ _ _ - - ~
l l
l f
l v7 1
that plant, and we are also satisfied that our l
I 2
inspection programs as described by Fred Maura is 1
3 sufficiently thorough to satisfy ourselves that j
4 licensees who have the primary responsibility aren't j
5 just simply manipulating numbers to come out with 6
the required answer.
We do have a thorough 7
evaluation of the work that is done by that method.
8 Now, we have been talking about 9
Commonwealth Edison plants, but I'll ask i
10 Commonwealth Edison to be patient, and I would like I
11 to turn the floor over to you, Dr. Reytblatt.
I 12 DR. REYTBLATT:
Okay, thank you.
t lv*
m ex.
13 MR. BERNERO:
Here are the two slides that I
14 Mr. Shapaker had.
I 15 DR. REYTBLATT:
Yes, and then I need that 16 fraudulent calculation that you did at the NRC.
I 17 MR. BERNERO:
You have a catching style of the l
18 use of adjectives.
t 19 DR. REYTBLATT:
I am responsible.
I am not j
20 saying about other things, but only about things 21 that are really so.
So in case somebody wants to 22 sue me, I have available.
i 23 I do not want even to talk about what
=r 24 Mr. Shapaker said.
What about Mr. Huang said, I can s
25 say that the first line was wrong.
And the way 61
=q 1
which students work, if we see it's wrong, than we 2
just abandon.
But I will cover the whole thing from 3
the beginning.
And the first thing, of course, I 4
want to show my request and responses, my complaint.
5 I complained that the equation that is-6 used in practice all over the United States, I 7
didn't mention any standards, is wrong.
I requested 8
that you return to the conventional physics equation 9
immediately.
I received no technical response from 10 the NRC.
The response had no equations, no 11 derivation, no calculations.
It stated that, well, 12 there is slightly mor'e, slightly less, there is a e
a 13 matter of opinion.
No, sir, this is nr
- he matter 14 of opinion.
This is a very simple thing that has 15 been known for 200 years.
16 Second, my complaint was that Zion' 17 weighting coefficients are wrong.
I presented the 18 NRC with a drawing, and requested to explain why 19 they believe these weight coefficients are such as 20 stated.
It is not addressed in the response of a
21 Mr. Denton.
Obviously, he cannot respond to this.
22 Third complaint was that the method 23 for weighting coefficients is bad.
I requested that 24 testing be suspended until the method is improved, s
25 I did not request an absolute suspension, only until 62
=g 1
you improve it.
It was not addressed in the 2
response.
3 Next, my complaint was that Com Ed 4
computer program and other programs either do not 5
calculate the sum or fraudulently may erase data and 6
double the weight coefficients.
I have actually the 7
listings of two computer programs that I will show, 8
and I will show where the fraudulent computer option 9
is in those programs, and I insist that these 10 programs be immediately withdrawn and debugged and 11 revalidated.
12 And then my last comp 1aint was that c>
13 the director's decision of March 16, 1984, is based 14 on false statement.
The falso statement was that 15 decision said we do not address Mr. Reytblatt's 16 report because it is proprietary, but we still 17 somehow reviewed it, and these are the conclusions, 18 and that was the blatant lie.
The report was placed 19 in the CFR seven or eight months before Mr.
I 20 don't know, Mr. Arndt or Mr. Shapaker, who knows, 21 prepared this shameful decision.
22 Laws do not allow falso statement.
No 23 one with falso statements in director's decision.
i
-~
24 Without any going into the details of this decision, e
j 25 I insist that this decision be immediately made null e
63
T 1
and void.
2 And now I want to be more specific and also I will cover a few 3
say a few words about 4
additional subjects.
Since Mr. Huang started this, 5
I think it's worth it.
Stabilizing, equalizing, 6
ventilating, irrigating.
This basically means for 7
cheating.
There is no scientific substance 8
whatsoever.
I am giving examples that show this.
9 And diurnal effect, short duration tests again also 10 basically means of cheating.
Reduced pressure tests, 11 the same thing.
And then I like still the Nuclear 12 Regulatory Commission to provide an answer to the v
l>
13 questAon that I raised three years ago in my report.
14 It has not been answered for three years.
Evidently l
15 you just don't know what to a,nswer.
16 Let's go to leak rate test.
The gas 17 law is very well-known.
Raw equals K coefficients, 18 P over T.
T is an absolute temperature at a given 19 point.
At every given point the density is 20 different.
From there you just obtain a very simple l
1 21 equation, which has been known for 200 years.
Only 22 the ignorance derives like excuse me, Mr. Huang, l
23 I just -- I am outraged.
It is ridiculous what you 24 said.
25 This is the wrong equation.
Look at 64
i q
1 this.
M equals KP over.
It is quite clear for 2
anyone who knows elementary calculus that the sum in 3
the bottom tends to the integral, all right, when 4
you increase the number of instruments and points.
5 So we can say that the first equation is an the more 6
approximation, because if you have 7
instruments you have the better results you obtain.
8 The second is not an approximation to 9
the correct equation, but is an approximation to thp 10 wrong equation.
The more instruments you have, the 11 more so to speak you have the result.
That's just 12 the simple scientific truth, m
fx 13 MR. BERNERO:
And I assume you're saying that 14 the use of the equation in the methods described 15 MR. REYTBLATT I will talk about this.
gives significant error?
16 MR. BERNERO:
17 MR. REYTBLATT:
Yes, yes, I will show the 18 figures, I will show the figures.
And I will show 19 the figures not that I received.
They have these 20 figures.
They never answered those figures.
But I 21 will show the figures from the Oak Ridge National 22 Lab, and I was there.
I picked up some of the stuff.
23 Also I have letters from Oak Ridge National Lab
=
24 through Arndt, where they say the whole thing and a
25 how they covered up.
I even have Jim Duggin's diary.
65
=t 1
Anyway, I would like really to investigate the 2
integrity.
3 If you do a variation of the equation 4
for the mass, you obtain the following equation for 5
the leak rate.
Where delta is the change, is a 6
symbol of change, and absoluto is supposedly small.
7 I have a paper which shows that it is not that small 8
to substantiate very short duration tests.
Like a 9
very conservative estimato shows that you have to 10 test for about 400 days in order to obtain really, 11 you know, top notch results.
Practically, probably, 12 I will talk about this, you have to test it for at 7
13 least weeks.
14 Please take a look at this equation.
15 Alpha sub 1,
sub 2, snd so on are weight 16 coefficients.
Usually it is just you pick up almost 17 any test, you will soo that dolta T sub 1, delta T 18 sub 2 may have different signs.
That means thoso 19 quantities, those terms, make and do cancel each 20 other at certain values of weight coefficients.
21 If ybu specifically orase some of 22 these weight coefficients, and there are computer 23 programs that do it, fraudulent computer programs,
=T 24 then you can manipulate leak rato the way you want, 25 and any value of loak rato can bo datormined from 66
=-
1 any set of raw data.
This was validated by work l
2 examples which are available to Mr. Huang, 3
Mr. Shapaker.
This was substantiated by actual 4
tests, negative leaks.
This was admitted by i.*
5 Mr. Arndt himself in one of his letters, a copy of 6
which I can provide you with, at the discussion at 7
ACRS.
8 Many prominent people were talking 9
about the same thing, that weight coefficients 10 affect substantially -- I will show the example.
11 Mr. Huang, Mr. Shapaker, they were attending many of 12 those meetings, didn't say a word.
Now they say L>
13 something without any substance.
No real examples.
14 We will talk about that.
15 Let us show something.
This is 16 unfortunately, I cannot show the slide, somehow it 17 didn't come out, but anyway, this.is LaSalle volume 18 subvolume 7, subvolume 7.
It has 4 gauges.
19 MR. BERNERO:
I think Fred might have it.
Just 20 one moment.
21 Di. RE T6 tLATT:
Yes.
Anyway, please take a 22 notice, put down, I want you to rerun the test I
23 because the weight coefficients, even if you r
24 calculate them the way Mr. Maura does, exceed
.1.
J 25 The total volume of this subvolume is.418.
If you 67 i
=r i
divide it by 4,
you determine -- divide it.105 or 2
something, which is in excess of
.1.
The law does 3
not allow to use such coefficients.
4 And let us take a look at what might 5
have happened with the data with the LaSalle data.
6 Let it be known that the data of LaSalle have been 7
requested two years ago.
Nothing on temperature 8
sensors have been received.
Nothing.
9 Mr. Thomanovich (phonetic) promised to send me.
He 10 lied.
Never send.
So I could not really verify 11 unfortunately your figures, Mr. Madra, unfortunately, 12 but if you submit this, and if you submit your v
13 computer program, because I know how to manipulate 14 your program, I will show to you that the 15 conservative estimate of the leak rate on LaSalle is 16 much much higher, and this is the proof.
17 Let's take a look at this again, 18 subvolume.
There are 4 sensors.
Their wrong 19 coefficients what Mr. Maura says.
The correct 20 coefficients are.105,
.105,
.105,
.105.
Yes.
More 21 correct realistic actually coefficients are, if he's 22 interested, I can show it to you how I obtained them, 23 is known that I do not know specifically what were
{
24 the reading of these sensors, because that was not 25 made available to me, but I know that the 68
1 temperature changes in the LaSalle were in excess, 2
correct me if I'm wrong, of 15 percent 15 degrees 3
overall, and spacial 40 degrees.
These are the 4
figures in your report, which were confirmed by 5
Mr. Luis Reyes.
6 MR. REYES:
Wait a minute.
You're going to have 7
to explain, when you say 8
- DR. REYTBLATT:
All right, explain.
9 MR. REYES:
No, no, you explain what do you mean 10 15 degrees from where and where, and in time.
11.
You're saying 12 DR. REYTBLATT:
In time 13 MR. REYES:
You say how much it changed from 7
14 where to where location wise, spacial location and 15 in time, because I don't 16 DR. REYTBLATT:
That I wouldn't remember and 17 that was not my purpose, because I am not studying 18 particular tests of which I have no data.
But I 19 have a knowledge that the temperature changes at 20 certain points, at least at one point, was over 15 21 degrees, and that the spacial variation was 40 22 degrees.
And if you object to this, you can attach 23 this objection, but these are bona fide, bona fide
=#
24 numbers.
I assumed looser numbers.
I just didn't sr 25 want to go to this.
69
=r 1
MR. BERNERO:
Dr. Reytblatt, excuse me, could the column you have listed as 2
you characterize 3
realistic weighting coefficients 4
DR. REYTBLATT:
Yes.
could you characterize how you 5
MR. BERNERO:
6 selected them?
In other words 7
DR. REYTBLATT:
Yes, yes.
8 MR. BERNERO:
-- what factors of those 9
subvolumes 10 DR. REYTBLATT:
Yes, yes, I can do this, because 11 I believe that of what I say.
- Probably, this is 12 sort of a little bit proprietary thing, but I 4.
13 believe that Mr. Huang is not going to make use of 14 it.
There is a so-called analysis, you know, a 15 theory of integration, quadrature formulas, and 16 knowing something about the nodes, you can derive 17 the weight coefficients.
These are simply 18 coefficients of quadrature formulas, and this stuff 19 is in very well-known books.
Unfortunately, 20 Mr. Huang is not aware of this stuff, but a n y w'a y, we 21 can learn a little bit.
So let us return.
22 MR. MAURA:
Which subvolume is this?
23 DR. REYTBLATT:
This is subvolume number 7,
=-
24 probably.
Yes, subvolume 7.
Subvolume 7.
It 25 exceeds
.1.
Actually, I told about this to 70
sj 1
Mr. Reyes, and I mentioned it several times.
I am 2
not for shutting down nuclear power plants on minor 3
things.
Please understand me.
But if people insist 4
on there infallibility, yes, please do so.
I 5
MR. REYES:
What do you mean it exceeds.l?
6 DR. REYTBLATT:
It-does exceed
.1.
It is
.105.
7 MR. REYES:
And you say the realistic one is 8
point.104 9
DR. REYTBLATT:
Yes, some one of those.
10 MR. REYES:
So what's wrong with that?
11 DR. REYT B LATT With what?
12 MR. REYES:
With.10 l ?'
?
I h
13 DR. REYTBLATT:
The wrong is that the law does i
l 14 not allow this.
15 MR. REYES:
You're incorrect.
16 DR. REYTBLATT:
Yes, I am correct.
When you i
L 17 look at the regulations, you will find it, that
.1 l
18 is the limit.
19 MR. REYES:
You're incorrect.
20 DR. REYTBL'TT:
Yes, I am.
Yes, I am.
A o
21 MR. REYES:
You're incorrect.
22 MR. MAURA:
You just agreed.
23 MR. REYES:
I have the record.
e 24 DR. REYTBLATT:
Show me this thing.
There will
?-
25 be another test which has more than
.l.
Okay.
71
what r
1 Let's go further.
So what was 2
would be the contribution to the leak rate if you 3
used the wrong coefficients?
Say one changed by 10 4
degrees, another also by 10 degrees.
You have zero a
5 contribution.
Now 6
MR. HUANG:
Is this your assumption or the real 7
number?
8 DR. REYTELATT:
Of course assumption, because I 9
do not have the data.
The data was fraudulently 10 withheld from me.
For two years despite FOIA's, the 11 data is still not available.
What is the reason?
12 Probably you just want this test to expire.
That's m-13 fine, but the new test, please, do it correctly.
I mean, I don't 14 MR. REYES:
Are you implying 15 understand your statement, so I need to refresh it.
16 DR. REYTBLATT:
What I imply is that at one 17 sensor, the temperature changed by 10 degrees.
18 MR. REYES:
During what time?
19 DR. REYTBEATT:
During the whole test.
20 MR. REYES:
During 24 hours2.777778e-4 days <br />0.00667 hours <br />3.968254e-5 weeks <br />9.132e-6 months <br />, 12 hours1.388889e-4 days <br />0.00333 hours <br />1.984127e-5 weeks <br />4.566e-6 months <br />?
21 DR. REYTBLATT:
Whatever it is.
At another 22 location it also changed by 10 degrees.
23 MR. BERNERO:
In the other direction.
a-24 MR. REYES:
In the other direction.
we 25 DR. REYTBLATT:
In the other direction.
72 t_.
t
1 MR. REYES:
How much distance between the two j
2 sensors?
j 3
DR. REYTBLATT:
Well, diameter of the the 4
diameter of the containment.
All right.
l 5
MR. REYES:
So you're saying that half of the 6
containment lost temperature for 10 degrees and the 7
other half 8
DR. REYTBLATT: That is precisely so, this is 9
what I'm saying, and it may not be precisely 10.
10 Maybe it is 7,
maybe it is 15, I do not know, but 11 these are figures that are 12 MR. REYES:
You believe that to be a real 13 physical phenomenon?
14 DR. REYTBLATT: What do you mean physical 15 phenomenon?
I just believe what you told me and 16 what the report says.
17 MR. REYES:
I would like you for the record to 18 state whether you believe that half of the
- e 19 containment can go up 10 degrees and the other half 20 can go 10 degrees in the opposite direction?
21 DR. REYTBLATT: Yes, I can say it.
I can say it 22 for the record that I believe that for the purpose 23 of obtaining conservative estimates, it is quite 24 possible to assume that one-half of not one-half, as-25 but a point in one-half of th e containment changed
~
73
l=q l
temperature by 10 degrees, and a point in another 2
half of'a containment changed a temperature by 10 3
degrees.
Okay?
4 Now, if you-do use the correct 5
realistic, maybe I should say, weight coefficients 6
you determine 7
MR. HUANG:
I would like you to give us that 8
reference.
9 DR. REYTBLATT:
No, let me finish the sentence.
10 Then I will answer your question..Then we d e t e rmin e 11 that the contribution of just these two sensors will 12 be point 14 percent per. day.
What is this?
This is um w
13 twice as much as the whole allowable Zion.
For I believe it's 14 LaSalle it is about, about I think 15 a little bit less than the total allowable error.
16 Can you allow such things?
17 I want to finish this.
If you use the 18 fraudulent LaSalle program of which I have listing, 19 and if you execute the option, I do not know what 20 kind of sensor did you lock out, Mr. Maura, I do not 21 know.
22 MR. MAURA:
What?
23 DR. REYTBLATT: You locked out.
This is called 24 erasing or locking out in your computer program.
25 MR. MAURA:
That's because I am a naughty fellow.
74
=r 1-DR. REYTBLATT: You locked out one sensor of 2
LaSalle.
You said you didn't like it because it 3
fluctuated.
So let us assume I-do not know which 4
one it was.
5 MR. MAURA:
I said in my presentation that I 6
used the sensor.
7 DR. REYTBLATT:
Yeah, well, I do not know what 8
sensor that was.
9 MR. MAURA:
I said I used it.
10 DR. REYTBLATT:
See, when I wrote this I didn't 11 know what sensor you meant.
So I assumed one of 12 those sensors.
Now, your computer program has a 13 fraudulent option of doubling weight coefficients.
14 What it does, it substitutes the value of 15 t e m p e r,a t u r_e pertinent to one of the sensors, by a 16 value which was picked up from another sensor.
And 17 in fact that simply means that the coefficient is 18 doubled.
19 Actually, I will talk about these 20 fraudulent programs, but let us assume that this was 21 done.
If you 22 MR. REYES:
I would like 23 DR. REYT B LATT :
Excuse me, let me finish.
I do f
24 not know whether it was actually done.
I do not say J
l 25 it was actually done.
But I say it could have been 75
=t 1
done using your weight coe f ficients,. and using your 2
fraudulent program.
Than this would be the result 3
of the undetected leak rate.
4 Now, I did some calculations using 5
fraudulent, or maybe not fraudulent yet, because it 6
was not used, but using who Huang's stabilization.
7 It turns out that it can be manipulated.
I have an 8
example, a numerical example, on Mr. Huang's things,
- 9 how it would manipulate the results.
10 The result would be 2.6 percent per 11 day wouldn't be detected.
That is, what, a hundred 12 times more than the allowable error for Zion.
This J-13 is, what, 4 times more th a n the total allowable leak 14 for LaSalle.
15 I invite your question.
Go ahead.
If 16 are there are no questions, I go further.
17 MR. BERNERO:
I have one question.
The word 18 stabilization appears here and from my notes i
19 DR. REYTBLATT: I will touch on this.
l you previously commented on 20 MR. BERNERO:
l 21 by the way, we will want copies of your slides.
I 22 DR. REYTBLATT:
Oh, yes, I have here.
I l
23 prepared for them.
Okay.
You said, if my notes are s
l 25 correct, you approximated stabilization and cheating.
15
1 DR.
REYTBLATT:
That's right.
2 MR. BERNERO:
In the physical sciences, are you 3
saying that any series of measurements which does 4
not attach confidence to the result or to the mean 5
the average result 6
DR. REYTBLATT: No.
7 MR. BERNERO:
Weighting for, you know, like 8
gravametric analysis, igniting crucibles?
9 DR. REYTBLATT:
No, I will explain precisely, I 10 will explain precisely what I mean by stabilization, 11 although Mr. Huang did not present really what he 12 meant, but fortunately there is a document.
This is n-a-
13 infamous Arndt's draft, which contains Mr. Huang's 14 stabilization, and I'm going to expose this 15 stabil_izat_ ion here and show what it is.
16 MR. BERNERO:
Go ahead.
Carry on.
17 DR. REYTBLATT:
I will do this.
18 MR. MAURA:
The only comment I want to make 19 about that slide is that you have treated one 20 subvolume with four sensorse.
You have placed some 21 arbitrary numbers which are in some cases maybe a 22 factor or two, not even two that I can see.
23 MR. HUANG:
Forty percent.
24 MR. MAURA:
Yes.
And you have ignored the rest 25 of the" containment.
So this, when you come up with e
77
r-1 a number like 2.6 percent or 2.8 or 0.13, that might 2
be just for that subvolume, and it does not 3
represent entire containment.
4 DR. REYTBLATT:
Yes, yes.
5 MR. MAURA:
The number two is I did a worse case.
~
6 I did not a factor of two like you did, but a factor 7
of seven.
8 DR. 'REYTBLATT:
No, no, see you didn't pick up 9
probably the right one.
I will show to you --
10 MR. MAURA:
I think you didn't pick up the right 11 one.
12 DR. REYTBLATT:
I will show to you how to use 9
13 your own problem.
But anyway, let me answer the 14 question.
These figures pertain to the whole 15 containment.
The way.I determined these 16 coefficients, of course, I have a computer program 17 to do this.
It's just mathematics, nothing more.
18 And I can assure you that for the functions that I 4
19 consider, and which is usual, is considered in 20 analysis or in numerical mathematics probably, I 21 have been teaching this course for a number of years, 22 for 30 years actually.
So I can assure you that I 23 am off no more than maybe one decimal there.
Maybe A
(
24 it's not point 07, maybe it is point 09 or point 06, w.
25 but they are pretty good.
78 i
6
q l
What's your question, Mr. Huang?
2 MR. HUANG:
I just have a comment here, because 3
number one, I did not invent stabilization.
4 Regulation requires stabilization.
And I want to
~
5 make a comment on your data T 1 and data T 2.
You 6
know, I'm just engineer, you know, I'm just trying 7
to substantiate 8
DR. REYTBLATT:
That is true, you are not a 9
mathematician, 10 MR. HUANG:
I would not continue my test when I 11 see one sensor goes up 10 degrees and the next 12 sensor drops by 10 d e g r e,e s, and I continue my test.
m 13 I only wanted to add that. comment to your 14 presentation.
15 DR. REYTBLATT: Let me respond to this comment.
16 In your --
17 MR. HUANG:
Because 18 DR. REYTBLATT:
No, weight a second.
In your s
19 proposal, in your draft regulatory guide, you say 20 what you would do.
You say that you would threw out 21 that bad sensor, and we will look at this scenario.
22 This is the infamous position 19, 20 of Arndt's rag 23 guide draft.
This is just a fraud.
I will show.
24 MR. BERNERO:
For the benefit of the audience, w-25 reference is being made to a revised regulatory 4
79
=t, 1
guide that would accompany the regulations on this 2
subject, and the regulatory guides are amplification 3
or description of how one might acceptably satisfy 4
the requirements of the regulations, m
5 MR. GOGEL:
Have they been published?
6 MR. BERNERO:
I'm not sure that they have e
7 DR. REYTBLATT: No, it's not going to be.
8 MR. BERNERO:
It is in the process, and there is 9
a draft reg guide, and the proposed regulation that 10 I referred to earlier, it's been discussed at the 11 ACRS, and other internal procedures of the NRC.
It 12 hasn't been published yet and is expected to be
- m. i.
13 published shortly.
14 DR. REYTBLATT:
No, I'm sorry, it is not because 15 it is not a reg guide.
It is a collection of bad 16 wishes, and that's what it is.
It doesn't have a' 17 format of regulatory guide.
The last that I know
~
18 that it is going to be just buried someplace in the A
19 NRC.
I talked to Professor Zees (phonetic).
20 MR. BERNER :
Dr. Reytblatt, you have 21 corresponded with us on that subject separately, and 22 that's getting a little off side.
23 DR. REYTSLATT:
Okay.
I would like to show now i
Pr l
24 how the results of testing can be fudged.
They
?
25 place. instruments.
The number of instruments and e
80
=g I
calibration is not sufficient.
That was shown in my 2
paper of which a copy I am to attach to the 3
proceeding of this.
There are done arbitrary 4
assumptions.
You know how.they find the weight
~
5 coefficients, don't laugh, you will crack, what they 6
do, they divide the total volume, but use of 7
subvolumes, and anything that gets into this 8
subvolume is assigned the same weight cdefficients.
9 That would be like comparative, like 10 say you look at the map of Europe and to every state, 11 be that Belgium, Luxemburg or Spain or Russia, you 12 would decide the same weight coefficients, anh then
=:
13 they say that it doesn't matter.
This example shows 14 that it does matter.
I will show more examples.
15 Then use of the wrong equation, then 16 stabilizing, equalizing.
I will show blowers, 17 showers.
What they do they irrigate, they power 18 water on their instruments.
They actually did it 4
19 for Zion, 1981, two verification tests in a row 20 didn't come out.
They went and they slept.
After 21 six hours, they came back and they did irrigation.
22 This is in your log of events.
23 HR. MAURA:
Lot me clarify one thing.
During 24 the 1981 test at Zion, I reviewed that and same with s-25 the
'83, and for the record the only people that 81
q l
took showers were the workers.
The containment did 2
not take nor the instruments took any showers during 3
that test.
4 DR. REYTBLATT:
What was it irrigation then?
5 MR. MAURA:
There was no irrigation.
6 DR. REYTBLATT:
What does it say in the report 7
then?
8
-MR.' MAURA:
What report?
9 DR. REYT,BLATT:
The report on Zion 1981.
How 10 could the third test come out?
Why did it came out and the first to fail?
11 12 MR. MAURA:
There is no s u c.h thing as an u-13 irrigation system except for the grass.
14 DR. REYTBLATT:
My question is to you, two 15 verification tests failed on' Zion.
The third came 16 out.
Why?
17 MR. MAURA:
Which?
1980 what?
18 DR. REYTBLATT:
1981.
19 MR. MAURA:
'817 s
20 DR. REYTBLATT:
Yes.
21 MR. MAURA:
There were 22 DR. REYTBLATT:
If ycu cannot answer, we will go 23 further.
This is the point.
All right.
Let's go y.
24 further.
Locking out, dispose of any data you don't
]
25 like.
You throw out pressure gauges.
You throw out
~
82
1 I
l l
which was done actually on LaSalle, as you said, one
(
b l
2 sensor, which was done actually at Zion.
It is not 3
clear how many total was changed, 4,
5 or 10.
Then 4
you can block leaking paths.
Then you do faulty 5
verification.
Then you do simple forgery.
That's 6
what you do.
That's what you do.
7 MR. BERNERO:
I would like to raise a question 8
on this slide.
You speak of dispose of any datt you 9
don't like, locking out, item 6.
In tests using a 10 large number of sensors in any engineering 11 application, are you really saying that locking out 12 or rejecting data is not permissible, or that the 4-13 criteria for deciding which data are flyers or 14 outlyers are not permissible or acceptable?
15 DR. REYTBLATT:
Can I answer this?
16 MR. BERNERO:
Yes, that's the question.
17 DR. REYTBLATT:
I have been in experimental 18 stress analysis for a quarter of a century.
I dealt 19 with very unprecise and with very precise 20 instruments.
Pressure gauges are extremely precise 21 instruments.
Also they are calibrdted.
RTDs also 22 are supposed to be calibrated.
There are 23 calibration certificates.
Y1 24 Now, on Zion 1983 when from 30 or 29, r
25 you will correct me, gauges you go to 25, which 83
j 1
means you dispose of 4 or 5 sensors.
That is in m.
2 percentage, how much, 25 percent or so, maybe a 3
little bit more or less.
Such instrumentation is no 4
good.
The test must be immediately terminated.
You 5
cannot do this.
This is a crime to conduct such a 6
test and then to assign somewhat different values.
7 I'll talk about those computer programs more.
8 MR. BERNERO:
In effect then I interpret you to 9
say that it is the criteria for rejection.
10 DR. REYTBLATT:
Yes, the criteria.
What was the during that 1983 11 criteria for looking out on 19 12 test.
We examined and there are these data m
i-13 someplace.
Let's take a look at'them.
14 Wh'at didn't you like about gauge 15 number 47 What didn't you like about the gauge 16 number 14.
It is somewhere between 63.7 and 66.3.
17 There is nothing wrong, at least to me.
Maybe there 18 was some reason.
I don't know.
19 I assume that simply what happened 20 that you pushed the button, and the number that 21 popped out of the screen was too large, and I assume, 22 I don't know this, but you have an option to do this, 23 then you push another button and your fraudulent 7
24.
computer program does erasing and does substituting 25 of the data that you don't like by another set of 84
1 data that you may like, and then you get another
,-s,
2 number on the screen.
I am not saying that this 3
actually happens.
I do not know that.
But I am 4
saying that there is a real possibility of the case 5
being just this.
6 MR. MAURA:
You cannot go by a strictly one data
~
e 7
set like that one, because it may be reading 65 each 8
time.
You have to look at maybe the last 20.
Maybe 9
DR. REYTBLATT:
I looked.
10 MR. MAURA:
Well, what did you find?
the last 20 maybe, 11 DR. REYTBLATT:
Not at the 12 but anyway, believe me, I observed the behavior of w
13 those gauges.
When I filed my petition, of course I 14 knew what actually you did there.
That computer 15 program, of course, did addition and multiplication.
16 That's obvious th i n g.
I know.
I just wanted you to 17 say in writing what you did and you said it, that 18 you declared them out of service.
Not that they A
19 were, but that you declared, and I think somebcdy
'20 must explain th is.
This is number --
22 DR. REYTBLATT:
Sonebody must explain this.
23 MR. MAURA:
The other comment I would like to P~
24 make is that the whole thing is immaterial.
The
- r-25 test was declared invalid, so what the hell is the 85
1 argument?
r, 2
DR. REYTBLATT:
Well, the argument is we don't 3
want this to occur in the future.
If we expose 4
those things today, my request was not to review at 5
this point, not to review all the past tests, but I 6
will do this, I will file such a petition.
My 4
7 request was at least to prevent it.
I even didn't 8
want to shut down Zion.
Excuse me.
I am against 9
this.
I want you to review your comp' uter programs, 10 to make th em descent.
This is I want honesty.
11 That's the only thing that I want.
12 How they sabotage, data unavailable, m
13 no response, some of my papers are sitting with
~-
14 Mr. Huang.
Of' course, if he cannot respond, he just 15 must admit this.
But they are sitting, and for 16 three years, with examples, with the theory, with 17 all these errors.
It's just a very simple thing.
18 Elementary.
. 4 19 MR. BERNERO:
Dr. Reytblatt, I said earlier and 20 I repeat that you have made many allegations of
~
21 fraud, perfidy, practically every crime one could 22 commit in the context.
We have referred that for 23 investigation.
I don't think that's appropriate 7
24 here.
25 DR. REYTBLATT:
All right, thank you, I will not 86
1 do this.
l
,-y\\
2
- Now, let me do conclusions of general 3
before going into all these particularities.
These 4
are the conclusions.
The'present testing is faulty S
to the degree that any value of leak rate can be 6
obtained.
7 All past tests must be reviewed.
The 8
bad testing must be band immediately.
9 And now I want just to tell you 10 something about the wrong equation.
- Well, first of 11 all, what the official Nureg 3549 says, and Oak 12 Ridge National Lab on this next question.
m
_L.
13, Page 49, and I quote:
"Extran 14 formulation - " Extran formulation is just a 15 conventiona1 physics gas law-nothing more of 16 the air mass equation is technically correct, while 17 the ANSI /ANS equation is not..."
18 Now, I want to show on page 48 there 19 is a table, A sub 5.
It has spacial variation just 20 5,- nine degrees.
Per cent difference in' average 21 temperature using the wrong equation is point 2 22 percent.
If the test duration were 12 hours1.388889e-4 days <br />0.00333 hours <br />1.984127e-5 weeks <br />4.566e-6 months <br />, at 23 some point the temperatura so to speak stabilized or FF
.2 24 equalized, that means there was no air, then the 25 error in the leak rate would be point 2 percent for 87
=q l
12 hours1.388889e-4 days <br />0.00333 hours <br />1.984127e-5 weeks <br />4.566e-6 months <br />, or point 4 percent per day.
Five times 2
more than Zion allowable.
3 I am not saying that this actually 4
happened at Cooper.
By the way, I dug out.
That 5
was not an assumed plan. That was Cooper weight
~
6 coefficients.
These coefficients are again in 7
excess of
.l.
I insist that you look at the values 8
of the weight coefficients at Cooper.
9 MR. HUANG:
May I make a comment on this?
10 DR. REYTBLATT:
Sure.
11 DR. HUANG:
I just wanted to reiterate that the 12 author who had made a comment during the ACRS T
13 meeting of this number apparently misinterpreted his 14 number.
When he said point 2 percent, he mean the 15 difference not in arithmatic difference, but a 16 difference of two number is 2.2 percent of the right 17 number.
18 DR. REYTBLATT:
Yes.
19 MR. HUANG:
So the difference extremely small.
20 It's not, if you assume now a leak rate of
.1 21 percent and the calculated number is
.3 percent, the 22 arithmatic difference is
.2 percent.
I follow that.
23 But in this case the calculated number is
.1
- percent, P
24 the other number is.1002, so that the difference is J-25
.2 percent.
6 88
l DR.
REYTBLATT:
Let's take a look at this table.
q 2
I am saying that this is just ignorance of yours.
3 Look at this table.
If anyone has a calculator, we 4
can verify this, and by doing this we will prove 5
what your expertise level is.
I have this I have 6
this publication.
This is just very good reason to 7
see.
If anyone has a calculator, we can look at it.
8 I have Oak Ridge report, no problems.
9 okay, let's do it.
Page 48.
We 10 probably can do it even in mind.
All right.
Extran 11 temperature term, 299.000 ANS temperature term, 12 299.6; percent difference in temperature,
.20380.
m 13 okay?
14 Now, suppose you had equalizing, 15 stabilizing, and the temperatures are equal toward 16 the end of the test or vice versa in the beginningi 17 then delta T contribution that enters to the leak 18 rate is precisely
.2 percent.
This is what it is 19 and do these cales.
20 MR. HUANG: 'This is where your --
21 DR. REYTBLATT: And then you say that you are an 22 engineer.
23 MR. HUANG:
This is where you miss the point.
24 You're using two different temperature equations s
25 which result in slightly different in temperature 89
1 calculations.
m 2
DR. REYTBLATT:
That is not slightly different.
3 MR. HUANG:
Then you attribute to 4
DR. REYTBLATT: This is what you don't understand.
5 When you understand it, then you will understand it.
6 Read my report.
7 MR. BERNERO:
Let him finish what he's saying, 8
at least 9
MR. HUANG:
We have demonstrated that to you in 10 our previous presentation.,
11 DR. REYTBLATT:
You did not demonstrate.
I have 12 your slide.
I show to you.
Give me your slide.
A 13 This is a very good reason.
This is a fraud.
Look 14 at this.
How many gauges was there?
25?
He has 15 only 5 nu e.b e r s.
What he used averages, but what is 16 the use of the average of the subvolume?
It is 17 against the conventional physics, my friend, to use 18 any averages, you have to use the so-called 19 thermonic average.
Your ignorance is awful.
I 20 think you better shut up.
21 MR. BERNERO:
Dr. Reytblatt, I wish you would 22 confine your remarks to the technical substance.
23 You make a habit of impuning people's character, w
24 their motives, and their actions, and I think it is a
25 wholly inappropriate.
9 90
1 DR. REYTBLATT:
All right, I apologize.
2 MR. BERNERO:
And certainly not scientific.
3 DR. REYTBLATT:
Yes, I agree with you.
4 MR. BERNERO:
As you purport to be.
5 DR. REYTBLATT:
But you must agree that this is 6
appropriate.
7 MR. BERNERO:
No, I resent that.
I resent the 8
constant impuning of people's characters.
Do you do 9
with the thousands of students that you have taught?
10 DR. REYTBLATT:
No, I am a very peaceful person.
'll MR. BERNERO:-
Do you say they are frauds?
12 DR. REYTDLATT:
Of course not.
13 MR. BERNERO:
Well, thank goodness for that.
14 MR. MAURA:
Well, here is a case where you have 15 roughly 2/10ths of a degree.
T, hat's what we are 16 saying.
17 DR. REYTBLATT:
A c c e.m t e d, accepted.
18 MR. MAURA:
Okay.
What is the difference in the 19 leakage?
20 DR. REYTBLATT:
Accepted, accepted.
21 MR. MAURA:
What is it?
22 MR. BERNERO:
What is it?
23 DR. REYTBLATT:
I will answer what happens that c=-
4 24 in some instances there is no error, it is always so.
un 25 In some instances there is no error, in some 91
1 instances there is an error.
I will show why and I 2
will show when it is acceptable, when it is not 3
acceptable.
4 MR. MAURA:
This is the real life.
You bring 5
things from the clear blue sky.
Tell me why.
6 DR. REYTBLATT:
I'm sorry.
This is Cooper.
7 This is real life.
This is Cooper.
It's Cooper.
8 This is real life.
9 MR. BERNERO:
Dr. Reytblatt, carry on, carry on.
10 DR. REYTBLATT:
Real life.
11 MR. BERNERO:
This is Excedrin headache number 12 four.
13 DR. REYTBLATT:
- Now, let me now just s'h o w what 14 actually you did on this wrong equation.
Of course, 15 mathematica1 equations cannot be wrong or correct.
16 If you write this equation S equals P over T, it's 17 neither correct nor wrong.
You can then manipulate 18 an equation in the wrong way, but modeling equations 19 in physics, economics, biology can be wrong.
You 20 cannot write this.
Whero S is best velocity and 21 time.
The ideal gas equation is correct only where 22 P is a uniform pressure.
T is a uniform temperature.
23 What you do is writing the same equation, and
~
24 assuming that these are totally different things.
s 25 This is a totally unscientific thing.
92
l I like : tow to show the effect of the
,q 2
wrong equation.
This number ORNL is picked up from 3
their actual simulations.
I also,did a lot.of 4
simulations.
I submitted my computer programs, 5
examples.
They show irrefutable evidence.
If it is 6
five degrees, then it is not a conservative estimate.
7 It is just the figure that very o[ ten happens.
It 8
is.02 percent.
Please notice that the allowable 9
error for Zion is
.02.
That was 10 MR. MAURA:
What?
11 MR. BERNERO:
Please clarify your units,
.02 12 percent of the calculation?
m 13 DR. REYTBLATT:
Point 2 percent leak per day.
s 14 MR. BERNERO:
Percent per day?
15 DR. REYTBLATT:
And this is precisely error in 16 the leak rate that they obtained -- that Jim Duggin 17 obtained.
18 MR. REYES:
Can you explain the five degrees?
19 You are talking about the different calculations 20 between the two equations?
21 DR. REYTBLATT:
No.
What I'm talking -- let me 22 explain what I'm talking about.
23 MR. REYES:
Yes.
- -r,
24 DR. REYTBLATT:
That Jim Duggin assumed spacial
,s 25 variation of five degrees. and time variation of 93
r 1
about half of this, and then they did random changes 2
in temperature, and registered the results, recorded 3
the r e's u l t.
This is so-called Monte Carlo.
He 4
didn't know how to do Monte Carlo in fact, but at 5
least he tried, and there were numbers point 00000.
6 Then there were numbers
.02, which means that it may 7
happen.
It doesn't have to happen, but it may.
8 MR. REYES:
And that's the worse case then?
9 DR. REYTBLATT:
Yes, it is considered the worse 10 case.
So what does this teach us?
That usually 11 what it says.
When you have a spacial variation of 12 less than five degrees, of course you can use the c
13 wrong equation.
When it is more than 20, watch it 14 up.
You may be all right or.you may not.
Nobody 15 knows.
Nobody knows.
16 By the way, it is not a proof to bring 17 a concrete calculation from one certain test and say 18 if it did come out, than all tests will come out.
19 The correct method in nuclear energy, let it be 20 known to you, is the method of conservative 21 estimates.
What we do is assume the wrong the 22 worst assumptions.
And then we obtain the balance.
23 And this was dono in my report which has been n-24 sitting on your desk, and it shows that absolutely s
25 you cannot use this equation, because you may get m
94
eq l
the bad results.
2 I appreciate your position, Mr. Maura, 3
that you want to finish this discussion, and just 4
return to the correct equation that all the 5
societies use.
6 MR. HUANG:
I still want to reiterate a comment, 7
because the author is not here, because the author 8
told us when at the meeting this error represented 9
percentage of error.
In other words, you can see 10 the error is not even a percent of the right number.
11 And Dr. Reytblatt interpreted it to be,the 12 difference in the leak rate calculation.
I want to 13 reiterate this comment for the benefit of the people.
14 DR. REYTBLATT:
Please divide
.6 by 200.
It's
.02.
15 MR., HUANG:
I only reiterated the comment made l
16 to you j
17 DR. REYTBLATT:
So you are referring to somebody l
18 saying.
Okay.
If you do not claim it's yourself i'
19 but blame it on somebody else, I accept that 20 somebody else said it.
21 MR. BERNERO:
Yes.
Let the report stand on its 22 own.
What it says it says.
23 DR. REYTBLATT:
All right.
Let's go further.
I r-24 want now to say something about the present j
e 25 verification.
This is a sham.
This is precisely i
95
I what.it is.
They say that we do some calculations r,,
2 and determine a number which is called leak rate 3
from the test, this number.
Then we superimpose 4
certain leak and calculate certain numbers, and we 5
call it superimposed, test plus imposed, we say, and 6
then when you subtract, then you should obtain the 7
imposed leak rate.
8 It is true that if you calculate 9
correctly the leak rates, then this equation will be 10 satisfied.
However, if both leak rate test plus 11 superimposed and leak rate test are wrong, this 12 equation also can be satisfied.
This is a simple
.. ~.,
i 13 exqmple.
14 Suppose that in fact 12ak rate was say 15 1 percent,,
but because of your bad faulty 16 methodology, you determine it is
.l.
Fine.
Then 17 you superimpose
.1, and in fact the leak rate will 18 be 1.1, but since you do not know how to determine 19 leak rate, you determine that it is
.2.
You 20 subtract and you obtain
.l.
21 This is the same thing as if you bring 22 to the bank say S1,000.00, and give it to a cashier 23 to count and he says, no, it's only 500.
No, I gave
?~
24 you.
Well, let's check it.
I am adding S1,000.00.
25 Let's check it again.
So he calculates, but because 96
1 he is a bad cashier, he finds that this sum is not cy 2
1500, but 20 so he subtracts and he says, see, you 3
just submitted 500 dollars.
4 You cannot determine two numbers from
~
5 one equation.
This is a fraudulent, absolutely 6
meaningless thing.
It is, also let it be known, e
7 that verification tests as done today are of 45 one 8
hour' duration some times.
This is what is specified.
9 There is no final bleeding specified in the 10 standards or anything else.
Maybe some plants do 11 this.
Maybe not.
But this is not true, and that 45 12 minutes is not sufficient, has been re' cognized by 9
A.
13 the NRC itself.
14 MR. BERNERO:
Dr. Reytblatt, you seem to suggest 15 that the superimposed leakage 16 DR. REYTBLATT Yes.
17 MR. BERNERO:
-- which is a measured quantity, a 18 flow rate measurement 19 DR. REYTBLATT:
Yes.
is not known by that flow rate 20 MR. BERNERO:
21 measurement, but must be determined?
d 22 DR. REYTBLATT:
No.
23 MR. BERNERO:
By the -- you know, you treat it E
24 as an unknown.
o 25 DR. REYTBLATT:
Yes, all right.
This thing is 4
97 T
T
?
1 unknown, leak rate test.
You want to prove that 7q 2
this is correct.
3
'MR.
BERNERO:
You're looking to verify that.
4 DR. REYTBLATT:
Yes, you want to verify that.
5 MR. BERNERO:
By superimposing a known increment.
6 DR. REYTBLATT:
Yes, so this thing is known..-
e 7
MR. BERNERO:
That's a flow rate measurement, 8
you would have confidence in that?
9 DR. REYTBLATT:
Right, right.
Okay.
So this is 10 not known and this thing is not known, because it is' 11 not known, because how do they determine it.
They 12 use the same wrong weight coefficients.'
They use m
~.
13 the same rigged ventilation.
So in other words, the 14 uncertainty in this number is the same as 15 uncertainty in this number.
What you simply measure 16 that the difference between these two numbers.
17 That's all.
Also you measure the degree of your. _
18 ignorance.
That's another thing.
But basically ~
19 only these two things.
20 MR. BERNERO:
Excuse me, sir 21 MR. NISSIM-SABAT:
Perhaps what you're_trying to 21 say, Dr. Reytblatt, is assume that the error is a 23 multiplicative one and this equation would ' check for I
24 multiplicative error, but if the error is an 25 additive one, this equation will not the 98
1 calculation will not check an additive error.
2 DR. REYTBLATT:
Sure, of course, of course.
3 That's a trivial thing.
By the way, I have a 4
document that was sent to NRC many years ago saying 5
that what is it.
Why do you use this thing.
Then I 6
have a letter f rom Arndt to the Oak Ridge National 7
Lab saying that you know what, some people say that 8
this is a sham, that this is just excuse me, I
9 like to use this word, B.S.
10 MR. BERNERO:
That's mild by comparison.
11 DR. REYTBLATT:
Okay, thank you.
And the answer 12 was, well, there is some merits.
There are some 13 merits, but it is merely just a measure.
It just
- m 14 tells you something about the instruments, which I 15
' agree.
I believe that it can be done good.
But the 16 way it is done today, it's not correct, it is not 17 telling you that your assumptions are wrong.
18 By the way, this is what 3549 says on 19 verification.
That verification is the measure of
~
20 how your assumptions on instrumentation, on weight 21
' coefficients, on everything are good.
It should be 22 a standard practice after the verification test 23 failed, to check what's wrong with your weight N
24 coefficients.
This is what you should have done, J
25 Fred.
You should have after the first test, failing e
99
1 verification test in 1981, just to look at those
-s, 2
coefficients and say, yeah, this is B.S.
Let me go 3
further.
I believe that you are an honest man and, 4
you know, do whatever you.believe is good.
5 I like now to say a few words about 6
the stabilization discussion.
of course, this is e
7 worse than B.S.
What it is today, what Mr. Huang 8
didn't say, but today what it is in fact is that the 9
present standards uses the integral 'of T over the 10 volume.
To say something about the integral over 1 11 over T.
And unfortun'ately, the integral eq'uation 12 involved is the integral equation of the whole of ml 13 the first kind, and its solution is not unique.
So 14 the attempt is ridiculous.
15 And in my e.xamples, I have examples of 16 which he is aware, I assume perfect stabilization 17 and still there is a leak rate 10 times, or maybe 18 about 10 times higher than allowable.
But I still 19 like there is an improvement.
It seems to be an 20-improvement now.
This is stabilization a la Huang, 21 and I use the draft, so to speak draft, of j
22 regulatory guide, so-called positions 19 and 20.
23 These positions call for verification 24 of inequality asterisk.
And what it requires, that a
25 the average temperature must satisfy asterisk, and
\\
100
~
' - - ~
2-.
1 that every individual sensor should satisfy asterisk,
-v
- 1 2
then conditions are satisfied.
Where T sub 0,
T sub
):
3 1,
and T sub 2,
are three consequent values at one 4
hour interval.
In other words, the second 5
derivative sort of should be limited.
All right.
6 No questions about this.
e 7
If I didn't interpret Mr. Huang, he 8
will probably speak up about this.
Let's take a 9
look at this example.
I have 30 sensors, but just 10 for simplicity I have 15 and 15.
I have for all of 11 them temperature increase in excess of the second 12 derivative being less than
.5.
So there is no Huang 13 stabilization.
You have to throw out this data.
A.
14 But this is B.S.,
because in fact the leak rate is 15 zero.
What Mr. Huang did, he qualified a good 16 containment as bad.
17 Comments?
No comments.
Okay.
Let me 18 show you -- we'll have time to do these examples.
19 MR. BERNERO:
You're saying that the temperature 20 is increasing with 21 DR. REYTBLATT:
Yes, it's increasing and so what.
22 This is ridiculous.
The mass involves both 23 temperature and pressure.
So temperature increases, 7
24 pressure increases, what's wrong?
.u-25 MR. BERNERO:
Yeah, but that doesn't say that 101 I
1 the containment is bad.
It would say the test might f s.
2 be bad.
3 DR. REYTBLATT:
That's what he says.
What he 4
will say, that s.i n c e there is no stabilization, then 5
you should run another test, but it is ridiculous, 6
because in fact the containment is good.
But it is 7
not detrimental.
It's not detrimental.
8 MR. BERNERO:
If you had the infinite number of 9
sensors and the precision to measure the absolute 10 leakage and mass conservation, I agree, you might be 11 able to have a good test.
But I don't understand 12 your point that if the test method refuses to accept m
~_
13 as qualification an unstable 14 DR. REYTBLATT:
I'm sorry, I didn't make it 15 clear.
P1 ease take differences between the 16 consequent values, all right, and then divide them 17 by 500, okay, and then compare to Huang's condition 18 about the second derivative.
You will see that the 19 second derivative at rny time instance, at any time a
20 instance, exceeds slightly exceeds, I just concocted 21 this example, exceeds his limit.
That means that 22 there is no Huang stabilization.
That means that 23 you have to rerun this test.
P. "
24 What does this mean, that Huang said s-25 the leak rate was zero?
He said you have to rerun.
I 102
4 1
In 'i( ;t the containment was good.
c.
2 M3
.ERNERO:
I would expect Commonwealth Edison 3
to c omp la i'n about that but 4
DR. REYTBLATT:
But I will show you another
~
5 example.
6 MR. HUANG:
Wait a minute.
Since my name was 7
mentioned a couple of times.
If you go back to my 8
slide, the stabilization was defined to be spacially 9
stable not temporally stable.
In other words, the 10 change in temperature you show here does not mean 11 whether you have a stabilizing containment or not.
12 Second comment I would like to make, m
13 in your example here you show a 15-degree increase 14 in your example again.
This by itself if anyone is 15 going to conduct a test, they.wouldn't like that, 16 because some heat transfer now goes'into a
17 containment, makes a 50-degree change 18 DR. REYTBLATT:
Okay.
I understand your 19 comments.
20 MR. HUANG:
The leak rate calculation just like 21 you show in your previous slides, the change of 22 pressure and the temperature.
23 DR. REYTBLATT Your calculations are ridiculous.
- =-t 24 MR. HUANG
I wanted just to correct you.
I 25 will not drop th is test.
The leak rate calculation
._3+
103
1 will drop the test.
2 DR. REYTBLATT:
That's you, but on LaSalle they 3
did.
4 MR. HUANG:
Don't put words into my mouth.
I
~
5 didn't say that.
6 DR. REYTBLATT:
All right.
Now, about spacial,
. D 7
for all 30 sensors they are the same.
There is no 8
spacial variation at all.
You do not understand it.
9 When you don't understand, don't ask questions.
10 The second question about 18 degrees 11 is again quite obvious.
LaSalle had it about 18 12 degrees.
It had 15.
You want it a little bit
=m 13 different numbers, you can have 15 not 18.
The 14 essence is the same.
15 Your stabilination is a sham.
You 16 didn't define it in your presentation.
You said a 17 few words, and these words are just words of an 18 ignoramous, excuse me.
But I want to say now what 19 your method does to --
20 MR. REYES:
Just a moment.
I need to address 21 one statement made by Dr. Reytblatt.
The data given 22 to him in the January meeting for the Unit 1 of 23 LaSalle 1982 Containment Leak Rate Testing does not 24 show 15 degree mean temperature increase.
So he's sr 25 incorrectly quoting the data submitted to him.
q l
DR. REYTBLATT:
No, I didn't say that mean 2
temperature increase was 15 degrees.
What I said, 3
that there was a sensor that showed 15 degrees 4
change.
Don't tell me, don't tell us things that I 5
told.
I told something.
And I know what I told.
6 MR. REYES:
I would like the record since I have p
7 the data in front of me, which I gave him, to show 8
again that he's incorrtet that 9
DR. REYTBLATT: Is this 10 DR. REYTBLATT:
subvolume number 6, subvolume 11 number 6, no sensor, not a sensor?
12 DR. REYTBLATT:
Oh, subvolume.
7 13 MR. REYES:
No.
You like to interpret your data 14 the way you like.
15 DR. REYTBLATT:
That means even more.
Okay.
He 16 said okay.
That is marvelous.
Thank you, Mr.
17 Rise.
- 18 (Laughter.)
~
- 19 Mr. Reyes said there was a subvolume, 20 which has a temperature change of 14.3 degrees, 21 which means that there were sensors in this 22 subvolume that had temperature change more than 14.3, 1
23 because otherwise you couldn't have had this average, w
24 right?
So th is simply reinforces my point.
Thank 2
25 you very much, Mr. Reyes.
105
._ _ _.___=
how z
1 And now let me slow what 2
Mr. Huang's stabilization will allow cheating.
This 3
is just marvelous.
Let me show the previous test.
4 It's just ridiculous.
Where was it?
5 Oh, okay.
What he recommends, that it 6
i r. written and I have a copy of it, that bad s
if 7
readings must be thrown out.
That you have 8
you have a sensor that does not satisfy one asterisk, 9
you just can throw out this reading, but he says you 10 have to throw it out from the very beginning.
Don't 11 do this in the middle.
All right.
Let's see what 12 will happen.
=
13 Let us assume this distribution of
~
14 sensors.
We can see that the temperature average 15 temperature is always 500 degrees.
That means that 16 always the first condition, Huang's condition, is 17 maintained.
Then you can see that sensors from 6 to 18 15 and from 16 to 20, they are bad.
They do not 19 follow his conditions, so they can be thrown out.
20 He specifies that you cannot throw out all the a
21 sensors.
Well, obviously that would be too much.
22 He said that you can leave as many as the law allows, 23 and the law allows to have 10 sensors.
So I left 10 24 sensors.
J-25 If you calculate leak rates, than leak 106
1 77 1
rate a la Huang is zero.
However, the real leak 2
rate is
.27. percent per day, which means Huang's 3
stabilization fraudulently qualified a bad 4
containment as good.
5 The leak will be, what, four times 6
larger than Zion, actual leak, and the error 10
'A 7
times larger than is allowed.
This is why they want 8
stabilization.
This is the oniy reason why they 9
want stabilization, and the diurnal effects, and let 10 me now go to the actual test.
11 MR. BERNERO:
Let me interject here.
Please, as 12 the chairman of the meeting, I must express'my own n
13 mortality and frailty.
I would like to carry on, 14 give you full audience here, but I.do want to set a 15 time limit.
It is approaching 5:00 o' clock, and 16 it's now about 10 minutes to 5:00.
What would you 17 foresee as the uninterrupted period of your further 18 presentation?
19 DR. REYTBLATT:
Well, I would say if you give me 20 15 minutes, I would probably be done with it.
I
?
21 just can do it very fast.
22 MR. BERNERO:
Well, I'm not trying to unduly 23 rush you, but at the same time I would like to draw w
24 a finite limit.
So let's carry on in that vein, and 25 I'll welcome it if commonwealth Edison is quite e
107
ng 1
limited in their remarks later.
2 DR. REYTBLATT:
All right.
Now, this is the 3
reason why the NRC needs stabilization.
There are 4
many, many tests.
They are all similar.
~
5 Interestingly enough about this type of curve, 6
Dr. Hill was talking two years ago at the conference.
7 This year nobody showed up but myself.
All other 8
papers were withdrawn.
Probably because people felt 9
they cannot argue their point.
Anyway, this is why 10 the NRC needs stabilization.
11 Look at this straight line.
This is 12 the ideal mass decay curve.
If you measure very m
13 well, if you have good sensors, many of them, if 14 it's expensive test, I don't know, you will obtain 15 this curve.
If you have a sloppy job, then instead 1C you determine the other curve.
17 Now, this curve is very convenient, 18 because you can discard the initial part of the 19 curve, say that this is stabiliza, tion.
We don't 20 want to do this.
Then you can neglect this 21 so-called diurnal effect.
22 What Palo Verde did, and it is an outrage 23 please put down, because I want to pay attention to e
24 Palo Verde test, they did an unofficial test.
They a
25 obtained this curve.
They entered where the -- with 108
7-1 the request to NRC to grant them a short duration 2
test.
That was a pre-operational test.
And they 3
had it granted.
They had granted a test of nine 4
hours.
I want this test to be tested.
Then you 5
determine something, and then you discard the rest 6
of the curve.
This is a diurnal effect.
You don't S
7 war *. it.
Then you draw this straight line.
You say 8
this is measured leak rate.
Is it?
No.
This 9
angle, large angle, is the true leak rate.
This is 10 the only reason why those people who cheat insist on 11 stabilization.
12 I attended conferences, and at those 13 conferences people would say, oh, gee, I just cannot
~.
14 stabilize, I cannot stabilire.
Why do you need 15 stabilizat, ion?
If you have everything correct you 16 don't need it, and there were people who really didn't need it, because it is 17 don't need 18 actually may be used as a fraud, and I think it has 19 been.
20 By the way, there is a notion that you 21 have to stabilize because of* gas absorption or 22 absorption or gas mass or gas dissolving in liquid.
23 What they do, they do a plateau thing.
Do you know 24 what plateau is?
Plateau is they raise the pressure r
25 over the testing pressure, hold the containment 109
=t 1
until all the air possible leak, increase in leak, 2
which may be smart, I don't know, dissolves, and 3
then they lower slightly the pressure.
This is what 4
they call plateau.
However, it is not helping very 5
much, interestingly enough, that the amount of air 6
that is degassed is very small.
So many, many smart 7
people, they just know that it's not going to do the 8
trick much, and they don't do this.
But some people 9
still do this.
10 Since I've almost exhausted my time, I 11 probably will t e rmi nat e, but I would like to read 12 aloud the draft of the resolution that I want this
-r 13 meeting tc formally accept and adopt.
+
14 MR. BERNERO:
Please do.
15 DR. REYTBLATT:
Okay.
The resolution:
16 The NRC staff meeting with 17 Z.
Reytblatt to discuss the emergency relief 18 decision DD-85-10 of July 3,
1985, Decision, has 19 unanimously determined that the decision:
20 One, neither contains calculations, it 21 does not contain, derivatioAs and/or proofs, there 22 is nothing of a sort in this decision, nor indicates 23 references in lieu of those.
There is no references.
24 Two, does not address some of the 25 petitioner's important statements.
For example, the 110
l claim that the Zion weighting coefficients are not q
2 correct is not addressed.
The claim that the 3
director's decision of March 6,
1984 is based on 4
false statements is not addressed.
That the same 5
decisions uses bad numerical examples, the same type 6
of examples as Mr. Huang used.
In other words, 7
instead of 25 something data, they used only 5 or 8
something.
9 Contains deceptive statements, for 10 example, and this is in your yard, Mr. Bernero, for 11 example, the' decision contains a deceptive 12 conclusion that the Volumetrics computer program 13 does not contain a fraudulent option on the basis 14 that particular test calculations were correct.
15 obviously.,
16 The fact that a certain computer 17 output is correct does not attest that the computer 18 program is correct.
Only by examining all the 19 options, you can state that this is correct.
I 20 request that Zion fraudulent computer program 9
21 listing be submitted for examination of the public.
22 I have requested this 10 months ago.
I received an 23 FOIA answers that NRC does not possess this listing.
7 24 The paragraph 9.4 obliges you, v/.
25 Mr. Bernero to obtain this listing, and to submit it e
111
7C 1
to Mr. Felton, so that the public could see what 2
kind of erasing they had.
If it's similar to 3
LaSalle and B y r'o n, fraudulent programs or not.
4 The meeting unanimously decided that 5
the director's decision must be amended in view of 6
the 'fo11owiag facts:
C 7
One, the method of air mass 8
calculations as practiced presently may or may not, 9
of course, induce errors exceeding the allowable 10 errors.
Recommendation:
Immediate return to the 11 conventional physics equation.
12 Two, some of the Zion 1 weight w
13 coefficients are i n c o r.r e c t.
Out of six verification 14 tests conducted at Zion in three years, three failed, 15 two were successful, as a part of fraudulent tests, 16 irrigation and doubling of weight coefficients, and 17 one was performed in view of analyzing the wrong 18 test assumptions.
And recommendation:
Find the 19 conservative estimate of the Zion 1 leak rate using 20 conservative assumptions on the weight coefficients.
21 I think that you did it.
You did a very good job, 22 it looks like, for LaSalle test, and old Zion test.
23 I will, if you want me to tell you, I did it.
I 24 obtained the leak in excess of the allowable, and I 25 will show you how to use your own program, what data e
112
7q l
you should erase, you know, and the program will do 2
the trick for you, all right, okay.
Let's do it.
3 okay.
Then the third, 3,
the method 4
of obtaining the weight coefficients as practiced 5
presently may or may not, may induce errors 6
exceeding the allowable errors.
Recommendation:
7 Immediately terminate application and/or development 8
of new weighting coefficients using unscientific 9
methods.
Only scientific methods must be used.
10 Four, the present verification tests 11 may not detect that the test assumptions are wrong.
12 Recommendation:
Immediately terminate all the s
a 13 verification tests until a reliable technique is 14 introduced.
15 For example, what Mr. Huang said may 16 serve as a basis of it.
There was good rationale in 17 that.
Unfortunately, it is not practiced, at least 18 widely.
19 Five, the Commonwealth Edison computer 20 program contains a fraudulent option which is called a
21 erasing the data.
I just have it.
I can show it to 22 you guys.
And doubling the weight coefficients, so 23 that those coefficients may exceed the reasonable
=
1 24 limit.
Similar computer programs, Volumetrics as an 25 example, are used all over the United States.
I 113
I do not have the listing, but I believe y
1 believe 2
I have some information on Volumetrics, Mr. Bernero, 3
and my suspicion is that it has the same erasing 4
doubling option.
I request hereby that the listing 5
of the suspected, it's not fraudulent until I look 6
at,i t, of the suspected of Volumetrics program be 7
submitted for examination.
8 Recommendation:
Immediately suspend 9
the use of suspected software, debug and revalidate 10 it.
Make available the listing and validation acts 11 for all the software used for the leak rate testing.
12 Six, the director's decision of March i
13 16, 1984, on,LaSalle contains falso statement.
14 Recommendation:
Make this decision null and void.
15 In addition to Dr. Reytblatt's 16 original request and appeals, the meeting has 17 considered the stabilization notion.
The meeting 18 unanimously decided that neither the ANS version nor 19 Mr. Huang's version have any scientific merit, but 20 may as the present examples clearly indicate, either 21 be harmless or mistakenly qualify a good containment 22 as a bad containment, or mistakenly qualify a 23 leaking containment as a good one.
24 The meeting similarly established that er 25 no scientific merits have been found in such means i
114
=..
1 of affecting the environment as irrigating, 2
ventilating, equalizing; whereas, on numerous 3
occasions these means were used for cheating.
4 Recommendation:
Immediately t e rmin a t.e equalizing, 5
stabilizing, irrigating, ventilating.
6 The meeting established that the 7
diurnal effect and short duration tests as the Palo 8
Verde Fall 1982 tests clearly indicate can be used 9
or may be used probably for cheating.
10 Recommendation:
Immediately rovoke all the 11 exemptions for the short duration tests.
No new 12 exemptions shall be granted.
The meeting supports mm J~
13 NRC attempts to terminate the reduced pressure tests.
14 Recommendation:
Immediately terminate the reduced 15 pressure testing.
16 The meeting urges the NRC to comply 17 with 10 CFR, and to prepare the full and open to the 18 public response of the present to the criticism 19 of the present faulty testing practices and the 20 faulty ANSI /ANS standard, which was submitted to the 21 NRC by Mr. Reytblatt in his report and at the 22 meeting.
That's it.
23 MR. BERNERO:
Two things, Mr. Reytblatt.
You F.[
24 read from a prepared document.
Is there a copy of v
25 it with the slides?
115
q l
DR. REYTBLATT:
Yes, I will submit.
I have a 2
prepared copy for you.
3 MR. BERNERO:
The other thing, and I welcome 4
correction or amplification from my own legal 5
counsel here, I believe that if you study the 6
Freedom of Information Act, it requires that we give 7
you within certain constraints material that the NRC 8
has.
9 DR. REYTBLATT: Yes.
10 MR. BERNERO:
However, if the NRC does not have 11 the material, and I believe you suggested in your 12 words that if we don't have it, we need to obtain it 13 DR. REYTBLATT:
That's right.
14 MR. BERNERO:
in order to satisfy the Freedom 15 of Information Act?
16 DR. REYTBLATT:
Precisely, 9.4.
17 MR. BERNERO:
And I don't think that's correct.
18 DR. REYTBLATT:
It is.
Read it.
19 MR. HOEFLING:
The Freedom of Information Act 20 only captures documents in the agency's possession 21 or agency records.
Obviously, if the agency does 22 not have the record, it is not an agency record.
23 DR. REYTBLATT:
Also raad 9.4.
If you don't
[
24 have it, I will give it to you right now.
J 25 MR. HOEFLING:
I appreciate your legal view.
f 116
=q l
MR. BERNERO:
Well, we will look into that.
2 May I consider that the concitsion of 3
your proposal here?
4 DR. REYTBLATT: Unless somebody objects this 5
decision, because I want it to be unanimous.
If 6
somebody does not object, that obviously means that 7
one accepts it.
8 MR. BERNERO:
Dr. Reytblatt, I can interpret 9
that you unanimously support that resolution.
I can 10 assure you that I don't unanimously support it.
I 11 will study it, as I will study the entire record o f' 12 this meeting, and we'll correspond with you, either m
13 I or Mr. Denton will correspond with you in 14 consideration, after we have had a chance to review some of the stuff you've showed here is 15 what 16 familiar to me from previous meetings.
Other 17 material you showed here is new to me, at least.
I 18 don't know if it might be new to others.
But I 19 think a deliberate consideration of it is in order, 20 and I do not intend and do not expect adoption of O
21 the resolution here today.
So let me take it as 22 your recommendation and put it into the record that 23 way.
24 And now what I would like to do is, I
s-25 would like to offer commonwealth Edison a chance to 117
1 speak, but 2
DR. REYTBLATT: Very well.
I would welcome it if your 3
MR. BERNERO:
4 remarks were quite brief, because of th e late hour, 5
and especially our two reporters here I think must 6
be feeling a bit of the strain from the pace of the 7
meeting.
8 Yes, Mr. Gogel?
9 MR. GOGEL:
After that, I would like to return 10 if we may for just a minute or two to the question 11 of the fundamental equation, because I think that is 12 a question on which we could achieve some resolution.
13 MR. BERNERO:
Okay.
Well, Mr. Farrar from 14 Commonwealth Edison.
That's F-a-r-r-a-r.
15 MR. FARRAR:
My name is Denny Farrar from 16 Commonwealth Edison.
For the record, obviously I 17 should first point out that commonwealth Edison does 18 also not support that resolution.
We, however, do 19 take Dr. Reytblatt allegations very seriously.
You 20 can see by the contention of people that I brought 21 with me today to listen to his presentation.
We 22 also take very seriously our obligation to maintain 23 containment integrity at all of our stations.
'Er 24 We have reviewed our procedures, and
%J 25 that review has reaffirmed our confidence that the 118
1 integrated leak rate tests have been conducted in 2
accordance with the applicable regulations, and that 3
the conditions under which the tests were conducted 4
have insured that the test results represent true 5
containment conditions, and that concludes my 6
statement.
s 7
MR. BERNERO:
Okay.
Thank you, Mr. Farrar.
8 And now, Mr. Gogel, is it?
9 MR. GOGEL:
Yeah.
10 MR. BERNERO:
Would you like to raise your 11 question, and I'll 12 MR. GOGEL:
May I actually use the blackboard m
13 for a moment?
~
14 MR. BERNERO:
Sure thing.
15 MR. GOGEL:
The mass inside the containment at 16 any particular time is equal to the integral of mass 17 at all the different little points.
Right.
- Now, 18 since we don't have an infinite number, we can't 19 determine every little one, we approximate.
We 20 divide the thing up into subvolumes.
So we have the 9
21 summation over J subvolumes, the mass in each 22 subvolume.
23 This in the appropriate thing to plug 24 the ideal gas law into.
What every one of you here y
25 today from the NRC has said is that you are plugging 119
you're calculating the mean 7-1 the ideal gas law 2
temperature for the entire containment at any given 3
point in time, and that is simply as I learned in 4
the last year of high school, that's simply invalid.
5 You have to use this equation.
6 MR. HUANG:
We start from there, yes.
7 MR. GOGEL:
So why don't you just come out and 8
say it is incorrect to calculate a mean temperature 9
for the entire containment; is that correct?
10 MR. HUANG:
If you had continued to write that 11 equation to a final form 12 MR. GOGEL:
All right, let's do it.
e 13 MR. HUANG:
Just continue, since you know that 14 now, please.
You can derive in terms of a 15 temperature, summation of temperature.
Go ahead, 16 start with perfect gas law.
That would be?
17 MR. GOGEL:
Pressure times volume.
18 MR. HUANG:
Use delta volume.
19 MR. GOGEL:
I'm not I can't understand that 20 word you are saying.
O 21 MR. HUANG:
All right.
You can start off on a 22 summation of delta mass.
23 MR. BERNERO:
Delta volume.
24 MR.. HUANG:
And then go back to the perfect gas a
25 equation.
120
in 1
MR. NISSIM-SABAT:
Mr. Huang, what you do, you 2
assume the pressure to be constant in the volume in 3
your calculation, and there is no reason to assume 4
that any more when you have reason to assume the 5
temperature is going to be constant.
6 MR. HUANG:
At a particular moment we have a 4
7 state of pressure and temperature.
At a particular 8
moment.
9 MR. BERNERO:
N'o, wait a moment. I think 10 pressure is constant.
11 MR. HUANG:
Constant.
12 MR. BERNERO:
Yes.
This is a pneumatically free L
13 volume.
The difficulty is that there are potential 14 temperature distributions within it, but it is 15 pneumatically free in communication.
All delta 16 volumes communicate with others.
There are no 17 compartments in the pneumatically --
18 MR. NISSIM-SABAT:
I agree with you, but the 19 problem is that you don't know the actual pressure.
20 You're measuring it.
You're measuring point by 0
21 point.
That's why you 1. ave more than one pressure 22 gauge.
You may have six.
You may have four.
You 23 may have eight.
24 MR. HUANG:
You have a different question.
25 MR. MAURA:
The only reason there is more than 121
=q 1
one pressure gauge, because if that one fails, there 2
is a back up.
3 DR. REYTBLATT:
From the inception at least two.
4 MR. NISSIM-SABAT:
My point about that is since o
5 you do not know the actual pressure, you only know 6
several pressures, each pressure-then will be 7
' applicable to appropriate subvolume, and you do -
8 MR. PAPERIELLO:
It isn't, though.
9 MR. MAURA:
That's not true.
see, you're 10 MR. REYES:
In one point in time 11 asuming pressure changes in time.
You assume the 12 pressure change we're talking about n
13 MR. HUANG:
Well, I think his comment is I
14 think I understand now.
He tried to profile the 15 pressure d,istribution now.
This may be 16 approximation.
We assume.
You can argue about it.
17 In this entire approach we assume pressure constant 18 at a particular moment throughout a containment.
19 That's our assumption.
You may take issue with that, 20 but that's our assumption.
21 MR. NISSIM-SABAT:
I do not take issue with that.
22 What I take issue with is since you do -- even it 23 the pressure is indeed constant, you don't know what 7
24 it is.
You don't know what that pressure is.
J 25 MR. HUANG:
We measure it.
We have a sensor.
s 4
122
nq l
MR. NISSIM-SABAT:
Well, to meaure it you have 2
to weight each measurement 3
MR. HUANG:
We have one pressure sensor.
4 MR. BERNERO:
Now, wait a minute.
5 MR. REYES:
Pressure is not weighted.
6 MR. BERNERO:
Pressure is not weighted.
If s
7 pressure is measured, there is a reference pressure 8
measurement and back up measurements, but there is 9
not a separate pressure imputed to each separate 10 subvolume.
11 MR. NISSIM-SABAT:
Well, then you have no valid 12 way -- the point about that is, if you have five r
13 pressure measur.ments, and you take the average of 14 those five, that may or may not be valid and that 15 MR. REYES:
It may not be accurate.
16 MR. NISSIM-SABAT:
It may not be accurate, 17 exactly, that's my point.
So what I'm trying to say 18 is that it's encumbent upon you to use the best 19 possible data.
If you have five pressure data, 20 there is no reason to lump them all five into one.
O 21 You would have to look at th e mass in each subvolume, 22 using the pressure for each subvolume, and the 23 temperature for each subvolume, and then find the 24 total mass and do this as time varies.
s 25 MR. REYES:
But if you are willing to' live with 123
=q 1
some accuracy by using an instrument, you already 2
have one.
Then 3
MR. MAURA:
This is what you have, the pressure 4
5 MR. NISSIM-SABAT:
The whole pressure gauge.
6 You told me you have more than one pressure gauge.
7 Where are the other four?
8 MR. MAURA:
There are no other four pressuro 9
If there were, they are all right here.
10 DR. REYTBLATT:
No, n't necessarily.
11 MR. MAURA:
What do you mean "no"?
12 DR. REYTBLATT:
May I say a few words?
Say D.
- C 13 C.
Cook, okay, D.
C.
Cook 14 MR. MAURA:
That's a different case.
15 DR. REYTBLATT:
Precisely.
It is a different 16 case.
That's what I want to say.
17 MR. MAURA:
It has three separate, but that's a 18 specific design.
19 DR. REYTDLATT Precisely, a specific design.
20 Let me just shed a light on this problem.
The o
21 pressure of course is increasing because there is a 22 head pressure.
Now, in the leak rate calculation, 23 all the change in pressure enters.
So if there in W
l 24 no air flow in the containment, then delta P will be 25 the same, and these, my friends, would be perfectly 124
1 right using only one pressure gauge.
Only one is 2
needed.
3 However, in fact as' studies show, 4
there are air flows, and the estimates of those C
5 shows that not only they are constant in time, but 6
they are changing in time.
So there will be 7
different delta P all over the containment, and 8
that's why I propose to use at least two pressuro 9
gauges just to guarantee that such a thing did not 10 acquire enormous proportion, because estimates show 11 that it may screw the results.
And Huang knows 12 about this.
You also know about this.
I told you.
13 Still, nothing has been done.
14 MR. BERNERO:
Excuse me, you're suggesting that 15 these convective air flows 7 16 DR. REYTBLATT:
Be checked out.
17 MR. BERNERO:
But you seem to be suggesting that 18 they can indeed generate major pressure differences.
19 DR. REYTBLATT:
Yes, yes, yes.
Look at my 20 report.
It is in my report.
The estimate is given.
21 I'll tell you more, that I consulted a prominent 22 authority in the field, Professor Levaughn 23 (phonetic) from IIT, and you can check with him.
24 That was about three years ago.
That those flow J
25 that those velocity speeds that were in my report, 125
=-
1 they may induce the error that is not acceptable.
2 I am not raising this question today.
3 MR. BERNERO:
Well, I thank you for that.
4 MR. GOGEL:
Is it correct or incorrect to o
5 calculate a mean temperature for the containment and 6
then plug that into the ideal gas law to come up 7
with the mass of the containment at any particular 8
time?
9 MR. HUANG:
All the way out to that summation 10 with that's where it is started, okay.
I can 11 show you on the blackboard.
I can bring you this 12 literature.
We start a whole derivations from here.
13 We start exactly --
14 MR. BERNERO:
That's one of your references, 15 isn't it?
16 MR. HUANG:
Okay.
I will give you this 17 reference.
18 MR. GOGEL:
If that's the case, then why do the 19 people from the NRC continue to reassert that 20 calculating a mean temperature for the containment 21 is valid.
It's not.
22 DR. REYTBLATT:
Of course, it isn't.
23 MR. GOGEL:
If you must start calculating, what 24 you should be calculating is the mass in citing j
25 subvolume, and you're treating it as if it wsro 126
=r 1
completely uniform inside the containment.
2 HR. SHAPAKER:
As the equation is derived, the 3
formulation for T is correct.
It's working to the 4
derivation of the equation.
5 MR. GOGEL:
I heard you folks saying mean 6
temperature.
Two ways of calculating mean o
7 temperature for the containment.
I see a head 8
shaking.
Will you admit that it is~ incorrect to 9
calculate the mean temperature for the containment?
10 MR. BERNERO:
No, I don't.
11 MR. SHAPAKER:
No.
12 MR. REYES:
Could I interrupt?
You're arguing 13 the accuracy.
That's wrong.
You're using correct 14 and accurate.
There is a difference.
15 M R..
GOGEL:
I'm saying correct according to the 16 laws of physics.
Remember, things change inside 17 that containment.
18 MR. HUANG:
Okay.
I will continue your 19 derivation here.
Let's see if I can do it quickly.
20 Up to here we have no argument.
That's the way it 21 is supposed to be.
But if you plug now this delta J, 22 I think you only need aJ here, if you want to be 23 consistent.
That's total.
Let's use total.
Forget 24 about a time J
25 DR. REYTBLATT:
This is bull shit.
~
127
- D 1
MR. HUANG:
All right.
Plus you plug this J 2
into that equation, and then eventually, let me see 3
how many: steps I have to take.
Let's plug that in.
equal to P,
right, continue 4
So now summation of 5
last step.
Just plug this in there.
6 MR. GOGEL:
Okay.
Yeah.
7 MR. HUANG:
Now, since at a particular moment we 8
do make this assumption, pressure is constant across 9
the containment, so we can take P and R,
they are 10 botn constant now.
So we can get the summation of 11 delta Wj and Tj.
12 MR. GOGEL:
Right, and there is no way you can 3.
13 get from there to your equation A.
14 MR. HUANG:
All right.
Let's step back a moment 15 now.
,I ca,n also say, hey, my total mass here equal 16 to the P,
I measure at that particular moment, and i
17 the total volune I got, and then I assign aT 18 temperature mean.
All right.
These two sides are 19 the same.
20 DR.,REYTBLATT:
He is writing the equation which 21 is not the conventional equation.
He writes his own 22 equations.
23 MR. HUANG:
You start off on here, right, you 24 start off on here, the Wt total is equal to the i
J.
th is one is right, this one is 25 integral of the 128
.a
. =,
1 wrong.
2 MR. GOGEL:
That's right.
3 DR. REYTBLATT:
These are two different 4
equations, come on.
5 MR. HUANG:
How about the 6
MR. NISSIM-SABAT:
The second one is a 7
definition of T volume, and then you equate one to 8
the other, and of course you get a T volume.
9 MR. REYTBLATT:
Yeah, but this is what we call 10 cheating in science.
11 MR. HUANG:
So we can literally for convenience 12 define a mean temperature to be all right, if ycu I.
13 cannot even understand this, I will stop right here.
14 MR. NISSIM-SABAT:
Mr. Huang, my question about 15 this has to do with not whether your pressure is 16 uniform or not.
And it is one thing about having a 17 law of physics which you believe.
It is another 18 thing when you have a set of measurements which 19 fluctuate about above and below around this 20 so-called physical law.
21 Now the point I want to make, since 22 you said the measurements fluctuates, if you have 23 two pressure gauges, unless you're going to assume 24 that the pressure gauges each see exactly half a 25 volume and, therefore, you take the arithmatic mean 129
it makes no r-1 of the two pressures, now that is 2
more sense to assume that than to take the 3
arithmatic means of a temperature.
4 MR. HUANG:
I think I understand your comment 5
now.
I did not want to respond to Dr. Reytblatt 6
because he did not think I understand, but I think I
~
7 understand you.
If you do have more than two 8
pressure sensors, they are different.
To be'more 9
scientificly correct, you should consider those two 10 numbers in your final pressure calculation.
I'm not 11 arguing about that.
But nobody in the common 12 practice will rely on one pressure sensor.
We use 13 one pressure sensor
-~
14 MR. REYES:
Let's go one step further.
If you 15 calculate both ways, and the results are within an 16 accepted accuracy, then if you use a basis of the 17 particular accuracy for the results is acceptable, 18 than either way would be acceptable.
Do you follow 19 what I -- we may disagree on our assumptions, but I 20 just want to explain to you why we view the use of 21 the pressure sensor versus the way that you think it 22 should be used, which is, if you calculate it both 23 ways, okay, using the average or mean, and the other 24 way, and the results come very close, the accuracy 25 is within what we will accept to be reasonable; do 130
r 1
you understand?
2 MR. NISSIM-SABAT:
Yeah, I understand what you 3
are saying.
Let me tell you why I wouldn't be 4
satisfied.
Because for one thing, if you have a 5
given vessel that says Zion 1,
you would want to 6
have a very accurate measure of the leak rate to see 3
7 if it changes from one year to the next, and 8
although both leak rates may be acceptable, a 9
worsening of the leak rate is something you should 10 watch out.
Therefore, I think it behooves you to 11 use your data in the most accurate possible way, not 12 just say it's okay, it's not okay, but see how the v
13 data allow you to give, perhaps predict what may 14 come down the line.
15 M R_. BERNERO:
I would just like to interject 16 here.
You're going into an issue that is at the 17 heart of much of the debate we're talking about.
18 Should we totally revise our method of containment a
19 integrity control.
And that is this, that when you 20 do an integrated leak rate test, you set up a 21 condition where systems are isolated.
You're 22 simulating by arrangement of systems drainage 23 filling, you know, where the fluid levels are,
~-[,
24 you're trying to simulate that hypothetical accident 25 condition, and you're setting up a condition that is 131 l
mr 1
not left alone while the plant operates, because in 2
order to operate the plant, you have steam in the 3
steam lines, and you have flows and so forth.
4 So that from one integrated leak rate 5
test to the next, to measure the leak rate at one 6
and to compare it even if you had infinite precision
~
7 to the other leak rate test is not really useful to 8
us, because I'm measuring an artificial accident 9
condition, set up in 1982 to the same condition set 10 up in 1984.
And the differences would undoubtedly 11 be found in the tightness of individual valves that 12 had been used in the two intervening years.
So I'm 13 not sure that that wou,1d be useful to us.
14 What we are more concerned, I said 15 earlier o u,r long range considerations are toward 16 assuring containment integrity for the fact of 17 closure, for substantial closure rather than subtle 18 differences in leak rate, because when we look at
~
19 actual public health risk, the subtle differences in 20 leak rate really don't affect very much even when 21 you exceed the regulations.
22 MR. REYES:
And there is another part of the 23 regulation that requires you that all the 24 penetrations to containment are tested individually
- s..
25 at a different frequency.
That way you keep up with L
132
1 the integration on the valves and packages and 1
2 things like that.
3 DR. REYTBLATT:
These measuroments are not, 4
accurate as much, and you can block these leaking 5
paths during the integrated leak rate test, and then 6
test them separately, which is I would compare to 7
testing an aircraft which failed in flight in the 5
manner that you test the tail separately and then 9
say, okay, ready, and go on fly.
Would you fly that 10 aircraft?
11 MR. BERNERO:
Ladies and gentlemen, I'm 12 exhausted.
I'm happy to be back in my native city, 13 but I'm tired.
~..
14 Now, I would like to ask, is there any 15 other member of the public here who would like to 16 make a comment or a statement?
If not, I would like 17 to declare this meeting adjourned, and Dr. Reytblatt, 18 as I said, after proper deliberation and conference, 19 we will be corresponding with you on the subject.
20 MR. GOGEL:
Mr. Bernero, I would like to get a 21 yes or no answer to the question, is one of the 22 forms you wish to call this mean temperature.
Is 23 one of the forms which is in use, there are two V
24 forms which are in use 4.
25 DR. REYTBLATT:
One is incorrect.
133
=q 1
MR. GOGEL:
One can be derived from that 2
equation.
Can others be derived from that equation?
3 MR. BERNERO:
To p u't it in simple terms, one is 4
theoretically correct.
The other is not.
They both 5
give equivalent results.
6 DR. REYTBLATT:
May or may not.
~
7 MR. BERNERO:
We say they do.
8 MR. GOGEL:
The other one is not theoretically 9
correct.
They cannot be derived from that equation.
10 MR. BERNERO:
It is a simplification.
It is not 11 directly derivable.
12 MR. PAPERIELLO:
I think I suspect, and I have 9
13 to p' lay again to the ar thme tic tonight, I suspect
'nT, it is a series of expansion...
i 14 for small changes 15 (Discussion off the record.)
16 MR. BERNERO:
Are you a member of the public, 17 sir?
18 DR. REYTBLATT:
Yes, sir.
I am Tim Wright.
I
~
19 am an attorney at Business and Professional People 20 for the Public Interest.
What I wanted to find out 21 was whether that response that you will make to the 22 Professor will be made available to the public or at 23 least the participants in this meeting?
- D-24 MR. BERNERO:
Oh, yes, certainly, certainly.
If 25 we have your name, you know, your name and address, 134
- . =-
.=
1 I will be happy to send it to you, if you have a 1
2 card.
3 MR. HOEFLING:
It will be made publicly 4
available, and it will be in the public document 5
room and if you want a personal copy, if you give us 1
6 your name and address 7
MR. BERNERO:
Yes, if we have your name and 8
address, we'll send you a copy.
9 Thank you very much.
10 MR. GOGEL:
Mr. Bernero, in the computer rate, 11 why don't you use the one that can be directly 12 derived 13 MR. BERNERO:
In fact I think the reg guide 14 actually says that.
15 MR. GOGEL:
The new one?
16 MR. BERNERO:
Yes.
17 MR. GOGEL:
Well, why don't you have that 18 promulgated immediately?
19 MR. BERNERO:
We're trying.
We're trying to get 20 the regulatory guide, but our argument again is, you 21 know, do we get equivalent results.
So that's it, 22 folks.
The meeting is adjourned.
23 (Whereupon, this meeting was 24 adjourned at 5:45 o' clock p.m.
s 25 this date.)
e 135
1 CERTIFICATE OF OFFICIAL REPORTER J
2 3
4 This is to certify that the attached 5
proceedings before the UNITED STATES NUCLEAR 6
REGULATORY COMMISSION in the matter of:
7 8
NAME OF PROCEEDING:
Public Meeting with 9
Dr. Zenovy Reytblatt 10 11 PLACE:
Glen Ellyn, Illinois 12 d
13 DATE:
' Thursday, October 17, 14
- 1985, 15 16 was held as herein appears,- and that this is the 17 original transcript thereof for the file of the 18 United States Nuclear Regulatory Commission.
19 20 21 22 410 M a rla @/. C u s a no, CSR, 23
. Official Reporter Sullivan Reporting Co.
.py 24 Chicago, Illinois 25 3
,,y
5 /4 +/92zh r/c NRC PRESENTATION o
INTRODUCTION /
SUMMARY
- J. SHAPAKER ALLEGATIO.N REVIEW AND CONCLUSIONS TO BE DEMONSTRATED TECHNICAL DISCUSSION OF ILRT METHODOLOGY - J. HUANG
~
o OVERVIEW 0F LEAK-RATE COMPUTATIONAL METHOD IMPORTANCE OF STABILIZATION 4
EFFECT OF TEMP, FORMULATION EFFECT OF WEIGHT COEFFICIENT SELECTION SIGNIFICANCE OF VERIFICATION TEST PARAMETRIC ANALYSIS OF ILRT AND DISCUSSION OF INSPECTION 4
o PRACTICES - F. MAURA j
o
SUMMARY
- J. SHAPAKER CEC 0 o
CECO PERSPECTIVE OPEN DISCUSSION WITH PETITIONER
INTRODUCTION o
PURPOSE OF MEETING ACCOMMODATE PETITIONER'S REQUEST FOR TEf.HNICAL DISCUSSION OF DIRECTOR'S DECISION UNDER 10 CFR 2.206 (DD-85-10) 0F JULY 3, 1985 PETITIONER'S REQUEST PROMPTED BY DECISION (DD-85-10) TO DENY Tile o
PETITION (LETTERS DATED MARCH 6, 1985, AND MARCH 8, 1985) SEEKING IMMEDIATE ACTIONS WITH RESPECT TO CONTAINMENT LEAK RATE TESTING.
THE BASIS FOR THE DECISION IS THAT THE ALLEGATIONS MADE ABOUT o
I ZION 1 CONTAINMENT INTEGRATED LEAK RATE TESTING DO NOT SUPPORT i
THE CONCERNS UPON WillCil THE PETITION STATEMENTS ARE BASED.
t S
l
-a MARCH 6, 1985 PETITION o
SOUGitT IMMEDIATE POSTPONEMENT OF ALL CONTAINMENT ILRT'S o
EXPRESSED CONCERN OVER THE DETERMINATION AND CONFIRMATION OF WEIGHT C0EFFICIENTS USED IN CALCULATING THE CONTAINMENT ATMOSPHERE WEIGHTED AVERAGE ABSOLUTE DRYBULB TEMPERATURE.
j I
I i
1 4
l 1
i 4
.m MARCH 8, 1985 PETITION SOUGHT IMMEDIATE BAN ON USE OF THE (VOLUMETRICS OR CECO) COMPUTER O
PROGRAM (I.E., SOFTWARE) TO PROCESS RAW TEST DATA, UNTIL IT IS DEBUGGED AND REVALIDATED 1
PRIMARILY CONCERNED WITH COMPUTATIONAL VALIDITY OF SOFTWARE T0 0
d AVERAGE TEMPERATURE DATA 1
e e
l l
l
'7 6 @ 94 e
- ^
9 ALLEGATIONS AB0UT ZION 1 TEST WHICH FORM BASIS FOR PETITIONER'S O
CONCERNS 1)
ZION 1 SOFTWARE DOES NOT PROPERLY COMPUTE SUBV0LUME TEMPERATURES 2)
ZION 1 SOFTWARE ALLOWS FOR THE MANIPULATION OF WEIGHT COEFFICIENTS TO OBTAIN ACCEPTABLE RESULTS 3)
THE EQUATION USED TO CALCULATE CONTAINMENT AIR M SS IS WRONG 4)
CURRENT PROCEDURE FOR WEIGHT COEFFICIENT DETERMINATION /
VERIFICATION MAY LEAD TO UNDERESTIMATING ABNORMALLY HIGH LEAK RATES TO THE DEGREE THAT LEAK RATES APPEAR ACCEPTABLE e
e 9
o o
- 1 ALLEGATION 1 ZION 1 SOFTWARE DOES NOT PROPERLY COMPUTE SUBV0LUME TEMPERATURES l
DISPOSITION DATA SET FROM ZION ILRT REFERENCED BY PETITIONER WAS EXAMINED FOR EVIDENCE OF SOFTWARE INABILITY TO ADD AND DIVIDE CORRECTLY 3
J STAFF CONCLUDED, BASED ON RESULTS OF_ HAND CALCULATION, THAT SOFTWARE i
DOES, IN FACT, PROPERLY COMPUTE SUBV0LUME TEMPERATURES f
J 3
2
ALLEGATION 1:
ZION 1 SOFTWARE DOES NOT PROPERLY COMPUTE SUBV0LUME TEMPERATURES DATA SET 302 0F ZION 1 DECEMBER 1983 TEST CITED IN PETITIONER'S EXAMPLE
- SUBV0LUME 2 M
SENSOR READING (*F)
] y' 2
64.71 T
=
3yg 3
65.12 4
005 [O8 FOR N = 7, 8
64.0Y
/gyq,z =
d3.34 f
9 63.71 WHICH IS THE SAME 13 61.99 SUBV0LUME TEMPCRATURE hb
{
14 005 REPORTED ON Tile DATA 18 60.23 SHEET i
19 64.63 443.40-i
- DATA SET SHOWN IN PETITIONER'S EXAMPLE IS NOT N0. 302; NEVERTHELESS, j
SOFTWARE ACCEPTABILITY CAN BE SHOWN USING ANY DATA SET
CECO IION STATION-INTEGRATEB LEAKRATE TEST UNIT 1
,.12/05/83
. RAW AND PRIMARY DATA QATASET4302 11ME8 52.3136111 HRS.{Dlk
.d *
- RAW DATA ***~
Y:
,j p ggy/(,E PRESSUREDETECT0ES Pl= 41 22 P2= 40.97 [ P. MANUAL = 0 TEMPERATURE CHANNELS 1 TO 39 UNITS =MV.
(ST )->S=SUBVOLUME e T= TYPE ('0-00Se !-WB e,2-DB) 1 (12)u 64.69
--*2 (22)= 44 71
--*3 (22)= 45.12 - = 4 (20)= 45 21 5 (32)= 65 41
~
6 (32)= 64.49 7 (12)= 62 37 --* 8 ( 22) = 4 4. 01 + 9 (22)= 43.71 10 (32)= 65.33 11 (32)= 64.05 12 (12)= 62.98 -*13 (22)= 61.99 -*14.(20)=. 44 48 15 (30)= 44.91 14.(32)= 45.67 17 (12)= 63.75 +18 (22)= 60 23 --* 19 ' (22)= 43 63 20 (32)= 43.45 21 132)= 65.420000122 (42)= 43.36
.23 (42)= 67.19 24 (52)= 44 37 25 (52)= 46.89 26 (52)= 42.47 27 (52)= 66.22 -
28 (52)= 45.26' 29.(50)= 44.22 30 (11)= 56.51 31 (10)a 8.96 32 (20)=
.73 33 (30)= 48 14
,,34 (21T='56.86 35 (30)=
.28 36 (20)= -3.54 37 (30)= 53.3 38 (41)= 57.52 39 (51)= 55.3
- PRIMARY UALUES ***
~
BUBYOLUME'
' ' AVG, SUBV_.. VAPOR PRESSU,NE. '. y,-- ;_.;-I_',,'.1-h_[-.:.: _~/J'
. Z.. 1. *' " ' -
~
-~
a.
.u g
,,2g,,,,; -
.^-- g g p-
.. y, _y. -fog
. z) 2
.229010384 3
.229010384 gg ggggg & p,gp/
4
.234523708
$ (// 8 6 6hl& OA 6 ~
5
.216428626 g:Gyy. pp ggRV/C6 (bO) tviRAGE VAPOR PRESSURE =.227856306 ff/giggy/ CE (2,2) e BUBVOLUME AVG. SUBVOLUME TEMP (DEG F.)
1 63.4475
-9=- 2 63.3428572 <
3 64.86.
4
,65.275 ~
.. '; - # ' [-
'J.
~#
5 5
65 442 r-bVERAGE CONTAINMENT TEMPI 44'."2099943 (DEG F.)
kVERAGECONTAINMENTDEWPOINT=56.716 (DEO F.)
{VG.CORRECTEDPRESS.=
0RRECTED PRESSURES: Pl= 41.4260475 PSIA P2= 41 4240475 PSIA 41.4260475 PSIA OMT. DRY AIR PRESSURE = 41 1981912 PSIA i
$ONT.
DRY AIR M A S S==
5763 LBS.
g
____________________________________54.039 lHANNELS LOCKED OUTt 4:14 15:29:31:32:33,35,34e37eP1(Pl=P2)
\\
}RY ALARM =.5 jppp4ppgge $
ET ALARMS 2 g
rRESS ALARM =.1 l
IRESSURE CALIBRATION CONSTANTS 1 01489 B1=
.1411 M2= 1.01489 32=
.1611, W
1 e
l
=
ALLEGATION 2 ZION 1 SOFTWARE ALLOWS FOR THE MANIPULATION OF WEIGHT COEFFICIENTS TO t
OBTAIN ACCEPTABLE RESULTS j
DISPOSITION 1
STAFF CONCLUDED THAT LEAK TESTING PRACTICES IN THE INDUSTRY, COUPLED WITH NRC REGULATIONS AND INSPECTION PRACTICES, PROVIDE ADEQUATE ASSURANCE THAT MEANINGFUL ILRT'S ARE BEING CONDUCTED A SUBSEQUENT PRESENTATION BY F. MAURA WILL CilARACTERIZE THE INVOLVEMENT OF NRC INSPECTORS IN MONITORING ILRT'S AND IN PERFORMING INDEPENDENT CALCULATIONS TO ASSURE A MEANINGFUL TEST j
1 e
a
l ALLEGATION 3 THE EQUATION USED TO CALCULATE CONTAINMENT AIR NASS.IS WRONG DISPOSITION MASS EQUATION USED IN THE AMERICAN NATIONAL STANDARD, " CONTAINMENT I
SYSTEM LEAKAGE TESTING REQUIREMENTS," ANSI /ANS 56.8-1981, IS NOT
" WRONG" MASS EQUATION:
W
= 144V g - Pyr.j 9
R Tg i
THE FORMULATION FOR CONTAINMENT TEMPERATURE (T ) THAT SHOULD BE USED IS AT ISSUE ANSI /ANS 56.8-1981 DOES NOT PRESCRIBE A FORMULATION FOR T s
l
j o
ALLEGATION 3 (CONT'D.)
TWO T FORMULATIONS ARE CURRENTLY IN USE
/
7~)~ =
n 2
a.
C 6
H,f i
- i STAFF AGREES THAT FIRST FORMULATION IS MATHEMATICALLY MORE CORRECT STAFF CONCLUDED THAT WITHIN THE RANGE OF TEMPERATURE PATTERNS EXPERIENCED DURING ILRT'S, THE DIFFERENCE IN LEAK RATE USING Tile SECOND FORMULATION HAS NG,ETY SIGNIFICANCE, SUBSEQUENT PRESENTATIONS BY J. HUANG AND F. MAURA WILL SHOW Tile IMPACT OF THE T FORMULATION ON THE CALCULATED LEAK RATE
u f
ALLEGATION 4 CURRENT PROCEDURE F.0R WEIGHT C0 EFFICIENT DETERMINATION / VERIFICATION MAY LEAD TO UNDERESTIMATING ABNORMALLY HIGH LEAK RATES TO THE DEGRE THAT LEAK RATES APPEAR ACCEPTABLE i
l DISPOSITION l
CURRENT INDUSTRY GUIDANCE ON THE DETERMINATION / VERIFICATION OF WEIGH 3
COEFFICIENTS IS ADEQUATE i
4 THE NEED FOR A PRECISE KNOWLEDGE OF THE WEIGHT COEFFICIENTS IS R
i DIMINISHED BY THE ACKNOWLEDGED IMPORTANCE OF STABILIZATION OF C0dTAINMENT ATMOSPHERE CONDITIONS THROUGH0UT A TEST'TO ASSURE THE t
j VALIDITY OF THE MASS EQUATION
)
SUBSEQUENT PRESENTATIONS BY J. HUANG AND F. MAURA WILL:
l 0
PROVIDE AN OVERVIEW 0F THE DATA ANALYSIS TECHNIQUE AND THE l
REQUISITE TEST CONDITIONS, DISCUSS THE SIGNIFICANCE OF THE VERIFICATION TEST o
l PRESENT THE RESULTS OF A PARAMETlilC ANALYSIS (USING ACTUAL TEST 0
DATA) TO SHOW THE IMPACT OF THF ' Fir,HT roccetercuT McTunnu
.c IMPACT OF T FORMULATION SONGS 1 ILRT OF JUNE 1985 METHOD OF FORM 0F CALCULATED ANALYSIS EQUATION LEAK RATE FOR T
(% PER DAY)
MASS POINT wT 0.05169 (24 HR) 1 0.05101
& /T TOTAL TIME cv T 0.05065 (24 HR) 1 0.04994 w /T O
/d- -A 3,
- l
}
l l
11 1
)
ILRT METHODOLOGY 5
1,.
\\!
li I!
I o
PROCEDURES FOR COMPUTING CONTAINMENT LEAKAGE RATE 3
I o
MAJOR ASSUMPTIONS AND PREREQUISITE REQUIREMENTS
]
o VERIFICATION TEST
'f I
2 t.
EQUATIONS FOR COMPUTING CONTAINMENT LEAKAGE RATE f
jP / c m N I
~
t i
0 EQUATION OF STATE (PERFECT GAS LAW) i P/= RT N a. M h
BY MEASURING P AND T v = RT I
P-1 AND i
V"U i
l ASSUMING V CONSTANT, THEN AT ANY TIME tt l
j
""N4 l
FOR ANY Ah, THE LEAKAGE RATE IS DEFINED
,g i
AH M = 1_ * %
at i
I
~
t w
W.
i!
lij j
o BECAUSE OF INHERENT DATA SCATTER, A LINEAR LEAST SQUARES ANALYSIS IS USED TO ESTIMATE THE LEAKAGE RATE (MASS POINT TECHNIQUE)
Ot=At+B t
I AND THE LEAKAGE RATE IS COMPUTED BY L,,,(% PER DAY) =-21:00 A I
1 WHERE, A n SLOPE OF LEAST SQUARES LINE l
B.m INTERCEPT OF LEAST SOUARES LINE 1
,I o
AN UPPER CONFIDENCE LIMIT (UCL) IS SET SUCH THAT i
THERE IS ONLY A 5% CHANCE THAT THE ACTUAL CONTAINMENT LEAKAGE RATE EXCEEDS THE UCL VALUE l
i e
g 3:
! 'l I
a
.i!
I MAJOR ASSUMPTIONS AND PREREQUISITE REQUIREMENTS FOR CILRT j,-
1i o
STABILIZATION:
SPATIAL TEMPERATURE GRADIENT DURING
!~
TEST IS ACCEPTABLE I
t i.!..
PROVIDED THE SPATIAL TEMPERATURE _ GRADIENT IS RELATIVELY
.l
!l
- }l STABLE, I.E.
GRADIENT CHANGES DCCUR SLOWLY -
j
..ii MORE SPECIFICALLY, THE TREND OF INDIVIDUAL d
TEMPERATURE SENSORS MUST FOLLOW THE GENERAL TREND OF THE
'i CONTAINMENT TEMPERATURE l
l 1
k i
I u
i s'.-
ti i
i i
nwwrue 1
7 l
i j
o OM.Y OE SET OF EIGHT COEFFICIENTS MAY E APPLIED TO TE TEST DATA i
i o
TE DATA E E TAkEN AT EGULAR INTERVALS NO MANIPULATION OR SELECTION OF DATA IS ALL0ED l
'l o
SifFICIENT DATA MET E TAEN FOR A VALID STATISTICAL j
ANAYLSISs LEAK RATE CALQ1ATIONS AE NOT SIWLE i
Ii T)D-POINT CALCULATIONS i
'i l
0 TE ACCEPTABILITY OF TE TEST ESILT MJST BE VERIFIED BY DECKING TE EASURING CAPABILITY AGAINST A kNOW l
M ME l
l
\\
I VERIFICATION TEST l
- l f
TE PURPOSE OF VERIFICATION 1EST IS TO VERIFY TE ENTIRE CILRT TEST PROCEDIE, INCLl0ING SELECTION OF INSTRWENTATION, LOCATION 0F INSTRWENTATION, ACCURACY OF INSTMENTATION, CONTAIWENT EIGHT COEFFICIENT EI)0RK, CONTAINGT ATMDSPERE STABLIZATION, COPPUTATION OF TE LEAK RATE, ETC.
i 3
e 4
e D
T.
~~
- /
1
~::;
/
ii
~ -
THE WAY TO C0f00CT TIE VERIFICATION TEST IS TO BLEED M QNTAlltelT AT A l(NM LEAK RATE (EASUREl) Fim A ii I
SEPARATE FLOW EASURING DEVICE) Af0 THIS IIPOSED LEAK I!
RATE ALONG WITH TE WASURED LEAK RATE WILL BE VERIFIED
.g g.u h
BY llE ILRT C0lHITAT10N PROCEDURES V;/
y '.
f a
a 1
~
, s.tilu e
I
[c
)l i
M
!i lo
!l lI.
L c.- Lam)
L ;.
4 /.25Ly p
l-
,f a-
{l l
i j!
!j VERIFICATIONTESTPROCEDURE
!j t
BLEED TlE CONTAIWENT AT A kNOWN LEAK RATE,USING AN AC-a CURATE FLOW EASURING DEVICE. NASUREENT OF TE IWOSED
'l 4
LEAK RATE (L ) AM) TIE C0WOSITE LEAK RATE (L h 1.E., IM-I c
FOSED PLUS ACTUAL CONTAIRENT LEAK RATE, MJST SAT!SFY TIE 1
!1 FOLLOWING RELATIONSHIP:
1 i
(Le - L W - L 6 0.25 L a 1
- k
,97/./ /,.
p,,- ifah i.
'/ / / / "/.
/
t
///iv,,ggs j
t
/
4 1
it
- ^
}t
\\
a M
s 6
l
=
t i
.e E
s s-5 le
=
s
=
g E E Edl mg A
r ;
!(lg EElf,"
8 a$s, m,s E
EEE i
Q~
- =**=====,m w-
--~-~s_~~~.-----*---~=.o,~.._=.--,,...
y 4
i l
CONCLUSION 91 THE EFFECT OF TD FERATURE EQUATIONS o
18 YEARS AGO, F.C. ZAPP (REF. 2), A2, PDRE ECENTLY, S. FRAE, ET. AL (REF. 8), CONCLll)ED THAT EQ(A) Am EQ(B) YlELD C0WARABLE RESil.TS FR(N TE LEAKAGE RATE CALQ1.ATION, IF CHANGE IN TE SPATIAL TENERATURE DISTRIBUTION IS INSIGNIFICANT. TE STAFF HAS ECOWIRPED THIS CONCLUSION BY PEWORMING LEAK RATE CALCULATIONS USINGACTUALTESTDATA.
I o
DESPITE TE MINOR IWACT ON TE CALOUTED LEAK RATE FR(M USING f
DIFFEENT TENERATURE EQJATIONS, TE 120STRY WILL E20RSE THE i
USE OF EQ(A) SINCE IT IS PORE MATEMATICALLY CORRECT, j
8 S
V-3 I
i t
9 WFid I
l 3
3 Table 3-3 i
COMPARISCH OF SPATIAL AVERAGING NETEODS 6
i 6
apper l
Averaged Confidence confidence Data Set Parameter Equation Leak Rate Interval Limit (UCL) e Number Varied Number
(%/ Day)
_(%/ Day)
(%/ Day)
% UCL 1
1 Temperature 3-5 0.036701 0.006917 0.043618
-100.0 f
3-8 0.036326 0.006919 0.043245 99.1 0
2 9
Vapor 3-6 0.035972 0.007044 0.043015 200.0 i-Pressure 3-9 0.036701 0.006917 0.043618 101*4
.}
3-10 0.035678 0.007070 0.042748 99.4
.g
.1 2
Temperaturs 3-5 0.296588 0.007249 0.303837 100.0 4
3-8 0.299886 0.007002 0.306888 101.0 t
j I
Vapor 3-6 0.29495'3 0.007741 0.302694 100.0 Pressure 3-9 0.296588 0.007249 0.303837 100.4 j
3-10 0.294060 0.007492 0.30155L 99.6 (i
i 1
h ',
i i
a
+
k
+
+
+
t b
'i't?'aI[
s s.
s k
4 g
4 g
A M
4 g
4 w
k (aO k*
E
~
2 4
4 t
x
~%
y u
8 3
4 g
g E
SE E
,c C E, I
m
+,
a%
s s
s s
EIR ER-
~ f 4
+
4 w
g e
o s
s 3
y q.
t e
k 4%w t
F l
n t
- ( &1
't n (
(9 4
4 3
w.
4w S
L
<M h.%.h y
s n
m h
Y i
St Q
t>. -
.N M-M M
I s
g g
a d
4 0
N h
k f
4 o
tu u
k k
N d
%. 4' %
4 i
q N
m%
i.
l 8
s 1
,,n e
a maam Mn
~
^
- f
~
~
EFFECT OF EIGif COEFFICIENTS
'E ON ILRT ESULTS 1
I:
MASS IS C0ffillED BY I
l!
Wc = P.V AT ANY TIE ti R7l TL = f2 "I&
WHEE o(j IS THE EIGiT COEFFICIENT l
NOTE:
o A EIGif COEFFICIENT DEFIES A FRACTION OF TE
- f.-
CONTAIPPENT VOLlFE WITHIN W11CH TE TEPPERATUE IS ASSlMD ESSENTIALLY lMIFORM DURING TE TEST i-I' o
TE CONCEPT OF USING A SINGLE SET OF WEIG{T
,j COEFFICIENTS ASSLKS THAT TE SPATIAL TEPPERATURE DISTRIBlfTION DOES NOT CHANGE DRAMATICALLY DURING 1
I
.p 3
4 I
~-
j CONCLUSIONS EGARDING TE EFFECT 0F EIGHT COEFFICIENTS
?.
0 TE CONCEPT OF USING A SINGLE SET 0F EIGHT COEFFICIENTS ELIES Gl A STABLE SPATIAL TEWERATURE DISTRIBlITION. FOR THIS EASON,
~
lE EGULATION, AS ELL AS IPI)USTRY-SPONSORED
['-
GUIDANCE, EN HASIZE STABILIZATION
.j
}
o PARAETRIC ANALYSIS HAS SHOWi THAT WITH STABLIZA-
~
TION, VARYING EIG# COEFFICIENTS HAS opt.Y A MINOR f
EFFECT ON CALCULATED LEAK RATE t
I i
f 9
80e e
~
R R
y
^
g
~ ~t 4
~
t S
g i
4 A
Il N
Q~
0 V) 4 14
$"k D
k Pj pk
??
g Is M
g.
xN}
qA N
lb (4
d t
g g
I
>h pkN {
o U
l l
9 0
._.._3
- 5
!i
?
4 M;
REFERENCE l
i.
1.
Reactor Containment In-Service Leakage Testing Criteria, Feb. 1, 1965 il 1
2.
y Testing of Containment Systems Used With Light-Water-Cooled Power Reactors Frank C. Zapp. ORNL-NSIC-26, Aug.1968 J.
n 3.
Air Leakage Rate Studies On the C.S.E. Containment Vessel, M. E.
Witherspoon 7 g. j. Rogers, BNWL-1028. September 1969 lI 4.
Leakage Rate Testing of Containment Structures for Nuclear Reactors ANSI j
N 45.4 - 1972 5.
ii Testing Criteria For Integrated Leakage Rate Testing of Primary Containment Structures for Nuclear Power Plants. BN-TOP-1, Revision 1
- 8 Nov. 1972
}
6.
Appendix J. Primary Reactor Containment Leakage Testing For Water-Cooled Power Reactors, Feb. 14, 1973 7.
Containment System Leakage Testing Requirement. ANSI /ANS 56.8 - 1981 i
8.
Containment Integrated Leak-Rate Testing Improvements EPRI NP-2726 Nov.
1982 1
?
Criteria For Determining The Duration of,. Containment Integrated Leakage 9.
Rate Tests EPRI NP-1393-5, June 1983 o
L 7
i
d r NbOT1 Lac e,,e as aswr rwr 6:7 F r. Revara _a o
}- ll'l $C 'tZ j ZION na n.a 77b'.
9 Sum mi.um e t
i i
I Mr' s
I i
i i
I
- 729 pgge t
i i
I l
I l
e 3
7/f#
l l
8 6,
6, 4,
s.n-e
,,s-4 l
A
.i z
5,r, i
s.
6, 6,
s,s, l
3 l
I I
67 [
j l
i I
473 66 [
l.
l I
I i
e I
sur 6uevotu m e sei 3
~
'6sr 8
l I
I
&&f I
I e
I I
i I
I l
l bli' q
t g,., '
O
- c..,
w D.
,, ~,,,.m,y m,o u,
ND FIN D COERTSpoNDeul, QbNME L,,
~
b - Draecru
- ND comittsPoupiMG CNawwrL.
~__
ZION druser Loc.ar.ou s BrmarEM WE E9e Fr ( 6 87 Fr. EL.
\\
O
% - comsroweiwe,cAurt.
b - DrWe.El.L AND 002EESFDMD8MU 0144MM EL i
?-
dad
% C.
{
Oug y0LMME duBWLumE rDP _
5 4
v,c w i
)
\\
\\
A*v
. v
- e A
e h
bli
- *=c e
Ae Eh v,e a
- $90?
l
\\
I
'v w
w
~v---
+-c.,. -
LLUD drwoor locorrows Berwrca r?/a Sag st / s po sr. ft.
O - rro.~, e m,oso.us cu~weu b - Descett. A wp c.out u roNoig h C Ra w u s t o
1 "TbP 6uBvot.u m E Sus mumE
~
v,.s sr 4
l t
sd
)
5'ae 0
O e
v;ew (t,8 '
m
...__y_.._
., ~..
..,-.-_-...,..,,,_c.,
-_-_.m.,
._-- - - <._-w...,,.-
-. m.
ZION 1 - 1983 Type A Test Weighting Factors 4
Sensor No.
Case No. 1 Case No. 2 Case No. 3 1
0.04500 0.040 0.080 2
0.04570 0.040 0.000 3
0.04575 0.040 0.080 4
0.00000 0.000 0.000 5
0.04575 0.040 0.010 6
0.04570 0.040 0.060 7
0.04500 0.040 0.010 8
0.04570 0.040 0.020 9
0.04570 0.040 0.060 10 0.0457G 0.040 0.010 11 0.04570 0.040 0.000 12 0.04500 0.040 0.020 13 0.04575 0.040 0.040 14 0.00000 0.000 0.000 l
15 0.00000 0.000 0.000 l
16 0.04575 0.040 0.100 17 0.04500 0.040 0.060 3
18 0.04570 0.040 0.000 i
19 0.04570 0.040 0.080 20 0.04570 0.040 0.030 21 0.04570 0.040 0.040 v.
22 0.04500 0.040 0.000 23 0.04500 0.040 0.080 24 0.01800 0.040 0.070 25 0.01800 0.040 0.070 26 0.01800 0.040 0.080 27 0.01800 0.040 0.000 28 0.01800 0.040 0.000 29 0.00000 0.000 0.000 i
30 0.180 0.250 0.100 31 0.000 0.000 0.000 32 0.000 0.000 0.000 33 0.000 0.000 0.000 34 0.640 0.250 0.100 35 0.000 0.000 0.000 36 0.000 0.000 0.000 37 0.000 0.000 0.000 38 0.090 0.250 0.500 39 0.090 0.250 0.300 o
1 i
2
i Zion 1 - 1983 Type A Test Subvolume Temperature Dew Cell Weighting Factors Number Sensor No.
Sensor No.
Case No. 1 Case No. 2 Case No. 3 1
1 0.045 0.040 0.080 7
0.045 0.040 0.010 12 0.045 0.040 0.020 f
17 0.045 0.040 0.060 30 0.180 0.250 0.100 31 0
0 0
2 2
0.04570 0.040 0
i 3
0.04575 0.040 0.080 4
0 0
0 8
0.04570 0.040 0.020 4
9 0.04570 0.040 0.060 13 0.04570 0.040 0.040 14 0
0 0
18 0.04570 0.040 0
19 0.04570 0.040 0.080 32 0
0 0
34 0.640 0.250 0.100 36 0
0 0
3 5
0.04575 0.040 0.010 6
0.04570 0.040 0.060 10 0.04570 0.040 0.010 i '
11 0.04570 0.040 0.
15 0
0 0
16 0.04575 0.040 0.100 20 0.04570 0.040 0.030 3
21 0.04570 0.040 0.040 33 0
0 0
35 0
0 0
37 0
0 0
4 22 0.0450 0.040 0
23 0.0450 0.040 0.080 38 0.090 0.250 0.500 4
24 0.0180 0.040 0.080 25 0.0180 0.040 0.070 e
26 0.0180 0.040 0.070 27 0.0180 0.040 0
28 0.0180 0.040 0
29 0
0 0
39 0.090 0.250 0.300 i
i I
j i
r 3
2 -
j L
Zion 1 - 1983 Type A Test i
Weighting.
emperature Averaging Method Factcrs Used T=IEfTW F1 x Ti T=1/EUW Fi x Ti)
Temp ("RJ Mass Llb)
Lam 95 UCL Temp (*R)
Mass (lb)
Lam 95 UCL Correct Weighting 524.17 576,319 524.17 576,324 Fact:rs-Case No. 1 523.61 576,181
-0.001 0.015 523.61 576,186
-0.001 0.015 All sensors have 524.31 576,152 524.30 576,157 equal weight--
523.77 576,052 0.010 0.022 523.77 576,058 0.010 0.022 Casa No. 2 Arbitrary weighting 524.70 575,709 524.69 575,714 fact:rs-Case No. 3 524.15 575,653 0.004 0.015 524.14 575,657 0.004 0.015 1
l l
e$
4
'l LASALLE
/W" Qu
/
D n
.w g
l 7
P i
s 2,
i I
pd#
/
i c
I
/
/
.i y
l q
l 4
- /
- 9
~~
i@
O o
e f
f i
f//////./
h/////)
/.
/
\\
,f
,/
' I '////
!,/
}
f
,/
/ e
/
e '
/
/
e /
/
/
i i
r/
/
/
j
/l//
l/
/
/
/
f l
/
/
k
///
/l, h
.e-1 I
i Elevation view of contaimment I
and subvolume locations 1
l LaSalle 1 - 1982 Type A Test Subvolume Temperature Weighting Factors No.
Sensor No.
Case No. 1 Case No. 2 Case No. 3 I
1 1
0.00515 0.03333 0.00515 2
0.00516 0.03333 0.00516 2
3 0.00873 0.03333 0.00873 4
0.00873 0.03333 0.00873 3
5 0.01033 0.03333 0.01033 l
6 0.01032 0.03333 0.01032 7
0.01033 0.03333 0.01033 i-8 0.01032 0.03333 0.01032 4
9 0.02407 0.03333 0.02407 10 0.02407 0.03333 0.02407 11 0.02407 0.03333 0.02407 12 0.02407 0.03333 0.02407 5
13 0.03516 0.03333 0.03516 i
14 0.03516 0.03333 0.03516 15 0.03516 0.03333 0.03516 16 0.03515 0.03333 0.03516 6
17 0.00449 0.03333 0.01019 1
18 0.00449 0.03333 0.01019 1
19 0.00449 0.03333 0.01019 2
20 0.00450 0.03333 0.01019 7
21 0.06442 0.03333 0.00644 i
22 0.06442 0.03333 0.00644 23 0.06442 0.03333 0.00644 1
24 0.06443 0.03333 0.00644 T"
8 25 0.06973 0.03333 0.10458 26 0.06973 0.03333 0.10458 27 0.06972 0.03333 0.10458 28 0.06973 0.03333 0.10458 29 0.06973 0.03333 0.10458 30 0.06972 0.03333 0.10458
]
i 5
i i
l h
2
. -.. - - -. -,, - -. -... - - -. - - - - -, - ~., - -,,. - - _
e t
LaSalle 1 - 1982 Type A Test Results Weighting Temperature Averaging Method Factors Used T= KfTW F1 x Ti l T=1/t(TW Fi x T1)
{
Temp ("RJ Mass I,1b)
Lam 95 UCL Temp (*R)
Mass (lb)
Lam 95 UCL t
Correct Weighting 556.100 104,684 555.785 104,743 Factors-Case No. 1 557.368 104,299 0.372 0.374 557.006 104,366 t
All sensors'have 562.770 103,443 562.477 103,497 equal weight -
564.367 103,005
.418
.421 564.040 103.065 412.415 l
Case No. 2 Arbitrary Weighting 551.067 105,640 550.754 105,700 Factors-Case No. 3 551.913 105>329
.299
.302 551.553 105,398
.291
.294 1
1 i
I i
[
t t
1 1
6 j
INSPECTION PROGRAM Review Test Procedure Select systems for valve lineup verification Vent paths affecting penetration Possible sources of in-leakage Historical problems Perform independent verification of selected valves and sensor location.
Obtain copy of weighting factors, other correction factors the licensee has entered in computer.
Perform independent calculation of at least one data set to ensure satisfactory computation of mass.
' Verify temp stabilization.
Perform leakage rate calculations during Type A test.
Observe sensor data for trends.
Review log of events and procedure changes.
Obtain independent superimposed leakage flowmeter readings.
Perform independent calculations to determine results of verification test.
k l
1
- \\
ll l
l
!e y
e d
Ja n
e r
t h
o u c, g
e s
o tA c n q.
mi o
s L
d t
e f
n nl v *lupM A
o t,
Ye s
S y
4 n o t
y e
W e u
'1 u.
&%W c
i
- R e
do Wl e lP e -
u s
s s
i'D O
R~t V
t r
. L Q
O
= Yl t s -
s
/
tO i
A yE
,a s f v S
t h
V A
C N
[P 6 Q F AA m b)d.
V
)A O
c sUap
.a % n me
%c R
M* h Cs mf W
r%/
kk A
t
- t. d ke 0
E cfr as
?
o a o ut._
lh
=
n T
t o n.
q t
9)e. aW cs ea F
8-0 g
e Ps v.
J r
r n
dA pis sd
.o j 9
l s aa s
M/uWs3 s
ai gtrt" a
si*
u a
)y a e a 1* ta s d. f s m
s
~
e t
o ei sd t
o r
to hPePo S
i aen o n I'
k.
9 3
,f l
ll\\
!l
'l L
W g
b O s
~%
c0 g
co 3u k$'
e I
P'k i.3 8,45 3
1
$ls J a
s4 emn}asl i
li i ijj
Sou j 5 ff,f.fy& # fm'g>,
k Sbe&dboakam a8 pteeWeek
. -i j
WGeh c4ctstachet OV4
% = v'P M
\\
l fo j$dy c w t!,'9 st a pa,g & %
s
\\
to>uth%c4.veaeaen.
ma u es & de %.s),
\\
eaL & e %
- % *Reym24 Q peeve n,
- "MD 4 M.
ggn ggggg;gygL 1 hd4gb a M THR0WAI OUTli rwoST B E m
p' 1 7,+ 7: 47 )< 0.s~waa l
(. 0 g c a u e w s,, e
- t g n c e tw woru~s a u sa m ped
\\
~
b Yg' f ' Y 444. &
CWA Q WY'LO A>A 4 > n..
intsxvat )
____.._.m
.._._._m_._
r i
o N
i 3
V
!4
~
F q
- A W
O b&
W 3
11 5 &
N~b 29 l os g e
. cs e
g-b5%
csj
, e e_
WQ.T W
!k 6-fe) h~Q g;-
lM 4
Y i
d
$$ di li. $e 1 @<
a 3
,3
l-
\\
\\
\\
i o.
GO-
\\
f
\\
% Q
\\
Q' 8
t
\\
Q
((
W
\\
a
\\'
?
Y.
e
\\~
b
($
~
i g
'g
- 4 d e a 6*=
oc g
g'
!k *o m.
,=
u o
\\
e
$s 8
s e
O N
d o
- g
\\
~
1 s
g O
W A
M
\\
C *
\\.
vb su$e k h
\\
a w p>
5;.
\\
d 4
m n
E g
7 a
_a yn
% -P
\\
- ~ ~
~ * ^ ~ ~
NOW THEY FUD$E.1T SAROR
\\,
- l. $NS TR UN5NT PL AHis NBNT.
x l0 A. # 0F INSTRWMENTs, CAGBRAT.ICW.
X IO u.
\\
3.ORSITRhAV R $ S UMTIONS.
x 100 j
- 4. US2 OF THE WhMt EQUA7JoM X l0
\\
- 5. "ssasitig ag aguacg ng"(N x too i
i
\\
.6 'loaINs 097"($l&$7'g',w) x too
- 7. "Btoex;ng taaxing pahg" x&on l
- 8. fwotty veri /;ca2%>
x y
l
- v. getry l
l
O r
4 m.%e e
e i
t4
,e d
isau i.
I6%
+ w 4
e.N
%e
.:. A 4
24 4
4
.,~
s De i
1 9
=
c
=
lq 4
4
.o,.
i, Uttd E
$e d se 1
e 4,
$ ax-3
\\)
r h
% p$
ci " * '
N-T$
. ~ '
) %
3
=,
3 w-7 i
n l
NW %%%
p-I a
$.II l
<1 7
=,
i2
.4 3+
T aity_
4 (7 4
1'h
$s $ [ 1 f.
l W
e e l
v s
L.
\\
l
\\
I. Tus pneseur resrzarc rs
'l
<./
FAULTY
' 1. Au pasr resrs nosr sa 1
RE VIEWED
\\\\
3.Tus sao rssrzuc nosr se
.\\
BAMED IMMEDIATELY
]\\0 W THEY S ABOTRCE I. DnTn uNA VR1L A BL E. 1 .2. No nesposse, nearzarcs orscossz a.
- 3. )Vo kerloN.
- 4. COVE R UP.
S. NITHOLDINC INFORNRTIoW
- 6. GRCCI6fC
- 1. P R o H I s s s'
- 8. L z s $, fo R G E R Y
- 9. Fm.sa "RESEARCH"
$0..lWCRERSE DF ALLOWRBLE LE&R RATES ("Ph0BRblQ$7]e" APPROACH)
a --a, p g B 4 4
- f
\\ W i w..f lt Qi m y m %, p1 L,,i 4 ob t 1.h(k '4 ( t ) -1 %{ v i E , CV - m ,j tra O D 6N \\ a, s @ @2 ~4 1 m.gtk\\s64~W w d i 4 { *w. J (D CC =h [L.M A N
- e
- v
$d$ st d = 44 e 4, teq% o3 .s 5%
- m. a15 ki,$,*$
e 'gg bg a t' 4
h3M -RRTE TE s LINC L q MONITOR lmt0$0RE:' r 3
- 4. % u v e e P
inru m m 2.Tenpanervar T-pmsta. es Leeat rasrzwe luresAet. resrlue
- nre 8,0 lunDEQualir crves R) FAULYY oprew FRBUDULGMY Tueoitsricas.
pouvoariou orieau anre nunsvaanmar. 7 1, p 9= x& - cas uw m=nydv l 14 5jy 3 to q' M = No &Ms+'*" W k (d +d +....) o Tne wnouc eevnrios gy KP % v x a m
- r.. -..
ma$1'0V
a a-wp]A.IN{l 1 7 44 8 4 1 4 a e as 414 "1 h T4 3 i st~448 a w% 3+ 14 9 % k b 0 sw e a Q [N p s {,$=E1%13 b 3 \\s 1 s.:}1v$4es a 3 y
- 31 32 2
+ a c +
&= soo dwN< (<.e.om) t ' I-S 4-Is 'Ib- 0 Al-Sol, nak o soo sos.on 4gt,9y soo Q f-l5 y gg.39 g, I soo so+.01 500 r } soo s10.S6 soo het hdet (2-) and Cvce 5 coo %wn eat, 4%a.seet"ee "okelanet out of wedice*] I &nwivi 3LS e# /6-f 0 \\' jn.af ku,aag9 (.2) od"#"* I y p b 50* S00,S00 < 5001 o e;ke a ed'll'ouy=o %g c hal &n" tatt= 3%lf(k +fg +hy,&;lsoo = 0.174 / day; Hvcung's staeiQath d'ca"d"b% &'f' of a 4ad 60'1tfajhhseh$ $ @ g
i'bei no Med lTO}ktletoon. THE To7AL or Booxs a -3095:5g, ThE wReMC soLUT10tJ: ,,m gyggggg pggg a i
- f($i+$10)= 8S: 50)Torpt e soons =.
- te
=? l
- & = 18 8c925 g
N LEOR RRTE IS TNA i - TIME c l SLOPE OFMAss-Y1ME l cuRtfE. [ ERR *.6P - 4o d'% ST,
- " ~ ' " * " '
- i
/AATE) P P, ~ds vi - +g> e - sueu. $0Att tosloW* AHY VALUE @ {.k. Cd BE ; j [ DE TERMINB 9 " PRQM #AfV Ser DF DATR. l WRLID&TED Sy WOR 14E D 2XRNPL E S. I SUBSTMTJRTED by RCTm MSTS. t w n - r,n , a u. ic nr,
~ -- 7 4 u N gG Et 4 f q w 3, 'yo 4 if ^4 1 JL / 5 5 qs 2n %;!r A4 i 9 U f3 Qe n' e 4 N .c0 W f GD t $ y 8 4~h,hh5 % Q b =. s$' Jge 3 [ } e N 4 4-L Y
Draft RES0LUTI 0N The NRC Staff Meeting with Z. Reytblatt to discuss the emergency relief Decision DD-85-10 of July 3, 1985 (" Decision") has unanimoustly determiner that the Decision (1) neither contains calculations, derivations and/or proofs nor indicates references in lieu of those; l (2) does not address some of the Petitioner'.s important statements ; 3 2 3 (3) contains deceptive statements, The Meeting unanimously deceided that the Director's Decision must be amended in view of the following facts:
- 1. The method of air mass calculations as practiced presently may induce errors exceeding the allowable errors. Recommendation: immediate return to the conventional physics equation.
- 2. Some of the Zion I weight coefficients are incorrect. Out of 6 "verifi-cation" tests conducted at Zion in 3 years, three failed, two were "suc-cessful" as a part of fraudulent tests ("irrigatioA' and " doubling of weight coefficients"), and one was performed in lieu of analyzing the wrong test assumptions. Recommendation: Find the conservative estimate of the Zion I leak rate using conservative assumptions on the weight 7-
. coefficients; if the leak rate conservative estimate exceeds the allowat leak rate, make repair of leaking paths and retest the containment.
- 3. The method of obtaining the weight coefficients as practiced presently may induce errors exceeding the allowable' errors. Recommendation:
Immediately' terminate application and/or development of new weighting coefficients using unscientific methods.
- 4. The present " verification" tests may not detect that the test assumptior are wrong. Recommendation: Immediately terminate all the " verification" tests until a reliable technique is introduced.
O I For example, the claim that the Zion. weighting coefficients are not correc is not addressed; the claim that the Director's Decision of March 16, 1982 is based on false stat uses bad numerical exam (ments, fp not addressed; that the same Decision es and perjurous (in nature)" Summary", is not addres 2 For example, the Decision contains a deceptive conclusion that the Volumetrics computer program does not contain a fraudulent option (s) on the basis that a particular test calculations were correct (obviously, the suspected fraudulent option might not have been called for execution during a particular job). I (_.
- 5. The Commonwealth Edison computer program contains a fraudulent option for erasing the data and doubling of weight coefficients so that these coefficients may exceed the limit of 0.1 specified in Appendix J to 10 CFR via ANS N45. -72. Similar computer programs (Volumetrics, as an example) are used all over the United States. Recommendation: Immediatt suspend the use of suspected software, debug and revalidate. Make avai-lable the listing and validation acts for all the sof tware used for the leak rate testing.
t i
- 6. The Director's Decision of March 16, 1984 on LaSalle contains false F
statements. Recommendation: Make this Decision null and void. 1 In addition to Dr. Reytblatt's original requests and appeals, the Meeting has considered the stabilization notion. The Meeting unanimously decided that neither the ANS version nor Mr. Huang's version have any scientific merit but may 24F(as the presented examples clearly indicate) either 4kL harmless or mistakenly qualify a good containment as a bad containment or mistakenly qualify a leaking containment as a good one. The' meeting, similarly, established that no scientific merits have been found in such means of affecting the environment as irrigating, ventilating, equalizing, etc., whereas on numerous occasions these means were used for cheating. 1 Recommendation: Immediately terminate equilizing-stabilizing-irrigating-ventilating. L-T The meeting established that the" diurnal effect" and short duration tests the Palo Verde Fall 1982 tests clearly indicatg can be used fc as - cheating. Recommendation: Immediately revoke all the exemptions for the q short duration-tr.sts. No new exemptions shall be granted. The meeting supports NRC attempts to terminate the reduced pressure tests. Recommendation: Immediately terminate the reduced.. pressure testing. O The meeting urges the NRC to comply with 10 CFR and to prepare the full and open to the public response of the present faulty testing practicies to the criticism and the faulty ANSI /ANS standard which was submitted to the NRC by i Mr. Z. Reytblatt in his " Report". ? l l 4 0 ,----,-,--g-e- -,----,m, -,--.-------,--,w-,u -y ~-gm- ---,---aemygm.w ggm, --,y--,m x}}