ML20205N674
| ML20205N674 | |
| Person / Time | |
|---|---|
| Issue date: | 10/26/1988 |
| From: | Advisory Committee on Reactor Safeguards |
| To: | |
| References | |
| ACRS-T-1697, NUDOCS 8811040280 | |
| Download: ML20205N674 (148) | |
Text
--.
- Aces 7=/697) p r q,\\, i J
O UNITED STATES i
i NUCLEAR REGULATORY COMMISSION i
f i
...........................................................j t
4 In the Matter oft
)
f i
ACRS Subcommittee on l
Mechanical Components i
I i
l
! O i
j l
l I
Pages:
151 through 25 f
I Place:
Bethesda, Maryland l
I l
Date:
October 26, 1988 l
ACRS CFce Copy-le':a"
...................................,..,..........,.....il M J L O T E 8 1 0 O !. L J. 0 u 0 '. E l
i, HERITAGE REPORTING CORPORATION j
1 l
osuarseasm
\\
j O
m n w.w = ae WanWeston, D.C. 200H ctd" E f
~
G (281) 62N008 r-u st 1
1 l
l l
l p)
(,
i PUBLIC tlOTICE BY THE 2
UllITED STATES tlUCLEAR REGULATORY C0KMISSION'S l
3 ADVISORY COMMITTEE ON REACTOR SAFEGUARD 3 4
5 t
?
The contents of this stenographic transcript of the 8
proceedings of the United States Nuclear Regulatory 9
Commission's Advisory Committee or4 Reactor Safeguards (ACRS),
10 as reported herein, is an uncorrected record of the discussions t
11 recorded at the meeting held on the above date.
12 No member of the ACRS Staff and no participant at l
13 this meeting accepts any responsibility for errors or
()
14 inaccuracles of statement or data contained in this transcript.
15 i
i 16 17 18 l
19 1
1 20 l
l 21
[
1 22 i
23 I
l 24 l
l 25
(
Heritage Reporting Corporation I
(202) 628-4888 l
[
k I
UNITED STATES NUCLEAR REGULATORY COMMISSION
(}
2 ADVISORY COMMITTEE ON REACTOR SAFEGUARDS 3
(s 4
In the Matter of
)
)
8 SUBCOMMITTEE MEETING ON
)
MECRANICAL COMPONENTS
)
6 7
Wednesday, 8
October 26, 1988 9
Room P-ll4 7920 Norfolk Avenue HD 3ethesda, Maryland II The above-entitled matter came on for hearing, 12 pursuant to notice at 8:30 a.m.
)
ACRS MEMBERS PRESENT:
14 MR. CARLYLE MICHELSON 15 Chairman Retired Principal Nuclear Engineer I'
Tennessee Valley Authority Knoxville, Tennessee, and, I7 Retired Director, Office for Analysis & Evaluation of Operational Data is U.S. Nuclear Regulatory Commission Washington, D.C.
MR. CHARLES J. WYLIE 38 Retired Chief Engineer Electrical Division 21 Duke Power Company Charlotte, North Carolina 22 MR. JAMES CARROLL 23 24
()
25 Heritoge Reporting Corporation 1
= =
h
CONSULANT P. Wohld Cognizant ACRS Staff Membert
+
4 E.
Igne NRC DR. Gerald Weidenhamer, RES NRC Presenterst
- x. DeWall R. Steele le 11 12 33 14 is 16 17 p
is it 30 21 0
23 24 O
25 Heritage Reporting Corporation i== -
151 1
AFTERNOON SESSION
(
2 (1:20 p.m.)
3 MR. MICllELSON:
It !.s time now to reconvene, and 1 I
i believe that Jerry is going to tell us about the valve 5
related research plans for particularly the coming year, I 6
guess.
7 Will your remarks be cantered just on motor 8
oporated valves or all kinds?
9 HR. WEIDENilAMER:
I think that it is going to be 10 everything.
1 11 MR. MICIIELSON:
Okay, good.
That is better yet.
12 MR. WEIDENilAMER:
But it is going to be very 13 general, because I think that I am going to tell you about i
14 some things that have developed since.
( ) 15 MR. MICl!ELSON:
Good.
16 MR. WEIDENilAMER:
This work will probably most 17 likely be funded under either that FIN number or maybe even 3
18 the new one, it does not really matter.
But Idaho is 19 probably a very key figure.
The Idaho National Engineering 20 Laboratory will probably be a very key figure in doing this, 21 because they are heavily involved or have been heavily 22 involved in research in the past.
23 These are going to be the major accomplishments 24 for fiscal year 1989, the year that we are currently in now.
25 There aso two major events that are going to occur, at least lloritage Reporting Corporation (202) 628-4888
152 1
two major offorts are going to occur this year.
(
2 We want to complete the steam flow interruption 3
tests on typical gate valves.
We will most likely the two 4
valves that we had for the hot water blow-down, and l
5 refurbish them and get them back in shape again, and now l
6 subject them to the steam environment for itPCI and the RICI l
7
- lines, i
8 And then this bottom bullet here is to completo
[
9 the evaluation of the llDR data on the very large seismic r
i 10 loads that we tested this year.
The specific work that we y
t 11 hope to get out of this is from the first blow-down test is I
t 12 wo want to know essentially resolve Gonoric Issuo 87.
Wo i
j 13 want to develop procedures for assessing whether valves were i
t l
14 closed against high steam flows and against the high water 1
( )
15 flows.
That is essentially the resolution of Generic l
16 Issue 87, c
t 17 But as part of that work also, we hope to evaluate
{
i 18 that formula that heard so much about this morning.
There f
l 19 are many variables in there, and we are not cure yet what i
4 I
20 some of those variables should be.
l
)
l 21 MR. MICIIELSON:
Is there any more work l
t 22 contemplated on butterfly valves?
l 23 MR. WEIDEHilAMERt We have to resolve some of the i
l 24 comments that you brought up, not only today but in the l
25 past.
I L
lioritage Reporting Corporation
{
(202) 628-4R88 i
I 3
153 1
MR. MICHELSON:
More in terms of fiscal year 1989?
2 MR. NEIDEN!! AMER:
That is a possibility, although 3
I have not identified it.
You are talking about the offects 4
of one valvo affecting the closure of another valvo?
5 HR. MICilELSON:
Nell, that situation and whatever.
6 MR. WEIDEHilAMElt This is only 1989 that I am 7
talking about.
8 MR. MICllELSON:
If it is not underway now, it 9
would be pretty hard to have it underway.
10 MR. WEIDE!!MAMER:
Perhaps you are right.
I will 11 address that in just a few minutos I think, maybe not
]
12 directly but something that you should know about.
And then 13 with regard to the ilDR data, the high seismic or largo 14 seismic loads testing data, we want to determine what tho
( )
15 real effects on the aged gate valve were, these very large i
16 8-SSE loads.
That is how high wo actually got to in 17 exciting that piping system.
And with the valvo, this aged 18 valve right next to it, by the way, one of the excitors.
19 And then we also want to unlorstand the behavior of pipe 20 support, struts, and snubbers to those very large loade.
21 So that essentially will complete the work that we 22 have already outlined for fiscal year 1989, There is 23 approximately, tied up with those two items, there 's 24 approximately 5550,000 for next year.
We do not know who 25 the winner of any testing will be for the steam floritago Reporting Corporatien (202) 628-4888
154 1
blow-downs.
That has to bo pat out in a RFP.
So that is 2
fiscal year 1989.
3 MB. MICl!ELSON:
Will the steam blow-down bo dono 4
at Wylie also?
5 MR. WEIDENilAMER:
We will have to put a RFP on it, 6
Carl.
We are not sure who will win that.
That will bo done 7
probably either next month or early December, to put out the 8
RFP.
We hope around the first of the year have a pretty 9
good idea who will win that.
10 The impact of the safoty and technical issues, 11 again just to repect, that work will contribute to the basis 12 of resolving Generic Issuo 87, which is that title thoro, 13 and also the other hot water cleanup line, the reactor water 14 cleanup line, and the RICI line.
That work will also
( )
15 provide a basic for evaluating the utility responses, 16 MR. MICHELSON:
When you do the steam blow-down 17 situation, are you going to interject water into the steam 18 flow also as a part of the testing?
19 MR. WEIDENilAMER:
We would like to be as realistic 20 as possible.
If we think that that is a real condition for 21 Generic Issue 87, yes, we would.,
22 MR. MICllELSON:
It is a real issue for Generic 23 Issue 47, the steam generator overfill and the reactor 24 vessel overfill.
Clearly, if you overfill, you are going to 25 end up with water in the HPCI turbine if it is running at fleritage Reporting Corporation
{}
(202) 628-4888 J
155 1
the timo, which it could very well bo just doponding on what 2
area.
3 MR. WEIDENil AMER :
Okay.
I will certainly look 4
into it.
5 MR. MICllELSON:
It is clearly the case for the 6
auxiliary feedwater turbines, PWR.
Because they can be part 7
of that scenario, and they do not trip out again.
8 MR. WEIDENHAMER:
Okay.
I was wondering could the 9
quality of the steam have anything to do with this.
10 HR. MICilELSON:
I do not know.
11 MR. WEIDENilAMER:
Woll, that is one of the 12 variables that I am thinking about.
13 MR. MICilEI. SON :
In looking at your other results, 14 it looks like the quality might be
( )
15 HR. WEIDENil AMER :
Quality is one of the areas that 16 I think would perhaps address some of the concern that you 17 have with injecting the water.
18 But we hope to also provido a basis for evaluating 19 some of the utility responses for complying with the 20 resolution of Generic Issue 87, and also perhaps for 21 Owen Rothberg's II.E.6.1.
No believe that whatever l
22 information that we got out of this work will certainly 23 benefit Generic Issuo II.E.6.1 ns well.
24 MR. MICitCLSON:
Is there not a generic issuo 25 rotated to the PRVs in their block valves, or is that one lloritage Reporting Corporation (202) 628-4880 1
l
[
156 1
all resolved?
2 MR. CARROLL:
That is still out there.
3 MR. MICllELSON:
I think that that is still out 4
there.
5 MR. WEIDENilAMERt Do you remember the number?
6 MR. MICllELSON:
I do not remember, but I think 7
that it is still out there.
8 MR. CARROLL:
I was reading something about it on l
1 9
the airplane yesterday.
10 HR. MICHELSON:
And the block valves, of course, 11 again get into two problems.
One is the situation 12 downstream with a high disturbance.
And the other is the l
l 13 water quality or sea quality rather.
l 14 MR. CARROLL:
I guess what it was, Carl, was that
( )
15 it was in connection with the environmental qualification l
16 PRA business.
That is what I led me to believe that it is 17 still about that.
10 MR. WEIDENilAMER:
The only problem with all of 19 that is, and that is all well and good, that our budget for 20 this year is something like $400,000 for that.
We have a 21 total of about S675,000 or $700,000 for this work for fiscal 2"
year 1989.
We could I t h to that.
But please keep in 23 mind that we certainly e imited on the scope of all of 24 this. Generic Issuo R7 in the paramount issue that we are 25 addressing.
Ileritage Paporting Corporation
{}
(202) 620-4888 i
L
-l t
l 157 i
1 And then obviously all of the key information that l
l, ()
2 we get out of this work, we plan to where possible to f
3 incorporato it into the ASME IEEE standarr's wherover we can, i
4 MR. MICilELSON:
One of the aspects of Generic Issue 87, of courso, that you do not hear much about, and t
6 you have been talking about whether or not the valves can j
i 7
close under those kinds of conditions, but the other i
e 1
0 question is how reliably to do they closo, in other words t
9 does the reliability change when you do under dynamic i
i 10 conditions and what is the likollhood of failuro from the 1
i l
11 rollability viewpoint given a pipe break a.1d the nood to l
l l
12 isolato, using the normal operating numbers so to speak and 1
13 not the accident operating numbers.
W9 do not know what the 14 accident operating nsmbers are.
I
( ) 15 llow do these extra loadings, which admittedly we t
16 got some extra loadings, how do they affect the anticipated i
17 probability of closuro?
j 1
i 18 MR. WEIDENitAMER:
Well, I think that the seismic I
19 loading I believo has a very small offect on it.
l 20 MR. MICllELSON:
I would not dwell on it too hard.
f 21 I am mostly concerned just about the change in reliability I
22 due to those dynamic loadings.
Hf questien in other words t
i 23 is the ute of PRAs of non-loaded valve riumbers to reflect f
24 how probabio it is that a valvo urder loaded conditions will I
I 25 work.
None of this work seems to help that.
l l
lloritage Reporting Corporation (202) 628-488H 1
I
i 158 1
MR. WEIDENilAMER:
No.
Okay, then what about c
2
-future years.
This is where I guess that I wanted to 3
expound a little bit.
A few months ago,-I had to identify 4
where the areas of work that perhaps we could be looking 5
into with regard to equipment operability, Generic l
6' Issue 87 type work.
7 But with these decreasing budgets that we have to 8
face every year, we go through all kinds of scenarios on 9
budget cuts and so on.
It is becoming increasingly more 10
. difficult to get any justification for work, unless there in 11 somebody within the NRC who can identify a serious need.
12 And certainly, that would apply to ACRS as well.
If there 13 is a concern, please by all means let us know, and I believe 14 that you are doing that today.
And we are attempting or
( ) 15 will attempt to look et these as best we can.
16 But these are some of tne areas that I have been 17 looking into or try to ident11'y'or have identified.
18 Licensing concerns that may develop from utility responsea 19 when complying with these two genoric issues.
Certainly, 20 something might como up where 6 licensing need may be 21 understand something more clearly.
That is a potential area 22 1 think.
23 HR. MICllELSON:
Well, that raises an interesting 24 point, of course, that I wondered about in looking at the 1
25 generic letter, where it says that the licensee has to Heritago Heporting Corporation
}
(202) 628-4888 a
159 1~
establish the design basis requirements for the valve and so 2
forth before he decides how to do the switch set testing and 3
whatever.
4 Where the licensee may lack sufficient information 5
with which to do a proper job of determining the design l
6 basis requirement, such as he he
. a valve that has an l
7 isolated pipe break, but he does not know what those pip.o 8
> break flow effects are on the va).ve, who is responsible to 1
l 9
determino what those effects are, so that he can properly do 10 his switch settings or whatever, is that an industry-11 responsibility or-? NRC responsibility?
12 MR. WEIDENilAMER:
That is an interesting point, 13 because that is a losue that I wanted to develop.
I might 14 as well mention it right now.
( ) 15 MR. MICl!ELSON :
Oh, okay.
16 MR. WEIDEullA!!ER:
EPRI is in the process of 17 starting a v.11ve research progrrm.
Are you aware of that?
18 MR. MICl!ELSON:
No, they are probably going to 19 tell us that tomorrow.
20 MR. WEIDEN!! AMER:
Well, I do not know if they will 21 or not. I do not want to speculate on it too much.
But 1 22 can give you a briof no much as I know about it.
They havo 23 put out a RFP for bidders to work on next year-Their 24 fiscal year is the calendar year.
And starting in 1989, 25 January of 1909, they are going to fund someone to go and lieritago Reporting Corporation (202) 628-4880
[
?
160 1
evaluate all of the valve research work,.related work, that 2
has been conducted thus far.'
3 And having learned about that, I went and talked 4
to the EPRI fellow, and I told him,that we have done a 5
considerable amount of work, and certainly you should be 6
aware of this.
And I told them about our purged work, the 7
purge valve work, everything that you heard today.
And he 8
was not totally aware of that, but I think that ufter 9
talking with him and I identified the~ documents.
10 Anyhow he will hir? this agency escentially to develop a matrix on what has been done.
And within that 11 12 matrix, he will probably identify some open blocks.
And 13 that is where they would hope to them devote their monies t
14 for the subsequent years.
?( ) 15 Now the objective of their program though is to f
16 develop the procedures or criterit that they could use to 17 show or have the utilities use to demonstrate that their 18 valves will close without insitu testing.
Now that is the 19 way that I understand their objectivi.
l 10 So if they wish to indeed do that, that implies 21 that it soems to be an analytical approach.
And so their
\\
0 22 subsequent money expenditure then would be trying to do 23 that, filling in that matrix.
[
[
24 I do not know how many millions of dollars that t
25 they are going to have.
But it is supposedly quite a bit of
[
lieritage Reporting Corporation
(}
(202) 628-4888 i
I
.. ~
,.-.- -.... -. -.. ~. - -. = _,
t 161 1
money, and perhaps he will be able to tell you tomorrow.
I) 2 MR. MICHELSON:
That would be worth a lot of money 3
if you could make it go.
4 MR. WEIDEN!! AMER:
Yes.
But I doubt that that is 5
going to be possible for all of the valves that he needs to 6
show.
So I want to tell you about that, and tell you people 7
also that with decreasing budgets that cooperation and 7
8 cooperative efforts must be done.
I mean we must pursue 9
these things, and it is expensive to perform tests.
And 10 even if we get $600,000 a year in our future work on valves, i
11 that is essentially nothing if we nave to do any testing.
12 MR. MICllELSON:
Well, the thing that would be a 13 little disturbing though about the EPRI epproach is that by 14
-the time that they get this matrix developed which takes
(]) 15 several months, by the time that they devise a means of i
16 verifying operability without testing that several yeara
.i 17 might go by.
18 MR. WEIDENilAMER Perhaps.
19 MR. MfC11ELSON:
In the meantina, the generic 20 letter is requirlog that you come up with itrm proposals and l
l 21 be carrying it out within about three years.
The tir. ting l
22 might be jest a little bit difficult to achievo.
But
{
i-23 certainly, they can give it a try.
24 MR. WEIDEHIIAMER:
Yes.
I am sure that they have 25 thought about it.
I again cannot speculate.
l lieritage Reporting Corporation y
n (202) 628-4888
{
U i
4
162 1
KR. MICilELSON:
A great deal of thought was given
^
2 to this subject for a number of years all the way back to 3
when that ASME was outstanding on how do you do these things 4
with a minimum amount of testing, and a minimum amount of 5
data even.
6 MR. WOHLD:
Gerry, can I ask a question there?
7 MR. WEIDENilAMER:
Sore.
8 MR. WOllLD:
When you say no insitu testing, do you 9
mean no. flow and differential pressure testing, but doing 10 the valve thrust determination?
11 MR. WEIDSNilAMER:
It is not clear thet that is e
12 even involved.
I cannot see how they are going to 13 accomplish that without something like that, Pete.
14 MR. WOllLD:
Okay.
l
( ) 15 MR. WEIDENIIAMER:
It was not clear as to what they 16 really meant without insitu testing.
I think that perhaps 17 there are various degrees of that, and certainly diagnostics 18 I think would be one of the desirable features or at least 19 solcothing that they would not want to overlook.
That would 20 be the easieut.
21 MR. MICllELSON :
I think that they are just trying 22 to minimize the insitu effort.
23 MR. WEIDENilAMER:
Yes, I think so.
24 MR. MICllELSON:
That is a highly desirable thing 25 to do.
Ileritago Reporting Corporation
[}
(202) 628-4888
il u
163
,1 MR. WOHLDr Sure.
I0 2
MR. MICilELSON:
But how do you do it, clearly you 3-have to do some amount of testing somewhere and then
'4 extrapolate or whatever.
5 MR. WEIDENilAMER:
Those are the words that are 6
written down.
And how that is interpreted, I am merely f
7 stating.
8 How do we extrapolate BWR to PWR environments, are 9
there valves that are similar to Generic Issue 87 valves 10 that we have to be concerned about in PWR plants.
What 11 about globe valves.
They are some globe valves that are 12 important.
Ana the same formula applies.
Ilow does one use 13 that formula, if certainly they are globe valves.
[
14 Determination of butterfly actuator requirements I
I ) 13 necessary for venting.
We talked about purging, and'what I
16 about venting.
And by the same token, what about some of 17 your concerns that you raised this morning on the effects'of 18 one valve on another and so on.
I 19 Determination of the effects of replacing pl.pe l
20 anubbers.
This is not valve work, so I will not mention 21 that one.
And thoro is some maintenance of liaison 22 activities that we would have to fulfill in keeping abreast j
23 of proL:.. sis, interacting perhaps with EPRI.
24 Those a.co just a few items that I think are 25 important in addition to some of the ones that were raised 1
i lloritage Reporting Corporation
[}
(202) 628-4888
164 1
this morning.
.,~
(_)
2 MR. MICllELSON:
Nearly all of these activities 3
though would appear to be directed towards the motor 4
operated valve question, is that correct?
5 MR. WEIDENilAMER :
That is correct.
Also I have 6
heard instances of air problems, air operator problems.
7 MR. MICllLLSON:
How about the case of the air 8
operated valves, do we have any kind of research program 9
underway or anticipated to be underway in fiscal year 10 1989?
11 MR. WEIDENilAMER:
No.
12 MR. MICilELSON:
And a butterfly valve perhaps a 13 little bit to the extent that it was touched on in here.
14 MR. WEIDENilAMER:
Yes.
But this butterfly work O IS w uld be n thing more than along the lines of trying to 16 identify what it is that we are going to have to do.
We 11 have done some work in butterfly valves.
18 MR. MICl!ELSON:
Some very fine work I thought was 19 done on tha butterfly valves, but i t, was In a way ;ust a 20 good start.
Certainly, it is the best that I have seen 21 anywhere, at least published on the butterfly valves.
13u t 22 there is a fair distance to go in that area.
23 MR. WEIDENilAMER:
I agree.
But if we have te get 24 into testing, this is where tho big money is needed.
25 MR. MICIIELGON:
Well, I can foresee.
You know, we lieritage Reporting Corporation (202) 628-4800
\\
c.
4 165
- will have to start exploring the air operated valve 1
2 situation.
It is cenceivable that it may also involve some a
3 extensive effort before it is done.
I just do not know.
4 MR. WEIDENHAMER:
The same here.
I do not'know 5
either.
I have been talking to some of the people.
6 Ed Baker for one, I am not sure what branch he works in, but 7
he has indicated that there are some potential problems 8
there in the air operated.
9 MR. MICHELSON:
A lot of what you are doing that 6
F 10 you are doing for motor operated will be equally usable on 11 air operated situations.
The operator itself becomes 12 somewhat the unknown in its performance in the case of the 13 air operators, because of quite a wide variety.
I am not i
14 thinking so much of the diaphragm operator as I am on some
() 15 of the piston operatore and some of the push pull air' 16 operators.
And of course, the solenoid air sets.
They are I
17 just as important as the diaphragm operator itself.
j 18 And there are a lot of disturbing LERs in the air j
19 system business.
And right nmi, we have no anticipated 20 research.
Hopefully today, we will hear a little about test 21 techniques, and tomorrow about test techniques for air 22 operated valves, so that we can better understand how well 23 they do work, and whether they are indeed misbehaving, i
24 Dec6use the motor operators were somewhat in the same ball
[
[
25 park a couple of years or more ago before we had good Ileritano Hoporting Corporation i
(}
(202) 628-4888 L
.,, -. -, - -, -. - - -. - -... - - - - -,, _. - - - -, - -..,,, - - -, ~,
166 1
techniques for determining how well they were working..
2' Is there any of this work related to what is going
^
3 on'or'has been going on at Oak Ridge on' motor operated 4
valves?.
5 MR. WEIDENilAMER:
Yes.
6 MR. MICHELSCBI It is a different FIN though, is 7
it not?
q l
8 MR. WEIDENilAMER: ' Yes.
But Dave Eisenberg and any 9
number of people within NRC have all been factored in and-10 put into this research.
[
11.
MR. MICHELSON:
But I am wondering what other j
12 valve related research is going on, in other words besides 13 this particular FIN number.
I assume that the Oak Ridge' 14 work is still continuing.
( ) 15 MR. WEIDENHAMER:
Yes, Oak Ridge is still 16 continuing.
17 MF.. MICHELSON:
On the same level?
i s
18 MR. WEIDENIIAMER:
Yes, I believe that it is about 19 the same.
It is under the aging program.
And there may ba 1
q 20
' sono other work.
21 MR. MICllELSON:
Are they-going to be doing 22 anything on air operated valves?
j I
23 MR. WEIDENilAMER:
I do not believe so.
f 24 MR. MICllELSON:
Just on the motor operators?
i 25 MR. WEIDENIIAMER:
Yes.
I 1
i i
fleritage Reporting Corporation f
(202) 628-4000 s
I f
,s
.--,-.~~-..,.....-----i
167 1
MR. MICilELSON:
Another smaller area relating to
(~\\
(-)
2 motor operators are these rotorque operators which 3
constitute about ten percent of the population, but have 4
some unique characteristics that might be important to 5
understand.
6 But we do not have any research activity underway 7
there?
8 MR. WEIDENilAMER:
No, not to my knowledge.
I do 9
not, no.
10 MR. MICilELSON:
There may be that no research is 11 needed, that one can get the understanding of the need just 12 by studying the mechanisms and detail?
13 MR. WEIDENilAMER:
Well, I am asking Idaho to start 14 looking into these various problems and identifying them, so
( ).15 that we can sit down and start putting our efforts into the 16 areas that we feel are important.
And if rotorque turns out 17 to be an area, certainly.
I guese that any work that we do 10 in *;he future, I certainly would like to bring it back here 19 and let you people know about it.
20 MR. MICl!ELSON; W'll, we certainly are anxious to 21 hear about it, of courso.
Jt is a subject that wa would not 22 be dealing with so much jf we were not really intorested in 23 it.
We are trying to be assured that things are moving 24 along well.
25 MR. WEIDENilAMER :
That is all that I have to say.
Iloritage Roporting Corporation
{}
(202) 620-4000
168 1
MR. MICllE: iOti:
And a somewhat related question I(,)
2 before we see what the other members have.
In the past, I 3
have wondered whether or not, or let me say it differently.
4 I am wondering how close to having a good test one has if he 5
simply c > ens the gate valves against the full differential 6
pressure.
7 How close to being a challenging load is that 0
compared with closing under a full differential pressure 9
condition?
I never this morning quite got the feel for the 10 difference.
11 MR. W E I D E t111A M E R :
It is significant.
I think that 12 au is something like twenty percent.
13 MR. MICiiELSOti:
A quite bit bigger load to close 14 than to open, is that any rule of thumb or feel for it?
Lj) 15 MR. WEIDE!111 AMER :
Yes.
'6 MR. KIESSEL:
When you say that it is opening 27 against DPRP with just a little jock pumpe to build up the 18 pressure?
19 MR. MICllELSOfit tio, no.
20 MR. KIESSEL:
I have actually got my rystem pumps.
21 MR. MICllELSott:
The RIIR, whatever, and apply the 22 full differential pressure and that breaks the valve open 23 which is the real worl(1 of ECCCS which havo really big 24 breaks that pumps up the system before the vc1ves are 25 necessarily opened?
1:ar.itage Reporting Corporation r~s (202) 628-4888
(_)
169 1
MR. KIESSEL:
Some of the very early information
()
2 coming out'of the Wylie test.
I feel awkward in saying 3
this, because I am trying to quote what they told us was 4
their very, very preliminary information. I think that he i
5 was saying something in the neighborhood of about twenty 6
porcent more thrust is requirod to close the valve than to 7
open it.
8 Now what all of these parameters involved, whether 9
or not the twenty percent included the reversal of the stem 10 rejection load and any of the thermal binding effects, I 11 really do not know.
12 MR. MICllELSON:
I think that there is a worthwhile 1
l 13 reason to think about this hard, and the reason is the I
l 14 opening against full differential pressure is a fairly easy
(]) 15 test to set up in the shop or insitu as opposed to closing.
16 It is a much easiet test to set up, because I do not have to 1
17 have large dynamic flows or r.nything to look at the l'
18 breakaway torque.
And if there were some kind of magic l
19 relationship that was fairly dependable like twenty percent, i
20 I could just tuho my full dif ferential oponing numbers and j
21 multiply them by 1.2, and I would know what thrust that I
)
22 needed from the motor operator, that sort of thing to do.
I 23 was just wondering if that was even possible.
24 MR. KIESSEL:
This, of course, the motor operator i
25 gots to fill in --
1 j
lloritage Hoporting Corporation j
(202) 620-4000 i
i
)
L
170 1
MR. MICllELS011:
That is right.
That is just C)
(/
2 something to think about.
In the very old days before 1 3
became educated a little bit, I was generally told by most 4
good valve manufacturers that it always takes more to open 5
against full differential than it does to close.
But of 6
course, they did not know, because they had not done the 7
test.
They just thought that it did.
And the fact is that 0
TVA for one opened the valves requiring the full 9
difforential opening test in the shop.
10 MR. KIESSEL:
Do not forget, Carl, I do not know 11 if this is buried in the twenty percent or not.
But when 12 you go to open, your motor has already come up to full 13 speed.
It is delivered as a hammer blow to the stem to get 14 it moving, so that you have broken away any binding that O 15 1 at ne tecurriov trem the ete to the vecx1uu, eaa eneo 9
16 the stem has a chance to start coming up to speed before it 17 delivero first the hammer blow to the disk which gives you 18 some nice momentum offects to produce that unseating.
19 The valve vendor is looking at how much werryy ed 20 how much force it in taking to pull that thing out of the 21 seat.
It is probably more than what was required to try to 22 bring it in.
Bringing it in, I have got the momentum of the i
23 whole system coasting in and crushing out.
And so I am 24 baing helped by that.
IJow I am just talking about looking 25 at the actual seating.
I am not messing around with the lloritaqo Itoporting Corporation (202) 620-4000 l
171 1
differential pressures and this phenomenon that we are
(()
2 seeing.
So from the valve vendor, and that may still be a 3
truthful statement that it takes more thrust to break the 4
valve through.
5 MR. MICllELS011:
I think that it probably does, 6
except for these closing effects.
That is why we are doing 7
the test to make sure that there were not any closing 8
effects.
And it turns out that yes, there were some.
The 9
fact is that they seem to indicate that it takes a great 10 thrust to close them.
11 MR. KIESSEL:
Obviously, they knew that the stem 12 rejection load was working.
13 MR. MICilELS0 tis But if that ratio were a constant 14 known number, then I would not have to do the closing test,
( ) 15 and I could just do the open test.
16 MR. KIESSEL:
I have a fooling though that since 17 you are talking about really three turr.s that are very --
19 and one them is a differential pressure turn, and one is r.
19 nystem pressure turn, and one is a constant.
20 MR. WEI DEllll AMER:
tio, that in not necessarily 21 truo.
Wo do get a differonco.
22 MR. KIESSEL:
I mean that it may not be possible 23 to como up with a magic number.
24 MR. MICl(ELSOll You understand what I am thinking 25 about.
And the question is sometime when you are doing some lieritage Iteporting Corporation (202) 620-40811
172 l
1 of this research,'maybe somebody could spend more time than
)
2 I have had to do it, and see if there is any way in which 3
you could do valid opening tests only, and know from that 4
what your closing thrust requirements might be.
I do not 5
know if it is even possible.
i 6
MR..WOHLD:
Carl, can I comment on it or do you 4
7 want to say something?
j 8
MR. MICHELSON:
Sure, go ahead.
9 MR. WOHLD:
There's a lot of, I expect that there 10 is a lot of data that is already available in the industry 11 as a result of the Bulletin 85.03 testing.
A number.of 12 plants have done full flow differential pressure tests, both 13 opening and closing.
And if you can look at the data, you 14 can make those correlations.
And I am familiar with one set.
q
-( ) 15 of valves which was the Davis-Bessie valves that cauced the
[
16 Bulletin 599 and 608.
There was a very nice relationship j
17 that came out of that.
And I think that it would be very
)
18 lateresting to look come more at that.
i 19 MR. MICHELSON:
Is it true though that probably it 4
a 20 is oasler to do the opening test than it is to do the full 21 flow closing test?
22 MR. WOHLD:
Definitely.
a i
23 MR. MICHELSON:
Because a lot of systems just are j
g i
24 not set up to do much flow.
t i
25 MR. WOHLD:
Well, that is an availability problem
(
i I
t 1.
Heritcar Roporting Corporation r
{}
(202) 628-4800 i
s
e 173 1
too.
You can do the opening tests with a small jocky pump
()
2 and not have to align the entire plant for the test.
And 3
Bessie did a test.
If you get past the cracking point and 4
look at the sliding friction point, it is only a very small 5
portion of travel with the jocky pump.
But you can have a 6
thrust requirement measured under that sliding friction 7
opening that goes away, but you can pick it off in their 8
case in the MOVATS trace.
9 If you add two times the stem rejection force to 10 that, you come very, very close to the closure requirement 11 of the valve.
And it was interesting.
12 MR. MICIIELSON:
But there was the closure 13 requirement under normal flow?
14 MR. WOHLD:
Under full flow.
( ) 15 MR. MICllCLSON:
Full flow, not under 16 blow-down necessarily?
17 MR. WollLD:
Not under blow-down.
But if you look 18 at the differential pcessure developed dur ng the closure, 19 that differential -~ wait a minute.
The differantial 20 pressure thing is the f.riction t.hing where you determirn an 21 opening and you assume that it is the same at closing, okay.
22 They measured it on opening.
If you add two times 23 the stem rejection force, you can get the closing force 24 requirement. It turned out very nicely on Dauis-Bossie's l
25 valves.
It turned out that the stam rejection force was tieritago iteporting Co rpora t. ion (202) 628-4000 O
-a
174 1
about 2500 pounds going from memory.
Twice that is
()
2 5000.
So it took 5000 more pounds to close it than it did 3
to open it in the sliding friction regime.
4 MR. MICHELSON:
So I would hope that the industry 5
would try to find some of these neat little correlations, if 6
they can be validated, and make full use of them to reduce 7
the amount of insitu testing that is required.
8 MR. W0llLD:
Right.
9 MR. MICilELSON:
Maybe EPRI will eventually get 10 around to it.
11 MR. NEIDENilAMER:
Well, I forgot to mention that 12 as part of the Generic Issue 87 that we intend to see what 13 we can do along those lines too, if there is some procedure 14 that we can como up with that we can use.
( ) 15 MR. MICllELSON:
Is that a part of this FIN?
16 MR. WEIDENilAMER:
Yes, it is part of 17 Generic Issue 87.
I just forgot to write it dowri.
10 MR. MICl!ELSON:
So INEL is going to look a little 19 hit at what does this data mean and perhaps what they glean 20 from it in terms of simplification of test requirements.
21 MR. WEI DEHil AMER:
Yes.
The opening and closing 22 relationship is certainly one of them.
23 MR. WollLD:
Gerry, I think that you could get some 24 of that data simply by asking some of the utilities.
I 25 think that they would be glad to provide it.
Iloritage Reporting Corporation (202) 620-4000
175 1
MR. WEIDENIIAMER:
We have the Davis-Bessie work.
f 2
We have been heavily involved in looking at that.
I think 3
that you probably already know that.
Kevin's or INEL's work 4
when they first started the Generic Issue 87 effort obtained 5
that data.
And I believe that.*nder the nom.-
_ flow 6
conditions that two times the stem rejection load was a good i
7 relationship.
But when you come to the saturation point or s
O get close to that saturation point and have a blow-down, 9
that is the concern that we have.
l 10 MR. WOHLD:
Right.
It would not account for that
[
t 11 difference.
[
12 MR. MICllELSON:
Then you have to work out some
[
13 other good factors to account for the blow-down phenomenon 14 short of doing an awful lot of testing, if you could.
O 15 MR. WE1DEalinMER:
Thee is r1 ht.
9 i
j 16 MR. MICHELSON:
Are there any other questions on
{
17 the research program?
l 10 MR. W0llLD:
I have got one, Carl.
Earlier this 19 morning, and this 10 going back and is not related to what 20 you are talking about right now, Gerry.
The determination 21 apparently was that the friction factor used in the old 22 thrust calculations was
.3, and it looks like in many cases 23 that it is.5 and even greater.
24 One of the things that has happened in the 25 diagnostic testing is that people have been able to set the lieritago Itoporting Corporation (202) 620-4000
176 1
thrust exactly to the manufacturer's recommendations which
(' )-
2 was
.3.
Unfortunately, that removes, if overything oise was s.
3 done correctly, that would remove some of the conservatism 4
in the way that the valves were set up in the past, where 5
there were conservative friction factors and so forth 6
assumed in the stem factor I think it is, I am not familiar 7
with the exact calculations.
8 But in setting it exactly to the manufacturer's 9
recommanded thrust value, thoro has been some conservatism 10 taken out in the volve capability, and that looks like a 11 major concern resulting from your tesuing data.
12 So for instance, in the extension of the bulletin, 13 what is the requirement going to be as far as setting the 14 thrust.
If a differential pressure test cannot be dono,
( ) 15 which is in most of the cases, what thrust value is going to 16 bo selected, what now friction coefficient is going to bo 17 assumed, and that is a mnjor concern.
18 I know in the bulletin the manufacturor's 19 recommended thrust values woro used which is the bottom line 20 for NRC acceptanco.
21 MR. MICilELSON:
They may not havo used
.5.
22 MR. WOllLD:
Well, no.
They used
.3.
Dick.
j 23 HR. KIESSEL:
It was not our intention to simply l
24 accept blindly the manufacturer's thrast values.
The 1
l 25 omphasis of the bulletin was to verify by test that you h.id llor i t.ayo Ituporting Corporation (202) 628-488tl
177 l
1 set the valves properly, period.
If you could not do that,
(~)T x..
2 we were then allowing engineering justification based on 3
other testing that had been done.
4 My understanding of the general technique as 5
proposed by Calloway which in sometimes called the MOVATS 6
technique was that they had established a trend line based L
7 on observed data of the influence of justification.
And 8
that this f act or this ef fect was then added to the kne- '
9 pressure cap as a result to come up with the value.
Simply 10 verifying that you have got your switches set so that you 11 can deliver the manufacturer's recommended numbers of thrust 12 was not the intent of an acceptable program.
13 (Continued on next page.)
ditasr14
( ) 15 16 r
17 10
{
19 20 21 22
+
23 24 f
25
[
Iloritago Hoporting Corporation (202) 620-4000 1
0
I I
178 T2tesr 1 MR. WOllLD:
I do not know.
f w)
(_
2 MR. KIESSEL:
I expressed that several times at 3
the meeting with industry that we are questioning the 4
validity of those calculated thrust.
5 MR. WOHLD:
I think Clinton la one plant probably 6
set their valves to the manufacturer's recommended thrust. I 7
am going from memory again.
8 MR. KIESSEL:
And you let them get away with it.
I 9
MR. WollLD:
What else did I have to go by?
They 10 did differential pressure tests, and presumably showed that 11 they operated on a differential pressure, but they used 5
12 jackie pumps.
And the torque switch, of course, is bypassed 13 during that testing.
So they said that they had 14 differential pressure testing using MOVATS, but they had no
( ) 15 measurements.
And the closure thrust and the torque 16 switches were otherwise set as I recall to vendor 17 recommended thrust values.
These are the kinds of things that you get into when you see the details of how the tests 18 1
19 respond, when both responses.are implemented in the field.
20 There are a lot of variations that you do not recognize in 21 the response.
22 HR. WYbIE:
Gerry, has there been any attempt to j
23 see what may be learned from the foreign valve test program?
f i
Yes.
No have been talking with
[
25 the CEGB, the Central Electricity Generating Board, from l
lleritage Reporting 9erporation (202) 628, J 8) 1
179 1.
England, the United Kingdom.
And they are heavily involved, 2
or they are involved in a large valve reliability program 3
for their Sizewell B PWR reactor.
And the objective of this l
4 large test program is to select the most reliable valves 5
that they can to make their shutdowns and so on, to minimize 6
their shutdown to make it a more efficient plant.
7 Go yes, they are the largest one right now that is 8
doing this kind of work.
As a matter of fact, I was able to 9
talk one of their people to como over here and make a 10 presentation at the Water Reactor Safety meeting that is 11 going on this week.
And they have agreed to information 12 exchange with us.
And I told that after we once get some of 13 this data down and reviewing data tapes that they can come 14 over here and we can talk to them.
()15 So yes, we have interacted with at least those 16 people.
As part of looking around some of the areas to see 17 who could do some valve testing, during the HDR testing in 18 Germany, I visited one of the large ".11ve test facilities, 19 and they have conducted a number of blow-down tests and are 20 recognized experts in this area.
And we fully intend to go 21 back to them and talk to them as well.
And yes, we are 22 trying to do that.
23 MR. MICilELSoll:
Did the CEGB do any work on air 24 operated valves, do you know?
25 MR. WEIDEllllAMER:
I cannot remember, Carl.
They lleritage Heporting Corporation
(}
(202) 628-4888
t 180 1
have got such a large test program.
()
2 MR. MICilELSON:
They must have a fair number of 3
air operated valves.
4 MR. WEIDEllllAMER:
They could very well have done 5
some of that, yes.
6 MR. MICllELSOtt:
It is typical at least.
7 Do you know of any other European experience or 8
possibly any other European work on air operated valves?
9 liR. WEIDEti!! AMER :
No, I do not.
You seem to be
{
10 directing a lot of concern toward the air operated valves.
1 11 MR. MICilELSOll:
Well, no, it is just that we have 12 an interest, not necessarily a concern.
We just want to l
a f
13 look into it.
14 MR. WEIDENit AMER :
Okay.
1
( ) 15 MR. WYI.I E :
As I recall, the Germans used a lot of 16 DC operated valves.
They should have a pretty gcod data 17 base on valve experience with DC operated valves.
Of 10 course, as recall, their voltages were quite lov, 19 40 volta and thi:1gs like that.
20 MR. MICllELSOft:
Okay.
I believe that takes care 21 of the questions then.
Thank you very much, Gerry.
i 22 According to our agenda sheet, I believe that l
23 Mr. Thibault of Wylu Laboratories is next
.o n, and he is 24 going to tell us about his valve testing techniques.
25 (Slides being shown.)
r l
Ileritage Reporting Corporation t
(202) 620-48011 O
f i
181 1
MR. THIBAULT Good afternoon.
My name is 2
Claude Thibault with Wyle Labs, and I am manager of nuclear 3
products.
Part of the activity in my area is to develop a 4
new system.
Wyle has been involved in testing generally 5
speaking for many years, I would say 45 and more years.
And 6
one of our testing areas on the valvo is on the SRV.
7 So the logical trend of Wyle was when the time 8
came to try also to get into the business of the MOV.
That l
9 was two or three years ago.
But before that, we were called I
10 to qualify those valves on an EQ point of view and 11 especially the actuators.
So the experience again of Wyle 12 in the EQ business at that time was highly involved.
We i
13 knew Limitor fairly well I believe.
And at that time, we 14 decided to try to implement a now system for diagnose this
()
15 valve.
16 Our diagnostic system today is comprised of two 17 complementary segment.
One that we call MCSA is the Oak l
18 Ridge system that was developed by the request of, I i
19 believe, the NRC.
20 I know the purpose of my talk here is not to talk
[
21 about MCSA.
However, I would like to refresh your memory on
{
22 that MCSA, and as I said this morning, the current signature 23 in two ways: on a time basis; and on a frequency basis.
24 The way the other system is analyzing the current 25 does not go into that depth of analysis.
MCSA advantages, I
fler i t.n ge Heporting Corporation j
(202) 620-4000 P
4
m 182 1
however, is it's remote, you don't need to go to the valve
<^
(,)
2 itself or in the containment or in some very bad area to 3
diagnose the valve, you do that from the MCC.
And it's a 4
very popular system.
And you can ane'.yze it after that.
5 The analyses, however, require a lot of i
6 experience.
And to me or to Wylie, I would say, motor 7
current signature analyses we have a tendency to give a l
l 8
trend more than a baseline.
So our feeling is to have a 9
system which will give us the baseline, the thrust that you 10 a.'o talking about, the setting of the switch, even it Oak 11 Ridge can detect also the setting of the switch.
That's why 12 we developed the valve motor operator diagnostic system 13 which we call the baseline.
14 I'm acing to go fnster on this one because 1 would
(])15 like to tolk more about the schematic of it.
We got mainly 16 five components cansors: the interconnect box, the cables 17 sets, the data acquisition moduie, and the 10 computerisoftwaro/ display.
19 That was rieveloped about three years ago anc! when 20 we went to the load washer we talked, we learned, we 21 discussed c lot, shall we go to load washer, load cell or 22 strain gauge.
23 We had to see also the practicality of it and we 24 decided to go to load washer which was probably the more 25 difficult part to go, because we had to decide where to put lloritage Reporting Corporation (202) 620-4880 O
v 183 1
them to be reliable, repeatable, and precise.
()
2 So for about two and a half years we had a long l
3 research program to find out the position of the load washer 4
and also how we are going to torque them up, and then to 5
calibrate that.
6 Let's look at the schematic of the baseline system 7
which is V-MOD.
It's going to be a little bit better that O
way.
I'm sorry.
9 V-MOD is comprised of two set of sensors; two sets 10 of interconnect box; and umbilical which can go to 250 feet.
11 WJth that acquisition module compute very fair what you want I
j 12 to put on.
13 Why we went to two sensors set, because during the i
14 time we mounted one valve one of our guy can prepare the
()15 second valve with the interconnect box and the sensors, and 16 when that valve is finished to be maunted, we take that off 17 and go to the third valve and we start mounting this one by 10 just switching the umbilical cable to the other interconnect 19 box and try to limit the time.
20 We are going to talk a little bit every of them.
21 I would say that th" sensors are comprised of armprobes, 22 like everybody does, they take the switching.
We have 23 armprobes.
We have as many as four armprobes.
And also we 24 have ammeter clamp to detect voltage drop.
25 If you look at the schematic of a limit actuator lloritage Iteporting Corporation (202) 6 2 0-4 81111
184 1
you will-have very often two cwitch in power.
Nobody tell
)
2 mo.that the two switch are going to fall together at the 3
same time or to switch together at the same time.
That's 4
why we have that redundancy in our system.
5 Two of the armprobe on the torque switch open and 6
close, and two on the bypass switch umbilical, the other 7
clamp on the limit switch.
8 The load washer -- I will stop to tell you a 9
little bit, it's a long story, but many there are charged 10 sensitive, so they need to charge them before.
And they 11 don't know what happened before.
They measure a transition.
12 So I apply a load in the load washer sooner or later it's 13 going to lose that load.
14 That means that we cro looking at -- what I'm
( ) 15 trying to say it's if we want to measure the thrust going 16 closed than going open, it's a different ball gama geing 17 open because you cannot free stress your load washur and la say, this will be my :ero and then I can go compression and 19 extension.
20 So we -- the exporlence again wo had, we said 21 let's investigato the closing on tho stom.
So wo decided 22 where to put them and how to put them and all those reports 23 aro confined in two Wy110's reports that wo uso always when 24 we go in the field.
25 Wo testeel repeatability on dif f erent cetuators, Horitago Hoporting Corporation (202) 620-4000 4
185 1
most of them SMB, I would say, because 6-to 90 percent of 2
the actuator safety are related SMB.
1 know that we have 3
some torque or othee one.
But that's the way Wylie went.
4 We wanted to look at the repeatability, and the 5
repeatability was accepted, tiow let's look at the 6
calibration and also the procedure and accuracy.
7 We advertise plus or minus 10 percent of the load 8
measurement knowing and confirmed by the test report that we 9
know it's plus or minus five, but we want to be on the safe 10 side.
Those load washers were calibrated against a load 11 cell which was traceable to the U.S. National Standard.
12 MR. MICilELSott What does a load washer look like?
13 MR. THIBAULT:
It looks like, I don't have a 14 picture of it, but it icoka like a wather about t hat thick
( ) 15 and the diamotor UD and ID vary depending on the load that 16 you expect to apply to it.
17 MR. MICilELSCd How did you mount it and where did 10 you nount it' 19 Mi rilIDAULT:
That was pat t of t.he test, we mount.
20 them -- let'r see if I have a picture of an actuator here, 21 maybe.
I don't know if we can see that on here.
tio, wo 22 cannot see it.
The top plate of the rim you got eight, 12 23 bolt.
We mount them -- well, it depends on the actuator, 24 the position dependn upon the actuator, but let's take SMB-2 25 which we test during the blowdown test.
We mount them on an lieritage Itepo rtir:g Corpcration
(}
(202) 620-4009
~
a_
186 1
axle that is part to the axle of the riotor.
(
Because you've got a lot of unbalanced weight in 2
3 cn actuator.
Then we torque them to a certain torque and 4
that torque has been exactly the size -- I would say, test 5
it to find out which --
6 MR. MICllELSON:
llow are the attached?
7 MR. TI{IBAULT We took two bolt out, two longer L
0 bolt and bolt them back on the actuator to that preset 9
torque for that type of actuator and load washer.
10 MR. MICilELSON:
And 100 degrees apart?
11 MR. Ti!IBAULT:
Yes, sir, most of it for the SMB-2, 12 SMB also I believe.
Jack?
13 I would like just to show you h ne the interconnect l
14 box look like.
This is a box which is closed from the valve
()15 where all the armprobe and the cla.np comes in, the load 16 washer cor,<os in the back.
And from here we got the 17 umbilical going to the data acquisition module.
18 That acquisition module is this one, and again, we 19 have the charge amplifier and we got the vult meter.
You 20 can set different calibration because you want to know hov 21 many volt per inch, ( c volt per pound or how much pound you 22 have and all that has been calibrated.
23 Let's say now that what parameter that we measure 24 with the baseline V-MOD, I would say like aost of our 25 ccmpetitor we measure the same thing.
We measure the stem lieritage Reporting Corporation
(}
(202) 628-4888
i 187 1
travel.
We measure the stem load.
And I said, directly not
()
2 through a third party which can be the spring pack that wo 3
also measure.
A strain gauge that is on the yoke, and I 4
would like to soo that strain gaugo personally on the stem 5
because that's where it belongs.
6 So I would say that the stem load that we meacure 7
is direct because -- but we measure it only one way, going 8
close.
The motor current is mainly an armprobe like 9
ovorybody.
10 We try to take care of one thing that you 11 montioned this morning which is an ovaluation of voltago.
I 12 cannot with that system accurately say that during the 13 closing of or the opening of the valve the voltage change.
14 Ilowever, I can say that with MCSA.
When you go to the swim
(])15 analysis with MCSA you can throw back your torque 16 characteristic of your motor.
It takes a lot of time to do l
17 that, but it can be done.
I 18 We moasure the spring pack displacement as a l
l 19 backup of the load.
I would say that the spring pack 20 displacement measure the load, that's what we were told and
(
21 we prefer to measure the load rather than the spring back; 22 however, we take the spring back also.
And like everybody 23 we measure the torque owitch, limit switch, and bypass 24 switch.
There is no mystery about that, overybody does the 25 same thing.
lieritage Heporting Corporation (202) 623-4888
('
180 1
I would like to share with you or show you -- I.
(
forgot to mention our computer which is a standard computer.
2 3
The softwaro we uso in the 14 channel.
The name of the 1
4 softwaro is commercial, it's CODAS I believe designed by 5
DATAQ.
And it allows you to acquire 14 channels 6
simultaneously and then when you play them back group them 7
by two, by four, f
8 Yes.
Oh, I'm sorry, I thought you were asking me 9
a question.
10 This is just to show you one of the -- doing our 11 test, valve that we were testing.
That's not the blowdown 12 test.
It's on the same channel.
And all those units can be 13 engineering unit and it's included in the software.
14 The load -- and I'm sorry, before I go back we are
( ) 15 going from closed to open to closed.
The load is measured 16 from the load cell.
And here you are going to tell me you 17 have the running load.
I will show you on the blowdown test 10 that we can see that running load increasing when the gate 19 goes down.
20 And the area of the gate exposed et the fica 21 increased.
And here we can put pounds and I will show you 22 later on.
We got the current signature here.
You can seo 23 your unseating.
We will blow out that during the closing, 24 and then the opening and the closing actica.
25 By the way, you can see that your current increare lloritage lleporting Corporation
(}
(202) 620-4000 i
189 1
who'n you close the valve.
Now you cannot put a value, but
()
2 you know that.
3 The spring pack displacement that should and would 4
confirm the load, liere I just displayed the open limit 5
switch, the close torque switch and the open bypass switch, 6
where you can relate your bypass switch to the ham:nerblow 7
and the unseating.
You have to cover that lapse time to be 8
sure that you don't trip.
9 I'm not trying to explain the valve, how the valvo j
10 works; I'm just trying to explain what we measure.
I t
11 Capability of the software to magnify both ways, X and Y, 12 the plot.
Here I show the current, the spring pack, the 13 open limit switch and the open bypass.
And again here you 14 can see bottor that your timing here is okay for that valvo.
15 Another set that I group, you group them by four, 16 This is the closing of that valve where you see on trace one 17 the load, the closing load and that was an SMB-2, I believe, 18 Jack That closing load was 37,512 pound.
And I say that l
19 is plus or minus 10 percont knowing currently that it's plus 20 or minus five.
Your current, spring pack, and cloao torque.
21 As a matter of fact, if you look at that valve we 22 closed on limit, we did not close on torque.
That was a 23 test.
24 I wanted to show you one of the blowdown test that 25 we gave to Kevin.
I want to tell you which test it is i
l Ilor t tage Reporting Corporation (202) 628-4800
190 1
because Kevin did not finish his report yet, but that was 2
during a blowdown when you nave the stem travel, you're 3
running -- the valve was about half open and we were going 4
closed.
5 You can see your running load and when the area of 6
the gates going increasing, you will see your running load i
7 increasing up to the point where you close.
8 And that's the main reason why I wanted to show 9
that slide because also it's reflected by the spring pack, 10 by the way.
But I have a direct reading of the pound force 11 without going to the spring pack.
12 The advantage that Wylie sees in our system is i
13 that it's a direct measurement of the closing thrust, as I i
14 said before, without referring to a third party.
It's
()15 always difficult when you have to go to a containment to put 16 something, but it's quite easy to install this quite fast, f
)
17 All the hardware is commercially available.
And I 10 believe, that's to me, the moret important thing is that 19 V-MOD will give us the baseline, but you don't need to go f
20 back to the valve overy time you have a checkup or
[
21 surveillance or maintenance stuff, you can go to the MCS, to i
22 the MCC and do your MCSA, because at the first time you do i
23 that you got baseline and MCSA.
The second time for 24 surveillance check.
For maintenance check you can go with l
l 25 MCSA and you'vo got all your parameters, and MCSA will give i
IIeritage Reporting Corporation i
(}
(202) 628-4800 s
I
191 1
you the training.
O2 1 tweex vo> verv eca ena 1 111 trv te ene er 3
your questions, et/2 4
(Continued on next page.)
5 6
7 8
9 10 11 12 l
13 14 O 15 16 17 18 19 20 21 22 23 24 25 l
llerita90 Reporting Corporation (2a2) 628-48aa O
I 192 t/3 1
MR. MICliELSON:
Now, what's new h, ore that isn't
()
2 already being -- wasn't already practiced by others, I don't 3
think, and correct mo if I'm wrong, that there is anything 4
new nere except putting in the --
5 MR. TilIDAULT:
The load washer.
6 MR. MICllELSON:
-- the load washer.
7 HR. TilIDAULT:
Yes.
8 MR. MICilELSON :
But other people measure that with 9
stem or --
10 MR. TilIDAULT:
Other people measure it through an 11 effect of it.
12 MR. MICllELSON:
Now other vendors of such 13 services, I don't think, have done the frequency analysis on 14 the current, necessarily.
But you haven't told me how you O 1s uee it.
16 MR. TilIDAULT:
Well, that was not part of Gerry 17
-- when Gerry called me and asked me to give you a 18 presentation of V-MOD, that was V-MOD, that was not MCSA 19 because MCSA was explained to you by Oak Ridge, 20 MR. MICliELSON:
Right.
And I understand how their 21 system works, but you haven't told me how you use that I assume you use it then to look for broken gear 22 except 23 teeth and other kinds of audities that they can pick up 24 through the current signature.
25 MR. TilIDAULT :
You can use it for broken tooth IIeritage Reporting Corporation
(}
(202) 628-4888
193 1
gear or whatever, but you can use it for everything after
)2 when you've got the base line.
You can use it for the 3
packing.
You can detect tapered stem.
4 For the frequency analyses it's done from the 5
noise in the current signature.
o 6
MR. MICHELSON:
But other than for the load washer 7
the rest of the technique has been usel in various forms
(
8 by --
9 MR. TilIBAULT:
There is nothing new, certainly.
10 MR. MICllELSON:
I just want 'to make auro I wasn't 11 missing something.
i 12 MR. TilIBAULT:
No, no, no. 300 are not missing a 13 thing.
And, Jack, correct me if I'm wrong, but I believe 14 that the main difference is the load washer.
( ) 15 MR. MICllELSON :
How what does the load washer do i
16 for you that you think is better then stem or better than 17 yoke.
10 MR. TilIBAULT:
It's a qu<3stion that was asked to i
19 me when I was asked by Wyle, how would you measure -- I come 20 from the valve business in the oil field, how would you 21 measure the thrust, the stem thrust.
22 MR. MICilELSON:
The stem was very difficult.
[
j 23 MR. TilIBAULT:
And for a non-auditing stem it's f
24 still okay.
My first anewer was to put a strain gauge on
}
25 the stem when I can.
l l
I fleritage Reporting Corporation
(}
(202) 628-4888 1
194 1
MR. MICilELSON:
Yes.
()
2 MR. Ti!IBAULT And it's, to me, because it's a 3
direct measurement.
We don't refer to, again, a strain 4
gauge put somewhere as the spring back which, in one of the 5
picture I've shown reflect perfectly the load.
But you have 6
a converting factor, you measure some tenth of an inch, you 7
have to go from there to pounds.
So it's a direct 8
measurement, that's it.
9 I'm not criticizing anybody, that's not Wyle's 10 purpose, that's the way we approach it.
It's a different 11 way than the other did, that's it.
12 MR. MICllELSON:
Now the load washer is a direct 13 measurement, isn't it?
14 MR. TilIBAULT Yes, it's a direct measurement,
( ) 15 yes.
16 MR. MICllELSON:
You precalibrate the load washer.
17 MR. TilIBAULT :
Yes.
We precalibrate the load 18 washer.
We have to know where to go depending on the 19 actuator, and how much to top them.
20 But I would say that Wyle feeling is more to use 21 the two system rather than using base line or MCSA.
22 MR. MICilELSON:
liow does the load washer take 1
23 account of the yielding of the bolt.
The supporting stud 24 that you've used is going to have a potential for yielding 25 during the operating cycle.
I IIeritage Hoporting Corporation (202) 6'P-4888 O
195 1
MR. TilIBAULT:
I had that question with some
}2 people from Morgan and Schlat.
I took care of it by the 3
calibration.
I cannot tell you how much yield I put on the 4
bolt, because that wasn't the purpose, but I can tell you 5
that the load that I measured on the stem is exactly what 6
the load will be.
7 And I'm not stressing the bolt more than it is 8
because I torque the bolt at the smaller torque than it's 9
original torque anyway.
10 MR. MICilELSON:
But if it were to yield it would 11 disturb the correctness of the result, wouldn't it?
12 MR. TilIBAULT:
It would, but it doesn't, 13 MR. MICllELSON :
But it doesn't because you're 14 keeping it a* very low stress.
()15 MR. TIIIBAULTt Yes, sir, absolutely.
16 M'. MICl!ELSON:
Are you using a special bolt for 17 this purpose as well?
18 MR. TliIBAULT:
Not a epocial bolt, a longer bolt.
19 MR. MICllELSON:
Just longer.
20 MR. TIIIBAULT It's special in some way that it's 21 longer, but it's a standard bolt.
And we put the other one 22 back after we took it off.
23 The thing is you have to -- what you have in the 24 market it's intrusive methods.
V-MOD is an intrusive method 25 like MOVATS, whoever; MCSA is not.
But you have to have an lieritage Reporting Corporation
(}
(202) 620-4088
f 196 1
intrusivo -- at least today an intrusivo method to -- and
()
2 that's again what I strongly believe to have a baseline.
3 And after -- until you change somothing in the valvo you i
4 don't need to have any more of that intrusive method, you S
can just go with the current MCSA.
6 And MCSA, by the way, can be used for many other 7
reasons.
You can put that in a pump, MOD driven pump.
So h
8 it can be used -- because the big problem with MCSA, I 9
believe, and that's why I was a little bit discussing this f
10 morning is, the motor seo exactly what it is driving.
11 Whatever you put -- it doesn't know, you have to interpret L
12 it and that's where the frequency analysis come in.
But the 13 motor is a transducer of what is happening after.
14 Now, there was a treat that was served by Oak I
()15 Ridge if the voltage lock during the closing or something, 16 how were we going to do it.
That's how they came to the f
f 17 analysis.
Whore you are looking at the very small delta, j
i 18 let's say from a 12.7 to 13.1 hertz.
So it's a lengthy way t
19 to do it.
[
20 MR. MICllCLSON:
I thought they were also looking
{
21 at analyzing the power input in the motor instead of the l
T 22 current, that way they had something more closely related to
{
[
23 it.
24 MR. THIBAULT:
That's what we tried to do with V-25 MOD by recalling the current and the power, because the
(
t Ileritage Reporting Corporation
(}
(202) 628-4888 i
l l
f
197
'l power is two times square of three and whatever.
But the
(~
2 licensing that has been done by Oak Ridge is at the current 3
signature, no power.
4 MR. MICilELSON:
I see.
5 MR. WYLIE:
llow do you measure power?
6 MR. TilIBAULT You mean the MCSA?
7 MR. WYLIE:
Yes.
8 MR. TilIBAULT You just go to the MCC with an 9
ainprobe.
10 MR. WYLIE:
With what?
11 MR. Ti!IBAULT:
An armprobe.
12 HR. WYLIO:
Okay, I understand.
13 MR. TilIBAULT:
And you just go to one of the 14 leading going -- you go to one of the lead going --
()15 MR. WYLIE:
Do you use a transducer or do you 16 use an --
17 MR. TilIBAULT:
Just an armprobe.
18 MR. WYLit:
I understand that.
But I mean, you 19 take the output of the amp motor --
20 MR. TilIBAULT That's where the licensing is that, 21 what Oak Ridge did.
It's mainly the process of a signal in 22 two ways.
The input from your armprobe goes into a 23 different channul, one is a time analysis and the other one 24 a frequency analysis.
And they put a feed there of 25 everything whict-is -- which is above 60 amps, they find out lioritage Reporting Corporation (202) 628-4888
y 19E; I
the above 60 amps on an AC is not related to anything.
()
2 Now if you go to a DC they use a whole effect 3
probe and change the frequency cutoff point.
So that means 4
that MCSA can be used and is used also in a DC.
MR. WYLIEt Well, that wasn't my question.
My 6
question was, you say you measure power.
My question wau, 7
how do you measure power?
]
8 MR. THIBAULT:
With V-MOD we measure power; we 9
don't measure power with MCSA.
i 10 MR. WYLIE:
MCSA.
i 11 MR. THIBAULT MCSA, I did not say that, 12 MR. WYLIE:
It says --
i l
13 MR. THIBAULT V-MOD, yes, I do measure power.
i t
l 14
!!R. WYLIE:
Okay.
( ) 15 HR. MICHELSONt I think he was asking how you did
]
16 it?
17 MR. THIDAULT You measure tension and current.
[
[
10 So you don't measure --
l 19 MR. WYLIE:
That won't give you power.
i 1
20 HR. THIBAULT:
Well, almost, because you have to I
4 21 take --
l I
22 MR. WYLIE:
Amps, it doesn't give you power.
I 33 MR. THIDAULT:
Yes, it gives you an active power,
[
24 it doesn't give you the reactive part of it.
So that*r. why l
f 25 I say, you measure power almost because it's U I square --
3 1
i i
I lieritage Reporting Corporation (202) 628-4888 i
I I
I
.+
199 1
MR. WYLIC:
You don't really meacure power, you O
2 measure volts amp.
ss 3
MR. TilIBAULT :
You measure the product Q time I.
4 MR. MYCllELSOti lie doesn't worry about the amps.
5 MR. TilIBAULT :
Bocause I believo it's important in 4
6 this, especially when ti..
rank of the tuotor.
7 MR. WYLIE*
You measure in volts and amps.
8 MR. THIBAULT:
- Yes, i
9 MR. WYLIE:
Then you calculate the power.
l 3
10 MR. TilIB AULT:
Yes, you just made a product.
11 MR. WYLIE It's not a measurement.
You say you 12 measure it, that's the reason I asked.
l l
13 MR. TilIBAULT:
I'm sorry, so I measure the current i
14 and the tension.
I don't have a watt meter.
But the watt
)15 meter will measure U and I anyway.
?
]
16 MR. WYLIE:
You don't do anything with that power i
17 measurement, do you?
18 MR. TilIBAULT:
tio, just display and show its --
19 yes, we compare it with the current.
You can have a i
l I
20 constant power under current increasing.
21 MR. MICllELSort:
Well, do you do a frequency 22 analysis of the core?
23 MR. TilIBAULT:
tiot on V-MOD.
110t on MSCA either.
i i
24 MR. MICllELSOti:
11 0.
But you're not doing anything 25 different than Oak Ridge is doing.
r l
L i
l i
r i
lioritage Reporting Corporation l
(}
(202) 628-4000 l
l l
l t
200 1
MR. TilIBAULT:
No, no, no, no.
We took their
()
2 system and incorporated it with MSCA as a complementary 3
system.
4 MR. MICIIELSoli It appears that you took their 5
nystem and added the washer --
6 MR. TilIDAULT:
That's correct.
7 MR. MICIIELS0ll:
-- the way of meesuring stress.
8 MR. TilIDAU'.T:
Yes, sir.
9 MR. MICilELSON:
Okay.
10 MR. WOllLD:
Let me comment on the power.
I think 11 that might be a good diagnostic tool, but there's problems 12 in measuring the power.
I think MOVATS has run into that 13 where, you know, the square of three V-line, high line --
14 MR. TilIDAULT Yes.
()15 you have to measure the CO sign, MR. WollLD:
r 16 and it also assumes the sign wave --
17 MR. THIBAULT That is correct.
18 MR. W0llLD:
These motors are wound to be in a 19 saturation region to koop the inertia down.
20 MR. TilIDAULT:
Yes.
21 MR. W0llLD :
And you don't got the typical sign 22 wave of current for a sign wave of voltage in, so I don't 23 know how you measure power.
That would be a difficult thing 24 to de as far as power actually delivered to the shaft of the 25 motor.
IIeritage Reporting Corporation (202) 628-4888
l l
1' l
l 201 1
MR. TilIBAULT:
I agree with you.
And I should not
()
2 say we measure power, we measure the product tension time 3
current <
l l
4 MR. MICIIELSON:
You must find it useful for I
5 something, though, l
6 MR. TilIDAULT:
Well, that came up, I believe, I 7
was not at Wylie at the time.
But it came up, what if we 8
find some current increasing -- let's suppose you got the 9
thrust that is increasing, the current increasing and then i
10 dropping back.
You can say I got a bad stem or something 11 like that.
So you can view that current signature to a i
l 12 mechanical effect that in fact it's not a mechanical effect, l
l 13 but can be evaluation of the tension d"ring the valve 14 stroke.
And that has been found by Oak Ridge, by the way.
()15 HR. MICllELSON:
Yes.
But how does the power help l
f l
16 you much there, because you're not quite sure you're looking 17 at power.
18 MR. WYLIE:
You can if you're actually measuring t
19 power.
l 20 MR. MICllELSON:
Sort of a currogate.
Yes, but 21 they're not actually measuring power.
5 22 MR. WYLIE:
Yes, I kno*:7, they're calculating volt
{
23 amps.
t 24 MR. MICilELSON:
They aren't dealing with --
i I
25 MR. WYLIE:
But it's a constant.
Volt amps is a I
lleritago Reporting Corporation l
{
(}
(202) 628-4888 1
202 1
constant.
If the voltage varies, the volt amps is still
()
2 going to be a constant because your current is going to go 3
up as your voltage drops.
So it comes out as a constant.
4 MR. MICllELSON:
If you're saturated.
No, that -- the problem is these are 6
strange motors, they don't follow the typical inductive r
7 motor volt amp relatior. ship.
As the volts go up you go 8
further into the saturation region and what you're reading l
9 is a short circuit on the windings.
You're not reading a I
10 normal inductive type of voltage.
11 MR. TilIBAULT:
That's a saturation of the --
f 12 MR. WYLIC:
It doesn't make any difference, if the 13 load on the motor is a constant, the power into that motor i
14 is going to be a constant.
()15 MR. W0llLD:
Right.
16 MR. TilIBAULT:
If the load is constant.
17 MR. WOllLD:
But if you're measuring two power 18 delivered to the --
19 MR. WYLIC That's what I'm saying, if you can j
20 measure true power, then if thero's something wrong in the
[
21 system mechanically, it will put more load on the motor, you l
22 can tell.
I think that's what he was saying.
23 MR. TilI BAULT:
That's what I was saying.
l
[
24 MR. MICllELSON:
But you can't measure true power.
l l
25 MR. TilI D AULT :
I misled you when I said power, l
l l
l linri tayo Reporting Corporation f
(}
(202) 628-4888 L
l I
[
203 1
which wasn't power.
()
2 MR. MICllELSON:
The question is, how closely can 3
you measure --
4 MR, WYLIE:
When you use a power transducer, for 5
example, which will do it for you.
You can do it, there are 6-ways of measuring power.
7 MR. TitIDAULT:
There could be a way to go.
We did r
8 not go --
9 MR. WYLIE:
You can't calculate it with just 10 measuring volts and amps --
l t
11 MR. TilIBAULT:
Yes.
12 MR. WYLIE:
-- because you don't know the power 1
l 13 factor.
[
r
]
14 MR. TlilBAULT :
At least you can have the relation f
()15 of U and I, hoping that is going to koop constant all the 2
16 time.
That's right, no.
l a
l f
17 MR. WOllLD:
The only thing I can say that it's a 18 complicated situation.
If you look at the typical reliance i
19 pump -- motor curve you see that current -- the measured
)
20 current goes up significantly for small increases in L
j 21 voltage.
i i
22 MR. TitIDAULT:
Yes.
Yes.
And that's how, I 1
J a
23 believe, how Oak Ridge discovered what happened in some of I
j 24 their tests, because they went to reliance.
I'm not Oak
[
25 Ridge, okay, I was trained by Oak Ridge.
When they f
I l
t I
fioritage Reporting Corporation
[
(202) 628-4888 i
i j
I l
l
l l
204 1
discovered that twico'in the curront, that normally in the
(
2 current they went to -- they thought about that and they l
l 3
went to reliance to the torquo curvo of the motor.
And l
4 that's how they found that the small increase -- well, 1
1 5
dopond whero you aro, also, in the curvo, as you say.
G MR. W0llLD:
Right.
7 MR. THIBAULT:
But I think it's in the lower part f
0 of it when a small increase in the current in a big increaso 9
in the torque.
Is that correct?
r l
10 MR. WollLD:
Right, in the low end, riejnt.
l t
11 MR. TilI BAULT:
That's how Oak Ridga discovered and
[
P 12 went to that analysis.
tiow again, that's me speaking, it's I
13 not Oak Ridgo.
r r
f 14 HR. WOllLD:
It's a complicated situation.
It's I
()15 not a simple sign wave multiplied times another sign wave.
l 16 I don't know if you get a second harmonic in there or l
t 17 whatover when the arm eaturates, but it's a strango way for j
f 18 them.
19 HR. MICilELSoll Any other questions?
20 (flo responso.)
21 MR. MICIIELSail:
I bellove that completed your 22 presentation.
23 MR. TilIRAULT Thank you very much, sir.
i 24 MR. MICilELSOll Thank you very much.
It's pretty
{
l 25 close to that time.
Yes, I think we'll just take a 15 Ileritage ieporting Corporation
(}
(202) 628-4888 l
1 205 1
minute break and reconvene at 3 o' clock.
2 (whereupon, at 2:45 p.m. a 15 minute recess was 3
taken.)
l t/4 4
MR. MICHELSON:
At this time we would like to pick l
l 5
up the first part of our agenda dealing with air crerated 6
valves and air operated valvo operating experience, and for 7
that purpose Mr. Ilayos of Toledo Edison has volunteered to l
8 tell us about their experiences with air operated valves; so l
l 9
that will be next on our agonda, f
i 10 (Slides being shown.)
11 MR. IIAYES:
My name is John Hayes, I'm from Toledo j
l l
l 12 Edison and what I have to say to you, I'm not necessarily 13 representing any industry or any owner's group or anything 14 af that nature.
I have just come to report to you what i
( ) 15 Toledo Edison is doing in the field of air operated valvo.
i l
16 I fool liko kind of a stranger, alA this motor
[
17 operated valve talk all day, but bear with me.
l i
18 MR. MICilELSON:
You're going to hear a lot about I
air operated valves as time goes on.
19 I
20 MR. IIAYES:
Well, I sure hope so becausa I'm very l
i 21 interested in it.
22 From our June 9th of '85 event that we had at L
23 Davis-Bessie, which I imagine all of you are aware, the air l
24 system is one of the systems of concerns, even though it 25 wasn't a safety related system, the system was completely l
l lieritage Reporting Corporation l
(}
(202) 628-4888 i
h
l l
206 i
L 1
reviewed and redesigned and the modifications from that has
()
2 just been put into place.
3 And from the experiences that we learned from the 4
motor operated valve program, we want to carry that over to 5
the air operated valve program, realizing that there are a 6
lot of hidden areas in that program.
So we're just trying 7
to get a leg up on everybody and carry on into that.
8 In the program itself we formed an air operated 9
task force which I'm the leader of.
10 MR. MICilELSON:
This is just for B&W owners?
11 MR. IIAYES:
No, this is just for Toledo Edison.
12 MR. MICllELSON:
No.
Okay.
13 MR. IIAYES:
It has nothing to do with the owners 14 group.
. ()15 MR. MICif ELSON :
Within Toledo Edison.
All right.
16 MR. !! AYES :
Y9s, sir.
17 The scope of the program was to take a look at the 18 short and long term activities.
Okay, both will involve the 19 acquisition of different failure data available to the 20 industry through NPRDS; previous problem evaluations; 21 drawings; maintenance records of any history documents that 22 we could draw together.
l 23 Now we're going to take this information analysis l
l 24 and evaluate it.
And then from that evaluation come up with I
l 25 our recommendations and implementations to our senior floritago Reporting Corporation
(}
(202) 620-4000 l
t 207 1
management people.
()
2 Then wo broi ' the -- we had to define what our i
3 program responsibilities were and that's to support the 4
resolution of any immwdiate air operated valve problems that r
i 5
wo may have in the plant itself; analyze and ovaluato air 6
operated valve problems at our plant; and evaluate 7
recommended program improvemonts to our senior management.
i l
I 8
MR. MICHELSON:
Now when you talk about air
[
t 9
operated valves, are you including all types of air i
t 10 operators?
l t
11 MR. H.
ES:
I'm including any valve that has an
{
12 air operated device on it.
^
f 13 HR. MICHELSON:
Okay.
14 MR. HAYES:
That's including diaphragm, spring,
(])15 that's lncluding, you know, piston operators, and the valve l
16 that's associated with that, you know, with that air l
l
]
17 operated valvo.
r 3
18 MR. MICHELSON:
Very good.
But you're not going l
19 into heating and ventilating unless -- how about ventilation l
20 dampers that are air operated?
21 MR. HAYES:
We're looking at ventilation dampers j
22 in the offect that we actually hiae butterfly valves that l
23 are acting as dampers, okay, we're looking at those.
l I
24 MR. MICHELSON:
Are you looking at your 25 containment isolation valve in this program as well?
I f
i
)
{
}
Heritage Reporting Corporation (202) 628-4888 i
t I
__,_. _ _.a
208 1
MR. IIAYES:
That's correct.
()
2 MP. MICilELSON:
Okay, good.
3 MR. IIAYES:
Then the activity of the program I 4
have broken down into short term and long term.
The shcrt 5
term aspect of the program which was one of the main reasons 6
why wo formed the tosh force was to evaluato and analyze 7
known problem valves including a design basis for those 0
valves, okay, and then como up with what short term or in 9
term fixes that we could with the current outage that we're 10 in.
11 And from -- and also any spinoff, any long term 12 actions that would be required with these valves.
And the 13 known problem valves that we looked at in-depth for our 14 turbino bypass valves, our service water valves on our
(]}15 component clean water heat exchangers, and component cooling 16 water valves, thoso are our isolation for our decay heat 17 removal and to our aux building non-eesential cooling, and 18 our main and startup food control valves.
19 We found a lot of interesting things in doing this 20 review.
And I can give you a brief synopsis of some of 21 these problems.
I'll leave this slide up.
22 The turbine bypass valves, they have been a 23 constent problem at Davis-Bessie.
Okay.
They were part of 24 a course of action from our June 9th, '85 incident.
Okay.
25 Through the course of the use of these valves, which they're 1
Heritago Hoporting Corporation j
(202) 628-4888
e.
209 l
1 six inch Fisher EP which is a -- Jt's an unbalanced valvo 2
using through a primary plug assembly, a primary pilot i
3 assembly through bolvol washers which turns an unbalanced j
i 4
valvo int o a balanced valvo.
J 4
5 Throughout the years we have busted actuator I
i 6
Wo have actually put the top of the actuator.
i i
7 Busted out the top of the actuator from forces that were 8
transmitted in the valvo and they have boon a constant 9
nagging problem at Davis-Bessie, f
j 10 MR. MICilELSollt I'm not real acquainted with the l
2 l
11 kind of valve you must have, but is this some kind of a drag j
j 12 valvo design?
i i
13 MR. IIAYES:
tio, no.
The drag valvo is utilized in I
I 14 what we call our atmosphere event valves.
l l
aR. alC><EbSort,
net in these eveess --
1 O 15 a
t 16 MR. IIAYES:
This is like a plug valve, '/asically.
l 17 MR. MICilELS0ti t You're taking a very '.argo I
i 10 pressure drop through them, d
t 1
19 MR. IIAYESt Well, the inlet prestare you're t
i 20 talking about is around U60 some pounds "jressure and it's i
21 going to a vacuum.
I 22 MR. MICllELSOll t Yos, it's going to a vacuum.
So l
]
23 you're taking a very large pressure drop across the valve j
i 24 and it's just sort of a globe valve.
j j
25 MR. IIAYES:
Well, it's a plug valve basically.
(
l l
lloritage lleporting Corporation j
(202) 620-4888 l
i
210 1
MR. CARROLL:
flow about a GE turbino stop valve, 2
is that a good oxample?
In principio at least.
3 MR. IIAYES:
In principle.
Okay.
It's a short 4
travel valve, it's very quick acting, you're talking.
5 MR. CARROLL:
The kind of valvo that equalizos.
6 MR. IIAYES:
That's correct.
Okay.
So the leak by 7
from the piston of the valvo is utilized to seat the primary 8
plug.
9 MR. MICIIELSON:
That I understand.
10 MR. IIAYES:
Okay.
11 MR. MICllELSON:
Good.
Thank you.
12 MR. !! AYES :
From our investigation we had Pisher 13 involved in it, the vendor, and we brought Fisher Service 14 Company out of Columbia, South Carolina and got them
()15 involved in it because of their reputation with Duke Power 16 and the recommendation that they gave us on those gentlemen 17 out of South Carolina.
18 And wo took our valves internals and shipped to 19 them and we found that the local valve representative for 20 our area, representative of the manuf acturer *.hought we had 21 a different valvo design than we actually did of that valve.
22 And as a result, for the 12,*ars that we have had 23 this valve in service they have been shipping us the wrong 24 internal parts for these valves.
25 MR. MICitELSON:
You've been putting them in?
lieritage Reporting Corporation
(}
(202) 628-4888
6 211 1
MR. IIAYES:
And we have been putting them in
)2 because we have no drawings to give us any tolerance on any 3
parts because the valve manufacture considers these 4
proprietary information.
5 The service company that the valve manufacture had 6
in our area, which is out of Mansfield, Ohio, didn't 7
recognize this problem over the a' years that we've had 8
these valves in service.
9 The Fisher people out of Columbia, South Carolina 10 recognized the problem and working with the application 11 engineer in Marshalltown and for the first time in the 12 service of those valves they are set up properly and have 13 all the right parts installed inside of them.
14 MR. MICllELS0!!:
This causing premature wear out or
()15 breakage?
16 MR. HAYES:
tio, it was causing the seats to be 17 broken out.
It was causing the valves to leak through 18 because of not letting positive seating at the primary plug.
19 The vendor manual inf ormation that they supplied us was 20 wrong as far as the primary plug spring pack compression.
21 And the advertised travel of the valve was in 22 error because Bailey, who they were supplied through, 23 stamped the wrong travel distant on the side of the valve, 24 and come to find out that our valves were advertised as one 25 of nine-sixteen inch travel, which they were set up for, in lieritago Itoporting Corporation (202) 628-4888
012 1
reality we had one and seven-eighths inch travel on the
'(
2 valve.
3 So sorting out all this information, finally, I 4
hope we'll go into this next run with a valve that is with 5
the right parts and with sot up properly.
6 MR. WYLIE:
Are those parts being supplied by the 7
original manufacturo?
O MR. !! AYES :
By the vendor.
Yes, sir.
9 MR. WYLIE:
They shipped you the wrong parts?
10 MR. HAYES:
Yes, sir.
11 MR. WYLIE:
Did they fit?
12 MR. IIAYES:
Well, sort of.
We took exception with 13 the manufacture.
And right now we're in correspondence witn 14 them trying to got a set of drawings from him so that when
()15 he does send us parts that we can verify what he sends us is 16 what goes in the valve.
And you can't get drawings, 17 detailed drawings from the valvo manufacture because, quota, 10
- it's proprietary information."
19 MR. MICHE LSON :
It's a problem --
20 MR. IIAYES:
Yes, sir.
And I think it** --
21 MR. MICHELSON:
You can't got drawings of the 22 internals of your valves.
You got outline drawings.
23 MR. IIAYES:
Well, we're asking for detailed 24 drawings and better vendor information, and I hope --
25 MR. MICHELSON:
But you apparently didn't order by licritage Report ir.g Corporation
(]}
(202) 628-4888
I i
f 213 t
I the serial number on the valve or something of that sort.
(
2 HR. HAYES:
You order through the vendor's locai 3
representative, okay, and he's supposed to have your valve j
4 sheet and valve information.
l 5
MR. MICl!EbSON:
The serial rumber Je supposed to r
6 protect you from getting the wrong one.
[
i 7
MR. 11 AYES :
That's correct.
And we do order by 8
serial number.
t
]
9 MR. MICllELSON:
And it didr.*t do any good in this i
d 10 case --
11 HR. IIAYES:
No, sir.
12 MR. M!CllELSON:
-- because that was catalogue 13 wrong --
b 14 MR. HAYES:
There was a design :hange made on '.he
()15 valves subsequent to ou. purchasing the salves, and they i
16 thought that design chango was includod in our valves and.it i
17 wasn't.
l l
18 MR. MICHEl. SON :
I see.
It wasn't subtle.
19 HR. IIAVES :
Yes, sir.
And we had no -- as the j
20 ut'lity had no way of knowing this, because every time that t
21 those valves were opened up for any work there was a i'. sher f
22 representativo at tho -- with the valvo.
i 11 I can focus this thing l
23 What I've got here is 24 somehow.
What I touched on ecrlier, this is on their 25 service water valves.
What we have is a valve stem that lloritaqo Reporting Corpora tion
(}
(202) 628-48G8
s I
i 214 I
goes through a butterfly valve.
This is out of a Hammerdahl
()
2 16 inch, a 150 pound class valve that we use as a 3
temperature control valve and an isolation valve on a 4
component heat -- component cooling / heat water exchanger.
5 Here is your valve body.
Here is a bearing insert 6
cartridge with a roll up teflon impregnated fabric back 7
bearing with a trade name of Slippies.
And right here is 8
the valve disk itself.
9 okay.
So you have the same thing through the 10 valve on the opposite side, okay, it means you have your 11 butterfly.
12 We had a problem with these valves in that if you 13 rebuilt the valve with a new bearing, substitute the full 14 differential pressure on th. valve itself it would work
( ) 15 fine.
And you come back a week later or two weeks later and 16 repeat the same test, the valve would not operate.
Okay.
17 The thing would fail to operate.
It would jam up and the 10 actuator wasn't large enough to overcome the forces in the 39 valve to operate it.
20 MR. MICHELSON:
Was the differential to close or 21 open it?
22 MR. IIAYES :
This is -- the valve is closed and 23 attempting to open.
24 MR. MICHELS0 tit Okay.
25 MR. IIAYES:
Okay.
The task force, we looked at it IIeritage Reporting Corporation (202) 628-4888 O
215 1
separate and we contracted Telsy Engineering, which I
(
2 mentioned earlier from the audience, to do an' independent 3
evaluation of what our findings were.
4 The problem that we had here is that previously 5
all areas of this valve could somehow be' interrelated,or 1
'6 related back to maintenance activities.
Because of this 7
bearing design all this thing is, it's a bearing that comes 8
in sheet, you cut the sheet.
You roll the bearing material 9
up, okay, and insert it into this cartridge.
And it's no a
10 indication of which end goes up or back and forth, so the I
11 poor maintenance person involved, in some cases, had 12 installed these things backwards with the wrong surface up.
l 13 Or if you don't get it just exactly right you can get some I
14 rolling of the bearing and this sort of thing.
l
()15 And there's very close tolerances inside the valve 16 itself, so that when you put the tamper pin through the 17 valve disk into the stem, okay, you can cause some i
18 distortion in that method and also packing problems and 19 corrosion problems in the valve and getting between the disk 20 and the valve causing a toggle effect and causing the valve 21 to freeze.
1
)
22 So always in the past we can always point at a
23 something, you know, at some symptom and say, this was the 24 problem.
We come to find out that thia slippies material is 25 designed for 10,000 pounds of compression forces.
The 1
IIeritagu Reporting Corporation
(}
(202) 628-4888 t
216 f
I calculations in our valve we could experience as high as
(_)
2 about 23,500 pounds of compression forces on the bearing 3
material itself.
4 MR. MICilELSON:
Is that from, what, just from 5
inserting the bearing?
6 MR. IIAYES:
No, no, no.
The forces from the cet 7
of the differential against the disk.
8 MR. MICilELSON:
Oh, okay, it's pressing on it.
7 MR. IIAYES:
Compression, yes, sir.
And there's 10 about a 4 mill permanent col that that bearing takes.
In 11 other words, it's in elastic, it won't return.
So when you 12 first put the bearing in service, okay, it would work fine, j
13 And then after you compress the bearing, then what you had i
1 14 here is between the stem and the body, you had about a 4 1
()15 mill tolerance and between the edge of the bearing cartridge 16 about a 5 mill tolerance.
l I
17 So if you took the set of the bearing against the 18 full differential pressure you were coming up with stem 19 contact between the stem and the body and the actuator 20 wasn't large enough to overcome those forces.
21 So what we went and did in this outage, is that we 22 bored this out, turned -- took this bearing design and threw 23 it away and put another bearing, one piece bearing cartridge 24 in there.
And we took on the top of the disk, we cut out 25 the top of the disk and into the body itself and we put a tieritage Iteporting Corporation (202) 628-4888
217 1
thrust washer, teflon thrust washer cent
'g device for the g) 2 disk to preclude any possible body to discontact, there's a
(,
3 very tight tolerance in there.
4 And we also changed the pocition of the valve in 3
its closed position to preclude any problems with the
- any 6
corrosion products of debris toggiing between the disk and 7
the valve body.
And we are in contact with Itammerdahl here as far 8
9 as what we considered to be a design problem with the valve.
10 And I don't know where that stands right now.
Our design 11 engineering department is working with that problem.
12 MR. MICllELSON:
llave any of your utility 13 colleagues run into the same problem?
14 MR. IIAYES:
That I don't know, sir.
We put this 15 out on the NPRDS.
16 MR. MICHELSON:
Did you get any responses back?
17 MR. IIAYES :
No, sir.
This is a unique design.
10 Okay.
Our component cooling water valves and our 19 service water valves had the same problem.
Okay.
That when 20 you took air away from the valve the valve would not stay in 21 its designed failed position.
Okay.
Either due to leakage 22 or long term effect.
23 MR. MICllELSON:
llow does it close to its design?
24 MR. IIAYES:
Well, we had -- in this case we had a 25 total of six valves involved.
You have a piston operator IIe ri t.a qo Iteporting Corporation (202) 620-4880 O
218' 1
and you have a cusliion air accumulator on either side of the
{)2 piston.'
So when the valve goes.to its failed position you J[
3 have a solenoid that. bleeds off one side of the piston, and 4
then the cushion air that's remaining on the opposite' side of the piston is enough force to drive it to its failed 5
e 6-position.
Okay.
7 But if there's any leakage and/or long term decay 8
of' pressure the valves will come out of their failed 9
position and then take a position in the flow stream, okay, 10 however the flow stream determined in the air dynamic 11 characteristics of the valve itself.
12 So our USAR didn't give us any definition of how i
13 long the valve had to stay in the'falled position.
So it's 14 one of these nebulous things, do we have a design problem or not.
And if you lost air, then how the air was lost will
()15 16 make a determination, too, if it's a gradual loss of air, 17 will the check valve in there seat.
If there's any debris 18 in the system will that interfere with the check valve from 19 seating.
So there's no guarantee that that valve will stay 20 in its failed position.
21 MR. MICHELSON:
Were you modulating with the valve 22 as well?
23 MR. IIAYES:
You have that capability, but not for 24 long.
25 MR. MICIIELSON:
If you start modulating and your l
lioritage Reporting Corporation (202) 628-4000
--.-,------,--n,n,-
n---
,_,p
219 1
air pressure proceeds to drop, then the system -- the 2
accumulator just comes on down with it.
3 MR. liAYES:
Now the generic letter that this came 4
out on is the very problem.
Okay.
About three weeks ago, I 5
don't know the number of it.
Okay.
Asked you to test this 6
feature.
It asked you to ensure that the valve will get to 7
its failed position.
But it doesn't put any requirement on 8
how long it shall maintain itself in the failed position.
9 MR. MICllELSOll:
It's something --
10 MR. IIAYES:
We tested this feature as part of our 11 recovery problem from June 9th, but we didn't try to see how 12 long it will stay there.
13 MR. MICilELSON :
Something has always been, 1 14 guess, fail safe that once you get there you stay there, it
( ) 15 doesn't come back out again.
16 MR. CARROLL:
The old by stable problem.
17 MR. MICl!ELSON:
Yes.
18 MR. WOllLD:
I think there's an information notice 19 out on the time, I'm not sure.
I assume there is.
l 20 MR. IIAYES :
I haven't read one, Pete.
I'm not l
21 aware of it, unless eny of these other gentlemen may be 1
22 aware of it from the stay.
23 MR. W0llLD:
I thought I read something, but okay.
24 MR. IIAYES:
What we went and did to solve this 25 problem is that we designed a detent system.
It's a simple lleritago lloporting Corporation
()
(202) 6 7 6 -4 01111
220 1
system.
It's so simple it's almost embarrassing.
On the
()
2 actuator itself we put a pneumatic versus spring piston.
3 And when the air goes away this piston retracts, the spring 4
pressure forces it 9t.
We put a striker plate on the 5
actual bell arm itself, these are quarter turn valves, 6
butterfly valve, so when air goes away the piston comes out 7
against the striker plate; then as,the valve rolls to its 8
failed position, then the piston comes in as an interference 9
to the operating mechanism to prevent the valve from coming 10 back, okay, from the dynamic forces from the flow and it 11 will keep the valve there.
When air is regained, then the 12 piston retracts and then the valve is free to move again.
13 It doesn't rely on any electrical signals or 14 anything, it's strictly a mechanical device with pneumatic
{ } 15 versus spring.
And we installed it and it works fine.
t 16 MR. MICilELSON:
llave you actually tested it?
17 MR. IIAYES:
Yes, sir.
18 MR. MICHELSON:
Pulling that guard pin back out 19 again is a little bit of a trick.
20 MR. IIAYES:
Well, there's also on the other 21 opposite end of the piston is a hand mechanism that you can 22 manually pull Lt.
So if you want to manually operate the 23 device with the hand jack you can pull it out of the way, 24 engage your hand jack and then the pisten or the detent will 25 ride on the striker plate.
It's a real simple device.
Ileritagn Itoporting Corporatton (202) 628-4888 O
n -_
221 1
MR. MICHELSON:
It keeps it in a fail safe 2
position.
3 MR. HAYES:
Keeps'it in a fail safe position.
4 MR. MICilELSON:
Did you find that problem was just 5
the component cooling water valves?
6 MR. HAYES:
No, the three service water valves; 7
two component cooling which are the isolat. ion on the decay 8
heat removal heat exchangers; and our aux building non-9 essential load.
10 MR. MICHELSON:
What prompts the valves to want to 11 drift back out once the holding force is starting to 12 relieve?
13 MR. IIA ~iES:
Is the flow stream itself and the 14 weight of the flapper.
()15 MR. MICHELSON:
The flow, when it's closed, it's 16 closed ar the failed position.
17 MR. HAYES:
Well, you have a lot of other problems 18 in there, too.
19 MR. MICilELSON:
Are these always closed?
Is the 20 failed position opened in most cases on these systems?
21 MR. IIAYES:
On five of the valves that failed the 22 position is opened, and the one valve it's failed closed.
23 MR. MICllELSON:
How in the failed closed case what 24 prompts it to come back out?
It only can be differential 25 pressure than not flow condition.
I!eritage Reporting Corporation
{}
(202) 628-4888
d 222 1
MR. HAYES:
But we've got a lot of different
-( )
2 forces in the flow stream, where the valve is in 3
relationship to elbows you may,have an impact greater on a 4
lower part of the valve disk than you do in the upper part 5
even though you had a symmetrical valve.
I mean, there's a 6
lot of different conditions.
7 The key part is that you cannot guarantee that 8
that valve is going to stay in it's flow position.
9 MR. MICHELSON:
This has happened to a couple of 10 other utilities, I think, recently too or at least the 11 problem is being highlighted through LERs.
12 HR. HAYES:
Well, I think it's a bigger problem 13 than people want to realize.
And like I say, you can test 14 the feature, it gets to the failed position you say, wow,.
(}15 it's in the failed position.
But I don't think anybody has 16 done a time test how long it will stay there.
And I don't 17 think anybody has done an analysis on how long it's really 18 required to stay there.
19 MR. MICHELSON:
Once it drifts back into the flow 20 stream that produces the flow is the problem.
Is this a 21 significant reduce of flow?
22 MR. IlAYES :
In a case of our service water valves 23 we had to at least maintain about 40 percent open to come up 24 with our design base flow of about 8,000 gallons per minute.
25 MR. MICllELSON:
Was it tending to drift even more lloritage Reporting Corporation (202) 628-4888
- l 223 1-closed than that?
13
. g,j 2 MR. HAYES:
Even more closed.
It would come to 3
about the 30 percent position.
4 MR. MICl!ELSON:
Certainly was putting a lot of 5
fluctuation in the flow as well.
6 MR. IIAYES:
That's correct.
And it was valve 7
dependent.as far as the pipe configuration, the K factor of 8
the systen, you know, how the flow stream was through the 9
system.
Okay.
It was not a predictable vacuum in the 10 valve.
You can come within,,.say, a 10 or 15 degree area 11 that it's going to end up in, but you can't predict.
12 And we were looking at, on the service water 13 valves, will the valves stay in a position where we get the 14 required flow and we can say, no never mind, and we couldn't
(}15 guarantee that.
So that's why we had to go with the detent 16 system.
17 MR. MICliELSON:
I gather chey were significant 18 it was a significant effect on flow?
19 MR. !! AYES:
Yes, sir.
Well, we have -- in those 20 valves we have a down stream orifice that when the valve 21 does go to its failed position it protects the pump from run 22 out.
And we still, even with that, we still couldn't get 23 our flow.
24 MR. MICIIELSON:
I see.
25 MR. IIAYES:
A long term portion of the program is lloritage lleporting Corporation (202) 628-4888 9
m
224 1
broken down into equipment _ oriented and --
( )'-2 MR. CARROLL:
Did.you mean to skip main and 3
startup feed control ~ valve?
l 4
MR. HAYES:
Oh, main and startup feed control
~
5 valve, I apologize there.
We did a review of those and the 6
only rual problem we found out with those valves was 7
outdated control equipment.
The positioners were no longer 0
in manufacturer, so we changed the type of positioners that
[
9-we had in the valves.
j 10 And we had a slight hunting problem in one of the
[
11 valves and we solved that with the positioner.
We think wo 12 solve it.
We will know when we start it up, start the plant 13 up.
We found no major problems in those valves itself, h
14 A long term, like I say, is broken down into
(
(}15 equipment oriented problems and program oriented problems.
i 16 In the equipment oriented problems we wanted to set our 17 priorities by the system and component importance.
In other 18 words, how does the valve effect the plant or how does it 19 offect the system.
20 So we did that by putting the rest of our group to 21 work.
Of the 474 air operated valves or components that we 22 had, we filled out a failure consequences analysis work 23 sheet, okay.
And our system engineering department filled 24 one out on each of their responsible valves and our nuclear 25 engineering department also filled one out based upon their Heritago Reporting Corporation (702) 620-4000 O
225 1
knowledge of the system; and then we compared the two and
(
2 came up with a priority for a valve.
A valve may have six 3
priorities associated with it.
All six categories may apply 4
or. it may have none.
If it has none, we just prioritized it 5
as not applicable for our data base that we have set up.
6 And priority one is just simply that, if the valve 7
fails you're going to lose redundancy or you re. going to go 8
into an alert period as far as an LCO condition.
9 And priority two, if a failure causes a forced 10 outage for repair or results in a limited power output of 11 the unit.
12 Priority three, if the failure results in a 13 challenge of safety systems or causes plant to trip.
14 Priority four valve is, failure causes undesirable 15 transients or decreases the reliability of the system.
16 Priority five is, failure requirements operators 17 to make undesirable exceptions.
Okay.
In other words, 18 faster lineups within the system, these sort of things, or j
19 realignments or invoke alternate procedures.
In other 20 worris, it puts a burdon on the operator and the operation of
[
21 the unit.
22 Priority six is, failure is chronic and requires 23 excessive maintenance effort.
Is this guy really taking a 24 lot of our time.
25 And then like I say, 11A is that none of these lieritag'e Iteporting Corporation
(}
(202) 628-4880
226 1
apply and it's a no, never mind valve.
C\\
(/
2 MR. MICHELSON:
When you were discussing the 3
various kind of valves that you had problems with, I was 4
wondering if in the case of, for instance, diaphragm 5
operated valves or piston operated, are you having trouble 6
with the air sets for them, the solenoid control valves, and 7
if so, what was the nature of your difficulty?
8 MR. IIAYES:
What we have now is the short term 9
aspect of the program is completed.
And we have out data 10 bases in place.
And we have the history files in place.
11 The next phase of the program, which I will show 12 you in the programmatic orientation of the program where 13 we're just getting into that.
14 The only preventive maintenance program that's
()15 established in any air operated valves was as a result on 16 the turbine bypass valves from the course of action from our 17 June 9th trip.
And any EQ requirements placed on the Q 18 related or "N-stamp" valves, to have a safety related 19 function.
20 And other than that, in our plant there is no, 21 quote, "preventivo maintenance program" on any of the t
22 components, okay.
23 Recently on the NPDRS inquiry, I forget which one I
L 24 it is that, I forget which unit or utility, you know,
[
25 questioned the industry, they wanted to know what preventive Ileritago Reporting Corporation i
()
(202) 620-4000 r
227 1
maintenance were in place as far as solenoid valves.
He got (s
2 one response and that was the EQ requirements.
3 MR. MICilELSON:
Was the solenoid valves giving you 4
any trouble?
5 MR. IIAYES:
I haven't studied it in detail enough 6
yet.
I've talked to Asco as far as recommended preventive f
7 maintenance items with the thing, and they are so vague that 8
they're not going to commit themselves on anything, because I
]
9 they're telling you that there's so many variables as far as l
10 what type of environment are they operated in?
Ilow many l
l 11 cycles those valves receive?
This sort of thing.
l l
12 MR. MICllBLSON :
Certainly if you had trouble with 1
l 13 these valves failing and thereby losing the function of the l
air diaphragm and essqntially the function of the valve that 14 1
()15 you -- that would certainly be -- you would know about it.
16 MR. HAYES:
Solenoids have been a problem, don't 17 get me wrong.
But we. haven't been able to study it in l
1 18 enough depth or detail to determine if it's a chronic '
l 19 problem; is it a generic problem; is it a generic problem 1
20 related to a particular model or make of a solenold?
l 21 I mean, you see bulletino on Asco problems.
You 22 see industry related problems with specific models of 23 solenoids.
But in our program thus far, we haven't got to 24 that point in the program where we can do an in-depth 25 evaluation analysis of going through 474 valves maintenance lieritage Hoporting Corporation
()
(202) 628-4800
228 1
histories and try to pick out, do we have a, you know, is
()
2 there a bad actor in here or is this due to wear and tear.
3 Also, we've got to look at that, if it's due to 4
wear and tear what's the porosity of the wear and tear, is 5
it every five year period of time?
Do I want to look at 6
these solenoids every five years?
Do I want to rebuild 7
them?
Okay.
Do I want to rebuild them on what priority of 8
the valves that I established?
Maybe just try one, two, 9
four valves?
Ilow many do I have?
What does it entail?
10 These are the things we have to get into.
11 MR. MICilELSON:
So you really just haven't looked 12 into that in great detail yet?
13 MR. IIAYES:
flot yet.
We've only been -- this 14 program -- I want to iterate, we've only formed this task force at the end of March, the beginning of April this year.
{}15 16 I think we've come a long way, you know, into it.
17 MR. MICi!ELSOil Are you going to tell us more 10 about containment isolation valves of the purge and vent 19 variety?
20 MR. SIAYES:
tio, sir.
21
!!R. MICl!ELGO!!:
That wasn't on your list?
22 MR. IIAYES:
flo, sir.
23 MR. MICIIEL30!I Did you view that as a particular 24 problem?
I 25 MR. IIAYES:
The only containment isolation valve I lloritago Reporting Corporation (202) 628-4000 O
1
)
l i
229 1
founct a problem with, which is on slate to be replaced and
()2 outages are for nitrogen system.
There's a valve that's 3
installed in there, not designed for the service.
It has 4
been a continuing galling problem with the valve.
5 And getting back to the manufacture, for the type 6
of trim that we have in that valve, for the type of service, 7
it's not designed for that because the flow stream is 8
intended to supply some lubrication to the actual valve 9
itself between the plug and the cage.
And there's no 10 lubrication in a dry nitrogen system, so the valve galls 11 because of the hardness of the two materials.
And that 12 valve is on site to be replaced.
13 MR. MICIIELSON :
llave you looked at the air 14 operator on the containment -- do you purge the containment
(}15 continuously?
16 MR. IIAYES:
No, sir.
17 MR. MICilELSON :
You bottle it up?
18 MR. IIAYES:
Bottle it up, yes, sir.
19 MR. MICllELSON:
You must operate them other than 20 for occasional surveillance.
21 MR. IIAYES:
That's correct.
22 MR. MICilELS0!!:
They are not a source of 73 MR. IIAYES :
Well, that's another problem that we 24 found is in the section 11 testing, as far as detecting a 25 problem with a valvo.
If you don't -- like the se*vice lieritage Reporting Corporation (202) 628-4800
- \\
230 1
water valves, we would have never found the problem with the t
y k../
2 failure to operate against the full DP if we hadn't put it 3
in that condition.
Okay.
4 Now the testing requirements outlined in there 5
don't require you to put, you know, the test of valve under 6
any condition.
So you can test the valve in a less hostile 7
condition and we would never have found the problem in this 8
case.
9 MR. MICllELSON:
And isolation valves you don't 10 generally -- they're not done either.
11 MR. IIAYES:
No, sir.
And air operated valves of 12 the actuators, their air operated valves, are a lot more l
13 sensitive than a motor operated valve.
I mean, you have to H
14 have a major problem with a motor operated valve for that
( ) 15 valve to bind up and lock that rotor up on a valve.
16 The margin between an A0V and an MOV, there's a 17 big difference.
With an AOV any small friction additive to 10 the valve is going to have a major effect on the valve.
l 19 MR. MICIIELSON:
You also have the problem with a 20 small drop in air pressure that might 21 MR. IIAYES:
Yon, sir.
22 MR. MICllELSON:
-- a noticeable effect on the 23 thrust.
Are you using largo -- just largo spring loaded i
24 pistons for your containment isolation operators?
25 MR. !! AYES :
It varies on what you're --
r lluritago Reporting Corporation
()
(202) 620-4888
231 1
MR. MICHELSON:
I'm talking about the larger ones
( )
2' like 24 inch.
3 MR. HAYES:
The larger ones?.
Most of them are
,4 piston operated valves.
5 MR. MICHELSON:
Are they air to close or just 6
spring to close?
7 MR. HAYES:
Well, on the piston it's air to close.
8 Some of them don't even have springs associated with them.
9 MR. MICilELSON:
Yes.
Some of them -- some of them 10 I'm acquninted with have springs as well as air -- the so-11 called air assist and it turns out air is most of the force.
12 MR. IIAYES:
All your questions earlier where 13 you're asking these research gentlemen if there is any 14 testing done with, you know, as far as their operated valves
( ) 15 go.
In that type of testing that they do you won't see any 16 air operated valves.
17 I don't think you'll ever run across an air 19 operated gate valve, personally, you know, to try to shut 19 off any type of flow like that.
20 MR. MICIIELSON:
No.
21 MR. HAYES:
Most of your air operated valves are 22 used for control process or a specific failed position.
23 MR. MICllELSON:
Yes.
That's containment isolation 24 is a big case that we worry about, and those are butterfly 25 valves as well, most cases.
Heritago Reporting Corporation (202) 628-4888
232 l
1 MR. IIAYES:
Most cases, yes, sir.
2 MR. MICIIELSOll I assume yours are probably 3
butterflies also.
et/4 4
(Continued on next page.)
5 6
7 8
9 10 I
11 12 i
i 13 14 15 16 17 10 19 20 21 22 23 24 25 lloritogo Itoporting Corporation (202) 620-40011
233 1
MR. MAYES:
Yes.
The last thing I'm looking at is
(}
2 the long term of this program.
And this is where we'er at 3
at right now.
That we're evaluating our data and records 4
collected, and experience in dealing with the problems, and 5
looking at the recommendations for improvements in all these 6
areas.
And the valve sets up an operator calibration.
7 That's a big problem.
I don't know how the other utilities 8
deal with tuis, but you normally have a control person, an 9
INC mechanic that's setting the controls on the valve, and 10 then you have a, well I'll call it flange head for nothing 11 else to say, that the hammer mechanic playing with the 12 bottom of the valve.
So, what this guy does doen here, he 13 has no idea how it affects this guy up here. And when this 14 guy up here has no idea how it affects the valve down here.
15 Okay, so most of your calibration procedures on an operator 16 valve are based upon a term that's called bent set.
17 Bent set is that the operator is disconnected from 18 the valve itself and you're measuring the pressure that it 19 takes to initially compress your spring in the case of a 20 diaphragm valve.
And the pressure to fully compress your 21 spring and then what's left over if for your seat load of 22 the valve.
1 23 So he sets it up on a bent set, couples it to the 24 valve, and the mechanic comes by and torques the packing 25 down and the valve won't operate.
And then he says this is Eeritage Reporting Corporation i
(202) 628-4881
~
234 1
the actuator's problem, no it's the valve's problem and t'(,j\\
2 there's nothing in between there to correlate that.
3 We're solving that problem at Davis *** is that 4
we're going to make the INC mechanic responsible for the 5
entire valve.
But it's going to take a lot of doing.
It's 6
going to take a lot of training to train these people to 7
what goes on inside the valve.
And I imagine there are a 8
lot of utilities that have that problem.
9 MR. MICHELSON:
You don't really have any, from 10 the kind of testing you've described, you don't really have 11 any appreciation of the margin that the operator is capable 12 of overcoming -- corrosion or whatever.
13 MR. KAYES:
We will get into that.
14 MR. MICHELSON:
Stem packing you can pick op if he
( } 15 gets that too tight, but what happens if it's just barely 16 just loose enough yet to work, you've passed your test --
17 MR. RAYEF:
We'll get into that tomorrow, Mr.
18 Walls' presentation.
19 MR. MICHELSON:
That gets a little tricky.
20 MR. RAYES:
Preventive maintenance for -- we have 21 a data base developed so that we can sort out the different 22 controls, et cetera, and determine what we're going to do 23 there, but like I told you now, there's none in place e.ncept 24 for EQ.
Corrective maintenance we looked at the procedures 25 involved and the actual set up of the valves.
Predictive Heritage Reporting Corporation (202) 628-4888
235 1
maintenance is going to go with our diagnostic tools that O2 w 'r e
1eg1=e-eee-ie
=c e
ei=e,toaer1-3 what testing is required.
4 Surveillance testing, we' re looking at that in 5
depth as far as how we can challenge these valves and come 6
up with a meaningful test for the valves to see what they're 7
actually doing.
Especially in your section 11 testing.
8 We're looking at that.
I 9
Training qualifications, that's a big area that to 10 me is one of the most important areas.
You can come up with 11 all of the programs and everything in the world, but unless 12 you teach the guys how to swim, you start to wonder why 13 they'er drowing all the time unless you do teach them to
[
14 swim.
(
15 Procedures is another problem.
QA/QC we have to 16 look at how involved we want to get these gentlemen into the 1
17 actual procedures and maintenance.
And what technical 18 support we can have.
19 MR. MICHELSON:
In the case of Davis-Bessie, how 20 much overview have you been getting from say, the resident i
21 inspector and regional inspector and so forth.
How much 22 attention have they given to that in your case?
1 23 MR. HAYES:
I'm glad you asked that question.
't 24 Paul Dyer and Mr. Koslof have been in my knickers a lot.
25 MR. MICHELSOtis Well, they've been watching it Heritage Reporting Corporation (202) 628-4888
Q:
236 1
closely.
()
2 MR. MAYES:
Well, they were concerned at the end 3
of our last run, prior to coming to this outage, with the 4
failure problem on the service water valves.
The inability 5
to stay in their failed position.
6 MR. MICHELSON:
Yes.
7 MR. HAYES:
They thought that we should come down 8
at that point.
We went and did a safety analysis and proved 9
that on a worse case condition, the LOCA, et cetera, that 10 our rad levels wouldn't be such that we couldn't take manual l
11 control of these valves and get them to their failed 12 position.
13 MR. MICHELSON:
How would you get them to their 14 failed position?
15 MR. HAYES:
By a hand jock.
16 MR. MICHELSON:
You'd lock it in?
l 17 MR. HAYES:
That's corroct.
18 MR. MICHELSON:
Into that position and it will 19 stay there.
20 MR. HAYES:
Okay, and also all the problems that 21 we've had with the service water valves failing to operate 22 or sticking.
They were very interested in those valves.
1 23 That's one reason why we chose those valves as one of our 24 problem valves in the program.
We did an extensive reveiw 25 of those particular valves.
What we've done with them isn't i
Heritage Reporting Corporation (202) 628-4888 1
237 l'
done.
We're on tap to replace those valves with aLdifferent
(}
2 type of valve.
3 MR. MICHELSON:
Non-butterfly type?
4 MR. MAYES:
Yes sir, we're going to go to a ball 5
valve.
We have a problem with it.
It's a valve that's 6
developed by Nelus, a company out of Helsinki, Finland, that 7
recently bought Jamemberry, and this particular valve is 8
called a Q ball.
Not Q related to the nuclear industry, Q 9
because of quiet.
And this is a valve that Dr. Kelsie found 10 for us that will meet the applications within our system.
11-The only problem we've got is figuring out how to 12 get an "N-stamp" on these valves.
But they're fantastic 13 valves that are being used.
14 MR. MICHELSON:
What type of operator will they 15 use?
O 16 MR. MAYES:
They will use a pneumatic European, 17 okay, it will be a piston spring operator.
It will take the 18 accumulator problem out of the picture.
We're going to 19 install three of these valves in our turbine plant cooling 20 water heat exchanger outlets, which are just, there's no 21 cold requirements on those valves.
22 MR. MICHELSON:
That shut off is generally not a 23 requirement either, is it?
24 MR. MAYES:
Well, you can get a classified shut 25 off with these valves with no problem.
Part of our long Heritage Reporting Corporat!on (202) 628-4888
238 1
term program is to evaluate diagnostic capabilities.
We
{' }
2 went out into the industry to find out who had diagnostic 3
capabilities for air operated valves.
There are a lot of 4
people out there that say that they're going to have --
5 Fisher is developing a thing that he presented over at an 6
EPRI conference over a year and a half ago on the idea or 7
concept that we've gone to valve manufacturers.
l 8
We went to Valtec out in Utah, to see what they 9
have in their lab, if they can translate that into the 1
10 field, and there's nothing really out there right now 11 available to the industry.
So we took, with a simple $100 12 program, a program that anybody could buy off the street, a 13 standard PC and a couple of sensors, and we're developing 14 our own program.
15 Pete will get into that tomorrow.
We've been
(
l 16 quite successful with that program.
We can tell you 17 everything you wanted to know about the valve, but were 18 afraid to ask.
Pete will go through some curves that we 19 have and what the curves mean, and how simple it really is.
20 And that's all I have.
Does anybody have any questions?
I 21 MR. MICHELS0!It Questions?
l 22 I think the record should appropriately show that
{
23 Pete Wohld does not participate in this particular 24 presentation because he has been working for Davis-Bessie.
i 25 I believe the members are allowed to ask Pete specific l
Heritage Reporting Corporation i
(202) 628-4888
()
l 239 1
questions if they wish, and he can try to answer them.
So,
(
2 that's the way that one is.
3 MR. HAYES:
I just want to say one thing, that I 4
collected everything that has come across on air problems in 5
the last say, year or year and a half, all the 6
correspondence.
And it's a stack of paper about this high.
7 With everybody's concerns about air system, their system, 8
valve problems, to do this, valves to do that. And my 9
concern is that there is a problem out there. And there's a 10 problem wa.cing to happen.
11 But unless the regulators, being these gentlemen 12 over here, give the guidelines and a good programmatic 13 approach for the industry to follow to solve their problems 14 or to realize that there is a problem, all the bulletins and 15 all tha generic letters and everything that you put out 1G aren't going to mean doodly.
17 Because the utility's going to go answer that 18 concern at the time, but he's not going to make sure that 19 that concern is continually addressed unless the management 20 of utility consider that to be a problem.
He's got all l
21 these problems that he's juggling where his money is going 22 to.
23 MR. MICHELSON:
This has been one of the 24 dif fir:ulties.
I think you have a correct observation, that 25 has been one of the difficulties with the informatien notice Heritage Reporting Corporation (202) 628-4888
i 240 1
on the process, and that is it keeps coming out peacemeal O2
-1exo e r 11v t#e er e a e cx11 9 e
e reic=1ar 3
problem area until we get to something li te the resolution 4
of a generic issue, and then we begin to see more, of the 5
picture.
But so far I guess air oper'.ted valves are not a f
6 generic issue.
Is that riyntt 7
DR. ORNSTEIN:
I'm not sure of the designation of p'
8 a generic issue or a generic problem with their operator 9
valve systems, and defined as such.
However, the agency has i
I 10 acknowledged the fact that air problems as they manifest i
i 11 themselves throughout the plants indeed are a generic f
12 problem.
13 MR. MICHELSON:
That is not a genwric iscue in the 14 sense of a list of items.
15 DR. ORNSTEIN:
Well, it's a generic issue, which 16 has been raised as such, and we feel we've got a handle on l
j 17 it ivr tha *aaalutic..
And it goec beyond what the i
j 18 gentleman here from Davis-Bessie has mentioned.
The first 19 stage of it is what we're into right now, which is to havs j
20 each plant try to examine the features of the plants, and i
21 assure that they've got a connect between the air system and i
22 the components that are fed by it.
l I
23 The second phase of the problem, which he very 24 astutoly observed, is a very important one that we' re f
25 concerned with, ind that is once you've done what you have
[
Heritage Reporting Corporation
[
(202) 628-4888 l
l
241 1
to with regard to assuring you've get e compatibility
{)2 between the air system and the equipment it serves.
And we 3
believe many plants will make quite a bit of changes to 4
their hardware to try to assure that they do have the kind 5
of compatibility.
6 The next stage is to assure that the system is 7
taken care of correctly.
And, th9 industry and the NRC is 8
presently involved in a ASME, AMS, or ANCE group to try and 9
put together a standard for -- I gueso, performance testing 10 of the air sys'3c.s.
11 MR. MICHELCON:
Which one would that be?
12 DR. ORNSTEIN:
It's presently being worked on with 13 OM 17.
I am a committee member.
We have had some 14 calancing, and we are very close 4.n approval.
The idea 15 being that on a periodic basis, depending upon which feature 16 of the air system you're talking about, you do check to make 17 sure that things are not degrading, and you do maintain it 18 properly.
19 For example, we've had a great deal of dialogue on l
20 how frequently do you want to check do point.
Mhat does the 23 do point tell you and can you detor.mine that things are 22 going astray.
Hew f requently --
23 tm. MICHELSON:
O&M in pitched towards operation 24 as opposed to design.
25 DR. ORNSTEIN:
That's correct.
As I said, the Heritago Reporting Corporatiun (202) 62 8-4 8 t"J
242 1
design part, we think we have a handle on from the
( )
2' standpoint of the present actions.
Where we go with 3
maintaining the systems properly is from the OEM end of it.
4 MR. MAYES:
The biggest problem with air systems 5
is no secret.
You have a dry particulate free system 6
supplying to your components, and if your components are 7
properly design and applied within the system application, 8
you're not going to have any problems.
9 MR. MICMELSON:
The Check valves are --
10 MR. MAYES:
That sort of thing.
That's because 11 you have a non-safety related system trying to perform a 12 safety function.
So you took a small portion of that system 13 and said okay, I'm going to upgrade that to safety related, 14 in this non-safety grade system.
You're fooling yourself.
15 I don' t know what they have to do.
16 MR. MICHELSON:
Does Davis Bessie have a safety 17 related air system?
18 MR. RAYES:
No sir, I don't think you'll find 19 anybody out there that does.
I think they' re all 20 grandfathered.
21 HR. MICHELSON:
Normal service water or service 22 air, and upgrading by dryers or something?
23 MR. RAYES:
Well, no. What we have is our 24 instrument header is red brass.
And our service header is 25 carbon steel.
And alJ. our compressors are oil free machinos Heritage Reporting Corporation (202) 628-4888
243 1
that supply both the service header and the instrument
()
2 header.
3 MR. MAYES:
Well, it comes right out of the 4
discharge of the compressors into the red brass system to 5
filtration systems, through air dryers, and to your 6
7 MR. MICifELSON :
Taking off instrument air right 8
off the compressors?
9 MR. MAYES:
That's correct.
I know a lot of 10 people use oil lubricated machines supplying their 11 instrument header, which is a problem. And a lot of people 12 have carbon steel systems in their instrument header, which 13 becomes a problem, because if you have any moisture within 14 the system, then you' re going to get particulate into your
( } 15 component. And when you' re talking about air-operated valve 16 components, you have components that have very close 17 tolerances that are causing balances or imbalances for it to 18 operate.
If you get one little spec of dirt in there, it 19 could cause that component to fail.
20 MR. M,tCHELSON:
Are you using belvel valve to 21 adjust the speod?
22 MR. MAYES:
Well, in your position, there's 23 another valve itself.
24 MR. MICHELSON:
You have to adjust the bleed off 25 rate.
Heritage Reporting Corporation (202) 628-4888
244 1
MR. MAYES:
You have to bleed off through a
() 2 balance beam.
Yes sir.
And if you have moisture in your 3
system, it's going to plug up filters in your air sets, and 4
the valve's going to start -- and it's going to cause 5
corrosion and various problems within the air set itself.
6 So the secret of a successful air program is that 7
if you have a reliable source of dry, particulate free air, 8
if you can do that, then half your battle's over.
9 The other half of your battle is to take all of 10 your components, and review all of those components to 11 ensure that they meet the design basis for your system.
And 12 that's the next hard part.
13 MR. MICHELSON:
Do you use water coolant in fear 14 cores?
MR. KAYES:
Yes sir, on all our machines, yes sir.
[}15 16 MR. MICHELSON:
You're sure they never leak?
17 MR. RAYES:
That's another problem too.
But then 18 once you have that portion of it done, take a look at your 19 components and watch them for degradation over a long period 20 of time.
And that's what we're trying to do with the 21 diagnostic system.
2'i Well, that brings up another point about dew 23 point.
You've got to look at what part of the country 24 you're operating at.
If you're sitting in Florida, your dew 25 point is not going to be as critical as if you're sitting in Heritage Reporting Corporation (202) 628-4888
.O
245 1
Ohio.
We run with a dew point of minus 40 degrees F for the
('} 2 worst condition.
If we have water drop outs we don't have U
3 freezing, or we don't have water drop out in our system.
4 In Florida, if you're trying to maintain a minus
[
5 40 degrees F dew point, it's going to be almost impossible 6
unless you use a refrigerant dryer in tandem with a descant 7
dryer.
You're never going to get there from here.
So, the 8
different characteristics and operations of the systems are 9
going to vary.
And air systems are no secret.
There are 10 ISA standards out there for them.
They've been used in 11 industry for ever and ever.
But, the fact of the critical 12 operations involved in one of these things versus a dirt 13 burner is different.
The consequences are a hell of a lot 14 different than on the other.
That's where the attention is 15 going to go.
O 16 MR. MICHELSON:
What's a dirt burner?
l 17 MR. RAYES:
It's a coal fire.
18 MR. MICHELSON:
You mean fossel fuel.
I 19 MR. RAYES:
Yes, sir.
i I
i 20 MR. CARROLL:
I don't necessarily agree with that.
21 We've got a super critical plant loses its air system, it i
22 geto pretty scary.
23 MR. RAYES:
Well, we had a trip from loss of air I
24 September 7, right before we came down.
What we expected 25 happened, or eliminated factors as far as the air wasn't i
(
I Heritage Reporting Corporation l
(202) 628-4888
[
i
246 l~
1 what actually happened at all.
We lost control of our plant
() 2 because our heater drains which caused us to go out on high 3
flux, which we never dreamed would happen.
4 I think that's another scary aspect that people 5
haven't addressed is what's going to happen to one of these 6
units if you have a loss of air, and you don't establish 7
some administrative trip when to dump your unit, and if you l
8 gain your air, what kind of transient's going to result from 9
that?
And I think that's something else that's got to be 10 really, they'll look at that and evaluate it.
11 MR, MICHELSON:
What could happen during an 12 earthquake?
Nothing is saying that those little lines 13 couldn't break.
What if there are other things falling on 14 them?
MR. CARROLL:
It's worse from that because a lot
}15 16 of multi-unit plants have a common air system.
17 MR. KAYES:
Well, Oconee for example, three units, l
18 one air system.
19 MR. MICHELSON:
Yes, well Browns Ferry has got all 20 theirs.
They went out one tiae, it blew the casing out of 21 one and the whole thing went down.
22 MR. CARROLL:
I agree with what you're saying, 23 that two or three on a plant, and what you think wouldn't 24 happen will happen.
Something really wierd, or not really 25 predictable because it really depends on how fast they Heritage Reporting Corporation (202) 628-4888
247 1
deplete and where.
()
2 MR. MICHELSON:
The precise analysis of loss of 3
air
-- we only worry about certain devices failing safety, 4
and then they do fait safety.
But there's been no analysis 5
of the transients that the system will have to face up to, 6
which have to be address'd by safety related equipment. The 7
system interaction situation has been analyzed, it's a 8
secret.
9 MR. CARROLL:
That's correct.
10 MR. MICHELSON:
It may not even be very easy to 11 nnalyze it.
12 MR. CARROLL:
That's correct, too.
13 MR. MICHELSON:
Any other questions?
Do you have 14 any questions?
I think Pete, you have actually watched 15 c17sely the work that Davis Bessie has been doing in this J
16 area.
l 17 MR. WOHLD:
John and I worked together.
l
(
18 MR. MICHELSON:
So, you would be qualified to 19 answer any questions that anybody might want to ask?
Seeing 20 none, then we want to thank you very much for your 21 participation.
I believe --
22 MR. HAYES:
I want to thank you for the honor for 23 inviting me.
24 MR. MICHELSON:
I didn't realize it was an honor, 25 but some people come here not quite giving it the same word.
Heritage Reporting Corporation (202) 628-4888
(
248 1
They describe it a little differently.
"'2 MR. MAYES:
I consider it such.
3 MR. MICHELSON:
We certainly like to hear from 4
people out doing the work.
Glen Reed used to say that 5
people in the trenches we hear about more often, and I think 6
it's very useful inpit.
7 MR. WOHLD:
I think you're about ready to close.
8 If I could raise one other question.
I think it's critical, 9
we might have gone over it a little too quickly.
That's on 10 the frictiot'. factor on the gate valves, where.3 has been 11 used in the past, and.5 appears to be more realistic.
12 MR. CARROLL: Or higher.
13 MR. WOHLD:
Or higher, right.
I'm going to 14 prevail on Joe Nadeau, if he's willing in back of me here, 15 Joe would you mind?
I talked to Joe right after the last O
16 discussion on this and indeed it appears that there are a 17 lot of plants that have set up their thrust values to the 18 calculations using the old.3 or the old valuca.
And Joe 19 had some very good insight on problems related to that, and 20 I'd appreciate it if Joe would share that with us.
21 MR. NADEAU:
Thanks again, Pete.
Joe Nadeau with 22 MOVATS.
I think what Pete is referring to is a comment that 23 he made where I felt that he really hit the nail on the head 24 when he said that the current level of diagnostics available 25 from MOVATS and vogues and Wiley and all the vendors right Heritage Reporting Corporation (202) 628-4888
249 1
now, are allowing us to set up motor operated valved
() 2 precisely to the calculation.
If it comes out that the 3
calculation is incorrect then what is really happening is 4
that they're going out and setting up redundant systems 5
incorrectly.
And over 20 years in the industry, when a 6-valve didn't work, the normal fix was to raise the torque 7
switch.
That's been accepted.
Or. change out the torque 8
switch and put another one in and keep raising it until the j
9 valve worked.
l 10 Well now you're getting in your testing motor 11 operated valves and you' re finding out that the valves are 12 over thrusting. And that in fact they're delivering more 13 thrust than they were initially set up to, especially for 4
j 14 the degraded motor voltage.
So, in order to bring them back
[}15 in line you're using diagnostic equipment to pull the 16 actuators back to deliver the thrust that the vendor's 17 telling you to deliver.
So, if we go out and test all motor 4
1 18 operated valves with a generic letter, then not only have 19 you eet up some safety related valves, but you've set up all i
?
20 safety related valves to an equation that's now being 21 questioned.
I 22 So, the question 1 presented to Pete was aren't j '
23 you putting the carrot before the horse?
You've really got 24 to address the issue of the equation.
25 MR. MICHELSON:
I think we agree with the i
j i
j Heritage Reporting Corporation (202) 628-4868 9
l
'l 5
250 1
equation, it's the factor that we use in calculations; is
() 2 that correct?
3 MR. NADEAU:
Yes.
4 HR. MICHELSON:
From your MOVATS experience, when 5
you go into set it up, you use whatever the manufacturer 6
says is the required thrust, yeu put your valve in and 7
you're assured it's all adjusted.
The operator provides 8
that thrust.
9 MR. NADEAU:
Normally, yes.
10 MR. MICHELSON:
You don't normally go back and do 11 your own calculations?
12 MR. NADEAU:
Yes, we do.
We do it two ways.
We 13 have our own calculation that was statistically developed 14 from industry data that we obtained over the testing of 15 motor cperated valver.
e 16 MR. MICHELSON:
You're talking the valves in the 17 plants?
18 MR, HADEAU:
Yes, we tend to bounce information 19 off of that data base, even though it is considered in the 20 industry to be too conservative a data base, so it is not 21 widely used.
Because it is much more conservative than the i
22 normally accepted calculations.
And we will go in and set 23 motor operated valves up to whatever we're told to set them 24 up to.
Essentially, the vendor calculation, or maybe an 25 independent as in the case of, oh, I can't think of one that I
Heritage Reporting Corporation (202) 629-4888 O
251 1
does calculations.
But several people will do calculations p) 2 more powerful.
(
3 MR. MICHELSON:
Is that generally the case that 4
you will set it up to other people's numbers, or how much of 5
the time do you set it up to what you think the setting 6
ought to be?
I guess that would only happen if the utility 7
didn' t give you a specific number to use.
8 MR. NADEAU:
Exactly.
9 MR. MICHELSON:
But if they don't give you a 10 specific number than you set it up ucing your best judgement 11 of what the setting should be?
12 MR. NADEAU:
Yes, then they'll ask us to run a 13 calculation against our data base, if they cannot get a 14 calculation of their own.
15 MR. MICHELSON:
Okay.
When you do it your way, 16 you're reflecting more the realistic friction factors as 17 opposed to the theoretical.37 18 MR. NADEAU:
Yes.
If we tried to model the 19 information we do have against a formula, we would again 20 point out the fact that the friction factor should be 21 between.5 and as high as
.7 in some cases.
22 MR. MICHELSON:
So, by that switch -- that's what 23 you mean by more conservative the people want to take.
24 Because you're going about
.5 to.7?
25 FR. NADEAU Right.
1 Heritage Reporting Corporation (202) 628-4888 O
252 1
MR. CARROLL:
But isn't the consequence of that
() 2 potentially broken yokes?
3 MR. MILLER:
Right, and undersized actuators.
4 MR. MICHELSON:
Undersized actuators?
5 MR. NADEAU:
Right, and that is a major concern, 6
because you've got a lot of double lot actuators, and the 7
module testing reconfirms this five years ago.
That the 8
calculetion, the friction factors were not conservative 9
enough.
And that the actuator sizing, certainly you people 10 are aware that there are double lot actuators that are being 11 asked to do 16 and 17,000 pounds of thrust with a rating of I
12 14,000 pounds, requalifying them to higher limits, because 13 of an actuator sizing problem.
And to go along with that, 14 that is not only limited to actuators because we took part i
{}15 in the blow down testing in Germany, in the cement's test i
16 loop, which is very similar to the testing that INEL is 17
- doing, t
18 And when the rotorque was delivered, the setting 19 it was delivered at, the close against the DP, we bounced 20 that against our data base and asked the question, what 21 would happen to the loop if the valves didn't close?
- Well, 22 they never even thought about them not closing.
They just 23 assumed it was going to close.
So they did, in fact, move 24 the setting to their maximum.
Not only did the valves 25 wouldn't close at their as-found settings, but they wouldn't Heritage Reporting Corporation i
i (202) 628-4888 i
253 1
even close at their maximum settings.
They had to jog them
{'] 2 closed in order to get the valves closed.
3 So that information is readily available too, and 4
it's all pointed to the same thing.
5 MR. CARROLL:
Okay, but if I want out to the 6
industry today, how many, what fraction of the MOV's are out 7
that are acceptable?
How many people would have gone and 8
done a diagnostic program?
9 MR. NADEAU:
I think the question is two-fold.
A 10 lot of them are using the calculation.
But also it's the 11 conservatism that is added on top of the calculation by the 12 utilities themselves that would have to be questioned.
If 13 they did a calculation and it came out they required a 14 certain thrust, if the actuators are capable of delivering 15 more thrust, many of the utilities are adding conservatism 16 to the calculation.
17 HR. CARROLL:
Even though they recognize as many 18 pairs of valves?
19 MR. HADEAU:
No, I didn't say that.
20 MR. MICHELSON:
I would think we're accumulating 21 the kind of information with which one could arrive at a 22 better judgment as to what the friction factors to use 23 should be.
Certainly from all of this 8503 work.
24 MR. NOHLD:
That's one of the problems.
I don't 25 think we're collecting the data to that extent for Heritage Reporting Corporation (202) 628-4888
254 4
1 evaluation.
( )2 MR. MICHELSON:
Their operator valves don't have 3
piston drive or something.
4 Do you have any comment on it, Earl?
5 MR. CARROLL:
Basically your pitch is that we've 6
got to understand what the hell we're doing before we tell 7
every utility that --
8 MR. NADEAU:
Yes, that's pretty much summing it 9
up, but you know, I really believe that they're heading in 10 the right direction but we better make sure that when we do 11 set it up that we're n'at overlooking the real issue.
12 MR. NADEAUt Thank you very much.
13 MR. WOHLD, Thank you, Joe.
14 MR. MICHELSON:
Are there any other questions?
( } 15 MR. WOHLD:
I guess the key thing on what Joe's
.t 6 saying is what is going to happen with this.5 view factor, 17 and perhaps there may need to be a re-look at what people 18 did under the bulletin.
As I recall, Clinton again, stande 19 out in my mind, I hate to keep picking on them, but, they've 20 got an enormous program to recover from their mistakes.
So, 21 they' re fixing themselves.
But they started out as I recall 22 with the standard calculation plus 10%, because of the 10%
23 tolerance in the MOVATS capability.
24 So, if MOVATS was reading more than it should, it 25 would read on the.3 friction factor.
Heritage Reporting Corporation (202) 628-4888
255 1
im. MICHELSON:
I believe a vendor recommended
() 2 thrust requirements based on a.307 3
MR. WOHLD:
I believe that vendor recommended 4
values.
.3 has been the standard in the past.
If they 5
increase that standard I don't know about it.
They set up 6
all their valves to that same standard. They went through, 7
they started with what they thought was a conservative 8
program and did diagnostic tests on every MOV.
9 MR. MICHELSON:
Conservative value for the 10 friction point?
11 MR. WOHLD:
Right.
12 HR. MICHELSON:
A non-defendable value.
13 MR. WOHLD:
Right.
So, that needs to be 14 addressed.
MR. MICHELSON:
We'll have to ask the staff again
()15 16 tomorrow about that.
You would think that would be one of 17 the items maybe ire the research program, but there's no 18 laboratory work involved.
It's just a saatter of spending a 19 little bit of money.
20 MR. CARROLLt This is true really only for valves 21 that you' re relying on to isolate a line work.
The one we 22 looked a this morning, it didn't seem like it needed that 23 margin, except for interrupted line work.
24 MR. WOHLD:
That's hard to say.
A lot of it 25 depends on how much open torque switch bypass a plant Heritage Reporting Corporation (202) 628-4888 O
256 1
adopts.
Some people say just past cracking is good enough.
2 Other people put 20 percent in IJke Davis-Bessie.
Some do 3
all the way.
So the dependents on the accurate setting on 4
the open torque switch depends on the bypae3.
Whether you 5
bypass that high DP section or not.
i 6
MR. CARROLL:
That's also been proven in testing 7
five years ago.
8 MR. NADEAU:
You're using the interpretation of 9
two valves that were done during the testing of a blow down 10 to reach conclusion and the Marshall testing which was done i
11 five years ago, and much testing that's been done in the I
12 industry will prove the point that the calculation at.3 is 13 questionable conservative.
14 MR. CAPROLL:
Whether you're opening or closing 15 that system?
16 MR. NADEAU:
Just closing.
l 17 MR. MICHELSON:
Even for non-dynamic cases?
18 MR. NADEAU:
Yes.
l l
19 MR. MICHELSON:
Fcr dynamic cases.
20 MR. NADEAU:
It pushes even more.
That's what 21 they're bringing out high in the INEL.
l t
22 MR. MICHELSON:
I didn't quite appreciate that l
23 point for sure.
1 24 MR. CARROLL:
Tell me more about the Marshall 25 testing, what was that about?
f Heritage Reporting Corporation (202) 628-4888 I
I I
_ _.... _ _ _ ~
257 1
MR. HADEAU:
A study that was done after the THI
() 2 incident on the PRW block valves.
And I think not so much 3
in the Marshall report but in a Westinghouse report that 4
came out, the Marshall report might have been a little bit 5
more lenient in how they term things.
But the Westinghouse; 6
report that was put out, which I can get you a copy of, was 7
very specific on the coefficient on the friction use.
8 MR. MICHELSON:
What did that say?
9 MR. NADEAU:
That the.3 friction factor that was 10 used on calculating those valves was incorrect.
11 HR. CARROLL:
What was that based on?
12 MR. HADEAU:
Based on the testing of the Marshall 13 testing.
14 MR. CARROLL:
What was the Marshall testing?
15 MR. MICHELSON:
There were blocked valves.
16 MR. NADEAU:
Right.
17 MR. CARROLL:
Marshall refers to Marshall steel?
18 MR. NADEAU:
No.
19 MR. MICHELSON:
Marshall --
20 MR. NADEAU:
Right, it's an EPRI study, yes.
21 MR. MICHELSON:
Did they recommend a number in the 22 Westinghouse report?
23 (Continued on next page.)
24 25 Heritage Reporting Corporation (202) 628-4888
258 1
MR. NADEAM:
They did recommend a number, if you
() 2 put the higher friction factor in, then it corresponded with 3
the data that they had.
4 MR. MICHELSON:
Did they indicate what they 5
thought the higher friction factors would be?
6 MR. NADEAU:
Yes, they did.
- And,
.5, thank you.
7 MR. MICHELSON:
These numbers are magic number.
8 Could you get us a copy of that?
Or get us a number from l
9 Westinghouse and we'll be happy to --
10 MR. NADEAU:
I can get you a copy.
It's floating 11 around the industry right now.
12 MR. MICHELSON:
So it is a non-proprietary 13 document?
14 MR. NADEAU:
No, I'm sure it's not, I'm sure it's
(
15 not.
16 MR. MICHELSON:
Would you send it to us?
17 MR. CARROLL:
This test was limited to the size of 18 valves, types of valves?
19 MR. NADEAU:
Right, very limited.
20 MR. MICHELSON:
We're trying to decide whether it I
21 blocked valves --
22 (Simultaneous discussion) 23 MR. MICHELSON:
I think we can probably pursue 24 this a little more with the staff.
25 MR. CARROLL:
I think we ought to.
Heritage Reporting Corporation (202) 628-4888 i
t I
i
259 1
HR. MICHELSON:
Because their research is showing 2
-- I think the answer will be, we've got to wait. I think if 3
there aren't any trore questions, I'd like to close the 4
formal record.
Recess until tomorrow at 6:30.
5 (whereupon, at 4:20 p.m.,
the proceeding was 6
adjourned) 7 8
9 10 11 12 J
l 13 1
14 O"
16 17 18
)
I 19 i
20 21 1
22 1
23 24 l
l 25 i
a Seritage Reporting Corporation l
(202) 628-4888
1 CERTIFICATE 2
3 This is to certify that the attached proceedings before the 4
United States Nuclear Regulatory Commission in the matter 5
of:
ACRS Subcommittoo on Mechanical Components 6
Name 7
8 Docket Number:
9 Place:
Bethesda, Maryland 10 Date:
October 26, 1988 11 were held as herein appears, and that this is the original 12 transcript thereof for the file of the United States Nuclear 13 Regulatory Commission taken stenographically by me and, i
14 thereafter reduced to typewriting by me or under the
(}
15 direction of the court reporting company, and that the 16 transcript is a true and accurate record of the foregoing 17 proceedings.
18
/s/
Oh M
19 (Signature typed) :
Joan Rose 20 Official Reporter 21 Heritage Reporting Corporation 22 23 24 25 tage Reporting Corporation
()
(202) 628-4888
~
m aus sus ame m
e ami e
sur sus immr aus mes sus zum muu e
amm ~ mus O
O O
o WYLE VALVE MOTOR OPERATOR MONITORING SYSTEM i
SCIENTIFIC SERVICES l
& SYSTEMS l
LABORATORIES' GROUP
{
P.O. Box 1008 Huntsville. Alabama 35807-5101 (205) 837-4411 P 10-80 48 (01)
o o
o i
l WYLE VALVE MOTOR OPERATOR i
MONITORING SYSTEM l
l i
l SCIENTIFIC SERVICES
& SYSTEMS l
LABORATORIES ~
GROUP P.O. Box 1008, Huntsville, Alabama 35807-5101 (205) 837-4411 i
e.io sua ton
I I
/l')
WYLE VALVE MOTOR OPERATOR MONITORING SYSTEM Wyle Laboratories appreciates the invitation by the NRC to present our approach to valve diagnostics.
Wyle's valve l
diagnostic system comprises two complementary segments:
MCSA (Motor Current Signature Analysis) has been developed by ORNL under the auspices of the NRC.
Wyle is the ORNL licensee for this technology.
MCSA consists of one clamp-on ammeter, signal conditioning unit, FM portable tape recorder I
and computer / software.
I V-MODS (Valve Motor Operator Diagnostic System) is utilized to establish base line parameters and was developed entirely o
by Wyle Laboratories' Scientific Services and Systems Group.
i 1
We do ao: intend to spend much time on MCSA due to the fact that it was very well described by ORNL at different meetings l
(IEEE, EPRI, etc.).
Howaver, just as a reminder, it is a non-intrusive remote technique based on the fact that the motor of l
the actuator acts an a transducer that reflects all perturbations occurring in thur. which it is driving (valve, pump, fan, etc.).
By design, MCSA moaltora and trends the indicative parameters.
I I
3o I
I V-MODS Pos Ass: 2n
)
V -N
. e'c.sists of several entities, mainly: (See V-MODS SCHEMATIC, pg. 6.)
o Sensors:
Clamp on ammeters, load washers and position transducers.
l o
Interconnect Box o
Cable Sets o
Data Acquisition Module o
Computer / Software / Display I
l Interconnect Box W
Limit and Torque switch cables are routed through this junction box which is located near thc valve and allows for 250 feet of umbilical cable to be run back to the Data Acquisition Module. The Motor current, motor voltage, stem travel, stpring o'-
pack displacenient and load washer cables also feed through thia box.
II Data Accuisition Module The unit contains a Kistler charge amplifier and Simpson voltmeter and acts as a signal conditioner for all inputs to the 1I computer.
All ranges are selected through this module and input to it comes from the interconnect box via umbilical cable.
End-
)
to-end checks of all signals are performed prior to test.
Li Load Washers I
Kistler series 9000 load washers are used to measutto actual O
1eedine of the stem duriae ve1ve c1oeure.
tecetion.
mountine 1
I torque and range of the load washers have been precisely
( )
determined from extensive testing on difforent actuators (SMB-000 and up).
Also tested were repeatability and precision.
The results are contained in Wyle test reports No.
82699-04 and No. 82699-06.
They show good repeatability and an accuracy of the stem load better than 110%.
The load washers are supplied with calibrated sensitivity by the manufacturer and were tested by Wyle against a load cell traceable to the National Bureau of Standards during the research programs reported above.
I Clamo-on Ammeterg There are lo clamp-on ammeters supplied with each system, which make up two complete setups for use with two interconnect boxes to allow "leap frog" working.
Four of the amprobes detect
]
current through the torque switch, bypass and limit switches and the fifth, and more accurate, measures motor current (AC or DC).
Linear Transducern A Celesco position transducer (string potentiometer) and Waters displacement transducer measure stem travel
".nd spring pack displacement, respectively.
Interconnect Cable Sets a
There are two switch monitor cable sets for use with the two lg Interconnect Boxes, i.e.,
for instrumenting the valves to the Interconnect Boxes.
l30 ll t
4 Unbilical cable l (]
Two hundred and fifty foot of umbilical cable is available for connection of the Data Acquisition Module to the Interconnect l
Boxes allowing for remote operation of the Data Acquisition Modulo / Computer.
I l
Carryinc cases A hard shell plastic transport case is provided to house l
both Interconnect Boxes, Interconnect cable Sets, the umbilical j
cable and transducers.
The Data Acquisition Module and Compaq t
computer are prov$ded with integrated carrying cases.
I Software l
The CODAS scientific software package developed by DATAQ (b'
Instruments, Inc.,
is used by V-MODS for data acquisition and i
1 engineering enalysis.
l CODAS features built-in screen, text and array editors, multiple graphic windows, array handling function, and plotting functions.
CODAS allows 14 channel simultanecus input and acquisition.
All required information is gathered in one l
complete cycle of a valvet close-open-close (or vice versa).
The l
display can show one or any combination of eight of the fourteen channels on a computer screen.
Sample output are provided en 1
Figures 1,2,3 and 4.
i l
'l' tgo
.I
I connuter O
we uee - co=raa "oae1 I co=9 uter vita coo ^s sortvere aaa Epson Printer.
All inputs are stored in the computer which utilizes the software package to provide comprehensive data to the operator.
All test functions are controlled by keyboard command.
Hardcopy output is provided by a r.
Epson printer / plotter.
'I t
I o i
I I
I i
I I
to I
mM W
mmM eW m
eam W
mm eeem O
O O
V MODS SCHEMATIC SENSOR VALVE A
{
Amprobes l
Load Washef l
I Data etc.
t Interconnect
//
l Acquisition Computer r
box
- }
Module l'!
//
l l
1
!l!
V l:l SENSOR VALVE B Peripherals Amprobes y
Load Washef l
etc.
t Interconnect
- l box "l
T m
SENSOR VALVE C l
Amprobes l
Load Washer l
e c.
Interconnect box O
P-to 8848 (03)
i I
I e
Q SYSTEM CAPABILITIES I
V-MODS verifies torque
- switat, torque bypass and linit switches settings, and stem load on valva motor operators by gathering valve operational data, eitL r during static actuation 3
or ander full differential pressure ca ditions.
It also provides information useful for predicting valve condition and for performing failure evaluation.
The syrcem measures eight operational parameters during valve one'..ing and closing.
Parameters Measurgd As the valve is operated frem fully close to open or, vice F
I Q versa the following parameters are measured:
L Bail 11renent Ou t p'.if, 1.
Valve Ster Travel Stta Travel (inches) 2.
Valve Stem !.m d Valve Stem Thrust (force h-pounds) 3.
Motor Current Arps h
4.
Motor Power Motor power (kv) 5.
Spring Pack Disp 1' coment Spring pack displacement (inches) 6.
Teresue S eitch Act.ivations current Change 7.
Linit Switch Activations Valtage change across I
c73 tacts c
8.
Torque By-Pass Switch Ctrrant Change T:nivations 30 m
I
0 Data Reduction
(]
All measurements are analyzed and graphic output can be displayed, plotted or printed.
Cursor control can enlarge any section of the graphic display and present elapsed time between I
any points on the screen.
Therefore,
- power, stem
- travel, current, switch activation and/or other parameters amplitudes can be evaluated on the sarre time basis using actual engineering units.
I Data Outout After on-screen data manipulation and evaluation, any selected information can be output to the printer / plotter.
Samples of V-MODS hardcopy output are shown on the following pages.
These plots are ind?::ative of the flexibility and
]
comprehensiveness of the V-MODS data system output.
5 Figure 1 Shows a full cycle of SMB-2 Limitorque actuator (close-open-close).
I Channel 1:
Shows stem travel in both directions (inches).
Stem position at any point during the stroke can be found from this signature.
'I Channel 2:
The load trace shows the load being transmitted to the valve by use of load washor(s). There is running load during 30
E the stroke, then the valve seats, the load increarcs and the O
torque switon eaute aowa the motor.
Channel 3:
Shows current signature from close to open to close direction.
The trace shows time and amplitude of:
I Close to Open--
o Inrush current o
Hammerblow time o
Unseating time o
Running current o
Open to Close--
C o
Inrush current o
Running current o
Seating current o
Suroke time I
channel 4:
Shows spring pack signature in both directions.
Spring pack displacement (inches), hammerblow and unseating tiae can be obtained from this trace.
I Channel 7:
Shows open limit switch actuation.
When the valve reaches the het point, this switch will trip and stop the valve (open stroke).
IO I
Channel 10:
Shows close torque switch actuation.
When the
(]
preset thrust has been reached, the close torque switch actuates I. %)
and stops the valve in the closed direction.
I Channel 11:
Shows the open bypass switch actuation.
This switch should cover the first halamerblow and unseating peak.
This trace verified the valve unseating.
I Figure 2:
This trace shows some of the expansion capability of the V-MODS System.
Current, spring pack, open limit switch and open bypass switch signature have been expanded for more precise analysis.
I Figure 3:
This trace also shows some of the expansion capability of the system along with analysis on 4 different signatures:
I o
Load trace shows 40588 lbf on this actuator in close position.
o Current trace shows inrush and end current (AMPS).
o Spring pack trace shows 0.18 inches compression of l
spring pack in closo direction.
o Close torque switch trace shows the closing stroke time.
I IO
'I
11 Figure 4:
Figure 4 shows partial open to close actuation of SMB-2 0
under 1400 PSI differential pressure and 530 F temperature during blow-down tests for Idaho.
I I
I I
.I I
'I 9 I
I
,I
'I llI llI lI lIo
- I i
.i.. -.mies
.. m uu volt Base line 1.56e+00 Volt CCMPR333 ION: G L
k 12 f
f TEF 11 10 7
4 3
2 1
TER g..
. g..
I (![
I
~:
~
q 4. 9 4 e + 00',
.m 1
,C.5se+00 i t, f.
j n
8 53e*00' 7.17e+00 I
i
- 1. 21 e +01l.
.l1. O s e +01 I
1
- 1. 57e+01!.
1.40e+01 1
I 1
.1
- 1. C 7e+01'
~.7Ce+01 2.29e+01:
t
.O.15e+01 2.e4e+01;
.I.
p.51e+01
.I.
I
!.00e+0!
0.872+vi t:
'I g. 27.wu 1
?.!4e+ '
3,
=
I I:T.53e*01 7.7:e-'.1 a. pi e -.,
i 0.44e+01 i
.1. J _1 3
- e d..~Ve+01 l
o a. 30,+.: 1 4.., e e -01 4,.
i 5.tfe+:t
".0!3*01 5.31e-01 l
'3.!?c+01 I
$,3 7,9 +t'. [
I t
e*UI f
I.
'a 0 4 i
2 3
2, 4,
9, 5
A 5
o a
3 s
a s
m a,
0 3
T r*
3 3
4 M
w.
cv O
5
- "3 I
a m
A D
es n
'A c
.4 H
FI;URT 1 I
i
.... s..s v r u s ~;
--u.A acate:
.e;e+00 Volt.
Base line:
' 4 e+ori vo1+
C h a n rs e t all */-Full scale:
3.4 v+og vol; Base liner 1 5 e+hh vol(
13 CCf1P A E 53 I ON : 0 I
f
- y --e 0.18 9--
34.34 A".PS 4 ---
(
'[*
TEF 11 7
.s TER I.
.c..
c.
l
..g.
}
}
'u l,
- . 52e +0<h 5
i
?
j
-) K
. a up..,
i
_.8.96e-01 I
x. J- :
o
{
.i 3:
i 0.40e+00 w J
.L.
.,5
- 1. :
. m.
it.79e+00
+
3
!?
w 3.:
Ng
~
i-i s
u i
I
- 4. 01 e+0cy en I
ji i
y'
.a.-
. i.
0.699+00
~.< - o
")
5.: 1 e.co;
.(
.l...
.p g.,
3.5ee+0c I
I d;
i li i
- 3. toe.oo j-
. -.. ?.9
.f.
-4.49e oo g
I:
I
- 7. coa +00:
i 4
t s-
.h3. Ge+00
{r I
7.90e+00:
i 1
a
.l.
.c.27e+00 y
3.Toa + r.' O J
,4, I
. t. e +'.."'
j g.
g$
e 7
i 4j' l
2,.c...-
I
?
l c-3,oc9+ou I
e o
j
@ l l
'o j
- 1. O t e +01,
- ..4de+00 t
7-w l
l
{,
v I
1.15e+o1-s 9i i
j J
7[
.p.}
'..s c e.. e 3
l 8'
I k
.t I
1.24e- -
i
.;n 1.....:s+01 c,
5:
3 O
4 3:
4 1* " a +'1
}
f.
1.!ce+01 I
E k
t. 4:e+.m 1
l a:
f' y
l.
e*0t a
Y I
1.51e+01t Q
y y
.1.
4c*v.i v
m l
- n d
- OON*O1l.
I
.f.
,].
, 3
,{,4 9+Qg 1
- a..c, v-
.r 1.$:e+Q1
~
f 1.7Ge+01:
,.n.
,1. 3 fe+01 2
-3 j,,
,l$
I y,1 5
e m
. G e e +o. l i o
- s
- s
.:a3 j :s 1.
..is -91 r+
tg 4
l 2
I i
I'
= re.o o u
n I
)
a
- 3
- a. :.y 1
1...-a-a et i
1
$.4
?
A s
. ( 'd ) +0 i i
4 TIGUPE 2 I
'24*'
I a
+
f i
2 1 s * ' '
i
i l
l I
i I
v-l
(
I I
w I
a.
L C
1
-i
.?-
5
-i o
i c.
5 i-e e
e e
e a
e s
a-t s
a o
e e.
s c.-
c.
.e t
1 c.
b a
e a
.a
.t e
a a
e e
-o
.e e
.e
.a.
g 4
e e
e
+
e e
e u
2
,4
=>
e.
te r;
r.
e c
4 c
2 s,
+
~
.--_ e.
n
_v
_. r. <.
r__
c
.z__e.
e-e e.
r i
r
,l i
4' ~
c:
e=
/
eu t
s.<
7:
ma l
J' Ioad M
y' os o
%eww..-- _* ' $}. " 'est*- - _-.
l
_T m.M* w.* nerrw eesme,tm as ces. verr 4** AMP 9FA,ine t.9hh'
. J J T;On - T.-..;.;; "fr
. aiserpeMerndmettew=*te J l
4374 Running load "If od
~
r n'
Oirrent 1
s' l
t I
I w'
t.
7
- - -j "J
l Spring Pack g
i I
I l
Close_ Torque ______._ _ __....___ __
j l
l l
o.
l I
1 I
___ 4 29.04_Sec.
j.
3
-i i
3 3
i
-I i
I i
i 3
I i
I I
I i
rs e
e
-a
~e-e a
.a
.e
-a a
e a
t e
s a
a t
a t
s e
t 6
a
.a e
e s
a a
e e
,a
,a.
.Y.
3 4
s.
f
- =
r r
c
~.'e e
e i.
e
[.
h
[
[
[
[
rs e
a-Y y
e e
e t.
=
+
T
?.
O f.
e e
Y T
Y Y
Y Y
Y Y
Y Y
[
g e
6 I
b 1
1 9
\\
in a
t
-r
~
~
i r
r r
r r
T r
w I
I I
M t
u a
o
.e s
e y
y e
.m 2,
.e.
c.
.,a
.e.
o a
3 e
s s
.s
,e a
a e
1,
~.
s
- )
x 4
c s
c s
s O
r.e 2
a.
4 s
r r
n
.. r
..e
.. t
.m...:__.2 o
- s
- 4 m
+. -.,.
~ __
f f f f.5 STD1 TRAVEL
.,' E
....?
w
, s.u..s..s, a.
s?
O :.,*. t.
4.45 ins.
i
.--t.-
_r
,, e s_
-n
????$
1 o-IDAD 7
mu I '-
$,., 0 E
,1 ?. i e,
.2
k. >WB. _s? c.
t 3254 lbf
,r n
m l
p-
~
k UEU UU n L1 y,yas n
bt a, A,
a !3 1 oroot c;
- :. n :, -
~
T L
a
.r 7..
,7.424 Amps T
5' 5 2 5~
-i
,e,e,.. i.
xy
- r. s
- m. :. 3.,. n
.L memn-r w'
SPPIfK; PACK
,8 ma p
ac
=
-.. =. =. =.. =
o.w a
P e
v u. u.u u
}-
z w
4 q
.a s
-.J J
-3303 3
- d 1 s, s, s, s.
2 x
a
-2 12;;;
CIDSE 'IUP(E.1E x
T T
.I 3
21.59 we-
-e.,..
u
=A T
- ----- 2
' =" :
=
=
=
=
=
b
=
=
=
=
=
=
=
=
=
2 eeeee I
I I
O s.
.e.
y a
e o
- i. *- f $$ c5 4-
'.4 n.
3
.o e
.a 3
e
.y
,e
,3
,e 3
c s.
p s
~*
s s
a e
=
c-t
?
D p
3 m.
c L
^u cc -
4 N.
D.
ra is te a
te
's s's
's
't
't
'e.
w i c79 9
t 6
y
?.
.=
v.
4 s
(a
-w J *.
p l
l i
I I
I i
L O
O O~
as
==
==
me
==
==
==
==
16 SYSTEM OPERATION OVERVIEW
(-
The following is a description of how V-MODS is connected to valve motor operators and how it is used to perform diagnostic testing.
I Eculoment The Interconnect Box, its cable set and instrumentation package (ammeters, load washer, string potentiometer and linear transducers) are set-up at the valve site.
The Data Acquisition Module, portable computer and, if desired, printer / plotter can be set-up remotely.
Sound powered headphone connection jacks in both the Interconnect Box and Data Acquisition Module provide for audio
]
communication between the two locations and can also assist with remote operation of the valve itself.
Installation once the Interconnect Box and associated equipment is at the MOV, installation is performed as follows:
o Valve is brought to mid-stroke using hand wheel.
o Spring pack cover is removed and linear transducer is attached.
o Stem adapter is clamped on stem and string potentio-meter is attached.
IO I
WYLE ANNOUNCES MCSA / V MODS THE NEXT STEP IN MOV DIAGNOSTICS Wyle's hiotor Current Signature Analysis (htCS A), can monitor h10V's (Motor Operated Valve's) from outside containment at the MCC (hioior Control Center). This means that this non intrusive diagnostics systein can perform faster, more efficiently and more accurately than other methods presently on the market.
Here's how MCSA / V. MODS works:
- V.hiODS establishes baseline data L
- hiCS A monitors variances from baseline via motor current frequency spectrum analysis
- htCSA diagnoses variances The Diagnosis includes:
. Stem Load
. Switch Settings
. Torque hfargin
. Spring Pack Condition
. Valve Stem Condition
. Worm Gear Tooth Wear q
. Packing Tightness
. Stem Nut Wear
. Line Voltage Effects
. Improper Worm and Worm Gear Lubrication n--,-.n,,,,-,-
- n ---m-n Advantages cf MCSA / V MODS
- ALARA i
- Red" d ^ " "
" COST REDUCTION 9n Intrusi e
- Realtime Analysis s i
If you are looking for an hiOV diagnostic alternative that gives improved results at reduced cost call Jack Bowm.m or Rich Little at (205) 837-4411.
Y G
L ABoRAToRIES GRTP 7800 Gowners Drive a P.O. Box 1008 Huntsville, AL 35807 5101
17 o
Bolts are removed where valve load can he measured and
()
load washer (s) installed.
o Actuator cover is opened and the folloking loca-tions are monitored by attaching alligator clips (No leads are lifted).
a.
Motor Voltage b.
Limit Switch Activation c.
Torque Switch Activation o
With cover open, the following are monitored by attaching clamp-on ammeters.
a.
Motor Current b.
Torque Switch Activation c.
Bypass Switch Activation o
The Interconnect Cable Set is then plugged into the
)
Interconnect Box which, in turn, is connected via the Umbilical Cable to the Data Acquisition Module.
The computer is then connected to the Data Acquisition Module.
Test Performance o
The load Washer (s),
Linear transducers, String potentiomenor, clamp-on ammeters and voltmeters are zerced-in and their limits set at the data acquisition module.
o For the test run, the computer is activated and the valve is operated in the close direction.
During the IO
18 closing stroke, the following most important parameters
(]
are monitored:
a.
Motor Current b.
Motor Voltage c.
Stem Displacement d.
Spring Pack Displacement e.
Torque Switch Activation f.
St em Load o
The valve is operated in the open direction and the following parameters are monitored:
a.
Motor Current b.
Motor Voltage c.
Stem Displacement d.
Spring Pack Displacement e.
Torque Switch Activation f.
Bypass switch Activation g.
Stem Load h.
Limit Switch Activation 5
Accuracy All V-MnOS equipment calibrations are traceable to ths National Bureau of Standards.
Laboratory quality instrumentation l
is used to provide the highest possible end-to-end system accuracies.
All transducers and the computer are readily avail-able commercially; they are not a special design.
'I IO
- I
i j
19 i
Advantaces Q
1.
V-MODS measures directly the actual closing force.
V-MODS l (J does not refer to any third party (spring pack, bearing area for strain gage, etc.).
2.
The load washer is easy to install 3.
All V-MODS transducers and computer are commercially avail-able.
4.
V-MODS will be used as a base line enhanced by MCSA. V-MODS needs to ce utilized only once until a major change is detected by MCSA trending during surveillance and preven-tive maintenance.
I o l
I I
lI l lO lI l
bd
,U PRESENTATION TO ACRS SUBCOMMITTEE ON AIR OPERATED VALVE PROGRAM BY JOHN H. HAYE3, TOLEDO EDISON O
ocT0sER ze, 1988 O '
3
~ AIR.0PERATED VALVE TASK FORCE.
e
' ()
1.0 SCOPE 1.1 SHORT AND LONG TERM ACTIVITIES BOTH OF WHICH l
WILL INVOLVE:
t c
- AC0VISITION OF NPRDS FAILURE DATA l
- PREVIOUS PROBLEM EVALVATIONS i
i
- DRAWINGS l
(])
- MAINTENANCE RECORDS 1.2 INFORMATION ANALYSIS AND EVALUATION l
1.3 RECOMMENDATIONS AND IMPLEMENTATION I
k i
I I !
i r
L
s 2.0 RESPONSIBILITIES O
2.1 SUPPORT THE RESOLUTION OF IMMEDIATE AIR OPERATED PROBLEMS 2.2 ANALYZE AND EVALUATE AIR OPERATED VALVE PROB!. EMS 1
AT DAVIS-BESSE 2.3 EVALUATE AND RECOMMfND PR0fRAM IMPROVEh?NTS TO SENIOR MANAGEMENT i
O l
t I
O t
3.0 ACTIVITIES 3.1 SHORT TERM A,
EVALUATE AND ANALYZE KNOWN PROBLEM VALVES INCLUDING DESIGN BASIS B.
RECOMMEND SHORT TERM AND LONG TERM CORREC-TIVE MEASURES FOR THESE VALVES 1.
SERVICE WATER VALVES 3.
COMPONENT CCOLING WATER VALVES
(}')
4, MAIN AND STARTUP FEED CONTROL VALVES
(~')
v
j 6
3.2 LONG TERM O-
- (_/
A.
EQUIPMENT ORIENTED
'j 1.
SET PRIORITIES BY SYSTEM AND-COMPONENT-IMPORTANCE 2.
ESTABLISH VALVE HISTORY AND INFORMA -
TION FILE FOR EACH VALVE 1
3.
EVALUATE DESIGN /SYSTEi APPLICATION 4.
RECOMMEND ACTIONS TO BE TAKEN
()
i e
l l'
I l
i
( :
I
.m k
B.
PROGRAM ORIENTED O
1.
EVALUATE DATA AND RECORDS COLLECTED AND EXPERIENCE IN DEALING WITH PROB-9 LEMS, THEN CONSIDER NEED FOR IMPROVE-4 MENTS IN:
A)
VALVE SETUP AND OPERATOR CALIBRA-TION B)
PREVENTIVE MAINTENANCE 1
c)
CORRECTIVE MAINTENANCE D)
PREDICTIVE MAINTENANCE l
E)
POST-MAINTENANCE TESTING 4
([)
F)
SURVEILLANCE TESTING G)
TRAINING /00ALIFICATIONS d
H)
PROCEDURES
!)
0A/0C J)
TECHNICAL SUPPORT i
2.
EVALUATE DIAGNOSTIC CAPABILITIES AVAILABLE AND CONSIDER AD0PTING ONE QR DEVELOPING CAPABILITY I
I i
FAILURE CONSEQUENCES ANALYSIS VORKSHEET s-DATE SYSTEM SYSTEM NO.
VALVE PRIORITT l FAILURE RESULT Y/N REMARKS
'1 1
Failure results in loss of redundancy throws unit into alert period for L.C.O.
I l
2 Failure causes forced outage I for repair or results in limited power output 3
Failure results in challange to safety system or causes plant to trip l
l t
l l
O 4
Failure causes undesirable l
l transients, decreases l
l reliability of system l
l l
l 5
Failure requires vperators l
to make undesireable exceptions,l l
realignments or invoke l
l l alternate procedures l
l L
I I
l i
l l
1 I
I I
6 l Failure is chronic; requires l
l excessive maintenance effort 1
l l
i l
i I
i l
I I
I I
I COMPLETED BY:
O REVIEVED BY:
Document No.1563 June 6,1955 Page6 s
STEM
)
T VALVE BODY mC i
M~
N y-BEARING I
SLEEVE N
N/1 L/
(316 SS)
/
/
gTEFLON FABRIC
/
/
BEARING s
s f
/
s s
/
l
/
N l
l
/
/
s s
/
N
/ s Q
/
/
\\
/
I
/
n N
s s
4'+.005 "I
POTENTIAL \\
MIN cDISC CONTACT LOCATION N
+,004
\\
x/
a
_ _M.LNf( _ _,
g
[
/
GOAD g<
~
"a% s FIGURE 1 POTENTIAL ST5M TO BODY METAL TO METAL CONTACT LOCATION O
- "' = ~ a ' ~ " a ' ~ = ' ~ =-
. b^o#
O O
EQU PMENT OPERABILITY FIN A6857 IDAHO NATIONAL ENGINEERING LABORATORY
O O
o EQUIPMENT OPERABILITY FIN A6857 Major Accomplishments (FY 89) a Complete steam flow interruption tests on typical gate valves.
- Develop procedares for assessing whether valves wE close against high steem flows and against high water flows.
- Complete evaluation of formula used to size valve actuators h steam and hot weter appications.
m Complete evaluation of data obtained from large seismic tests and effects ort
- Operability of aged gate valve.
- Behavior of pipe support strutsjind snubbers j
O O
o EQUIPMENT OPERABILITY FIN A6857 Impact on Safety and Technical issues a Contribute basis for resolving GI-87, Tature of I-PCI Steam Lines Without isolation."
u Provide basis for evaluating utility response for complying with the resolutions to GI-87 and to GI-il.E.6.1. tsitu Testing of Valves.'
s AI test results wil be used to upgrade appropriate ASME and EEE qualification standards.
O O
O
~
EQUIPMENT OPERABILITY FIN A6857 Follow-On Effort l
Because of decreasing budgets, research in the equipment operability ares is being phased out. However, where additional work may be needed include, j
E Resolution of licensing concerns that may develop from utility responses when complying with GI-87 and GI-II.E.6.1 requirements.
I a Extrapolation of effects of BWR eiwironment to effects of PWR environment on valve closing ability i
j E Determination of butterfly actuator requirements necessary for venting of BWR containtnents under accident conditions.
l E Determination of the effects of replacing pipe snubbers with either energy j
absorbers or shock suppressors. (MA ke caecaw=*ed A L f; t e r"***ck).
I l
E Maintenance of liaison activities.
d
_