ML20155A849
| ML20155A849 | |
| Person / Time | |
|---|---|
| Issue date: | 04/01/1986 |
| From: | Advisory Committee on Reactor Safeguards |
| To: | |
| References | |
| ACRS-T-1503, NUDOCS 8604100083 | |
| Download: ML20155A849 (249) | |
Text
~
OR1GTNAL O-/fdJ O UN11ED STATES NUCLEAR -REGULATORY COMMISSION IN THE MATTER OF: DOCKET NO:
ADVISORY COMMITTEE ON REACTOR SAFEGUARDS SUBCOMMITTEE ON RELIABILITY ASSURANCE (VALVES)
O .
LOCATION: WASHINGTON, D. C. PAGES: 1 - 177 DATE: TUESDAY, APRIL 1, 1986
.e
)$$$OEYISE$0b[
go XotRemove from ACaSORice 1
ACE-F.E.DE,3RAL (_ R,,EPORTERS, INC.
O" ,
, u. -
OfficialRbporEh h {
7 444 North CapitolStreet Washington,.D.C. 20001 i
(202)347-3700 )
a50ucoce3 e60 col
. .:gy png
CR26378.0 REE/sjg I
,_.s UNITED STATES OF AMERICA !
' s
\'-'/ 2 NUCLEAR REGULATORY COMMISSION '
3 ADVISORY COMMITTEE ON REACTOR SAFEGUARDS i
4 SUBCOMMITTEE ON RELIABILITY ASSURANCE (VALVES) ,
5 Nuclear Regulatory Commission Room 1046 6 1717 H Street, N.W.
Washington, D. C.
7 i
i 8
Tuesday, April 1, 1986 9
The meeting of the subcommittee convened at 8:00 a.m., f 10 Mr. Carlyle Michelson, chairman, presiding.
11 ACRS MEMBERS PRESENT:
12 i rm \
() j3 MR. CARLYLE MICHELSON l i
MR. CHARLES J. WYLIE f 14 !
MR. GLENN A. REED j 15 i i
16 q W. LIPINSKI, Consultant RICHAR MAJOR, ACRS Staff Member i 1
18 , !
i; i i 19 l,l ti h
20 1 .
l 21 l
22 l 23 Aa-ruinal Reporters, inc 25 1
i I
e
PUBLIC NOTICE BY THE
'{ }
UNITED STATES NUCLEAk REGULATORY COMMISSIONERS' ADVISORY COMMITTEE ON REACTOR SAFEGUARDS TUESDAY, APRIL 1, 1986 The contents of this stenographic transcript of the proceedings of the United States Nuclear Regulatory Commission's Advisory Committee on Reactor Safeguards (ACRS), as reported herein, is an uncorrected record of
- the discussions recorded at the meeting held on the above date. -
No member of the ACRS Staff and no participant at
- () this meeting accepts any, responsibility for errors or inaccuracies of statement or data contained in this transcript.
i e
I
%w -
y-- ,w - - , - - , , --- ,, , - - - , - -w - - . - , _ . , - - , , , ,
.,,,-,,-~,---.---wv 'a, s-,-W--**- -***-,w---*'vw-- -
26378.0 2 REE n
i i x_f 1 PROCEEDI NGS i
2 MR. MICHELSON: The meeting will now come to 3 order. This is a meeting of the ACRS Subcommittee on 4 Reliability Assurance and I am Carlyle Michelson, chairman 5 of the Subcommittee . Other ACRS members present today are j 6 Charlie Wylie . who is the cha ir -- who is to chair the 7 Subcommittee in the future and Glenn Reed, and Jesse 8 Ebersole I don't think will be here. Also present is ACRS 9 consultant Walter Lipinski.
10 Richard Major on my right is the assigned ACRS t
11 Staff member for this mee ting.
12 The purpose of this meeting is to discuss the
(} 13 results of EPRI NP-4254, " Improvements in Motor-Operated 14 Valves." Also to be considered will be- NUREG/CR-4380, 15 " Evaluation of Motor-Operated Valve Analysis and Test 16 System to Detect Degradation, Incorrect Adjustments, and 17 Other Abnormalities in Motor-Operated Valves," and i 18 NUREG/CR-4302, " Aging and Service . Wear of Check Valves Used 19 in Engineered Safety Feature Systems of Nuclear Power
- 20 Plants."
4 21 A transcript is being kept of the entire meeting.
22 It is requested that each speaker first identify himself or 23 herself and speak with suf ficient clarity and volume so 24 that he or she can be readily heard.
25 We have received no written statements from the
() '
ACE-FEDERAL REPORTERS, INC.
202-347-3700 Nationwide Coverage 800 316-6646
_ _ _ _ _ _ _ _ _-______-______-___-____-______-___-_-_____-______-_-__..__2_-__-._._______.
26378.0 3 REE i (~)
(.- 1 members of the public.
,- 2 We have received no requests for time to make 3 oral statements from members of the public.
4 Do any members of the Subcommittee have 5 questions or comments they wish to make about the proposed 6 agenda?
7 We will proceed with the meeting. But before we 8 proceed, we have had a number of changes in our 9 Subcommittee chairmanship assignments. Charlie Wylie will 10 chair this particular subcommittee at the end of today and 11 thereafter. I think at the end of the day, if we have a
, 12 little bit of time, we would like to discuss just where we i
{} 13 should be going from here , what kind of report should be 14 made to the full committee and any other business of that i
15 sort. We will have to end promptly at 1:30 or perhaps even j
16 a little short of that if we can. So the summary session 17 still should have adequate time to sticp to the agenda.
4 18 We will proceed _ with the meeting. I would like
} 19 at this time to call Mr. Boyd Brooks in EPRI who will make 20 a presentation on the work that they are doing.
21 MR. BROOKS: Good morning.
22 I wonder if we can -- can everybody see the 23 screen clearly? Give it a try. See how it'looks.. Can 24 everybody see that now?
25 Anyone that can't hear me, please sound off.
xs 1
ACE. FEDERAL REPORTERS, INC.
202 347-3700 Nationwide Coverage 800 336 6646
26378.0 4 REE
~
1 My name is Boyd Brooks. I am a project manager 2 with EPRI. I have had a project ongoing, RP2233-2, which 3 we are taking a look at some ways of making improvements in 4 valve operator performance. I am very appreciative of 5 being able to make this presentation to the Subcommittee.
6 As the title sheet so says, the objectives of -
7 the project are to identify root causes of improper 8 performance of valve motor operators, to develop means to ,
9 effectively correct the performance deficiencies. By " root
, 10 causes," that would be something that would be not just
- 11 trying to put band-aids on, but trying to get right down to l 12 the bottom of the thing and determine what is causing the l
{} 13 problem. And the n to develop means to ef fectively correct 14 those deficiencies, and to provide for the technology 15 transfer. This could take the form of better maintenance 16 practices, better lubrication practices, a device that 17 would replace certain components.
j 18 W'e are trying to take a look at all aspects of 19 the thing. Finally, to provide for the technology transfer 20 and commercialization of the developed corrective actions.
21 We think that is going about it in the right way.
22 (slide.)
23 In the early phase of the work, we did do some
- 24 literature searches. The OPEC data base was a lot of help 25 to us. The work that the folks down at INPO had done was
. ()
4 ACE-FEDERAL REPORTERS, INC.
202-347-3700 Nationwide Coverage 2 336-6646
26378.0 5 REE O)
(_ 1 certainly a real boost to us as far as trying to pinpoint 2 just where the problems are. The INPO report, I think most 3S T3 folks are familiar with that, runs down s' tere the problem 4 areas are and seems to be the lion's share -- seems to be 5 in this area here where we are picking up something in the 6 order of 32 percent of total failures. There is a little 7 thing we ought to be looking at here, too, it is one of the 8 nuances, I guess you would say, of data bases. A man has 9 to make up some interpretation of what is going on and 10 certainly some areas here of bent stems could most 11 assuredly be as a result of a torque switch failure or 12 torque switch out of adjustment, same way with damaged
/~T 13 valve seats.
O 14 Failure to operate could be something -- we have 15 read a nunber of reports that seem to indica te that the 16 torque bypass switch, if properly adjusted, could also 17 result in it. So there is a lot of things that are leading 18 us towards the taking a good hard look at what we have in 19 the torque switches and the limit switches and see what we 20 could do there. Electrical failures seem to be one of the 21 things, too; that could be motor burnout, as a result of 22 applying lock rotor current for whatever reason and just 23 main ta ining it for the time necessary to do the damage.
i 24 That also seems to be a sore point.
25 Project results, trying to emphasize what are
! ()
ACE-FEDERAL REPORTERS, INC.
202-347 3700 ' Nationwide Coverage 800-33M646
l 26378.0 6 I REE l l
\
(~)/
N_ 1 the major performance deficiencies, which are identified as ;
2 failure of the electromechanical switches, and dif ficulty 3 in achieving the correct adjustment of those. I have read 4 over some of the literature on it, the instruction manuals, 5 and it is not an easy thing to do, to properly adjust these 6 things and have the unit work correctly. Additionally, the 7 inadequate thermal overload protection, where we are kind 8 of dependent upon an indirect type of action, rather than 9 taking some -- and in a timely fashion taking some direct 10 action that would prevent the burnout of a motor.
11 Looking at all of these things, we decided that, 12 my contractor and I decided that it would be a good idea to fG LJ 13 start taking a look at some sort of a system that will 14 perform the function of the torque switches and the limit
, 15 switches that seem to be causing so much of a problem. We 16 have developed a microprocessor system that does that, 17 performs those functions and additionally it does provide l 18 diagnostic information in the process of doing it, the 19 primary function of which is intended as a control function.
20 MR. REED: Be fore we ge t o f f on the 21 microprocessor, I am looking at your root causes and 22 problems and I am thinking of the long history in valves.
23 I am thinking of gross problems, not the things that you 24 are talking about, such as bent stems, which are smaller q
25 problems. I am thinking of gross design problems as ACE-FEDERAL REPORTERS, INC, 202 347-3700 Nationwide Coverage 800-336 6646
, 26378.0 7 REE 4
(~)/
\_ 1 examples that I have run into, or gross misapplication 2 problems.
1 3 MR. BROOKS: I~have a Vugraph on that coming up.
i 4 MR. REED: I am thinking of cases where the 5 manufacturer delivered a valve that the stems were one inch 6 too short so they couldn' t achieve full travel. All kind 7 of gross things like this have happened in the industry.
8 They get reflected in the fa ilure rate?
i
- 9 MR. BROOKS
- You bet.
10 MR. REED: I just wonder. what attention is given
[
11 to these gross misapplications and misdesigns.
12 MR. BROOKS: As I say, I do have a Vugraph that
?
(} , 13 I wanted to cover that. It is in -the process of how we --
- 14 when we get into our in-plant demonstration, we want to i
l 15 start taking a look at all of the things right down the
{ 16 line. I will certainly get into that. Please ask i 17 questions as I go through that. ,
l 18 MR. MICHELSON: Before you go on, let me ask, in j 19 your project objectivas, one of your objectives was to I
l 20 develop means of ef fectively correcting some of the 4
l 21 performance de ficiencies. Clearly one of the means was'to 22 use a new type of control syatem, which you are going to
.t 23 tell us about. But the re is certainly perhaps a number of 24 corrective measures that could be taken to take care of the i 25 equipment that is already there in a little different
- ()
i
! ACE. FEDERAL REPORTERS, INC.
) 202 347-3700 Nationwide Coverage 800-3E6646
_ - , _ . _ , . , - - - . , _ - , - _ . _ . _ . . . _ - - . _ _ . . - - , _ , - - - . - _ - .~ _ ._ . - . . . - , -
26378.0 8 REE 1 manner and so forth. In reading the report, there seemed 2 to be little or no mention lof that sort of thing.
3 Are you writing another report in which you 4 indica te how you will handle the present equipment as 5 opposed to a new type of design?
6 MR. BROOKS: That is correct. We will see that 7 in the Vugraph, the one that I wanted to spend most of the 8 time on.
9 MR. MICHELSON: Are you going to write a 10 separate report on that?
11 MR. BROOKS: Ye s . There will be a final report.
12 The report that was issued, NP-4254, is the interim report.
(} 13 That will be duplicated in its entirety and then the additional work that we do, as far as how we want to 14 15 implement these things, how we want to take a look at all 16 of the other aspects. We have put a lot of time and ef fort 17 into deveivping that microprocessor. We think that we are 18 going to pick up significant additional activities just in ;
19 the process of setting up our in-plant demonstrations.
20 MR. MICHELSON: Maybe I am still not making my 21 point clear. My concern is that developing microprocessors 22 and so forth is an ir, te ro s ting idea, but it takes a lot of 23 time and development and proof-testing. In the mean time, 24 the plants are going, ac they are today. Was there 25 anything in the process of your studies so far to indicate O
l l
ACE. FEDERAL REPORTERS, INC.
202-347 3700 Nationwide Coverage 8(53E6646
26378.0 9 !
REE l 1
- 1 corrective measures that could be taken on today's plants i 2 to improve ' the reliability of the valves while we are i i i
3 waiting for something like perhaps a microprocessor in the
- I I 4 future? l
- 5 MR. BROOKS
- I think one of the reasons we did j
6 take a look at the MOVAT system in the interim work was --
- 7 MR. MICHELSON
- What I am really interested in,
! 8 have you gone out into the industry to find out and examine i
9 more carefully the kinds of problems they run into and are 10 you going to write down what you found out, so T can read a
j 11 about it?
j 12 MR. BROOKS: Yes. I think it covers that in the l
l
(} 13 14 report, the fact that we did go out to supplement the
{ information that we presented here with in-plant i 15 discussions with the people who are actually operating the 16 equipment.
l
- 17 MR. REED
- I think what Carlyle is driving at is j 18 something that came up in one of our meetings. A i 19 Philadelphia Electric man made some statements which were 20 very fine s ta teme n ts .
.l l
21 MR. BROOKS: I have a copy of that.
! 22 MR. REED: It has to do with the fact that the i
l 23 valve industry is a very fragmented industry, very i
j 24 compe titive industry. I am trying to get at what you are I
- 25 driving at. Part of the problems in valves have - to do with j
i l ACE-FEDERAL REPORTERS, INC, 202 347 3700 Nationwide Covernse 800 3 4 6646
26378.0 10 REE j j 1 this competitive , fragmented, not coalesced industry and 2 the products that are dumped onto -- and they used to be
! 3 dumped onto steam plants, and nobody seemed to care very 4 much. Now they are dumped onto nuclear plants. There is ,
j
! 5 no underwriters' codes or standards for these things. I 6 think what you are asking ist Has EPRI looked into that l 7 aspect?
1 I 8 MR. MICHELSON: That wasn't what I was asking, 4
9 but that is another good comment.
?
I 10 MR. REED: Now, Philadelphia Electric person i
11 le f t us with the idea that perhaps he is going to go back 12 and look into some of the fragmented problems of the valve 13 industry. ,
l 14 MR. BROOKS: Well, as far as making a correction
. 15 there, that would be a rather monumental task, to solve all l 16 of the problems in the valve industry. And there are 17 significant ones.
18 What I think we need to look at --
! 19 MR. REED: When you talk about root problems, l j 20 those are root problems, not attaching more instrumen ta tion 21 to a valve .to say that it has got all those problems. You 22 go root problems.
l 23 MR. BROOKS: Well, one of the root problems I 1
24 think you are very clearly describing is the fact that i 25 perhaps there was not correct engineerira done on an !
- O i
ACE. FEDERAL REPORTERS, INC, I
202-347 3700 Nationwide Coverase M&336-6646
26378.0 11 REE 1
n v 1 application initially. One of our objectives in the final 2 report will be to advise the utilities on the correct way 3 to evalua te . Has someone done their homework and correctly 4 applied the actuator to the valve? Part of that also, has 5 the architect engineer used the correct delta P across that l 6 valve which it will experience? Has he - designated the 7 correct voltage that would apply both for the drive motor 8 and for the control system? Do the two match up? The se l
q 9 are some of the things that we are going to be looking at
- 10 as we go into our in-plant demonstrations. We will make 11 very clear the fact that these are the things that a person, 12 that a utility has to look at if he wants to assure success
) 13 in the operation of a valve motor operator.
I 14 It is like saying, I know that the transmission i
j 15 just fell out of my car, but I am going to go down to the 16 garage and have the guy put in a new set of spark plugs and 17 all of my problems are going to go away. That is not i 18 realistic. I think the same situation exists here, that we 19 have to look back and see if the thing has been correctly 20 applied for the correct function that it has to serve.
21 That is the slide that I plan on getting into and spending j 22 most of the time on.
l 23 MR. MICHELSON: I think it will probably come 24 out later in the discussion.
i 25 MR. BROOKS: I hope so. If it doesn't --
!O
! l
! ACE. FEDERAL REPORTERS, INC.
202 347 3700 Nationwide Coverase 800 336 6646
26378.0 12 REE 1 MR. MICHELSON: We will holler again.
- 2 MR. BROOKS
- Good.
3 (Slide.)
4 Let me continue on a little bit f ur the r ,
. 5 discussing just what the microprocessor does. I am sure <
l t
6 most of you are familiar with how a valve motor operator
] 7 works. And I would presume you are also familiar with the 4
8 fact that you are pretty blind as far as what that thing is 9 really doing in midstroke. Or if you push the button, to t
i 10 say, valve, I believe that you are open , I am going to push 4
4 11 this button, I believe that you are going to close. For l
l 12 perhaps 30 second or a minute , perhaps even lancor than i
l
{} 13 14 that, the operator really doesn't know what is happening.
To get around that problem, we do provide the 15 microprocessor with a number of things: What is the actual I 16 motor voltage being provided, the motor current and the t
l 17 phase angle be tween those , so that we can come up with a 1
! 18 calculation of what the motor load is on the valve , the l
i 19 motor current gives you some indication. But the voltage '
20 has a significant ef fect on that. We think that the use of 21 a device which will calculate what motor load is and then 22 give us a signal and tell us, hoy, you are exceeding some
! 23 value that you have preset into the microprocessor to say, i
j 24 you are going to be doing damage to the valve in a very 25 short order if you don't do something. The setpoint that i
t 0
i
- ACE FEDERAL REPORTERS, INC. !
202 347 3700 Nationwkle Coverase 800-336-MW6
-_ . . .= . _ - -. . -
- 26378.0 13 REE ;
I had been preestablished would try to do that. ,
i 2 We will also be picking up valve stem position 3 on a continuous basis and valve stem load which would be, 4 the valve stem load in compression or intention, le t 's
! 5 assume we have a gate valve here , we will be determining, 6 based on the deflection of the torque switch spring pack.
i 7 That information is evaluated in an algorithm 8 such as this.
j 9 (Slide.)
10 Le t's assume we are in the open position. Wo I
. 11 want to go into closed, to the closed position. We i
! 12 evaluate stem thrust and future stem thrust. I will get
(} 13 14 into that in a little bit. If stem thrust is Loo hign, for 3
something may be blocking the valve, we can bypass the
! 15 setpoint that we have established for that, go immediately 1
] 16 to a stop to prevent damage to the valve .
17 MR. MICHELSON: There is a little philosophical j 18 problem in this particular logic, as is true of many logics I
i 19 in the nuclear plant. That is, in the event that the valvo 1 20 must close to perform its safety function, like, let's say l 21 a pipe was just ruptured and the water is pouring out and 1
- 22 you have got to get tho valve closed, one school of i 23 philosophy says I will do whatever I can do to get the l 24 valvo closed, because if it doesn't close, I am in deep i
i 25 trouble. I certainly don't want any kind of protective l ACE FEDERAL REPORTERS, INC. t l 202 347 3700 Nationwide Coverase m33HM6
- _ - - _ . . _ - , _ . _ . _ _ . _ _ _ . _ _ . - . _ _ _ . _ _ _ _ _ _ _. - . _ _ _ _ ---. _ - _ ~ _ . - . . . _ , . _ _ _ _ -
26378.0 14 REE (3
\_J 1 devices now to protect the motor. The motor of the valve 2 is not my concern. The water on the floor is my concern.
3 So I don't like to see logic circuits that stop the valve 4 simply because the thrust is too high.
5 The other school is, I better protect the valve 6 because I will need to have it function again, perhaps 7 during this same scenario at a later point in time or 8 something and, therefore, I want to protect it now for 9 later usage during the same accident.
10 How do you reconcile these dif ferent schools.
11 MR. BROOKS: One of the advantages that I see in 12 the use of the microprocessor and in the other information
("T 13 tha t we develop in setting it up, the setpoint that we LJ 14 establish for stem thrust can be , depending upon the 15 application of the valve , can be up to the very point of 16 virtually doing damage to e i the r the mo tor , to the valve or 17 to some of the other structure.
18 MR. MICHELSON: If the motor is about to stall 19 anyway, then you might as well stop.
20 MR. BROOKS: Yes. Or if --
21 MR. MICHELSON: Sotpoints generally aren't 22 selected that way on valvo protective circuits.
23 MR. BROOKS: One of the things that we do 24 evaluate, let's say this is a valve stem. You remember old 25 Mr. Oiler's equation that said that depending on where the Ov ACE FEDERAL REPORTERS, INC, 202 347 3X0 Nationwide Cmcrage Exk33MM6
... , .-. . ~. . _ . - . . . - -
26378.0 15 REE i
1 guides are on this thing and you put a certain column 2 loading on that thing, you can buckle that valve s te m .
! 3 That may be the weak point of the valve.
{ 4 And we do that, we will be evaluating that in 5 our process to determine what is the setpoint for that stem 6 thrust. It may be dependent upon the application. It may 7 be more desirable to go to motor load as the limiting i 8 factor, if the valve is very very strong, much stronger i
! 9 than the actuator itself. That can be evaluated and used i
10 to select the se tpoint that we put into the microprocessor.
- 11 We also take into account the fact Lhal we 4
i 12 realize that there are things like inrush current which are i
(} 13 quite high, although momentary; we recognize that there is such a thing as a hammer blow ef fect when the valve is i 14
- 15 starting to close. You will got some very high loads there.
16 We have means of ovaluating those things and taking them 17 into account in the use of the microprocessor.
18 Going in the other direction, we look at 19 essentially the same things. Again, whatever the i
j 20 application of the valve would be, we could select tho
- 21 setpoints an appropriate to achieve the function. '
l! 22 Gotting into the slido-that I did want to spend i
23 the most time on.
f 24 (Slido.)
25 Everyono -soo that okay? We do have currently
{
l(2) l
- ACE FEDERAL REPORTERS, INC.
202 347 3700 Nationwide Coverage 80n.336 6646 A . _ ____
. . - ._ __ _ _ _ _ _ _ _ _ . . _ .._m . .- - _ _ _ , . . _ . _ _ . ,
i i 26378.0 16 i 'REE ,
i !
I going on right now negotiations with a utility to have j 2 in-plant demonstrations of the microprocessor application. ;
j 3 The various things that we are going to do to be assured of >
! 4 success are as follows: First of f, determining what the ,
I j 5 valve performance requirements are. I am in full agreement ,
i I
- 6 with you, Mr. Michelson. There are cases where valves are <
1 7 pulled off the shelf as a catalog item, because it j 8 certainly, it may turn out to be a price thing. Somebody I !
j 9 can furnish a valve at half the price and they just put it :
10 in. They don't really perhaps know totally what it can do.
{ i 11 We want to determine what the valve performance f
i i 12 requirements are, how fast does it have to open and close. l 1
4 13 MR.. MICHELSONt. Are you going to do this by .
i 14 going back to the basic process that is being performed to j 15 determine that the architect engineer even understood what f 16 the performance requirements might be? 1 1
g 17 MR. BROOKS: As far as we can go. There may be ;
i 18 areas where we can't go back all that distance, but we are ,
) r i 19 going to try to determine what are all of the functions t
20 that are required for that. ,
l 21 MR. MICHELSON: You ought to be able to trace [
! t
! 22 back the design requirements for critical application, all l
} ;
i 23 the way back. ,
[
I 24 MR. BROOKS: I hope you are right. It will take
! i
- f 25 a lot longer if we have to do it the hard way. ;
l () i i :
1 I
i i ACE FEDERAL REPORTERS, INC. !
j 202 347 3700 Nasionwide Coverage 23Em46
26378.0 17 REE
()
i 1 One thing I would like to mention, for the i
2 in-plant demonstrations, we do not plan on using the 3 microprocessor initially on any critical applications. We 4 want to have it in an area where it can be accessible ,
J 5 where we can run history of performance on a certain valve.
6 We want to have it apply to a valve that could be actuated 7 perhaps even on a daily basis or a weekly basis to 8 de termine the buildups of cycles of operation, or to see if 9 the thing does degrade over a long period of time. l i
10 That one can be an easy one. I know that there i 11 are -- there is usually a lot better documentation for the 12 critical applications of valves. You start getting down at
{} 13 14 the balance of plant and things start to get a little bit mushy down there. But we will try and nail down what the 15 performance requirements are. ;
16 MR. MICHELSON: Are you in a position yet to 17 indicate what utility might be doing this testing?
18 MR. BROOKS: Duke power looks to be the one that 19 is closest to our -- we don't know what plant it might be 20 going into, but we are hoping that we will get in there.
21 That is subject to their agreement. I am a guy with my hat 22 in my hand.
23 MR. MICHELSON: Wo just wanted to know who to 24 watch to see what was happening.
25 MR. BROOKS: Mr. Eissenberg has been in contact O i ACE. FEDERAL REPORTERS, INC.
202 347 3700 Nationwide Coverage a n33 6646 l
.. . ._ . - _ . . - - _ _-~- - - -- .-
i 26378.0 18 REE w
I '
I with our contact at Duke Power. So there is a tie-in there ,
i 2 already.
i 3 MR. REED: I am interested in a statement you 4 made a bit ago about performance requirements and the fact l l
} 5 that valve manufacturers maybe go to the shelf and pick a t
6 valve off the shelf.
! 7 MR. BROOKS: I don't know that.
8 MR. REED: I think there is a lot of truth in i
! 9 that. I think therein lie the problems. It is architect 10 engineers and qualified designers are fairly close to be ing
, 11 able to write a specification that is pretty good for the i
j 12 environment which, the valve is supposed to function, but I i
l
{} 13 think that the f ragmen ted industry, the compe titive ,
14 fragmented industry, then, will road the spec and go grab -
) 15 some valve and maybe te st the load valve after they have 4
l 16 polished it up and got it in proper position, and got somo I
17 sort of data sheet that says it moots it. But it is not a f l J
18 representativo valvo. It is a competitively produced ;
i
< P9 design. It is from this fragmen ted industry that doesn' t i
1 20 understand what their valvos do out in the field anyway. i 1
j 21 Therein lies the problem: That the valves that 22 are delivered -- wo just saw a horribio example of some j 23 chock valven on SONGS 1, five check valves all failed at 24 the same timo , all key valves. They woro absolutely
- 25 misdosigned. But I am suro for the first day or two, they C
- ) 1 i
h
- ACE. FEDERAL REPORTERS, INC.
202 347 37(M ~ Nationwide Coverage stW) 33H646
_ . . . . _. . _. . . . _ - . _ . . _ . _ . . . _ - - _ . _ . _ _ . _ _ . ~ . ._ . . . - - . _ . . .
L i
26378.0 19 REE l 1 met the requirements. Therein is the basic and gross 2 problems to the valve industry. I am wondering -- thank j 3 you.
j 4 MR. BROOKS: Maybe -- let's explore that. Kaybe 1
i 5 we can go on through here. You are bringing out some very L
{ 6 good points. Maybe -- I am just postulating. Maybe what 1
7 has to be done is to assist utilities, architect engineers,
) 8 whoever, in the correct manner to prepare- a specification.
J j 9 MR. REED: I said I think we are very close to a ,
J l 10 specifica tion that is meaningful, but ~I don't think we are i -
l 11 getting the product f rom the shelf that you are talking ,
j
] 12 about.
i f (} 13 MR. BROOKS: You are bringing up an interesting l 14 point. One of the projects that I have coming up is going i
15 to be a valve design handbook, coming up in the next --
) 16 whatever budget is going to make the funding available to l
17 do that. l 18 MR. REED: And that is going to be for use by i 2
19 valve manufacturers? <
! [
{ 20 MR. BROOKS: It could be used by anybody that ,
4
} 21 wants to use it. I am aiming, because my charter says the !
\
j 22 utility is my hero, I want to try and satisfy my hero -- !
l 23 MR. REED: You want utility companies to be j 24 going into the valve manufacturing business? l 1
l 25 JtR. DROOKS: No, to be able to de term ine , has a ,
l(2) i
\ ,
i f
r i ace FEDERAL REPORTERS, INC. [
202 147 3700 Neilonwide Cove, age MS)M4646
. ~ . .. - - ..~ .. - . .~. .--- -.. - .. - . - . . - .- .
s i
i 26378.0 20 ,
j REE l 1 valve been correctly designed?
i 2 MR. REED: You want utility company engineers i
t 3 then to do the detailed design of valves, l
! 4 MR. BROOKS: Looking at the industry as it 5 stands now, there is relatively little --
1 ,
j 6 MR. REED: What I am trying to load you into --
1 l 7 MR. BROOKS: Thero are very few valves being :
8 purchased today, and if the utility industry is aware of ;
}
1 j 9 correct design principios, they may be able to prevent +a i 10 problem. ;~
}
i 11 MR. REED: I am trying to Icad you into the fact -
t j 12 that there is a problem in the valve design and .
13 manufacturing industry. And you represent EPRI and 14 utilities. So I am not leading you in very well. Thank ,
i 15 you. !
j !
- 16 MR. BROOKS
- Okay. The next item that we want !
4 l 17 to do is to verify that the motor and control power ,
18 supplies are correct for what the application is intended 1
l 19 to be. Sometimos motor voltages, if the valvo is way down j
- 20 at the end of a power supply, might not be getting adequate !
} !
- 21 power supply. Voltages might be way below what they are ,
- 22 anticipated to be at the motor control centor. Very 1
i 23 possible. Uo plan on taking a look at that. Also, the i s
] 24 control power supply, one of the things that we nood to i
1 25 have more for the microprocessor is a supply of 120 volt AC, :
l0 I
f
- ace. FEDERAL REPORTERS, INC. f' i 202-347 3700 NationwideCo erage 800.336-4646
! r
26378.0 21 REE
( 1 reliable supply of that. .
2 MR. MICHELSON: The design of the motors of
- 3 course can be varied somewhat to compensate for some of the 4 electrical power problems, for instance? Are you going to 5 go into that area?
6 MR. BROOKS: Well, that could be a fix, 7 dependent upon what we dotermino here. If there is some !
8 modification necessary, what we may have to do is that.
9 MR. MICHELSON: I don't know to what extent 10 Limitorque will vary the characteristics of the motor to 11 suit the customor_or do they just have a standard one or 12 two types of motors that come of f the production line or --
(} 13 MR. BROOKS: I don't know that I could speak 14 directly for Limitorque. I am aware of the fact that 15 motors are provided with valve motor operators. To be 16 fully operable with nominal voltage, up to 110 percent of 17 nominal or down as low as 75 or 80 percent, thoso ,
18 specifications are normally required at the time the 19 equipment is --
20 MR. MICHELSON: But the torque characteristics 21 can bo very significant by changos in motor design.
1 22 MR. BROOKS: But that would mean a change in the 23 motor.
24 MR. MICHELSON: In other words, how much slip, 25 for instanco, they can experience before they breakdown, O
ACE FEDERAL REPORTERS, INC.
202 347 3700 Nationwide Covernee sin)SeM6
26378.0 22 REE
( 1 things of this sort. These are -- the problem with the 2 valve business, I think, from the utility viewpoint, has 3 been that the valve and its operator have been supplied i
j 4 somewhat as a black box, with very minimum knowledge of the
! 5 design basis for the internals of that box, such as the 1
l 6 motor characteristics. You can get the motor curves, but 7 the utility engineers don' t generally know quite how that 8 fits into the total program. You have to understand more 9 than the motor. You have to understand the gear train i
10 design, the whole bit, if you really wish to know how that I i i.
j 11 black box works. i i
i 12 MR. BROOKS: That is what we are trying to do
(} 13 with this whole laundry list.
i 14 MR. MICHELSON: I assume the manual you are 15 working on would be mostly educational. I wouldn't expect
] 16 a utility to design the valve, but they would maybe be I 17 almost as knowledgeable as the suppliers, so they can 18 better communicate with the suppliers as to what they need. !
- 19 I think suppliers in general are agreeable to do things, if i j 20 you tell them what you want. But if you don't tell them l 21 what you want, then they do it the way they have always 22 done it.
23 MR. BROOKS: They will normally furnish catalog 24 number so and so. That may be adequate and it may not be.
25 So af ter we have done these two things, then we i
() <
ACE. FEDERAL REPORTERS, INC.
t 202 347 3700 Natlonwide Coversee 86336-6M6 .
'l l
26378.0 -
, 23 REE l I
1 want to take a look at the mechanical and electrical 2 installation and integrity. Going over a number of the 3 reports from that OPEC data base and the INPO report, even 4 all the electrical conduits. We have seen a number of 5 cases where failure to operate is as a result of the .
6 conduit being a conduit for water to go into the limit 7 switch connection, things of that nature. We also want to ;
8 take a look at the mechanical 4.n s ta lla t ion . Have we -- has 9 the unit been put toge the r in a workmanlike fashion. Are 10 all the attachments correctly made?
11 One other area that we are reading, some of the 12 reports as far as mechanical is concerned, there is the 13 drive nut; the stem nut frequently has been shown to come
[}
14 loose and t he things that hold it together are threaded 15 connections which are staked at assembly. To me that might 16 not be the best type of assembly factors. We wa n t to look 17 at the whole drive train in the area to be sure that wo 18 will not be putting something in that is going to fail.
19 This can be a rather time-consuming thing, 20 becauso wo may be trying to -- let's continue to look at 21 those things as the microprocessor setup. These are the 22 things that we plan on doing with our in-plant installation, 23 to be sure that wo are loading ourselves to success.
24 We wanted to calculato what the required valvo 25 stem load is. And there are in existence rather simple O
i ACE FEDERAL REPORTERS, INC.
202 347 3700 Nationwkle Coverage 80lk))6 6646
__- --___ t -_ ____-__. _ _ - - _ _ _ - - _ - - _ _ _ _ - _ _ _ _ _ _ - _ - _ _ _ _ _ - - _ _ _ _ _ - _ _
4 26378.0 24 REE
(_x 1 calculational means to determine if you have a motor of so 2 many foot pounds output and you have a gear train of so and 3 so, a stem with a nut on it, you should be able to 4 calculate what kind of stem load you need or are capable of t
- 5 producing and then compare that with what has to be i
provided in order to operate the valve against the delta P
~
6
! 7 tha t we think that it is going to see and the time that it
, 8 has to perform, to open or close.
9 I understand that there is somebody down near 10 the San Diego area that has a little computer code that you 11 can just plug in a few variables and this comes out quite 12 easily. Once we determine what the required valve stem l
(} 13 load is, then we want to confirm the operator's performance 14 capability along those lines. Looking at all those other i
l 15 factors, depending on what the motor supply is, the voltage j 16 supply to the motor is, can it still function and give us
! 17 the things that we expect out of the unit.
I 18 We also want to calculate the valve stem load, l
19 structural limit. It may not be the valve stem as the
- 20 limiting factor. It may be something in the yoke .. It may 21 be something in the bonnet connections that, under pressure, 22 you would tend to be limited by some other factors. It 23 could be the connection between the valve disk and the l
- 24 valve stem. Is it capable of withstanding the load
- 25 necessary to pull it out of the valve seat? Again, going l ()
ACE FEDERAL REPORTERS, INC.
, 202 347 3700 Natlonwkle Coverage 80n 3364646
\ _ __ _ __ ____ _-_ _ _ - _ _ _ __ _ _ _ -_- _ ___- - -__ - _ _ _ _ - - - _ _ _ - _ _ _ _ _ _ _ - _ _ _ _ _ _ _ _ _ _ _ - _ _ _ _ _ _ _ _ - _ - - _ _ - _ - - _--
26378.0 25 REE 1 back to the design of the valve. How has it been set up
! 2 originally?
3 MR. MICHELSON: A major consideration, of course, 4 is the gates to guide friction -- that is of ten one of the 5 major loads -- and what work is going to be done ,- if any,
! 6 on determining the true value of that friction, 7 particularly under dynamic conditions such as when you are j
.l 8 trying to isolate a very high flow such as perhaps a broken 4
- 9 pipe or that sort of thing. That is what sets the sizc of 10 the operator. It may very well set the size of the t
11 operator needed. It is something that I have rarely been 12 a ble to get much information on. Do you have any plans 13 afoot to study tha t, at least on reduced scale?
(~))
N._
14 MR. BROOKS: To try to determine the valve motor
{
15 operator performance capability? I don't know exactly how i 16 we plan on doing this in the plant. If it is feasible , if 17 it is possible to do, I would like to operate the valve i
i 18 under actual system conditions, recognizing in a power 19 plant though, operating equipment under -- when the plant
- 20 is in operation can be dif ficult to do and perhaps be even 21 unsafe. I would like , we will be instrumenting the valve.
22 We have to instrument the valve in order to determine what 1
23 setpoints we want to apply on the stem load. It.is 24 possible that we could do that at the same time under 25 dynamic conditions. But whether we can or not is something ACE FEDERAL REPORTERS, INC.
202 347 3700 Nationwide Coverage 800 336 6646
26378.0 26 REE
' 1 that the utilities are going to have to --
2 MR. MICHELSON: You probably operated large 3 valves under dynamic conditions or watched them operate and_
4 you know what happens as they begin to throttle towards the i
5 end of the closure. They jump around quite a bit and get --
6 it gets all very exciting. At that point you better make
, 7 sure the valve operator itself and this piece of equipment 8 is going to continue onto function to the completion of the i
9 closure.
i 10 These are -- I thought that would be kind of a
! 11 major consideration. But unless somebody is going to take 12 steps to start understanding what is happening during the 13 dynamics of operation at that point in time , I am not sure
/}
14 how much good this other work does. It is useful and it is 15 good, but it is not going to leave me with a very
- 16 comfortable feeling that we understand the problem.
. 17 MR. BROOKS: One of the things that we are
. 18 a ttempting to do right now in the process of using 19 conventional analytical techniques to determine what the
, 20 required valve stem load is, we are attempting to go back 21 through the valve manufacturers, the valve motor operator 22 manufacturer, and determine what kind of tests have you 23 folks done, the valve manufacturets. Have you done 24 something that demonstrates that, by gosh, that i
25 calculational technique is so conservative that I really O -
ACE FEDERAL REPORTERS, INC.
202 347 3700 Nationwide Cmcrase 800-336-6646
26378.0 27 REE b
s/ 1 have no concern?
2 If -- and we are in the process of doing this.
3 If we find that there is a significant lacking in that area 4 -- you know, it is difficult to accumulate this information 5 --
if there is a significant lacking in that area, then I 6 may have to do a little bit more emphasis there. Some of 7 the techniques that are used are just simply to -- in the 8 shop -- is to close the valve, apply the maximum delta P 9 across the closed valve, pressure rising one side and no 10 pressure on the other side , p'Wh the button and see if the 11 thing opens. If you have dynamic conditions, that is even 12 be t te r . But this work, I think, is going to ' be ve r y
(} 13 revealing. I would think that the industry does have some 14 test data that is going to indicate that those 15 calculational methods are indeed very conservative.
16 MR. MICHELSON: The breakaway friction, I think 17 they have got some pretty good numbers on. The sliding 18 friction is not at all clear.
19 MR. BROOKS: Could be more than, you know,_once 20 you crack the disk of f the seat, the delta P really goes 21 down pretty fast.
1 22 MR. MICHELSON: The delta P goes down, but now 23 you have got a frictional factor that you need to deal with 24 under these conditions. You would like to be assured that 25 you can continue to open the valves.
n U
ACE-FEDERAL REPORTERS, INC.
202-347-3700 Nationwide Coverage 800 336-6646 -
26378.0 28 REE p)
(_ 1 MR. BROOKS: One of the parallel projects that I 2 have going on with this work is with valve stem packing.
3 And we are doing some work there to try and maintain a more 4 consistent stem load necessary to move the stem through the 5 stuffing box.
6 MR. MICHELSON: One of the f undame n ta.1. questions 7 people ask is, if I can open the valve against full delta P, 8 is that all I need to know? Are all the other loads 4
9 thereafter small by comparison?
10 Have you done anything that assures you that the 11 -- keeping in mind the hammer blows and the other aids that 12 occur during this time, if I can open a valve against full
(} 13 delta P, does that assure me it will always work fine for 14 any other condition?
15 MR. BROOKS: No. Not always. We have seen 16 cases where in order to -- we have got a valve over there 17 that has the stuf fing boxes, which is leaking badly and we 18 have got to continue to keep the lights going. So we send 19 somebody down there and he tightens up on the grand nuts in 20 order to stop the leak. Indeed, it does stop the leak, but 21 particularly with some of the old asbestos packing that has 22 been used, the stem load can significantly increase.
23 And that can be a factor that will -- that is 24 why I mentioned the other project that we have that is 25 running in parallel with this, where we would be replacing O \
(_/
ACE-FEDERAL REPORTERS, INC. j 202-347-3700 Nationwide Coverage 800 336-6646
26378.0 29
-REE 1 asbestos with graphite and possibly applying what we call 2 live loading, so that the amount of stem load necessary to 3 move that stem through the stuf fing box remains relatively 4 constant over a long period of time and at a value that you 5 know what it is. The stem loads for the graphite , some of 6 the things that we have found so far are perhaps one-third 7 or less than a third of the stem load necessary to move 8 that same stem through asbestos. Those are just general 9 numbers. If you are in te re s ted in that report, . it is 4255.
10 It is also out on the street now.
11 MR. MICHELSON: Is that a fairly recent report?
J 12 MR. BROOKS: Yes, sir. That was published in
(} 13 December of 19 -- I think the valve motor operators came 14 out in October or November, this one came out in December, 15 I believe.
16 Going on this with this one, the valve stem load 17 would certainly pick up the load necessary to move the stem 18 through the stuffing box. We hope to be able to maintain a 19 little bit more consistency there and determine our 20 performance capability. Is the operator capable of 21 providing that kind of performance with the motor voltage 22 that is supplied to it? If we can do that under dynamic 23 conditions, hurray. If we have to find some other means of 24 coming up with a good assurance that we are going to be in 25 pretty good order, we may have to go that way.
(D v
i ACE-FEDERAL REPORTERS, INC.
202-347-3700 Nationwide Coverage 800-336-6646
26378.0 30 REE 1 Once we have done all of that homework, then we 2 will start in on the task of correlating the torque switch 3 spring-pack displacement. This is the spring pack that is 4 in there already, the displacement of which will operate 5 the existing electromechanical torque switch. We are going j l
6 to correlate that displacement with the stem load. Of ;
7 course, you have to do it in one direction only. Coming up, 8 you open the valve and you are determining what that s tem 9 load is in compression. Then you would apply that same 10 knowledge for a comparable displacement in the opposite 11 direction which would then be for valve stem in tension.
12 We do have a continuous valve position sensor 13 and we would correlate that position sensor with the actual
(}
14 valve stem travel. Here again, how good can you do this?
15 Can you determine , by looking in the open end of the valve ,
16 that yes, it is fully closed or, yes, it is fully opened?'
17 I don't think we will have that luxury. It will be a valve 18 that will be ins talled in the system. And we will have to, 19 by perhaps turning the hand wheel, determine that the valve 20 is fully closed by the cessation of any flow or that it is 21 fully opened as a result of making contact with a back seat, 22 let's say. And then, from the other engineering 23 information, determine that, yes, I. measured that the valve 24 did travel 6.1 inches. I know from the engineering data 25 that the valve stem travel should be 6 inches or is rated n
V ACE-FEDERAL REPORTERS, INC.
202-347-3700 Nationwide Coverage 800 336-6646
26378.0 31 REE S
\~) 1 to be 6; it will give us a good feeling that we know pretty 2 much where we are.
3 Then we will utilize that information to 4 establish the microprocessor setpoints. That would go back 5 to this Vugraph here.
6 (Slide.)
7 Here we are attempting to de termine , is the stem 8 thrust too high, motor load too high? So forth. Taking 9 into account all the other factors.
j 10 (Slide.)
11 The last point is applying the appropriate level
- 12 of predictive control to avoid setpoint overshoot. We do l
{} 13 have a means in the microprocessor to determine 'how we are 14 approaching a setpoint, looking at the ef fects of inertia, 15 of rotating parts. Some existing equipment can be modified 16 to put the se spring packs on the top of the drive nut so 17- that you can accommodate some kind of overshoot and have a 18 cushioning effect that doesn' t damage the valve. We think 19 we can do it by other means, just by using this predictive 20 control system. That would be the next slide that I have.
21 (Slide.)
22 Any questions that we have gone through here?
23 We can come back to it if there is anything that you would 24 like to talk about.
25 We will be looking at sampling. The O
ACE-FEDERAL REPORTERS, INC.
202-347-3700 Nationwide Coverage 800-336-6646
26378.0 32 REE g
(_) 1 microprocessor continuously is sampling data and evaluating 2 data. Le t's say we have a control variable of valve s tem 3 position , could be one that we could adapt this onto. We 4 would be going through a sampling interval here. We take --
5 the processor would take a value here and a value here and 6 de te rmine the slope of that line and then predict what that 7 would be in some time into the future. And if we were to 8 continue on at that, from here to here to here to here 9 (indicating), we could then predict, without predictive 10 control, what would happen at some time in the future. And 11 then by applying the predictive control mechanism, or 12 electronic system, we can then say, all right, we can apply 13 that and rather than have that motor go until it hits the (J~T 14 setpoint, at which point it would coast to here, we can say, 15 we know that we are approaching that setpoint very rapidly.
16 We may shut the motor of f here so that it will coast into 17 its final position with a minimum of overshoot.
18 MR. MICHELSON: If you had an unstable fluid 19 flow condition like you are trying to isolate a broken pipe ,
20 things were shaking around, pressure drops were varying 21 rapidly, loads on the stem were varying rapidly, how do you 22 filter some of that out so that the protective-control can 23 look at this true rate of change?
24 MR. BROOKS: That would be a point of 25 application. Again, going back to what we are looking at
() l i
l 1
ACE-FEDERAL REPORTERS, INC. l 202-347-3700 Nationwide Coverage 800-336-6646 j
26378.0 33 REE g
k.
1 here , determining what the performance requirements of the 2 valve are.
3 (Slide.)
4 We are going to attempt to utilize all of this 5 information to' establish the setpoints that we put in the 6 valve, into the microprocessor valve. You might say, too, 7 well --
8 MR. MICHELSON: It may not be possible to get
- 9 good readings under dynamic conditions though, so if they 10 don' t mee t these idealized rate of change curves?
11 MR. BROOKS: Very possible . And if there are 12 any concerns on the thing, it would be possible to leave
{}
13 the predictive control out of the picture.
14 .MR. MICHELSON: The difficulty is, of course, we 15 think we jnowhowthingswillworkundertheseunusual t
16 conditions, even though we never experience them until the 17 real accident occurs. Perhaps when the real accident i
18 occurs, though, the microprocessor now doesn't realize what 19 is happening and prematurely shuts off the motor and 20 wha te ve r . It is a concern for something like this, which 21 is a little more sophisticated than a simple switching f22 arrangement.
23 MR. BROOKS: A little more sophisticated, but 24 very, very simple and very -- with great knowledge of what 25 you are doing when you change a setpoint.
A V
ACE-FEDERAL REPORTERS, INC.
202-347-3700 . NationwideCoverage 800-336-6646
26378.0 34 REE f"N
's_) 1 MR. MICHELSON: It is nice --
2 MR. DROOKS: You don't.have the one , two, three 3 four of a torque switch and then never know what is going
! 4 to happen when- the thing trips out. You will be able to 5 dial in what the actual stem load is and you will be able 6 to compare that with what the capability of the valve is to 7 withstand that load.
8 When I was preparing the Vugraphs here, I am 9 going to ask myself a question that I thought you folks 10 would have asked me. The question would be: If you do all 11 of these things, what are your chances of even needing the 12 microprocessor? If you do all of these things right, that
/~ 13 valve motor operator should work right and'you don't need a C}
14 microprocessor. But the microprocessor, once you have done 15 all of these things, will be able to very accurately track 16 any degradation that occurs and apparently it does occur 17 over a certain amount of time. You will know exactly what 18 is happening, that you will be able to avoid malfunctions 19 that we see do occur.
20 MR. MICHELSON: I am sure you understand my 21 point though. That is, that under dynamic, severe dynamic 22 conditions, a microprocessor may misinterpret what it is, 23 because it has never been tuned to that particular 24 condition.
25 MR. BROOKS: I don't know exactly whether it O
- ACE FEDERAL REPORTERS, INC.
202-347-3700 Nationwide Coverage 800-336 6646
26378.0 35 REE 1 will happen this way, but you do have the capability, by 2 the methods that we are planning here, we will know what 3 the valve capability is. And if we have any concerns over 4 operation, we can set, we can establish our setpoints just 5 as high as we safely predict that we can se t the m.
6 MR. MICHELSON: Or you may bypass that function P
7 entirely during an accident?
8 MR. BROOKS: That is possible to do, too.
4 9 MR. MICHELSON: Which is the practice sometimes 10 done for certain types of protective devices?
11 MR. BROOKS: I can't predict what all of the 4
12 answers to these things would be based on what the
(} 13 application would be.
14 MR. MICHELSON: We are just trying to be the 15 devil's advocate .
, 16 MR. LIPINSKI: I have a question. You used a 17 linear predictor for your system where it is a simple 18 second order system with inertias, stick friction, sliding i
19 friction, and in looking at your report, had you used the 20 second order model, you would have been able to do exact 21 predictions for final values. As was asked earlier, as the 22 friction changes over the life of this valve, your 23 predictions are going to be incorrect.
24 You could have gone to a more elaborate system 25 that would involve a common filtering or where you can i
ks) 1 l
l ACE-FEDERAL REPORTERS, INC.
l 202-347-3700 Nationwide Coverage 800 33&66 4
i 26378.0 36 REE
. (_
O) 1 estimate the friction factor at each point of operation, 2 but your simplistic model seems to leave a lot to be 3 desired in terms of being able to predict the performance 4 of this valve.
5 JiR. BROOKS: It is the first model of a device 6 that may be becoming significantly better, depending on 7 what we find out at our in-plant demonstrations.
8 MR. LIPINSKI: Had anybody sat down and looked 9 at the second order model that describes the sys tem?
10 MR. BROOKS: I don't think so.
11 MR. LIPINSKI: You are stepping in the right 12 direction, but I don't think you have gone far enough.
{} 13 MR. BROOKS: That is possible. There is an old 14 saying about research, that the toughest part of a research 15 project is to know when to stop. Maybe we are stopping too 16 soon, but we haven't looked at what you are describing.
17 In the predictive sys tem , we recognize that c 18 friction is going to act as a dampening system, to tend to 19 reduce overshoot.
20 MR. LIPINSKI: But it is continuously changing, 21 because once you set this valve up, it is going to be set 22 up for a particular value of friction. And if the friction 23 does not change, you will always get the same performance .
24 But a friction changes. If you do this in the lab'with no 25 flow in the system, you will have one value of friction. l l
(m) m l
ACE-FEDERAL REPORTERS, INC. l 202-347 3700 Natiorwide Coverage 800-336-6646 l
l l
26378.0 37 REE 1 If you put it in the plant, you will have flow conditions 2 and you will get a new value of friction. Even over the i
3 life of the valve , I expect the friction value to change.
4 MR. MICHELSON: If you leave it sit for a month, 5 you will have a different value.
6 MR. LIPINSKI: Yes. And this prediction factor j 7 is very sensitive to the friction.
8 MR. REED: And if you have 10 ppm of boron you-9 will have one.
10 MR. BROOKS: Let's look at the predictive 11 control. It is attempting to prevent an overshoot of a
- 12 setpoint. Le t me re pea t tha t . It is attempting to
(} 13 prevented an overshoot of a setpoint. You are not changing 14 the se t point. If this predictive control is set up in the 15 fashion that would, let's say, shut the motor of f here ,
16 then you coast into the final position. If you never reach 17 the setpoint, then something will happen to make you reach i 18 the se tpoin t. The predictive control is primarily aimed at 19 avoiding overshoot of a setpoint, not remaining below a 20 setpoint. And that, I think, would tend to cover many of 21 the ' concerns that I think f ou are expressing.
i 22 MR. LIPINSKI: One of the things you did is 23 bypass the mechanical limits and why, I don't know. You 24 could have left those mechanical limits in there as backups 25 with higher set values.
i
()
ACE-FEDERAL REPORTERS, INC.
, 202-347-3700 Nationwide Coverage . 800-336-6646
26378.0 38 REE
('~)
(/ 1 MR. BROOKS: Mechanical limits of what?
2 MR. LIPINSKI: On the torque switches; your 3 electrical diagram shows you have bypassed them as opposed 4 to leaving them in as backups with higher settings, such 5 that if your electrical system fails, then at least you 6 fall into your old mechanical limits.
7 MR. BROOKS: I don't think I would want to do 8 that. We do have the capability on the microprocessor to 9 go from the microprocessor control back to the conventional 10 con t'. ol s . As far as being a backup, I don't think that 4
11 would be advisable , because you would never know where you 12 are at with the existing torque switch.
I
(} 13 For example, then you are leading yourself back 14 into the path of saying, I know that my torque switch is 15 set at position 4. What does that really mean as far as 16 valve stem load is concerned? I don't know the answer to 17 that. I don't think you do either. You don't know. I 18 think we are much safer to stay, you know, if you are 19 attempting to control the position of the valve,'rather 20 than go back and using those old controls as a backup, I i
21 think wo are better to stay with the microprocessor itself?
22 MR. LIPINSKI: And bypass the old values?
23 MR. BROOKS: Ye s .
24 MR. LIPINSKI: I don't follow that. Beca use if 25 I have a value that is se t twice as high, you are te lling
- ()
i ACE-FEDERAL REPORTERS, INC.
202-347-3700 Nationwide Coverage 800-336-6646
26378.0 39 REE I'1
(_/
1 me that it may be half of what I think it is?
2 MR. BROOKS: Well, le t me re peat the sta tement I 3 made. Trying to correlate what your actual valve stem load 4 is with a given torque switch setting, without proper 5 instrumentation, if you merely adjust the Limitorque 6 actuator to have a position 3 or position 4, you really 7 never know what that means as far as stem' load is concerned.
4 l 8 You don't know.
9 MR. LIPINSKI: I could track it with your system 10 because I can put your system into a higher setting and 11 match them as a pair.
12 MR. BROOKS: But then what you are reading
(} 13 yourself into is really -- it goes back to what you select 14 as your setpoints for the microprocessor.
15 MR. LIPINSKI: Correct.
16 MR. BROOKS: I can set that stem load value --
17 let's postulate -- up to 90 percent of the structural 18 capability of the valve. Let's say as a conservative 19 engineer I would not want to go beyond that. Then based on 20 the reliability of that microprocessor, I would then say, 21 he y , that is the limit that I will achieve. And I will 22 with pretty good reliability. I don't think going back, 23 using the existing torque switches as a backup or limit 24 switches as a backup would be a good idea, not ~ in the 25 process of controlling the valve. Maybe we could take a (p.
x)
ACE-FEDERAL REPORTERS, INC.
202-347-3700 Nationwide Coverage 800-336-6646
l l
l 26378.0 40 REE
/~T !
(/ 1 look, i f you have anything written up on that, I would be l l
2 happy to take it back. My contractor and I can evaluate it. ,
1 3 I plan on asking that same question. I hope he gives me l l
4 t he same answer I gave you.
5 Okay. Any other questions on the predictive 6 control and how it works? How it is intended to work?
7 MR. WARSING: I was going to ask --
8 MR. MICHELSON: Would you identify yourself 9 first? ,
, 10 MR. WARSING: I'am following up on the question 11 that was asked earlier; that is, the part of this program 12 that I don't think I understand. You keep talking about 13 measuring valve stem load and you are going to do that by
.}
14 evaluating spring-pack movements. What is your assurance 15 that the coef ficient of friction between the stem and the 16 stem nut hasn' t changed so where you say that over a period 17 of time you are constantly evaluating stem load, you are 18 really evaluating spring-pack load without having any 19 knowledge of what the valve stem load really is.
20 MR. BROOKS: We will be making a correlation on 21 the one set of conditions. We will be making a correlation 22 be twee n the displacement of the spring-pack and the 23 measured stem load. That will be measured with a load cell.
24 MR. WARSING: What does that tell you with 25 respect to time and the aging of that stem nut, if you O
ACE-FEDERAL REPORTERS, INC.
202 347-3700 Nationwide Coverage 800-336-6646
26378.0 41 REE 1 don' t continuously go back and redo that te st?
2 MR. BROOKS: That is correct. We do, however, 3 we do have the capability of taking off information from 4 that valve periodically to determine , are there any changes, 5 does it appear necessary that we go back in and redo that?
6 The opera tor, the valve operator, the man who has to make 7 that equipment work will be informed several orders of 8 magnitude more than he is right now on the status of his 9 valve.
10 MR. WARSING: I am trying to make the point that 11 everyone he re seems to be assuming that measuring of the 12 spring pack gives you verifiable data on the stem load. It
( 13 gives you output torque, not output thrust. You keep 14 talking about thrust when you are measuring torque.
15 MR. BROOKS: Are you familiar with the MOVAT 16 system?
17 MR. WARSING: Ye s , sir .
18 MR. BROOKS: You recognize that when a stem, 19 when you apply a load cell onto the stem assembly and you 20 ac tua te the valve, you can then de termine what the amount P
- 21 of load is by a direct measurement and correlate that with 22 the spring-pack displacement. It is one data point. Can ;
1 23 it change? Sure, it can change.
24 MR. WARSING: That is the point. It is one data 25 point.
O V
ACE-FEDERAL REPORTERS, INC.
202-347-3700 Nationwide Ccverage 800 336-6646
l l
l l 26378.0 42 4
REE 1 MR. BROOKS: What is the magnitude of that 1
- 2 change? Is it very significant? ;
t i
3 MR. WARSING: That was my question for you.-
! 4 MR. BROOKS: I don't know. You are Limitorque . ,
i 5 I would think that you would probably know that better than i 6 I from the history of your own equipment.
j 7 MR. WARSING: Let's look at an example.
t j 8 MR. MICHELSON: I think we will have to proceed.
9 What I would like to do though is, I would-like to 10 encourage discussion f rom the audience. Let's do it at the 11 end of. the speakers' time and then identify yourself as i 12 time may permit.
(} 13 Are there any other speakers this morning that i 14 represent EPRI?
15 MR. BROOKS: No, sir. Just myself.
16 MR. MICHELSON: And you have until 10:00. So we 17 will have a fair amount of time to discuss it. I have a 1
] 18 number of questions on the microprocessor itself. I guess l 19 you are the person who is going to be here to answer it.
20 MR. BROOKS: I think I can handle virtually 21 everything that might come up.
j 22 MR..MICHELSON: As time permits , other-
. 23 interested members of the audience thtt may - want to make a 1
24 contribution, if they will identify themselves, then we can-j 25 have you until 10:00. Okay? Thank you.
! ACE-FEDERAL REPORTERS, INC.
- ' 202 347 3700 Nationwide Coverage 800-3 4 6616
. . _ . _ - . _ . . _ . . , _ - . _ _ , ~ . . . . _ , _ _ , _ . _ . . - - _ . , -.. , _ , . _~ . . - _ _ . . _ , _ _ - . .
i 26378.0 43 REE A
k,-) 1 MR. BROOKS: Le t me take this one off of here 2 and the -- we can get back to it if we need to.
3 (Slide.)
l 4 I will discuss brie fly some of the things that 5 we have accomplished so far in the project.
6 MR. MICHELSON: Le t me a sk you , on -- you
! 7 indicated that you have one utility interested in this. I 8 think that is predominantly, if not entirely, a light -- a l 9 pressurized water reactor utility. Are you making any 10 attempt to get boiling water reactor into this, since they
. 11 involve dif ferent, somewhat dif ferent kinds of valves, 12 certainly different water conditions?
(} 13 14 '
MR. BROOKS: It would be nice to be able to go i out and have a broad application. I would like to have 15 gone to more utilities. But the application that we are 16 considering, as I mentioned before, is going to be in the 17 balance of plant area. It is going to be in an area where 18 we can walk up to it and tweak it every now and then and 19 say, hey, you are doing fine. And we recognize that the 20 balance of plant, secondary side on a PWR, is pretty much 21 the same pressure , temperature and water conditions as 22 the re is on the boiling water reactor. The two are 23 rela tively comparable .
24 MR. MICHELSON: Relatively. Of course the water 25 chemistry contr'o1 for .a boiler is quite dif ferent since the I
ACE-FEDERAL REPORTERS, INC.
202-347 3700 Nationwide Coverage 800-336-6646
26378.0 44 REE fs_- water has to go back to the reactor.
1 And so the conditions 2 in the feedwater trains, condensate systems might be a 3 little bit dif ferent.
4 MR. BROOKS: But the secondary side of a 5 pressurized water reactor is pretty much the same as a 6 boiling water reactor system. Not precisely but close 7 enough.
8 MR. MICHELSON: Are you going to pick carbon 9 steel valves, stainless steel valves or you don' t know yet?
10 MR. BROOKS: Don ' t know ye t.
l 11 MR. MICHELSON: That could also make a
, 12 difference in what you might learn.
(}
13 MR. REED: There is a large difference in the i 14 water chemistry of a BWR and PWR secondary side. That is 15 oxygen. Oxygen certainly does a lot to friction. So I 16 would be concerned about that.
17 I am surprised that you are. going to find a nice 18 location for this valve so that you can tweak it all the 19 time. You are thinking of microprocesses for the real 20 world. You should be thinking about safety systems and 21 system of poor access and reliability and the harsh
- 22 environments and the failure rates.
23 MR. BROOKS: Very clearly. But, you know, if I 24 put myself in a position of being a utility, I am not going 25 to put in this new device which has never been put into any ACE-FEDERAL REPORTERS, INC.
202-347-3700 Nationwide Coverage 800 336-6646
26378.0 45 REE 1 other power plant and say, put that on my most critical 2 safety application. No way.
3 MR. REED: Well, you are going from the 4 laboratory. It is nice and easy to do things in the 5 laboratory. But when you get into the real world and the 6 power plant, sometimes laboratory applications don't work 7 out. This very obviously to me is a laboratory thing. It 8 may work in the laboratory. I just wonder about the real 9 world.
10 MR. BROOKS: That is what we hope to find out l 11 with our in-plant demonstration, but those would be in 12 applications that would be of a noncritical nature.
(} 13 MR. REED:
But you have all your tweakers, 14 giving all your tweakers access which may not be realistic E 15 with respect to performance.
16 It is just a comment. As a person who has 17 walked the gutters for 30 years, I --
18 MR. BROOKS: I am sure that there will be 19 controls placed to avoid someone going in and just 20 unilaterally making some changes and setpoints on the 21 device. I am certain that that would be the ca se . That is 22 the case right now with existing equipment.
23 MR. MICHELSON: Is your anticipation to put this 24 at the motor control center in the plant for_ this test?
25 MR. BROOKS: That seems to be a logical _ place to O
ACE-FEDERAL REPORTERS, INC.
202-347-3700 Nationwide Coverase 800-336-6646
26378.0 46 REE (m.s-) 1 put it, yes. It gives you a lot of advantages where you 2 could, if you have more than one actuator, controlled by an 3 MCC, you could have several of these , one for each of the 4 valves, and glean the information from the microprocessor 5 from a number of valves right at that point. I also look i
6 at that as being a real advantage when we do get into 7 valves that are installed inside containment, where you 8 would be receiving information from that valve and can weed 9 it out at the motor control center which is outside 10 containment and is usually not fully accessible , but 11 certainly more accessible than the valve itself.
12 So we are in the process of trying to set up our 13 in-plant demonstration. I have covered where we are at
{~}
14 with that activity, licensing agreements for 15 commercialization. A number of folks have expressed 16 interest along those lines. It is something that we want 17 to be able to of fer to the utility industry, so long as it 18 can be demonstra ted to be of real help to - them. If we find 19 we fall flat on our face in the in-plant demonstrations, we 20 will not let it stand there. We will find out what went 21 wrong and make corrections. We think we are doing all our 22 homework so that doesn't happen. W'e are going to make that 23 be a success.
24 And then the licensing agreements could be 25 something that we would of fer this unit to the utilities, O
ACE-FEDERAL REPORTERS, INC.
202-347-3700 Nationwide Coverage 800-336-6646
l 26378.0 47 REE 1 perhaps install it for them. I don't know. That is i 2 something that our licensing people can set up.
3 The final report that I mentioned earlier, we 4 an ticipa te that would be the fourth quarter of 1986 and it 5 would include in-plant, the results of the in-plant 6 demonstration, the results of any other -- the analytical 7 techniques that we described for the specific valves that 8 we plan on adapting the equipment to.
J 9 I would like to do a little advertising here.
I 10 We do plan on having a valve technology workshop and 11 seminar down at the maintenance equipment application i 12 center. That is the EPRI NDE center in Charlotte , North 13 Carolina. That will be December 3 and 4. We will have a
- 14 working model down there , together with some valves that we 15 purchased on surplus.
16 MR. REED: Back to your in-plant det'onstrations 17 and even to your laboratory work, I recall about five years 1
18 ago, EPRI did an evaluation or testing in the laboratory of ;
19 internal pilot-operated relief valves and I saw the 20 information coming out of that and disagreed with it 21 strongly, because you did demonstration on all your work in 22 pure water without boron , I felt that boron was the major 23 environmental swinger in that activity, that an internal 24 pilot-operated relie f valve , if properly tested in the real 1
) 25 world, and in a boron condition, might not ever function at
()
l ACE-FEDERAL REPORTERS, INC.
202-347-3700 Nationwide Coverage 800-3E6646
26378.0 48 ;
REE 1I all. What you are doing again is you are going to go out 2 in the secondary system, that is, not where the 3 safe ty-related valves are , there are few, you are going to 4 go out into the secondary system of a PWR, next step from ;
5 the laboratory, where you have certainly used pure waters, 6 I an sure, on all these things, and you are going to not 7 have a real test. You are not going to have a real-world 1
8 test.
9 MR. BROOKS: Again, perhaps we have to take 10 small steps tie fore we take giant s te ps .
I
- 11 MF. REED
- But you come out with conclusions i
12 that get di stributed to the public, like on the internal 13 pilot-opera r.ed relie f valve , that may mislead the industry 14 I at large.
15 MI. BROOKS: Well, I don't know. On the safety 16 valve, the relief valve test program, these were the 17 pressurizer safety valves and safety relief valves. On the 18 pressurizer, Where you are maintaining a steam atmosphere,
.; 19 I don't know.
20 MR. REED: It is 16 to 20 percent by weight- in 21 the steam space is boron. There is more boron in the steam
. 22 space proportionally than there is in the wa te r .
j 23 HR. BROOKS: I don't know that that was not 24 demons tra ted .
25 MR. REED: It wasn't.
O ACE-FEDERAL REPORTERS, INC. .
202-347-3700 Nationmide Coverage 800 336-6646
~ ._ . . -._ __ _ _ - _ - - _ - - - - --,- _
i l
^
26378.0 49 REE i
- ( l MR. BROOKS
- I see.
2 MR. REED: I wonder about, in going from your
! 3 laboratory step, I can understand the laboratory step, if
- 4 you don't do things like the real world. But the next step, i
5 , in order not to get people excited about something that
! 6 a might not apply, I would think you would try to give it the 7 l worst case even.
]
i 8 i MR. BROOKS: As far as the microprocessor is 9 concerned, I wonder what would be the ef fect of the boron 10 in the process media.
i l 11 MR. REED: I use that for an example, but 12 certainly it will affect stem friction on the s tem .
x 1
{} 13 MR. BROOKS: I don' t know the answer to that.
1 14 Re pea t ing , the applications that we can have access to in a l I j 15 , power plant have to be determined . by the utility. I would 16 think that they would be taking a little bit too much of a 17 risk on putting a brand new piece of equipment into a
- 18 critical area. Boron solutions would be in the primary i 19 circuit; isn't that correct?
20 MR. REED: Or in safety systems that are te s ted 1
21 weekly.
22 MR. BROOKS: Le t's take this one hurdle at a 23 time and see how good it works out. You know as well as I
- 24 do, there are people, there are men in the power plant that l 25 have 30-pound hands. Let's find out what that has to do i
ACE-FEDERAL REPORTERS, INC.
202-347 3700 Nat.onwide Cove age 800-33H646 b
b 26378.0 50 REE i
1 with it, too. We can't cover everything all in one shot.
2 We are trying to cover it as reasonably as we can in 3 appropriate step-by-step activities.
4 MR. MICHELSON: How do you tell the 5 microprocessor that somebody has tightened down on the stem 6 nut?
7 MR. BROOKS: You can determine that by --
8 MR. MICHELSON: You do occasionally have to 9 tighten them to keep the water from leaking out.
10 MR. BROOKS: You can do that by -- I think we 11 discussed that in the report --
12 MR. MICHELSON: It wasn't real clear that there i
(} 13 was any kind of simple screwdriver adjustment to make that 14 change.
15 MR. BROOKS: You can determine what the cause of 16 the trip was --
17 MR. MICHELSON: I mean if I have to tighten the 18 stem packing because it is leaking, I need to inform the 19 microprocessor so it looks at the process a little 20 differently perhaps. Depending on -- you haven' t always 21 adjusted these up to the limits of everything.
22 MR. BROOKS: That is correct.
23 MR. MICHELSON: Maybe I have some margin left, 24 so I can adjust my limits now and everything will work fine 25 and I can go on to the next outage. How do I -- do I have O :
ACE-FEDERAL REPORTERS, INC.
202-347-3700 Nationwide Coverage 80 4336 6646
26378.0 51 REE k'_'S) I to reprogram the microprocessor to do that?
2 MR. BROOKS: No. It is on a -- just thumb 3 wheels for adjusting.
4 MR..MICHELSON: I d idn ' t de te c t that there was a 5 thumb wheel available. Maybe I missed it. On the i 6 microprocessing card or where?
7 MR. BROOKS: You program it right into the
) 8 setpoint of the unit. I have a copy of a report back there 9 I can show you.
10 MR. MICHELSON: I have a copy of it here also.
11 MR. BROOKS: Take, for example, if you have set 12 up your microprocessor so that the setpoint on valve stem
{} 13 14 thrust is at the maximum limit that you want to apply to i
tha t stem, then you change, let's say you reach up on the 15 grand nuts to the , point where it takes a lot more s tem 16 thrust to operate the valves, your microprocessor would not 17 be able to provide that because it is set at the limit 18 already.
19 MR. MICHELSON: You have already used up your 20 margin.
21 MR. BROOKS: But you would know what your margin 22 is.
23 MR. MICHELSON: It is not clear that you have 24 set the microprocessor up to the margins of the equipment --
25 I mean, to the limits of the equipment.
- ACE-FEDERAL REPORTERS, INC.
202-347-3700 Nationwide Coverage 800 336-6M6
, 26378.0 52 REE i
( 1 MR. BROOKS: That depends on the application.
, 2 MR. MICHELSON: Clearly, if you don't set it to j 3 the limits, then you want to use and you want to do some r 4 adjusting. I didn't realize it was quite as simple as a i
- 5 thumb adjustment. If it is, then I have a lot of' questions l 6 about maladjustments.
7 MR. BROOKS: I can predict what that next i .
8 question would be. W'e better put something on top of that 9 to lock it up so that the plant superintendent _ i:s the only-10 one who can change - those .
f l 11 MR. MICHELSON: Is that provided in the design?
f j 12 MR. BROOKS: It is not provided now, but that 13 capability can readily be provided.
l
-( }
l 14 MR. REED: One last comment or,~that slide. I 15, hope that in your valve technology seminar ,.. workshop, that >
16 there will be a few real good maintenance foreman types at ,
, 17 the workshop, rather than all desk engineers.
1
! 18 MR. BROOKS: Clearly. They are . my target.
i l 19 MR. REED: It would be interesting to see what j 20 their reaction would be.
j 21 MR. BROOKS: Clearly, that is my target, ye s .
l l 22 (Slide.)
9 l 1 23 MR. BROOKS: The EPRI report did come out in l i
i 24 De cembe r . We also have an EPRI journal article, an R&D
- 25 status report that came out in December also. We have been l () -
ACE-FEDERAL REPORTERS, INC.
f 202-347-3700 Nationwide Coverage 800-33H646
26378.0 53 REE s/ 1 quite activ,e in presenting papers on both the 2 microprocessor, the valve motor operator improvements and 3 the valve stem packing improvement, both with the ASME and 4 with the American Nuclear Society. And the copy that I 5 gave you was the report that was presented by my contractor 6 at the ANS meeting on maintenance which was last week in 7 Salt Lake City.
8 That is the last slide I have got. How do you 9 want to proceed f rom here?
10 MR. MICHELSON: Well, first I guess would be any 11 questions that the members of the Committee have and --
12 MR. REED: I read the report and I have listen 3d
(} 13 to the conversation. Being a guy who has had a different 14 career trail than others in the nuclear industry, I am 15 concerned about something as complicated and as 16 proliferated as microprocessors might be, because I look at 17 the history of power plants and what has stood the test of 18 time in power plants. I always go back to the simple 19 rugged formula, that that is the way you make performance.
20 I look at this and*I see instrumentation and small wires 21 all over the place and load cells and other things and 22 micro , multiplexes and this, that, and the other. I say 23 to myself, this is something that is viable in the real 24 world where insulation and other things get thrown around 25 and removed and put on,,and there is dust and there is o .
'\ 3 I i ACE-FEDERAL REPORTERS, INC.
202 347-3700 .y NationwideCoverage 800 33 4 646
l l
. I 26378.0 54 REE -
l l
(~')
\ms 1 harsh environments and heat and so on.
2 I recently read a report that said that the 3 failure rates, the failure problems with I&C-type equipment 4 is where the hig he st failures are. If you want to get lots 5 of failures versus a pump failing, you put in lots of 6 instrumentation and control.
7 When you put in this microprocessor for some 8 technician or engineer to read out and try to in te rpre t ,
9 you are transfarring -- you are making more work and you 10 are transferring work from, let's say, the maintenance
- 11 people, who ought to do it right with suf ficient tools and 12 aids, who ought to do it right, you are transferring it to 13 technicians and instrument technicians and other 14 maintainers. What are we doing, making for a larger 15 station complement or are we looking at the real problem 16 and in the real direction that we ought to be going, namely, 17 simple rugged equipment with competent maintenance people 18 with proper tools applying their craf t?
19 End of speech.
20 MR. BROOKS: I think to direct an answer to one 21 of your questicas as far as the ability to withstand a 22 harsh environment, the microprocessor can be totally sealed.
23 It can be hermetically sealed. The heat output from it is 24 negligible. That fact alone can eliminate a lot of 25 problems. We do anticipate that the -- it would be one of
()
ACE-FEDERAL REPORTERS, INC.
202-347-3700 Nationwide Coverage 800-336-6646
26378.0 55 REE O
(/ 1 the features tha t woult' be provided.
l 2 All of the steps that I described earlier would 1 3 be done by qualified people, people who know what they are 4 looking at. We believe that the microprocessor would be 5 installed correctly, that the unit would be operating 6 correctly, that it would be a healthy unit af ter the 7 installation is finally made.
8 The need for continuing activity on that 9 microprocessor would exist if you were to, le t 's say, 10 de termine the fact that my valve stem loading is increasing.
11 I actuated this valve two months ago. The stem load was so 12 and so. I actuated it just yesterday and the stem load is
(} 13 up to some higher number. I would know that.
14 It will give the maintenance people information 4
15 that they don' t have now that would say, hey, something is 16 happening with that valve and by gosh, when we come down 17 for a power outage or refueling outage, assuredly we will
- 18 look at that one.
19 MR. REED: I ~ think we can look -- first of all, 20 we have to accept the fact that there are bad valves, 21 misdesigned valves, and misapplied valves. Now that we 22 have divided the problem into three areas as far as 23 operating the valves, we divide the m in to , let's say, a 24 tip-top maintenance organization who can take the present 25 designs, more or less, and make them work and maintain them
~j i
ACE-FEDERAL REPORTERS, INC.
202-347-3700 Nationwide Coverage 800-33HM6
26378.0 56 REE 1 and so on.
2 The next thing you might do is take the present 3 -- a good maintenance organization or a fair maintenance 4 organization and give them an aid such as a refueling 5 testing piece of gear like the MOVATS, gear and they do 6 evaluations and improve their skills.
7 The third thing you might do is, let's say you
. 8 have a lousy maintenance organization. You will transfer 9 electrical mechanical maintenance over to instrumentation l 10 and control and put in a microprocessor. That might be the 11 three things or you might arrange them differently. Say 12 you will have a tip-top qualified maintenance organization
(')
%./
13 with a microprocessor and so on.
14 But quite frankly, my feeling is that I would 15 try to be at the number two level. I would not want to 16 have all this kind of spaghetti all over the place. We got 17 too much spaghetti in nuclear plants already.
18 MR. BROOKS: Let me go back to this slide here.
19 (Slide.)
i 20 Depending on what happens in these areas right 21 up here, it is possible that none of the three cases that 22 you described would work. If this thing here , if we 23 determine that there was a major. disconnect between this 24 calculated valve stem, required valve stem load and the 25 performance capability of the unit, I don't care who you C) o ACE-FEDERAL REPORTERS, INC.
202-347-3700 Nationwide Coverage - 800 336-6646
26378.0 57 REE t
(_I 1 have maintaining that equipment, it is always going to be a 2 troublesome piece of equipment because you may be grossly
- 3 overloading it every time you actuate it. You never know i l 4 that unless we go through this step. That is why we are 1
1 l
5 doing it.
6 MR. REED:- I started out my three operating
! 7 organizations with saying that you could have design I
8 problems and somebody ought to get to those.
' 9 MR. BROOKS: Right. W'e are trying to. Maybe 10 there is a major problem area right in here , that we better 11 look at that first be fore we put the microprocessor into 12 place.
(} 13 MR. LIPINSKI: If you confirm that step, you may 14 not have to go any further.
15 MR. BROOKS: That is true.
16 MR. LIPINSKI: You raised the issue earlier --
17 MR. BROOKS: Once you do all of these things, if 18 everything is sweet and pure, hey, what do you need the-i 19 microprocessor for? Maybe you don't. Unless you want to i
- 20 have the additional advantage of providing the operator 21 with information that will prevent any future malfunction 22 of the thing.
23 MR. LIPINSKI: But if you verify design
, 24 conditions, then the only future malfunctions you are going i
25 to encounter are stem forces not seating properly. i
(~T i
\_/
ACE-FEDERAL REPORTERS, INC.
202 347-3700 Nationwide Coverage 800-336-6646
26378.0 58 REE
['N
(_) 1 One of the things that your own report pointed 2 out is that to install this system, you need an additional 3 wiring. It is not ordinary rugged wiring, it has to be 4 shielded wiring that will not interfere with other power 5 systems coming in and out of the plant. And if those 6 cables are not available , then you will have to go through 7 expensive installation procedures.- So now you have a 8 cost / benefit issue.-
9 MR. BROOKS: Everything -- the sum payback for 10 everything that you do -- we can go back to that INPO 11 report that showed that over a period of 1979 to 1983, 12 something in that order, there was some thing in the order
(^}
13 of greater than 500 events associated with valve motor U
14 operators, and the lion's share of those are associated 15 with torque switches and limit switches.
16 Now, I don't know what that costs, but I can 17 imagine that 500 events is a very, very high number if you 18 look at plant availability costs. It would be a very high 19 number. I think one of the major reasons that we are faced 20 with that type loss is the fact that the man who was using 21 the equipment really doesn' t know the status of that
- 22 equipment at any particular time. With the microprocessor, 23 he would.
24 MR. REED: Is the status of the equipment the 25 monitoring of the status of the equipment important or is o
\_)
ACE-FEDERAL REPORTERS, INC.
202-347-3700 Nationwide Coverage 800-336-6646
26378.0 59 REE (s-)/ 1 it the maintaining of the equipment by competent personnel 2 the important issue? A, if you maintained it properly, 3 your equipment'is just superfluous. All it is doing is 4 monitoring a nice scene. If it was designed right, and it 5 was maintained right , and we saw that very clearly in this i
6 check valve incident, it could have been corrected by the 7 maintenance department. If they had been on the ir toe s ,
8 they would have seen this' flawed design right away. But I 9 like to go back to root causes and root issues and solve 10 root problems rather than to tag on dams and other gear 11 downstream.
12 MR. BROOKS: That is, I believe , what we have
(} 13 done here. We are looking also at root causes. We feel 14 that there is significant deficiency in the existing 15 equipment. Either that it is so difficult to maintain or 16 to adjust or to operate that it is leading us into the 17 number of failures that we experience in --
18 MR. REED: Are you saying that your 19 microprocessor is going to do or can do or should do lots 20 more than, say, a periodic evaluation with a test rig?
21 Like a MOVAT system?
22 MR. BROOKS: It would provide the same 23 information.
24 MR. REED: The same information?
25 MR. BROOKS: But it would be as a permanent
() !
ACE-FEDERAL REPORTERS, INC.
202-347-3700 Nationwide Coverage 800-336 6M6
4 26378.0 60 REE
(-s) s_/ 1 installation and be used -- its primary function is as a 2 control device, not a diagnostic.
3 MR. REED: How many motor-operated valves do you 4 think might be hooked up in a particular plant?
5 MR. BROOKS: I don't know. Let's assume that 6 there are - what would be a good guess -- there are 7 probably close to 500 motor-operated valves in a power 8 plant. I don't know. Maybe there would be 10 percent of 9 those might -- you would be able to have a cost-effective 10 benefit for the installation of this thing. I am sure that 11 whatever utility decides to go that route , they will make 12 that evaluation.
{} 13 14 MR. LIPINSKI: One of the key questions is, if you set a valve up and you have a test device that verifies 15 it has been set up properly and you walk away from it and 16 it operates, will things change with time that require me i
17 to have an on-line verifier or can I do just as well by 18 adjusting it, te s ting it and walking away from it?
19 MR. BROOKS: I don' t know that answer. That is 20 dependant upon the application. If you have a very bad 21 match-up of this one and this one, you are probably going 22 to have changes occurring in that thing over a very close 23 frequency, I would guess.
24 MR. REED: But shouldn't the designer have 25 discovered that flaw before he shipped it?
b)
ACE-FEDERAL REPORTERS, INC.
202 347 3700 Nationwide Coverage 800-336-6646
26378.0 61 REE 4
1 MR. BROOKS: Sure. You know, it would be very 2 nice to say that every General Motors car that comes off 3 the production line doesn' t have any flaws to it. I can't 4 address that.
5 MR. MICHELSON: These are flaws that are i
6 introduced by the installation and the operation. These 7 are stem packing nut adjustments, these are nozzle loads 1 8 that the piping people put in because they had to bend the 9 pipe to get the valve in. There are lots of reasons that 10 the se loads can change significantly.
11 MR. REED: I think we have gone beyond weld ,
12 d istor tion-type things, and that we do provide so that you
{} 13 don't get dictortion. It seems to me that that is part of 14 design. You make the nozzle ends long enough so you don' t 15 get distortions.
16 MR. BROOKS: Has that always been done?
17 MR. REED: No. It hasn't in the past. That is 18 why I get back to the fragmen ted industry that really ought 19 to have some production criteria.
20 MR. BROOKS: Let's look at where my major target 21 ha s to be , recognizing that there are probably no valves l
22 being shipped today from the valve manufacturers to nuclear 23 power plants. They have all been delivered. A major 24 portion of them have been installed and the plants are 25 o pe ra t ing . Is a utility willing to take that thing and cut O
ACE-FEDERAL REPORTERS, INC, 202-347-3700 Nationwide Coverage 80 4 336-6646 .
26378.0 62 REE l
(~)
s_/ 1 it out and make a boat anchor out of it and put in some new 2 thoroughly engineered product? Maybe he will. That will 3 be a cost-effective analysis type of thing. But by and 4 large, most of the equipment that is in place today, there 5 is such a large capital investment in that that it has to --
6 the utility has to be able to solve the problem with it and 7 keep going with basically that same piece of equipment.
8 MR. REED: Certainly San Onofre 1 is cutting out 9 the five check valves and replacing them with something 10 that is designed in the design ball park, I assume.
11 That is a judgment decision in the plant on the 12 frequency of repair and how grossly is it design defective.
13 MR. BROOKS: Well, I think by and large my major
[}
14 target is going to be trying to improve the performance or 15 existing installed equipment.
16 MR. MICHELSON: Along that same line, it looked 17 to me like there is a two-step process, the first step of 18 which is to try to make it work right. I am not really 19 convinced that EPRI is putting a lot of effort in that area 20 as opposed to the ef fort in the lengthy processing area, 21 just from what I have read. Maybe the next report will 22 make it much more convincing.
23 MR. BROOKS: What would you suggest that we do 24 more that we are not doing?
25 MR. MICHELSON: The first thing is, can you make O
ACE-FEDERAL REPORTERS, INC.
202 347 3700 Nationwide Coverage 800 336-6646
26378.0 63 REE I)
(/
l 1 what you have function adequately, with adequate 2 reliability? And what do you need to do to accomplish that 3 objective in terms of training of people, adjusting of the t
4 switches, maybe making certain kinds of measurements, 5 better understanding how these things work? You keep 6 saying you are doing this. I just want to read about it.
7 MR. BROOKS: Le t me say this : I shouldn't have 8 pointed to these while you were speaking. W'e will show the 9 calculation in the. final report, we will be showing what 10 these calculational procedures are. Many people are not 11 familiar with them.
12 MR. MICHELSON: That would be important to know.
(~} 13 MR. BROOKS: Then the de termining -- aga in , by v
1 14 the calculational method -- determining the valve 15 performance capability and methods of determining what are 16 the valve stem structural limits -- these factors alone --
t 17 MR. MICHELSON: Have you reached the point where 18 you are reasonably convinced that what we have today won't 19 work, that we have to go to something like the 20 microprocessor controller? Are you reasonably convinced 21 that that is where you have -- that is the end that you are 22 headed for?
j 23 MR. BROOKS: I am reasonably convinced that 24 there are enough applications in a power plant where 25 existing equipment, torque switches, limit switches and so
, O -
ACE. FEDERAL REPORTERS, INC.
202-347 3700 Nationwide Coverage 800-336-6646
26378.0 64 REE A
() 1 forth, do not provide the operator with suf ficient 2 information for continued year-after-year good performance 3 of that equipment.
4 MR. MICHELSON: Maybe you are saying what I 5 needed to hear. What I wanted to know is that you are 6 convinced that even with proper understanding, proper i 7 maintenance, proper retesting and so forth, it still isn't 8 good enough?
9 MR. BROOKS: There are applications, I don't 10 know how many of them there are --
11 MR. MICHELSON: I wasn't clear, but I wanted to 12 understand that that is what your final -- you really have
{} 13 to go this way to get what you think is an adequate product.
{ 14 MR. BROOKS: To get the information that you 15 need.
16 MR. MICHELSON: Saying it in a dif ferent way,
- 17 you believe that even if we have properly trained 18 technicians, proper equipment, maintaining this equipment 19 in a proper array and adjustment, that that won't be good 20 enough? Is that what you are saying?
21 MR. BROOKS: I think what you may be describing 22 there , Charlie , will be a situation that will never ever 23 really occur.
24 MR. MICHELSON: It may not. I realize that.
25 MR. BROOKS: There are people who could take a n
U ACE-FEDERAL REPORTERS, INC.
202-347-3700 Nationwide Coverage 80 4 336-6646
l 26378.0 65 REE i 1 Swiss watch and make it work down inside a mud bath. They l
2 can do that because they are very, very capable people.
3 There are the same people that could look at that Swiss 4 watch and it would break.
5 MR. MICHELSON: There are those who I would 6 think would probably disagree with your views. I wanted to 7 make sure I understood your view correctly.
8 MR. BROOKS: I think there are enough 9 applications that this microprocessor would be very 10 desirable at the power plant. -
11 MR. LIPINSKI: Let me supplement the question.
12 It is a question of whether it is a permanent installation
(} 13 or whether you can do it with a specific test device on a 14 setup and then walk away. Do you view that you have to 15 have a permanent installation? I think you will agree that +
16 you could use a more sophisticated device to set up a valve 17 and adjust it and then walk away from it. But do you have 18 to have it permanently connected with all the associated 19 wiring?
20 MR. BROOKS: It all depends on what you do with 21 that temporary installation. Do you go through all of 22 these various steps with the temporary application?
23- MR. LIPINSKI: We agree, I believe, that you 24 would like to verify the design and the application and the 25 equipment to the task at hand. And that has nothing to do C)
ACE-FEDERAL REPORTERS, INC.
202-347-3700 Nationwide Coverage 800 336-6646
e 26378.0 66 REE 1 with installing a microprocessor. You verify that the i
i 2 installation is the proper installation.
3 MR. BROOKS: Who does that?
l 4 MR. LIPINSKI: Well, your program does.
5 MR. BROOKS: This is the pe rmanen t installation.
6 Who does all of that?
I 7 MR. LIPINSKI: Le t's just draw it above where we 8 say calculate required valve stem load and confirm j- 9 performance capability, calculate valve stem load
' \
10 structural limit.
11 MR. BROOKS: Do you want to stop there?
12 MR. LIPINSKI: No, keep coming down.
(} 13 MR. BROOKS: Then we get into the correlation of 14 the torque switch spring pack with stem load.
I 15 MR. LIPINSKI: Le t's move up.
16 MR. BROOKS: Okay.
1 17 MR. LIPINSKI: You have done all of this. The 18 next part leads us into the installation of the 19 microprocessor equipment. We have two situations. We do j 20 this, go to the microprocessor, or we do this and we go to 21 ef fectively what the microprocessor would do. You would i
j 22 have to modify the equipment in order to get some of these i
j 23 measurements in order to do a one-time setup, but we can l 24 carry it away. We don't have to wire it in permanently.
{ 25 MR. BROOKS: Is that a particular advantage to a O
\_)
i i
I ACE-FEDERAL REPORTERS, INC.
202 347-3700 Nationwide Coverage 800 33MM6
26378.0 67 REE 1 utility? Could be?
2 MR. LIPINSKI: You avoid this permanent wiring, 3 because in the permanent installation you have got the 4 wiring involved to lead you from every one of these valves 5 back to your motor control centers, as opposed to having a 6 sophisticated device to set up and check out the-7 performance of a valve and then walk away from it.
8 If I changed the stem loading on a valve, I may 9 be obligated to go back in there and reverify it with my 10 se tup equipment. So the question is, is there really an 11 advantage to hooking up the system permanently with the 12 associated wiring?
() 13 14 MR. BROOKS:
make that judgment.
I think the utility would have to It would be dependent, too, on what 15 the available penetrations are. That can be --
16 penetrations of the containment. If they have pene tra tions 17 available, tha,t would tend to simplify things. If they 18 don't have any, that can tend to complicate it.
19 MR. LIPINSKI: But then you can fall back onto a 20 more sophisticated setup device verifying that a valve is 21 performing like it is supposed to.
22 MR. BROOKS: The other thing that we did find, 23 from the work that was done and reported in that INPO 24 report, we did find that. there were a number of switch 25 failures, mechanical, electromechanical switch failures O
ACE. FEDERAL REPORTERS. INC.
202-347 3700 Nationwide Covernse 800 3'#6646
I 1
26378.0 68 REE ,
i l that were attributed to the oxidation on the contact.
2 MR. LIPINSKI: How does your system prevent that?
3 You rely on those switches.
4 MR. BROOKS: No, we don ' t. That was the point ,
i 5 that I made earlier.
6 MR. LIPINSKI: How are you picking up position 7 information?
8 MR. BROOKS: We will be transmitting that by a 9 po ten tiome te r.
10 MR. LIPINSKI: That is subject to the same 11 failure. It may not make contact. You don't have i 12 statistical data on your potentiometer.
(} 13 14 MR. REED: I noticed this potentiometer design.
This looks like an example of frailty and built-in failure.
15 MR. LIPINSKI: Yes. If you look at failure 16 rates, you would find that a slide wire potentiometer has a 17 higher fa ilure ra te than a contact.
18 MR. BROOKS: I think the possibility of putting 19 devices in there that are of a very highly reliable nature ,
20 I think that NASA has done a lot of work in that particular 21 area that demonstrates the reliability.
22 MR. MICHELSON: Would you care to comment on the 23 vulnerability of the microprocessor to elevated temperature?
24 Clearly, you could encapsulate it against moisture 25 intrusion, although you do have termination problems that ACE-FEDERAL REPORTERS, INC.
202-347-3700 Nationwide Coverase 800 336-6646
26378.0 69 REE
() 1 you somehow have to also protect. But how about the 2 elevated temperature conditions? What temperatures would 3 this microprocessor be able to withstand?
4 MR. BROOKS: As it stands today, the level of 5 development today is, it is a laboratory device; it has not 6 been qualified for the loss-of-coolant accident today.
7 MR. MICHELSON: Generally -- I don't know which 8 microprocessor -- I think you indicated -- do you recall 9 which one you are using in here for the memory chip on the 10 microprocessor board?
l 11 MR. BROOKS: I have forgotten what that is.
12 MR. MICHELSON: What kind of temperatures would 1
{} 13 you normally expect to start giving screwy results? Walt,,
14 could you tell me what temperatures the microprocessor 15 might be able to withstand?
1
! 16 MR. LIPINSKI: You get numbers like 120 degrees 1
17 C and you can go to -- ,
18 MR. MICHELSON: You mean Fahrenheit?
- 19 MR. LIPINSKI
- C. These are point in ternal 20 tempe ra ture s .
21 MR. MICHELSON: I am talking about ambients that 22 you would have to maintain around the microprocessor.
23 MR. LIPINSKI: If you have got them in closed 24 cabine ts and 80 degrees Fahrenheit --
25 MR. MICHELSON: Wa are not talking about ACE FEDERAL REPORTERS, INC.
202-347 3700 . Nationwide Coverase 800-3E6646
. . . . . - - . -_ - . - ~ .
, -26378.0 70 REE
}
1 loss-of-power conditions.
i j 2 MR. LIPINSKI: I think you have to get to 200 l
3 degrees Fahrenhe it. If you select your internal, PIC 1
4 military spec on a processor, you get up to 200 degroes 5 Fahrenheit on a case before - you get trouble with it.
6 MR. BROOKS: There again , . depending upon the 1 7 application, whether a utility decides they want to go to a 1
1 8 microprocessor -- let's assume that they do -- I would .
i 9 think that the most advantageous place to locate tha t - ;
]
10 microprocessor would be at the motor control center outside g 11 the containment --
i i j 12 MR. !!ICHELSON: Yes.
(} 13 MR. BROOKS: -- where it would not experience
) 14 the loTa-of-coolant accident event. That would be my
- 15 opinion. But the utilities could look at that. I am sure l 16 tha t -- I am not familiar with the exact details of t
17 temperature capability to electronic components. But I am ;
j 18 sure that there is some design feature that could be l, j 19 provided that would get around that. ;
20 MR. MICHELSON: I guess it depends on which 21 expert I talk to. I talked to other microprocessor experts t 22 who name much lower temperatures than 200 degrees for j 23 malfunction.
24 MR. WYLIE: Are you talking about a one-time
, 25 pea k?
t h
i ACE. FEDERAL REPORTERS, INC.
202 347 3700 Naslonwide Coverese 800 336-6646
26378.0 71 ;
REE !
4 i
(~)/
s_ 1 MR. LIPINSKI: No. But the junction runs still l l
2 higher than that.
l 3 MR. MICHELSON: You can get higher rated l I l 4 components, but unless you pay the money, you normally get 5 a much lower --
]
6 MR. LIPINSKI: That is why I was talking about a 7 military spec. i 8 MR. MICHELSON: I was thinking of 120 degrees, l 9 130 degrees Fahrenheit ambients would start the i 10 microprocessor to malfunction in an unpredictable way, but 11 if you lose ventilation in the room, for instance, compact 12 electronic equipment -- I know one application where there
(} 13 were 15 kilowatts of power going into the room and the loss I 14 of ventilation reaches a very high temperature in five to 15 10 minutes.
l 16 MR. LIPINSKI: That could be a problem.
! 17 MR. MICHELSON: You wouldn't want'your -- this 18 is not accident. This is just simply loss of ventilation
- 19 in a room.
20 MR. BROOKS: As I mentioned earlier, the 21 microprocessor itself does not have a high heat output.
22 MR. MICHELSON: That is not the problem. It is 23 the equipment in the room.
! 24 MR. BROOKS: Okay. Then if you have an 25 application where you are continuously going to see 200 0
1 ACE. FEDERAL REPORTERS, INC.
202 347-3700 Nationwide Coverage 800 336-6646
I 26378.0 72 ;
REE l
I
. k_s) 1 degrees, 300 degrees, I would imagine --
2 MR. MICHELSON: The point is that it has got 3 transformers in a room and their heat inputs on loss of
, 4 ventilation to the room, it heats the room up, and the 5 microprocessor may ask the valve to ope n .
6 MR. LIPINSKI: Right.
7 MR. MICHELSON: So it is not an accident of the I 8 normal sort you are thinking of, but rather simply a loss 9 of room cooling?
10 !!R. BROOKS: That is something that should be 11 looked at in the application.
12 !!R. MICHELSON: You don't want the wrong
{}
13 combination. There may be several controllers in that room.
14 In fact, there is -- of tentimes you have whole banks of 15 motor control centers in one room and if you start -- you 16 can get some strange combinations of openings and closings 17 if you were to have a malfunction,of a microprocessor.
18 MR. LIPINSKI: One of the other things is that 19 these microprocessors are very sensitive to spikes. If 20 they are being used in rooms with power switches, you are 21 getting spiking, you can destroy your microprocessors 22 unless they are protected with voltage spike suppression 23 surges on the front ends.
j 24 MR. BROOKS: That is part of the application.
25 MR. LIPINSKI: But each one of these would have 4
ACE. FEDERAL REPORTERS, INC.
202 347 3700 Nationwide Coverage 800 3 4 6646
26378.0 73 REE 1 to have a spike suppressor on the front end of it.
2 MR. MICHELSON: You also have the problem of 3 electromagnetic radiation from walkie-talkies or electric 4 arcs or whatever coming in --
5 MR. LIPINSKI They have to be properly shielded.
6 MR. MICHELSON: -- be ing interpreted as signals 7 of some sort. It is quite a problem to protect these 8 devices against external interference, be it temperature or 9 electromagnetic.
10 MR. BROOKS: All the things that would have to 11 be looked at in the application, the installation vehicle.
12 MR. MICHELSON: I was looking back at the 13 drawing in your report where you mentioned the little
{
14 thumbscrews. I thought these setpoints were read outs that 15 were put in when you programmed the computer and it just 16 read out what it had been programmed to do. You can 17 actually go with your finger and fiddle with L it. That is 18 never good.
19 MR. BROOKS: Unless protective devices are 20 installed over them.
21 MR. MICHELSON: Unless great protection is given 22 to assure that people haven't been fiddling with it.
23 HR. BROOKS: If a mechanic can get inside of a 24 cabinet and make adjustments with a screwdriver, it is the 25 same thing.
O ACE FEDERAL REPORTERS, INC, 202 347 3700 Nationwide Coverage 800 336-6646
l l
26378.0 74 l REE
/~%
(-) 1 MR. MICHELSON: A limit switch valve on a box is i-2 not a simple thing to get into. A motor control center is l 3 very simple by comparison to get into. If this is right l
l 4 inside the cabinet, people bent on doing you in can also ,
i 5 easily get into these cabinets and readjust all this. So l 6 the sabotage fea ture is kind of undesirable.
I 7 MR. BROOKS: We would have to lock the thing up t 8 to the point where it would not be changed spuriously or l
l l 9 inadvertently.
l j 10 MR. REED: Keep in mind when you talk security 11 now, we all know the Davis-Besse incident where locking 12 things up defeated the operators' ability to get to the
(}
- 13 14 equipment to make it function when it.needed to function.
So security is a double-edged sword.
15 MR. BROOKS: No question. But in this 16 particular case, by having something available at the motor 17 control conter -- let's assume wo have our device hooked up 18 to a valve inside the containme-c. By having the device 19 out of the motor control center, the plant operator would 20 have the capability of doing something with it, whereas he 21 would not have the capability of doing something with an 22 existing piece of equipment.
23 MR. REED: But I think the point is being made 24 that motor control centers are a lot more accessible than 25 inside containment. So --
O ACE FEDERAL REPORTERS, INC.
202 347 3700 Nationwide Coverage n00 3% 6M6
26378.0 75 REE l 1 MR. MICHELSON: If you would like to entertain <
2 questions from our audience of a couple of minutes. If 3 somebody wants to ask a question or make a comment, be 4 pleased to identify yourself and proceed. Do we have --
5 use the microphone, if you will, to help the reporter.
l l
6 MR. CURRY: I am Brian D. Curry from 7 Philadelphia Electric Company. One of the comments made l 8 about this is that the microprocessor gives you a very l 9 accurate se tpoint control. That is true if you have l
10 calibrated your microprocessor. Until you calibrate it, 11 you are in the same position as you are with the 12 conventional controls on the valve operator furnished 13 initially.' If you calibrate the existing controls, you
( 14 have that same degree of accuracy as you would with your i
15 microprocessor. I haven' t bought anyt.hing or installed any 16 new equipment.
17 MR. BROOKS: What you say is very true. That l 18 was the point I made earlier. Once we' have gone through l
19 all of these things here, our chances of having a device l
20 that is more reliable than the devices that we have seen 21 which have produced something in the order of 500 events j 22 over a period of four years, if all of these things were 23 done correctly, you probably -- we probably would have 24 experienced less than the 500. I don't know how many.
25 MR. CURRY: You come up with anything to do all O
ACE-FEDERAL REPORTERS, INC, j 202 347 3700 Netionwide Coverase 800 336-6646
26378.0 76 REE 1 those things? You said that they should be done. Have you 2 checked that anybody has done them and seen the results of l l
3 their doing those things?
4 MR. BROOKS: That would be a very, very large 5 scope of work to do that in total. What we are planning on i 6 doing is we will do this level of activity for those in 3 1
7 plant demonstrations that we are planning. !
8 MR. CURRY: But your number of in-plant l 9 de m strations is a very small population. And --
10 MR. BROOKS: But at least we will be showing the 11 utilities a manner in which their own people -- perhaps ,
12 their own engineering people can look at some valves, valve 13 operators that are having problems with. Here are the 14 various steps that you can use to evaluate your own ;
15 equipment. That is a possibility, too. What you are 16 describing, Brian, is a very, very large scope of work to {
i 17 look at a large number of them. l 18 MR. CURRY: But the thing that I am seeing is I 19 that that report, when 1 read it, almost proved that the 20 existing controls are extremely reliable if you install 21 them and set them up correctly. They don't have a lot of 22 variation. They are there. And your microprocessor is 23 about the same accuracy level. Yet you are telling me that 24 I need to put out, spend a lot of money to install this i 25 microprocessor control which is a whole, brand new package O
ACE. FEDERAL REPORTERS, INC.
202 347 3700 Nationwide Coverage M3366646
I i
26378.0 77 REE
( 1 that can have its own unique set of problems that I can 2 learn all over again. I still have my existing problems 3 that are back in there in the operator to begin with.
i 4 MR. BROOKS: One of the major problems that l
5 exists today is the fact that the operator, the person - i 6 operating the plant, the plant operator does not know the
! 7 status of the equipment. He is operating blind. Once you 8 bring -- let's say you bring a valve motor operator up to a
! 9 very healthy condition, the valve and the valve operator, .
l l 10 and you know f rom the fact that you have done all of . your 11 homework, things are going to work properly. And they do 12 for a certain amount of time, then you start developing i 13 problems and you really never know why those problems are
(}
14 occurring. I believe that there are applications where the ,
1 l 15 information provided by the microprocessor will prevent ;
l 16 future events, events similar to those 500 that occurred l ~
l
?
l 17 over a period of roughly four years. That is my-point. i i
18 MR. CURRY: The thing I see ' is that you are -- ;
19 MR. MICHELSON: I think we might have other 20 questions. So let's -- any other -- can go ahead. !
21 Identify yourself and -- !
l l 22 MR. EISSENBERG 1 am Dave Eissenberg from Oak I Ridge.
f l 23 I would like to make a comment and then have a i i
24 question. The. comment is that it seems that this device !
25 has two separate functions. One of them is for a control ;
(
L ACE FEDERAL REPORTERS, INC. !
202 347 3700 Nailonwide Comage 800 3M4646 < ;
26378.0 78 REE A
(_) 1 device and the other is a diagnostic device. As a control 2 device, it has certain improvements over existing control 3 devices and as a diagnostic it has certain improvements. I 4 wanted to keep them separate because they -- you can' t jump 4
5 from one to the other and have a good discussion.
6 With regard to its use as a control device, one 7 advantage that has been claimed is accuracy in setting up.
8 Brian, I think, addressed the issue of accuracy in setting
- 9 up. That is, proper setting up with the existing device j 10 seems like it could provide the accuracy that is proper 11 setting up with the electronic device. ;
12 That was a judgment. I agree with that.
J
(} 13 The other advantage that was claimed is the j 14 issue of inertia overshoot. That is -- that capability of l 15 handling inertia overshoot is not handled the same way by '
- 16 the existing device as the device that you are proposing.
17 MR. BROOKS: That is correct.
18 MR. EISSENBERG: I would like to address that.
19 Now my question: You provide data in your report in which 20 you show the amount of reduction of inertia overshoot which I
i 21 results from using your device.
22 MR. BROOKS: Yes, it does. It indicates that 23 you can have various settings that will . reduce overshoot to 24 some relatively small number. It also recognizes that the 25 point wo discussed earlier, friction, can have a major O O ACE FEDERAL REPORTERS, INC.
202 347 3700 Nationwide Coverase soo.346M6
26378.0 79 REE 1 ef fect on that.
J 2 MR. EISSENBERG: In fact, you give data which 3 shows that the overshoot is reduced from 44 percent to 26 i
4 percent now. The comment and question I have with regard
, 5 to overshoot is its e f fectiveness. The test data given in 6 this report -- that is all I have access to -- indica te s j
7 that the test valve was set up with a torque switch setting
) 8 of 1. Is that correct?
9 MR. BROOKS: The torque switch -- when we were 10 using the valve in that fashion, the electromechanical j 11 torque switch would be wired out of the circuit. It would j 12 not be in place.
{} 13 14 MR. EISSENBERG: What I am saying is that the l equivalent level, because you say -- you do state in here 1
15 that torque switch setting was equivalent to a setting of 1.
, 16 That is a low setting. I don' t know whe ther the setting --
17 MR. BROOKS: One poin t, that would be a low l
]
i 18 setting. It could be a very high setting. You never know
, 19 until you go through all these steps.
20 MR. EISSENBERG: That is a low setting in 21 general for a torque switch, because the torque switch can s
22 go to 1, 2, 3, e t ce te ra . There is a maximum. The maximum, 23 as far as I know, is above 1.
24 Another comment which we have observed in our i
25 tests of MOVs is that the higher the torque switch sotting, O
ACE FEDERAL REPORTERS, INC.
202 347-3700 Nationwide Coverase 800 336-6646
i i
26378.0 80 REE 1 the lower the inertia overshoot. So that setting -- the 2 tests that were described here, which were done at a low
, 3 torque switch setting or equivalent low torque switch 4 setting, emphasized or gave a larger value to the overshoot 5 problem than you would get if the torque switch were set at
! 6 2. That is, if it we re se t a t 2, the inertia overshoot 7 with no correction by your device would be less than 44 8 percent. And if it had been set at 3, it would be even 9 less. By "less" I mean significantly less. So my question 1
10 is, have you run in your tests this test valve at a torque 11 switch setting greater than or equivalent to 1?
12 MR. BROOKS: No, we have not.
(} 13 MR. EISSENBERG: So my conclusion is that the 14 evidence provided, which indicates ef fectiveness in 15 reducing torque inertia overshoot, was taken at a place at 16 which you maximized the amount of overshoot. And that if 17 you went to more realistic settings -- I don't know what i
18 the realistic setting of that particular valve was; I know I 19 it on other valves -- then you might find that the inertia 20 overshoot is not as big a problem and that takes away the 21 second advantage of using this device as a controller. It i
22 doesn ' t a f fect its use as a diagnostic. That is a separate 23 ist .te . I just wanted to point out that the data presented 24 really don't provide evidence that'it reduces overshoot in 25 all cases. !
t' l N s) l A !
- ACE. FEDERAL REPORTERS, INC.
202 347 3700 Nationwide Coverage 800 3 4 6646
~,. .
a e q ,
.c .. .
26378.0 81 REE e-1 MR. BROOKS: This is an interim report. It is 2 the level of development that we have taken the device to 3.3 right now. We have, by the test data that we have 4 accumulated, we ha se determined that tha. device, yes, it 5 does work. It does provide us Shj s level of protection.
I 6 think when we get i.nto our in-plant demonstrations, we will o V 7 -- again, we ' have an ad hoc situation. There is going to
- s. s 8 have to be this valve right here, how does that predictive 9 device protect that valve under the conditions that we plan 10 ,to opera te it.
11 MR. MICHELSON: I would like to .give the Staff 12 in opportunity to ask questions also. Questions? Comments?
13 Identify yourself.
14 MR. ROTilBERG: 'I am Arlen Rothberg from DSRO. I 15 am in the engineering issues branch'>at NRR. My question 16 was, are valve performance requirements to be determined
' ^
17 for the design basis eventh In other words,--
18 MR. BROOKS: T,e t"me try to address that one 19 first. I don't know',x Thh real conditions that we will be 20 looking at here would be delt.a Ps, stroking ime and so 21 forth. As f ar as the design, bar.is event, if that happens i 27; to ; t,e the worst case , tha t id:tho~one we would look at.
, c -
23 -
MR. ROTh8 ERG The problem is this: There is a 24" potential conflict betwoqn the design basis event' and the 25 settings that you might choose for that normal operating m
4
's, 4
ACE. FEDERAL RuPORTERS, INC'.'
1 l 202 347 3700 Natlonwide Coverase /KS3364M6
26378.0 82 REE l tm
(_) 1 conditions, especially for a dual function valve, a-valve 2 that is -- let's say, has to close for the design basis .
3 event and then reopen for some safety function, like an 4 ACCS valve. I see with any of these devices the potential 5 that you are never going to get the perfect setting because 6 there is no perfect setting.
7 MR. BROOKS: That is possible.
8 MR. ROTHBERG: When you determine this design
, 9 basis load, I would like to know how you are going to do it, l
l 10 whether it is going to be done analytically or by testing l
11 and then how verified. And if the design basis event l
! 12 settings, like, for instance, well, the torque switch to l
l
{} 13 14 open, conflicts with the normal operation, in other words, you are causing a high valve wear, I don't see that any of 15 these devices are really going to do anything much about 16 that. I was curious to see how you are going to determine 17 the loads for the design basis event.
18 MR. BROOKS: The only way we plan on it right 19 now was just to take a look at the -- what we determined to 20 be the performance requirements that the utility or the 21 architect engineer or the valve manufactuter, whoever we 22 have to go to, says, that valve was set up for those 23 conditions. If those are the valid conditions that would l 24 apply, that is something that I would have to get f rom the 25 utility.
bs_/
ACE FEDERAL REPORTERS, INC.
202 347 3700 ' Nationwide Coverage R&336 6646
26378.0 83 REE
/"'T
(_) 1 MR. ROTHBERG: From a utility or from the valve 2 manufacturer or both?
3 MR. BROOKS: It would be -- the utility would be 4 the ultimate say-so, if the correct information was given 5 to the valve manufacturer originally --
6 MR. ROTHBERG: You will just take the da ta that 7 the utility and the valve manufacturer gives you for the 8 design basis load --
9 MR. BROOKS: That is right. An we will try to 10 find some way of verifying those conditions. I don't know 11 exactly what they would be . I am sure it is going to be a 4 12 matter of searching through records.
() 13 MR. ROTHBERG: Second question is, right now the 14 IST programs, in-se rvice te s t ing , section XI te s ting is 15 virtually silent on the operator. There is an implicit i
16 requirement that the operator works along with the valve 17 when you stroke it. Will you have at the end of your 18 report any recommendations for verification or any other 19 tests that might be in the IST programs?
20 MR. BROOKS: Cou ld be . -That is something that 21 we would hope for. As I mentioned earlier --
22 MR. ROTHBERG: That is more of a suggestion than 23 a question.
24 MR. BROOKS: This particular one here, calculate 25 the required valve stem load, that is something that we (3
x_/
ACE-FEDERAL REPORTERS, INC.
202-347-3700 Nationwide Coverage - 800 336-6646
26378.0 84 REE 1 plan initially on doing just like the existing 2 calculational methods that are in place right now.
3 MR. ROTHBERG: That goes back to my original 4 question.
5 MR. BROOKS: Somehow, somehow, somewhere, 6 someplace hopefully someone has done some testing work that 7 demonstrates the conservativeness of those calculational 8 techniques. If that has been done , it gives you a nice 9 warm fee ling . Does it get you -where you want to be? Maybe 10 not. But we are not there yet. W'e hope it will provide us 11 usable information.
12 MR. ROTHBERG: The third question was -- this is
(} 13 something that might be important later on -- you are going 14 to essentially assemble a huge amount of data when you test 15 these things. You will be continuously testing. You 16 should have an idea of what records you are going to keep, 17 beca use if there is ever an event or an accident, that data 18 is going to become extremely important. .There is going to 19 be a tremendous -- it looks to me like a tremendous amount 20 of data that is going to be collected.
21 MR. BROOKS: Well, I don't know that a-22 tremendous amount of data would be collected.
23 MR. ROTHBERG: You are going to be continuously 24 monitoring these valves. You are going to be continuously 25 monitoring a large number of valves.
A U
ACE-FEDERAL REPORTERS, INC.
202-347-3700 Nationwide Coverage 800-336-6646
26378.0 85 REE b
s/ 1 MR. BROOKS: But we will not be continuously 2 actuating it.
3 MR. ROTHBERG: Every time you stroke the valve, 4 every time you open the valve, every time you close the 5 valve, you are going to develop a certain set of data.
6 MR. BROOKS: We will write down the information f 7 that is providud on the microprocessor.
8 MR. ROTHBERG: But that is not going to be 9 recorded electronically?
[
10 MR. BROOKS: I don't know the exact details of
]
11 the agreement that we have put in place. It is possible to 12 plug into those -- there is a corrector on the
{} 13 microprocessor where we can run a plot on these things. We 14 might be able to do that. That is something that the 15 utilities are going to have to permit us to do, to run a 16 trace of the performance of the valve in the process of 17 opening or closing.
18 MR. ROTHBERG: But there is no plan for any 19 specific list of data to be kept?
20 MR. BROOKS: There is a preliminary list that is 21 part of the agreement that we have with the utility. It 22 may not be the final one.
23 MR. ROTHBERG: Thank you.
24 MR. MICHELSON: As I understand it, these are 25 probably going to be nonsafety-related valves.
O ACE-FEDERAL REPORTERS, INC.
j 202-347-3700 Nationwide Coverage 800 336-6646
26378.0 86 REE g s,
(_) 1 MR. ROTHBERG: My concern was ultimately the 2 final thing.
3 MR. MICHELSON: With the final one, I had a i
4 number of questions on how they are going to store-- is the 5 microprocessor going to digitize and store the information, i 6 and once in a while you plug in and it dumps it?
j 7 MR. BROOKS: No. As it is se t up now, it 4
8 doesn't do that.
9 MR. MICHELSON: It will lose its memory from the 10 previous cycle when it goes through the next cycle. So you 11 would have to plug in every time and ask it how things went.
12 It won't store it.
{} 13 MR. BROOKS: If you want to run a plot on stem 14 load in the process of closing the valve , you can do that.
15 MR. MICHELSON: During an accident, if it moves 16 to close a while , that closing history is lost when it {
17 opens again.
18 MR. BROOKS: The only storage would be on what 19 you would store from an oscilloscope.
20 MR. MICHELSON: You asked -- unless you ask it 21 to store or digitize and store the information for several 22 cycles at a time.
23 MR. BROOKS: We have not. It is not designed to 24 do
- bat.
25 MR. MICHELSON: It could be , I would think.
N ACE-FEDERAL REPORTERS, INC.
202-347-3700 Nationwide Coverage 800-33H646
26378.0 87 REE 1 MR. LIPINSKI: It can be.
2 MR. MICHELSON: It can do almost anything.
3 MR. BROOKS: It might be done , ye s. It is 4 currently not set up to do that.
5 MR. MICHELSON: Are there any other questions on 6 the part of the S ta f f ? If not, then we will break for 10 7 10:20.
miggten and come back at 8 (Recess.)
9 MR. MICHELSON: Le t's proc 0ed. I believe - the 10 next item is the discussion by the people in research on 11 Oak Ridge National Laboratory. Does Research want to 12 introduce their speaker or make any comments ahead of time?
13 MR. VORA:
{ My name is Jet Vora. I am with the 14 center for engineering technology in the Of fice of Research.
15 To introduce my speaker, I would like to say that all of 16 our hardware-oriented engineering research programs related 17 to valves are being addressed in the division of 18 engineering technology of the Of fice of Research.
19 Specifically, t, hose activities are being addressed in the 20 electrical engineering and instrumentation and control 21 branch and the mechanical and structural engineering branch.
22 Our activity in the electrical engineering 23 instrumentation and control branch .related to valves is a 24 part of the aging program which is being handled at the Oak 25 Ridge National Laboratories and is primary toward the aging O
\
ACE-FEDERAL REPORTERS, INC.
202-347-3700 Nationwide Coverage 800-33MM6
26378.0 88 REE n
k_) 1 cha rac te r iza tion , the detection of the facts, and the 2 degradation of valves which had been addressed at Oak Ridge 3 National Laboratory. Our activities include check valves, 4 solenoid valves and the preliminary assessment of PORV 5 valves. In the mechanical and structural engineering 6 branch, our equipment director and the leak test on the 7 check valves will give you a broad idea about the 8 activities within the Of fice of Research in the division of 9 engineering te c hnolog y.
10 As a part of the aging program, the one which 11 Oak Ridge has been designed to do in the motor-operated and 12 check valves, we have completed the phase 1 studies on
(} 13 14 those components and I think Dr. Dave Eissenberg, who is the project manager, would like to discuss both of those 15 projects, including the report and update of the evaluation 16 of the MOVATS as an analysis technique to de te c t the aging, 17 degradation and other abnormalities on motor-operated 18 valves.
i 19 MR. MICHELSON: You have until 12:30 on our 20 agenda for the entire presentation. I will let you adjust 21 according to your knowledge of how much time each of them 22 takes if you wish. But we will have to be finished up at 23 12:30.
24 MR. VORA: Also with me is Dr. Weidenhamer who 25 can talk to you about valve operability for valve isolation m
, U ACE-FEDERAL REPORTERS, INC.
202-347-3700 Nationwide Coverage 800 336-6646
26378.0 89 REE ks 1 and the program he is planning to undertake for this fiscal 2 year.
3 MR. MICHELSON: That wasn't shown on the agenda.
4 Are you going to cover that towards the end, then?
5 MR. VORA: Yes.
6 MR. MICHELSON: Be sure to leave time. Because 7 you might also think about the question at least I have, 8 and the other members can indicate if they have something 9 they want you to think about. The question that I want to 10 pre sen t is not in research but in NRR: Where is the good 11 valve work being contemplated? Or valve studies, what is 12 the program of the agency, or don't we think we have any
{} 13 14 problems? If we do think we have problems, if we do, what is the plan? Research kind of responds to the needs of NRR.
l 15 So I want to hear what your program is, but I want to hear i
16 what NRR's plans are for the future, if any. And I would 17 like to hear who is in charge. So that will be my question 18 later.
19 MR. EISSENBERG: Thank you for the opportunity 20 to make a presentation today.
21 (Slide.)
22 I will be talking on the nuclear plant aging 23 research activities at Oak Ridge, and my name is Dave 24 Eissenberg. I am in the engineering technology division of 25 Oak Ridge National Laboratory.-
)
' ACE-FEDERAL REPORTERS, INC.
202-347-3700 Nationwide Coverage 800-336-6646
26378.0 90 REE (vD 1 (Slide.)
2 I have a few -- here is what I intend to cover 3 during the presentation. Briefly cover the NPAR prooram.
4 That puts our work into the context, why are we doing it 4 5 and what are goals are. The components to be studied were 6 mentioned. I will provide an update on motor-operated 7 valve diagnostics and also the aging essessment on check 8 valves. We are not as far along on check valves as we -are 9 on motor-operated valves. And the work is being carried 10 out by Oak Ridge with funding from NRR.
11 (Slide.)
12 I just want to -- this Vugraph is somewhat 13 different from that which I presented at the last meeting.
(}
14 I want to se t the stage or vocabulary. The objective of 15 the program is to recommend to the NRC diagnostic methods 16 for assuring operational readiness of selected safety 17 systems, and components and operational readiness is l 18 defined in terms of this question -- this is our working 19 definition of operational readiness: Will it function when 20 required at a future time and under the extreme postulated 21 conditions, for example, as given in IEEE 382, which deals 22 with motor-operated valve equipment modification. That is 23 not necessarily realistic, but that is one example of 24 extreme operating conditions.
25 And at a future time is where the term b) v ACE-FEDERAL REPORTERS, INC.
-202-347-3700 - Nationwide Coverage 800-336-6646
26378.0 91 REE
(_/ 1 " operational readiness" I think is important. We.can te st 2 the valve today and we know that it is operational. If we 3 te s t it at succeeded times, we know that it continues to be 4 operational, but we don't know that it will work the very 5 next time, particularly if that time is months later. So 6 one of the elements of operational readiness is the 7 prediction capability.
8 The diagnostics,- the purpose is -- what 9 abnormalities are present in the motor-operated valve , how
! 10 have they changed with time since we last checked them and 11 how do they af fect the functional performance? That is, 12 can we extrapolate to the point at which the valve will
(~' 13 fail?
(-) /
14 (Slide.)
15 This is a kind of a generic scope of how the 16 NPAR program considers all components. I put it up here to 17 mainly show where we stand with regard to motor-operated 18 valves. We have completed this characterization of 19 aging-rela ted failure modes and defects. And we have 20 issued a report. And.this report does not lead to any 21 answers. It just simply states the conditions that we are 22 looking at. It defines the valve and defines the operating 23 conditions and defines failure modes based on searches of 24 the data bases. It defines potential diagnostic methods 25 that can be used to trend the defects that have been found.
O G
ACE-FEDERAL REPORTERS, INC.
202-347-3700 Nationwide Coverage 800-336 6646
l 26378.0 92 REE
(_) 1 This se ts the stage.
2 We are in this mode for motor-operated valves, 3 identifying, characterizing and validating diagnostic 4 monitoring methods for detecting, differentiating and 5 trending all the defects that could lead to failure.
6 The next step, which we are scratching the 7 surface on -- it is a very difficult one; it was alluded to 8 in some of the comments earlier -- that is, okay, we now 9 know that there are de fects and we see them change with .
10 time. What can we do about it, which is the acceptance 11 criteria? We are addressing -- we are beginning to address 12 the issue of acceptance criteria. That is, acceptance 13 criteria based on the diagnostic monitoring methods.
V('T 14 And therefore, in order to develop acceptance 15 criteria, we need to first characterize what we are 16 monitoring, what it te ll s u's , how sensitive , et ce tera ,
17 before we can then determine when do we do something about 18 it.
19 Finally, to provide recommendations for methods 20 of diagnostic monitoring to the NRC.
21 (Slide.)
22 Here are some of the motor-opera ted valve 23 activities. As I mentioned, we have completed this 24 identification and characterization and issued-this NUREG.
25 We carried out a study with two purposes: to identify and O
ACE-FEDERAL REPORTERS, INC.
, 202-347-3700 Nationwide Coverage 800 336-6646
26378.0 93 REE
- (.) 1 characterize abnormalities as found in situ in a number of 2 motor-operated valves installed in four nuclear plants.
3 This study was carried out in conjunction with an 4 evaluation of the MOVATS method for providing diagnostic 5 information. That is, the same study accomplished these 6 two objectives. It identified just what kind of 7 abnormalities were present in valves which were operable.
8 That is, the valves passed the in-serv ice inspection based l
9 on section XI of the ASME code, which is the governing 10 testing method.
11 Then we evaluated MOVATS as a method to tell us 12 what more does it tell us; what more can it do that goes
{} 13 beyond the section XI requirements?
14 And finally, we are carrying out in progress the 15 issue of identifying and characterizing diagnostic 16 monitoring methods. That is, we are looking at MOVATS, 17 MOVATS-like and other- diagnostics that could be useful in 18 predicting options at readiness. This is in progren" 19 Still to come is the issue of in situ testing of tt 20 monitoring methods.- And finally, the very dif ficult one of i
21 establishing or trying to establish criteria. Criteria -- )
22 I put it in those words because it probably will never end
. 23 up with a single set of criteria. It will be 24 plant-specific and system-specific.
25 (Slide.)
O ACE-FEDERAL REPORTERS, INC.
202-347-3700 Nationwide Coverage 800-3366646
26378.0 94 REE gx
(_) 1 MR. MICHELSON: To what extent are you going to 2 provide process-like conditions to the valve that you have 3 in the test stand?
4 MR. EISSENBERG: As part of this -- this phase 5 is to identify monitoring methods. As far as the next 6 bullet goes is where we will -- that is, we will do in situ 7 testing of valves under field conditions at utilities or 8 other places.
9 MR. MICHELSON: You won't do that at your 10 laboratory, then?
11 MR. EISSENBERG: We don't have a pump-around 12 loop to do that.
f~} 13 MR. MICHELSON: You don't need a pump-around N_-
14 loop to apply full dif ferential pressure, though?
15 MR. EISSENBERG:' You are right. I was 16 considering the actual field conditions.
17 MR. MICHELSON: I was thinking of just simpler 18 things likc full differential pressure?
19 MR. EISSENBERG: We will be doing things like 20 applying hydraulic resistance to the operator and the valve.
21 We will be doing similar tests. We are artificially
-22 inducing de fects. But the bottom line still has to be 23 field tests. We still have that ahead of us.
24 MR. MICHELSON: Is that what you call phase 2, 25 then?
/^
V}
ACE-FEDERAL REPORTERS, INC.
202-347-3700 Nationwide Coverage 800-346N6
L:.
1 26378.0 95 l REE I i'
("N
(_) 1 MR. EISSENBERG: No, phase 2 is really here.
2 MR. MICHELSON: It is a part of phase 2, though?
3 MR. EISSENBERG: It is a part of phase 2, yes.
4 MR. MICHELSON: Okay. Thank you.
5 MR. EISSENBERC: The last presentation I made to 6 this Subcommittee, I did provide some information on the 7 MOVATS system. I am not going to re pea t it. I think that 8 Art Charbonneau can answer questions on what the MOVATS 9 system is and how it works af ter the presentation. I just 10 wanted to summarize from the MOVATS report that we issued, 11 some of the results. This particular Vugraph shows the 12 defects, the abnormalities that were identified by the 13 MOVATS method on the 36 valves that were tested in the four
{~'
14 nuclear plants: These results were cited in IE Bulletin 15 8503. They indicate a very high percentage of the valves 16 which we te s ted to have this kind of abnormality. However, 17 in addition we found a number of valves, not a significant 18 number, that had this kind of aging, this kind of wear.
19 MR. MICHELSON: This batch of 36 included both 20 safety-related and nonsafety-related applications. Do you 21 know what percentage of the 36 were safety-related 22 applications?
23 MR.-EISSENBERG: In carrying out the study, we 24 asked MOVATS to work with selected utilities to identify 25 the valves to test it, provide us with the test data . We O
ACE-FEDERAL REPORTERS, INC.
202 347 3700 Nationwide Coverage 800-33MM46
26378.0 96 REE ek k_s) 1 did not ask them for further identification. I believe 2 that the valves were primarily safety related. Is that 3 right?
4 MR. MICHELSON: Could you give us a percentage?
5 How many of those 36 valves were in safe ty-related systems?
6 MR. CHARBONNEAU: As I recall, 75 percent.
7 MR. MICHELSON: The sample that I was familiar 8 with, when item by item, it turned out only one was safety 9 related out of the seven. But I wondered if that ra tio .
10 pe r ta ined to all the rest, and you are saying no, as a 11 whole, 75 percent were in safety systems.
12 MR. CHARBONNEAU: That is as best as I can
(} 13 recall. Let me be conservative and say that it was between 14 50 to 75. What we tried to do, as was pointed out earlier, 15 was get hold of a safety-related operator of a plant and 16 that is a heroic effort. And so we, in essence, made a 17 deal. We said, if you give us healthy safety-related 18 operators, then we would also be willing to te s t some o f 19 your nonsafety. So that is the reason we did not end up 20 with 100 percent safety related.
21 MR. MICHELSON: When you got your listing, did 22 the ut'ility tell you whe ther the valve was safety related 23 or not?
24 MR. CHARBONNEAU: Absolutely.
25 MR. MICHELSON: You left it off of the data O
ACE-FEDERAL REPORTERS, INC.
202-347-3700 Nationwide Coverage 800-336-6646
26378.0 97 REE a
' (~h
(_/ 1 sheets, though, that were in or were apparently unaware of 2 the application.
3 MR. CHARBONNEAU: That piece of data was 4 provided to Oak Ridge when we submitted our final report 5 that was submitted.
6 MR. MICHELSON: So you know then what --
7 MR. EISSENBERG: Well, I don't know offhand.
8 MR. MICHELSON: I got two answers.
2 9 MR. EISSENBERG: I will have to resolve that 10 question as to --
11 MR. MICHELSON: I was surprised because the ones 12 I was looking at on one utility, those were in the
~
13 feedwater main steam or main steam and feedwater system, 14 not in a safety system, with one exception. So I wondered 15 if the sample was done dif ferently with dif ferent utilities.
16 MR. LIPINSKI: Be fore you take that off, I have 17 a question on that loose stem locknut. Are these attached 18 with a nut and a lock washer, relying on that to hold the 19 system together?
20 MR. EISSENBERG: This stem nut is held in place 21 with a locknut which then should be pinned.
22 MR. LIPINSKI: Do they use a castle nut to make 23 sure it doe sn ' t ro ta te that?
24 MR. EISSENBERG:. No, they do not.
25 MR. LIPINSKI: Then why should it turn loose?
O ACE FEDERAL REPORTERS, INC.
202-347-3700 Nationwide Coverage 800 3364646
26378.0 98 REE
(_) 1 MR. EISSENBERG: Well, if it were not pinned 2 correctly. It should not loosen, you are correct, if it is 3 correctly pinned.
4 MR. WYLIE: It is staked.
5 MR. EISSENBERG: " Staked" is a better word.
6 MR. MICHELSON: Would it be possible, then for 7 you to send to us a listing of, for your report, which 8 valves were safe ty related and which weren' t?
9 MR. EISSENBERG: Yes, if I have the information.
10 And if I don't, if I can get it from'MOVATS, I will provide '
11 it.
l 12 MR. MICHELSON: Because that doesn't identify
(~T 13 utility or anything, but I couldn't tell from the report.
V 14 MR. EISSENBERG: Okay.
15 In looking at the abnormalities we found, we 16 could divide them into groups. I think they are important 17 groisps. Some of the abnormalities were incorrect 18 adjustments that might go so far as to say design problems, 19 maybe which included incorrect adjustments. But primarily 20 the incorrect adjustments are not time-de penden t 21 degradation. They were a step function introduction. The y 22 either were there or not there. And these incorrect 23 adjustments then themselves could^ cause accelerated wear --
24 some incorrect adjustments coulb provide accelerated wear.
25 Some incorrect adjustments, on the other hand, just stay ACE-FEDERAL REPORTERS, INC.
202-347-3700 Nationwide Coverage 800-3 4 6646
26378.0 99 REE (j 1 that way. They don't cause any harm.
2 On the other hand, there are some incorrect 3 adjustments. For example, the bypass switch misse tting ,
4 which causes a defect which does not change with time , but 5 under some operating conditions could lead to the valve not 6 performing, that is failing. That is, under some
, 7 conditions if the bypass switch is not set right, the valve 8 will operate fine because the hammer blow is -- the ,
9 secondary hammer blow isn't large enough. So incorrect l
! 10 adjustments can lead to aging or they can lead to instant 11 failure under some op3 rating conditions, or they can just 12 be there and not cause any problems.
, 13 Time-dependent degradation is the area that the 14 aging program is concerned with, that is, de fects which 15 progressively get worse with time. They can be due to 16 normal service wear, they can be due to environmental 17 conditions. They can be due to incorrect adjustments.
18 However, this is the issue , because this is the 19 issue of operational readiness. You test it now and it i
20 works. The next time you try to operate it, it doesn't 21 work. That is what we are addressing.
I 22 (Slide.)
23 I . have tried to separa te , then, the 24 abnormalities which were detected by the MOVATS method into ,
e 25 these groupings. For the benefit of the aging program, 3
(v ACE-FEDERAL REPORTERS, INC.
202-347-3700 Nationwide Coverage 800 336 4646
26378.0 100 REE p
(/ 1 where this is really the kind of thing that we are 2 interested in monitoring, I note that we picked up, of 3 course , many more of these abnormalities, these se tup 4 problems, if you will, or pose maintenance problems. Some 5 of these I identified as under some operating conditions, 6 they could cause a valve not to work. These are , . there fore ,
7 serious, but they are not something that you need to 8 monitor. You need to determine in a one-time test that you
! 9 have that problem or you don't. These are the kind of 10 things that you need to monitor..
11 (Slide.)
I 12 And I have indicated here that the way that the 13 diagnostics handles the two types , the incorrect
(}
14 adjustments and the aging, then depends on the type. That 15 is, for incorrect adjustments, one needs a one shot te st 16 scheduled af ter installation and/or af ter maintenance.
17 They are operability oriented. That is, you de termine that 18 the valve operates because it is more of a no-go/go 19 situation.
20 And finally, the criteria is based on the 21 adjustment, that is, is the torque switch magnitude set 22 according to specifications? Is the limit switch covering 23 the bypass area correctly? That kind of statement. The 24 criteria, then, are objective criteria based on the design 25 information, in general.
O ACE. FEDERAL REPORTERS, INC.
202-347-3700 Nationwide Coverage 900-336 6646
1 26378.0 101 REE i
(~)h
(, 1 However, for following aging over a period of l l
2 time, you need periodic tests to. develop trends between the 3 maintenance periods. It's operational readiness oriented.
4 You are trying to predict the future, and the criteria must 5 be based on some extrapolation. You need a criteria which 6 is not objective in advance. It needs to be. If you trend 7 it and this particular degradation is getting worse, when 8 do you do maintenance? It is a much more difficult 9 question because you are also extrapolating at the time.
10 (Slide.)
11 Here is a list of diagnostic monitoring 12 parameters which we have examined for potential use in
(~T 13 detecting and trending time-dependent degradation. I have
\_)
14 indicated on this list the particular diagnostic parameters 15 which are used in MOVATS device and in the EPRI Foster-Miller 16 diagnostics. MOVATS uses spring-pack deflection, motor 17 current and the limit switch position indicator, i '
18 The EPRI device also uses spring-pack deflection 19 and it also uses the light indicator and it uses motor load, I
20 which is interpreted as motor current corrected for voltage, 21 if you will, or voltage and phase angle, if you will. The 22 primary measurement for motor load is motor current. That 23 is the thing that is changing. But it is a corrected value.
24 So in a sense , except for this difference, which in general 25 will be minor , the MOVATS device and the EPRI Foster-Miller
()
ACE-FEDERAL REPORTERS, INC.
202-347-3700 Nationwide Coverase 800-336-6646
26378.0 102
- REE 1 -
1 ~ device , as far as diagnostics go, are the same.
2 Now, again, as I mentioned in my comment earlier, 3 the EPRI device goes beyond the diagnostics and goes into i ,
4 the controls area and we are not -- our ' program is not 5 concerned'with improvements in valve design or in valve 6 control.
7 I also indicated that the ASME section XI, 8 subsection IWV has one diagnostic indicated, that is valve 9 stroke time. We have looked at all of those --
! 10 MR. MICHELSON: Exc use me . On the valve stroke 11 time, have you looked into the slip rating of the.various 12 motors, in other words, the torque versus slip --
l 13 MR. EISSENBERG:
{'} Yes, sir.
14 MR. MICHELSON: -- to determine where the breakdown occurs?
15 And what have you found?
16 MR. EISSENBERG: Okay. Number one, it depends 17 on the particular motor. We have gotten information ;
18 regarding motors characteristically used in MOVs. ,
19 MR. MICHELSON : How many different motors are l 20 characteristically used?
21- MR. EISSENBERG: How many-dif ferent designs?
22 MR. MICHELSON: Yes.
23 MR. EISSENBERG: I don't know the full -- we may 24 have gotten about eight or 10. But I think the fellow from 25 Limitorque probably can answer that better.
l ACE-FEDERAL REPORTERS, INC.
202-347-3700 Nationwide Coverage 800-336-6646
26378.0 -103 REE
- r'g
(,/ 1 MR. WARSING: 50 dif ferent motors.
I 2 MR. MICHELSON: And the slip range might be from 3 what to what?
4 MR. EISSENBERG: These are AC motors.
5 MR. WARSING: The general design criteria is
,' 6 similar for most of them. They would normally run from 5
- 7 to 8 percent.
1 8 MR. MICHELSON: You don't use any 20, 25 percent 9 slip motors as far as you know? Yet the ASME code talks 10 about 25 percent slip, I think, before you declare the
[
11 valve is inoperable.
I 12 MR. WARSING: I am not that familiar with that
{} 13 14 part of the ASME code.
MR. MICHELSON: Isn't that a speed reduction of 15 25 percent?
j 16 MR. EISSENBERG: I will address the --
4 17 MR..MICHELSON: With the AC batches of motors 18 that you apparently are aware of.
i 19 MR. EISSENBERG: I would like to address the 20 ASME code issue a little later.
21 MR. MICHELSON: I have asked this question and 1 22 have gotten. about the answer you gave me. 'But I keep 23 puzzling that maybe somebody else knows something.
?
i 24 MR. EISSENBERG: Wa have obtained also data from i
25 a utility which gives values of stroke ; time taken during
()
ACE-FEDERAL REPO.RTBRS, INC.
202-347-3700 Nationeide Ccverage .8t& 324 co46 j
_ _ __ _ . . _ . _ _ . _ _ . _ _ . _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ . . . _ _ _ _ _ _ _ _ . . . _ _ _ _ _ _ . _ . _ _ _ _ _ _ _ . _ _ _ . __.______________.___________3
26378.0 104 REE
' _/
_ 1 in-service inspections, plotted as a function of time for 2 many different valves. We find a great degree of 3 variability, random variability, as well as in some cases a 1
4 trend. We have not gone back to the utility to decipher i 1
5 the reason, because it exceeds the values that we would 6 have expected based on the motor charactoristic.
7 MR. MICHELSON: You are trying to figure out how 8 that drop to --
9 MR. EISSENBERG: It may have to do with the way 10 they take the te s t . Let me get on with this.
11 In atidition, then, to these diagnostics which 12 have been under study by MOVATS and commercially available
(^]
LJ 13 through t10 VATS and also EPRI, we havo been looking at thene 14 other indica tora as well, looking at these very carefully 15 to see what they tell us about time-dependent dcoradation 16 or trending, Kind cf a philosophical statement, l
17 che,racteristico of a good diagnostic taonitoring system, it )
18 phould detect and dif terentia te 811 the cignificant 19 u degradaticus leading to failures. It should provide 20 reproduceible and trenoable data. It should provide data l
21 , during plant o9er?.tton. It should provido criteria for t.hn 22 maintenance actions based on operation 61 readinesh under 23 tha worst anticipated conditions. It should be easily or 21 petianer.tly installed. It should be usable by utility 25 oodr.4tions and maintenanch staff, and it should be
(~'s l 1
j >
l
}
ACE-FFDERAL REPORTEP.S. INC.
202447-3700 Nationwide Covenge 800-1M %
26378.0 '105 REE
(~
, (_)) 1 cost-effective in terms of risk reduction as compared to 2 the alternate approaches.
3 These are a general set against which we coald 4 measure various diagnostics. That is an ideal set. Some 5 conclusions regarding the MOVATS system as measured, for t
6 example, against the criteria , the first two items, this i 7 tells what it can do. There is no question that it can he 8 detect and dif ferentiate many or maybe all incorrect 9 adjustments. It can detect and to some extent 10 dif fe ren tia te the time-dependent degradations. That is, 11 some of the characteristic MOVATS curves will tell you that 12 there is a degradation. It may tell you that it is getting ,
13 worse with time, but it may not tell you wha t it is unless k's}
14 you do additional diagnostics such as disassembly and 15 inspection. It is not a complete se t for detecting and 16 dif ferentiating time-dependent degradation.
17 Trie next three items ~ have to do with what we 18 would like to have it do. Can it trend degradations? I 19 haven't seen datg at this point from MOVATS where it has 20 been applied suc/essively over many periods of time and we 21 can see that there are trends. That is, I know it is 22 qualita tive ; th? question is, is it quantitative?
23 At this point, it is my understanding that 24 MOVATS is not a system that can be lef t in place. Again, 25 it potentially can be done , but I put a question mark. And
~~ -
()3 ACE-FEDERAL REPORTERS, INC.
ll 202-347-3700 Nationwide Coverage 800 336-6646
26378.0 106 REE rN
(_) 1 I don't know of any criteria for operational readiness for 2 the aging e f fects.
3 Now, the criteria for operational readiness for 4 incorrect adjustments, as I.said, are generally objective 5 criteria. And when you correct it, you then have corrected 6 that issue. But when it comes to degradation, the ability 7 to predict, I don ' t know of any information from MOVATS 8 which will allow us to predict when maintenance should be j 9 done. It is not convenient to install and remove. It 10 requires many man-hours per valve to install and remove, 11 and it does involve radiation exposure since it involves 12 people going to the valve. And it is relatively expensive.
(} 13 Per valve it would cost -- considering the number of valves 14 involved, it seems to to me to be high.
15 (Slide.)
16 We have been , as I no te d , investigating i
17 potential diagnostic methods using two motor-opera ted 18 valves mounted in te s t stands. These valves are not in 19 flow loops. They simply can operate. They are attached 20 -- operators are attached to valves as indicated here. The. ,
21 first valve we tested was a 6-inch globe valve with an +
22 SMA-2 opera tor. It is an old style operator. It was the 23 valve that we could get our hands on and start te s ting 24 diagnostic methods early, the logic being that the same 25 diagnostic methods as tested here could be applied to a O
a ACE-FEDERAL REPORTERS, INC.
202-347 3700 Nationwide Coverage 800-336-6646
26378.0 107 REE
) 1 nuclear grade, more modern design operator.
2 And in fact, we have also obtained more recently 3 an SMB-1 operator on an 18-inch gate valve. This is a 4 nuclear-qualified valve which we obtained from surplus from 5 TVA Hartsville plant. We now have both of these valves 6 installed and we are in a position to test diagnostic 7 methods on both of these, flipping from one to the other.
8 (Slide.)
9 Not in your handout, I have a picture of myself 10 and a couple other folks testing or posing like we are 11 te s ting this. This is the 6-inch nonnuclear valve and this 12 is the TVA valve . These are the two operators. And you
{} 13 can see some of the diagnostics installed on this valve and 14 some of the readout devices. We don ' t -- these are 15 overkilled from the standpoint of practical effort, but we 16 are in the phase of understanding what tLe diagnostics tell 17 us. We have gone beyond that. I guess we have another 18 table over here we are looking at -- we are unraveling what 19 the diagnostics tell us.
20 Based on our investigations to date -- and we
- 21 are really. winding down this phase of the investigation, 22 that is, we think we know what are good diagnostics; we 23 reserve the right to change our mind -- these look like 24 they are the key diagnostics
- Motor current, stem thrust, 25 directly or as torque switch angle. And I agree with the
()
ACE-FEDERAL REPORTE.JS, INC, 202-347-3700 Nationwide Coverage 433Mi646
26378.0 108 REE
(~h
(,,) 1 comment made by the Limitorque represantative , that stem 2 thrust directly is preferred. And we are going in that 3 direction also, looking at strained gauge devices, strained i 4 gauge bolts and nuts, washers that can continuously monitor 5 the direct thrust. We haven' t given up on torque switch 6 angle. We may be able to understand a lot about stem 7 thrust without going to one of these devices. But this
- 8 would be preferred. There seems to be no way out of
! 9 wanting to do this.
10 Motor stator winding temperature can tell us --
11 MR. BROOKS: One of the things in using just 12 motor current alone as a parameter, recognizing that some 13 of the power plant voltages can vary as low as 75 percent
~}
14 of graded on up to 115 percent of graded, that makes a 15 significant difference in the current that will be measured 16 also.
17 MR. EISSENBERG: You are right.
18 MR. BROOKS: I would suggest that you consider 19 motor load.
20 MR. EISSENBERG: I agree with your comment that s
21 in order to know the power input one must know voltage , but 22 the diagnostics go in a different direction than the 23 magnitude of the motor current. They have to deal with the 24 information content of the signal. And the information 25 content of the motor current signal is very powerful and we
[ \
L_)
ACE-FEDERAL REPORTERS, INC.
202-347-3700 NationwideCoverage 3 800-336-6646
A i
26378.0 _
109 REE (T
(_) 1 don't need the voltage.
2 Now, we do need the voltage to tell us the power d
,3 i input. I agree with thaE. But diagnostics go beyond motor
- i. ,
4 load.
5 The winding tsmpera ture is important because it 6 tells us the state of the motor separate f rom the sta te of e
7 the operator. We could, from winding temperature , increase 8 this -- I am thinking here -- winding tempera ture increases 9 per stroke or per 10 strokes as an indicator of the load 10 that the operator has seen. Also, it is a diagnostic that 11 tells us something is the matter with the motor. And it 12 also is ultimately a better way to measure the overload of 13 the' motor than a motor overload device which is a current
(~')
u-14 integration device.
15 MR. MICHELSON: But you don't have the pleasure 16 of 10 strokes, necessarily, in the real world application 17 at one point in time. So unless you have several strokes, i
g-18 you don't begin to see some -- get some real numbers.
19 MR. EISSENBERG: But one could do an in-service 20 te s t . The requirement could .sayi' Strokes the valve five 21 times. It is third on our list._ This is kind of~more or 22 less in the priority order.
23 This gives us also information about how near 24 the motor is to burnout. And it is a better ' indicator in 25 some ways than motor current, tha t ,is , using thermal O * '
's )
ACE-FEDERAL REPORTERS, .INC.
202-347 3700 Nationwide Coverage 800-33MM6
't
-26378.0 110 REE
())
(_ 1 overloads based on motor current. However, there are a lot
, 2 of problems with this'also.~ I think to unravel the issue 3 of whether temperature or motor current, integrated motor 4 current is the better criteria for an overload of a motor, 5 is a debatable issue. I know that Limitorque and its 6 competitors have used dif ferent philosophies with regard to 7 motor overload.
8 I know that NRC uses the integrated current.
9 MR. MICHELSON: How many data points do you take 10 for stator winding?
11 MR. EISSENBERG: Two. We make two measurements.
12 MR. MICHELSON: I mean how many-thermocouples do
(} 13 you put in --
j 14 MR. EISSENBERG: Two. At this. point -- this is 15 a primitive device at this point. We have one which is on 16 the outside of the stator winding but inside the casing, i
4 17 and one on the outside of the casing just to see what it f 18 measures. It is a weak diagnostic. However , we have 19 talked with a thermocouple' manufacturer who makes' exotic
! 20 thermocouples for nuclear safety research, and he makes 21 thermocouples that perhaps could be wound into a motor 22 winding.
23 MR. MICHELSON: The problem is the variation of 24 temperature around the casing and whatever, and I .just 25 wondered how --
O ACE-FEDERAL REPORTERS, INC.
202-347-3700 Nationwide Coverage Sno-33H646
26378.0 111 REE i 1
.xn 1 k,) 1 MR. EISSENBERG: We have not taken this except i
2 to note that a winding tempe ra ture is a very good parameter. )
l 3 How you can measure it accurately and use it is another i 4 question. I acknowledge your criticism.
5 MR. MICHELSON: Could you use the resistance of 6 the winding as an accurate measurement of temperature?
7 MR. EISSENBERG: I thought about that. Put a DC 8 voltage across it and measure the resistance, and I have 9 used that in other research to measure the average 10 temperature of a piece of metal.
11 MR. MICHELSON: That is what you are looking for.
12 MR. EISSENBERG: For a f a ilure , you might be 13 looking for hot spot.
()3 14 MR. MICHELSON: You might, but you have to have 15 a lot of thermocouples.
16 MR. EISSENBERG: I agree. It is a useful 17 diagnostic. It tells you information that you can' t get 18 from these other sources. That is, it is a good parameter, 19 but I don't know exactly how to measure it best. Motor 20 gearbox vibration will measure such things as loose bolts 21 and loose parts in general in the operator or in the valve.
22' So it is another diagnostic which you can' t pick up from 23 these other ones.
24 So this se t , which again may not be complete , to 25 us represents a pretty good set of diagnostics and seems
()
ACE-FEDERAL REPORTERS, INC.
202-347-3700 Nationwide Coverage 80 4 336 6646
26378.0 112 REE (O
s_/ 1 like a practical set.
2 (Slide.)
3 Howeve r , looking into this se t , we have reached 4 the conclusion that the motor current was potentially the 5 most valuable diagnostic parameter for monitoring MOVs. I 6 note it can be used by itself as well as in conjunction 7 with the other parameters. It will accurately detect 8 time-dependent loads imposed on the motor during valve 9 actuation, with the stipulation when proper signalling 10 conditioning is utilized.
11 The diagnostic equipment can be simply 12 permanently installed at the motor control center without
(~% 13 lifting leads, using a clamp on current indicator. It is L) 14 inexpensive, very inexpensive. It can be used in 15 conjunction with other diagnostic parameters, such as the 16 other ones I listed, to define and dif ferentiate many types 17 of degradation and service wear.
1 18 We see motor current as being the primary 19 monitoring device. That is, that is the one that you I
20 always monitor and if you see a deviation from some 21 established baseline, then you can apply some of these 4
22 other diagnostics. But since motor current can be measured
- 23 without anybody entering containment and can be in fact
- 24 easily monitored every time the valve is stroked at no 25 great problem -- once you can do it once, you can do it O
ACE-FEDERAL REPORTERS, INC.
202-347-3700 Nationwide Coverage 800-336-6646
26378.0 113 REE
/O (m) l continuously -- it in itself provides the diagnostic. With 2 regard to the details of the information --
3 MR. REED: I am glad you came up with that, 4 because if you check out with a plant that has a 5 well-skilled , compe tent maintenance group, you will find 6 that invariably, they always use motor _ current for their 7 fine tuning and their analysis.
8 MR. EISSENBERG: We ll, at the last meeting I 9 presented some data that we obtained with regard to motor 10 current. We have gone much further in sophisticated 11 analysis of the fine structure of motor current and have 12 been amazed at the information content that it tells about 13 the mechanical condition of the valve.
}
14 I would like to present the da ta , but it was 15 such a good bit of data that the DOE patent people are 16 looking into it as to whe ther there is some potential 17 pa ten ts involved. They have clamped the lid. In fact, 18 they have clamped - the lid af ter the fact in a sense, after 1
19 we showed them some of the more recent data. So it is a 20 very promising method. I can provide information to the 21 ACRS, but it is not available at this time.
22 MR. REED: What is this, patents in the --
23 standing in the way of safe operation? Probably most good 24 electrical foremen know this anyway.
25 MR. EISSENBERG: You are right. However, the y Q>
ACE-FEDERAL REPORTERS, INC.
202-347-3700 Nationwide Coverage 800-336-6646
26378.0 114 REE
, 1 may not know just how much they can get. Some of them do.
2 And I am not saying that I.think it is patentable . I am 3 saying that we had the lid . clamped on us by the DOE.
4 MR. MICHELSON: They may not have a 5 microprocessor in their brain to convert all those wave 1 6 forms into inter.esting information.
- 7 MR. EISSENBERG
- We think that the foreman does 8 not have to be an electronics specialist but that the 9 information content is remarkable.
10 Okay. So what I am proposing as our conclusion 11 is that motor current is a fine primary diagnostic and that 12 the other items listed are secondary diagnostics to do the
{} 13 14 dif ferentiating and to do some of the quantification.
(Slide.)
15 MR. MICHELSON: Will some of your work include 16 doing similar testing under various degrees of reduced 17 voltage or transient voltage conditions?
18 MR. EISSENBERG: You jumped ahead.
19 MR. MICHELSON: Okay. Excuse me.
20 MR. EISSENBERG: What we are currently involved 21 in now, for this set of diagnostics, is to evaluate its 22 potential. W'e are looking at reproducibility, sensitivity, 23 selectivity and trendability. How quantitative is it, et 24 ce tera? How useful is it in a real world noisy system, et 25 cetera?
O
\_) .
ACE-FEDERAL REPORTERS, INC.
202 147-3700 Nationwide Coverage 80 4 336-6646
26378.0 115 REE p
's ) 1 A data base of degradation measurements of a 2 variety of MOVs in dif ferent plants will be collected in 3 cooperation with the utilities and other industrial plants.
4 This next s te p , we have the set, we need to 5 evaluate the se t to just tell us just what it can de.
6 The criteria issue. The criteria for MOV 7 operational readiness is, by definition almost, operation 8 under some worst case condition. -
9 By worst case condition, I mean such things as 10 the maximum mechanical loading of the valve, either delta P 11 or pressure; maximum environmental temperature and the 12 impact that has on the motor and on the electrical 13 equipment, and on thermal expansion; and reduced voltage to 14 the motor.
15 At the present time, the IEEE 382 standard for
- 16 equipment qualification of motor-operated valves includes a 17 requirement that it operate two cycles at reduced voltage, 18 taken to mean, I think, 80 or 75 percent voltage. Whe ther 19 that is a realistic worst case or not is beyond the scope 1
l 20 of our study. There is some indication that it may be i
21 unrealistically harsh. That is, that the reduced voltage 22 shouldn' t be expected. 75 percent voltage is not good for 23 plant operation. It may be just under LOCA conditions when i 24 the diesel generator is put on and loads are being picked i
25 up. But that is beyond the scope of our program. But our Q
b ACE-FEDERAL REPORTERS, INC.
202-347-3700 Nationwide Coverage 800-336-6646
26378.0 116 REE
/~'s
\_) 1 program, then, is basically obligated to include under 2 operational readiness the reduced voltage requirement.
3 MR. MICHELSON: One of the interesting problems 4 with reduced voltage is that if you have to make a transfer 5 to the diesel engines in the time -- the largest motors 6 start first on the diesel and so forth, but the valves 7 generally are not transferred in time sequence. They just 8 come on as soon as the diesel comes up and the breakers 9 close on the load. So a number of these motor-operated 10 valves may be trying to opera te or are in some stage of 11 operation at the same time the 2000 horsepower motor is 12 loaded on. That isn't a 70 percent drop _now, it is a very
{} 13 large drop, momentary drop on the line. Are you going to 14 look to see what such large voltage fluctuations of very 15 short duration can do to the equipment?
16 It could be a particular problem. Of course , on 17 things like microprocessors which I don't know, I didn't 18 get a chance to ask the gentleman, Mr. Brooks, how that is 19 handled -- but these things will see severe voltage drops 20 in the source of power unless it is separa te from the motor 21 control center to the microprocessor. It sees a voltage 22 drop that is way down.
23 MR. BROOKS: I sensed that as a question to me.
24 That was the reason I was describing earlier that you do 25 have to work with a variation of motor voltage between 75 O
ACE-FEDERAL REPORTERS, INC.
202-347-3700 Nationwide Coverage 80 4 336-6646
_ . ~ ,_ _ _ . . - _
! 26378.0 117 REE n
- (_ 1 ' percent and say 115 percent, and your current. stays 2 constant, you have a significantly dif ferent value of motor 3 load.
J 4 MR. MICHELSON: But you realize that for 5 milliseconds, the voltage is much lower than 75 percent to 4
6 the microprocessor, because when a big 2000 horsepower 7 motor closes to an a diesel engine, the voltage really dips.
8 Its recovery is very fast, but it really dips.
9 MR. EISSENBERG: I wanted to mention that in 10 discussing the issue of reduced voltage; again, I would 11 like to say that it is beyond our scope to tell us what 12 this is. I think your guidance to us is a good guidance,
{} 13 but it is guidance to us. We can't infer it. But I did 14 talk with a representative of a utility that has done some 15 looking into it. That is Brian Curry. Brian and I talked 16 about the issue of reduced voltage. He might want to add a 17 word.
18 MR. CURRY: We did a voltage study at our 19 Limerick plant and found out that we went below 80 percent.
20 voltage for a total time period ' of six cycles. So that it 21 is not a significant period that it is on.
22 The other thing that happens is you have got to 23 look at how the loads are applied, that your diesel comes 24 up and your first load that comes on is your RHR pump, 25 which is your larger load. So many seconds af ter that, O
ACE-FEDERAL REPORTERS, INC.
202-347-3700 Nationwide Coverage 800-336-6646
i
> 26378.0 118 j REE 1
V 1 your load center closes in, which picks up all your valves i 2 and your valves start stroking at that time. There are two i 1 3 motors tha t ' come af te r tha t, but they are smaller. But l
j 4 your RHR gives you a voltage source when the voltage - dips.
5 MR. MICHELSON: The assumption is you lose
! 6 of f-site power times ::ero and then that' sequence you 7 outlined is quite right. If you lose of f-site power , then 1
8 you have got to start other interesting combinations
)
9 because now the -- depending on how the sequencing is 10 arranged, of course --
l 11 MR. CURRY: For our plant -- I can only speak a
j 12 for our plant -- under LOCA conditions, we stripped a bus j
, (JN 13 and we come back up again, right in sequence with the
- 14 diesel loading.
15 MR. MICHELSON: You never start your
- 16 motor-operated valves in combination with the 2000 17 horsepower motor , then? ,
18 MR. CURRY: No. It has to be sequential.
19 Otherwise , you will lose your diesel, i
20 MR. EISSENBERG: I think that the issue --
1 l 21 MR. LIPINSKI: Let me make a comment at this i 22 point on the microprocessor. They can always put filters 23 on the front end. They normally run at 12 volts, depending i
24 on the size of the capacity you want' to put on there. You 25 could run for 10 seconds without an appreciable voltage O
ACE. FEDERAL REPORTERS, INC.
202 347 3700 Nationwide Coverage 800-336 6646
26378.0 119 REE
/ 1 drop.
2 MR. MICHELSON: So it wouldn' t be a problem if 3 they put adequate filtering on it.
4 MR. LIPINSKI: Adequa te filtering.
4 5 MR. MICHELSON: But it would be something they 6 would want to be aware of.
7 MR. LIPINSKI: Right. If they have a severe l 8 number of beyond 75 percent drop, they should take that to j 9 their designers.
10 MR. MICHELSON: They have to be designed for the 11 75 percent drop without doing strange things.
12 Are microprocessors usually able to handle 25 13 percent drop in voltage?
14 MR. LIPINSKI: Yes, you can put a voltage 15 regulator on the front. That is part of your voltage 16 regulator design. I want 5 volts on my microprocessor with 17 this kind of swing on the AC. Or if I want it to survive a 18 dip over --
19 MR. MICHELSON: Power supply design. Okay.
20 What happens if you lose the power comple tely to 21 a microprocessor?
22 MR. LIPINSKI: With time it will die unless you 23 put in a battery. You could put in a --
24 MR. MICHELSON: Keep in mind that it takes 10 25 seconds to get power back.
O,.
l l
ACE-FEDERAL REPORTERS, INC.
202-347-3700 Nationwide Coverage 800 336-66 4
26378.0 120 REE 1 MR. LIPINSKI: You could put in a battery 2 backup as part' of your specs, a chargeable battery. That
, 3 thing could sit there for an hour or two.
4 MR. MICHELSON: Well, then you are all right.
5 Maybe that is what they will have to do if this thing 6 becomes a useful device.
7 MR. LIPINSKI: You could put in a chargeable
- 3 battery.
9 MR. MICHELSON: Then you have batteries to worry 10 about. Thank you. Excuse me.
11 MR. EISSENBERG: I think that this conversation 12 addresses an issue which I think needs to be addressed.
(~} 13 Given no other ictormar. ion, we need to te s t the operational v
14 readiness of aged operators with its other defects at 15 reduced voltage. That would be taken to mean 75 or 80 i
16 percent voltage, unless otherwise told by NRC. That may 17 represent an overly conservative design. It may be worth i
18 looking at. But in the meantime, we will be expecting that 19 the valve should operate over its complete cycle.
20 (Slide.)
21 I would like to wind up with some mention of 22 other activities going on with regard to MOV operational 23 readiness that we have been involved in, particularly an-24 informal workshop to discuss MOV diagnostics and activities 25 with the ASME O&M committee.
i ACE-FEDERAL REPORTERS, INC.
202-347-3700 Nationwide Coverage 804 336-6646
26378.0 121 REE 1 (Slide.)
2 We recently held or convened informal -- I would 3 like to stress that word -- informal MOV diagnostics 4 workshops. Seven utilities participated in an exchange of 5 MOV operating experiences and a discussion of diagnostic 6 monitoring options, including both motor-operated 7 monitoring and also valve leakage monitoring. The valve 8 leakage monitoring technology was provided by Philadelphia 9 Electric Company. And it was one of the seven utilities.
10 The workshop provided the basis for future 11 cooperative approach. If you will, a users group, a 12 pseudo-users group approach to improving operational
/~T 13 readiness. That is, this type of group can provide a data V
14 base from nuclear plants for evaluating diagnostic 15 monitoring options. Data base of valve information, It 16 can provide expert judgment in evaluating. This is the 17 point that was made earlier, that we need to have 18 main tenance people on board. There were maintenance people
, 19 at this meeting. And Brian can attest to it, not only from 20 his utility but-f%om all of the other utilities. There 21 were nuta and bolts people.
22 And we then can also provide recommendations to 23 the ASME O&M committee on the new standards that are being 24 developed . This can provide a consensus of the utility 25 side of what can be done to improve OM 8 and OM 10 which k
. ACE-FEDERAL REPORTERS, INC.
202-347 3700 Nationwide Coverage 800 3 % 6646
26378.0 122 REE i?~) ss 1 are the standards which address motor-operated valves.
2 They could be incorporated in technical specifications by 8 I 3 reference. Right now the technical specifications 4 re ference the section XI. But I assume at some point the y 5 would re ference O&M standards 10 and 8.
6 MR. MICHELSON: Does the NRC have a 7 representative on the O&M committee?
i' 8 MR. EISSENBERG: Yes, sir. I think there are 9 are several representatives behind you.
10 MR. MICHELSON: Who is that? Several on one.
i 11 comm it tee .
12 MR. BOSNAK: Joel Page is our principal
(} 13 representative. He is on the OM 6 and 10, which is the 14 pumps and valves. One is pumps and the other is valves.
i' 15 MR. BOSNAK: Right now they are undergoing a 16 rewrite of this whole area, particularly with respect to l 17 valves. The OM 8 we are not too well plugged into. That
- 18 is the one on operators.
i 19 MR. MICHELSON: .In the valve committee , they are 20 including all kind of valves or just particular --
21 MR. BOSNAK: They have all kind of valves.
22 Perhaps Joel can tell you a little bit more.
i 23 MR. EISSENBERG: I am going to cover some of 24 that material. I will ask Joel Page to add anything that I 25 may have left out.
O -
ACE. FEDERAL REPORTERS, INC, 202 347-3700 Nationwide Coverage 800 336-6646
i 26378.0 123 REE 1 I want to wind up with this Vugraph. This 2 workshop was successful. We did have - seven utilities. The 3 seven utilities were chosen by this process. I happened to 4 have representatives from each of these utilities who 5 expressed an in te re st in improving MOV diagnostics. I 6 asked if they wouldn't attend a workshop to get toge ther I 7 and talk about it. We did get all of the seven. In fact, 8 we got all of the five that I invited and two kind of 9 inv ited themse lve s . I think it was a very useful meeting.
10 The objective from our viewpoint is primarily 11 here. We need a data base and we have some ten ta tive 12 commitments from some of the utilities that if we would
(}
13 supply them with what measurements we want, they would 14 supply us with tape-recorded data. We could then use ours.
15 I would like to now turn to the ASME O&M 16 committee activities. Our contribution has been primarily 17 in these three areas. We do have a representative on the 18 ASME O&M main committee and the Subcommittee on Performance I 19 Testing which governs these two standards. That is the 20 same fe llow, Bill Greenstreet of ORNL.
1 21 One of the first things that Bill Greenstreet 22 did was to recognize that the standards were very loose.
23 That is, the standards varied from one' to the other. lie 24 recommended and was asked to lead a subcommittee to prepare 25 a standards format and content procedure. And that has O
ACE-FEDERAL REPORTERS, INC.
202-347 3700 Nationwide Coverage 800-3 4 6646
_ . _ _ . . _ - _ _ _ - . -_ _ -. _ _ _ _ - _ _ _ _ - _ _ _ - _ - . . - - - _ - - -=- - - - . _ _ _ _ - - _ _ - _ - _ _ - _
26378.'O 124 REE 1 been issued. It does tighten up just what the standards 2 should contain. It makes them uniform and it makes them
- 3 stronger.
4 I don't know if you were involved in that, Joel, l i
5 or not. It has been -- it was accepted by the Subcommittee.
{
2 6 OM 10 is requirements for in-service testing of 7 valves and this includes, as you point out, Carlyle, I
- 8 think, that this includes all valves: Hydraulic, pneumatic 9 and motor operated, solenoid, et ce tera.
10 Embodied in OM 10 is basically the same criteria ,
11 the same diagnostic that is in section XI, that is, stroke i
3 12 time. That is, the criteria against which OM 10 determines f 13 operability of the valve.
14 OM 8 refers to -- OM 10 is in its final stage, I l 15 believe. And I guess I have a Vugraph on that. So let me i 16 get to it. But OM 8 then is requirements pertaining i
17 specifically. to the valve assemblies. It is really looking 18 at the . ope ra tor . OM 8 is in a somewhat more primitive 19 s ta te .
20 We also, not part of this presentation, are I 21 involved in and participating in tim working group OM 6
. 22 which is looking at pumps.
i 23 A word about administrative procedures, 6200.
24 It provides for consistent standards format and content, l 25 and it provides guidance in preparing standards, wording to O
Ace-FEDERAL REPORIERS, INC.
202 347-3700 ' Nationwide Coverage 800-336-6646
l 26378.0. 125 i REE (R-
\_) 1 be used, which are consistent and comprehensible.
l 2 (Slide.)
l l
l 3 Some of the standards are practically l
4 I nonstandards. They really don't require very much. They 5 leave it up to the owner. That is not a good standard. CM 6 10 working group is -- which just had its meeting at San 7 Diego about a month ago; Bill Greenstreet and Joel Page l
l 8 attended -- they are considering modifications to the l
l 9 stroke time criteria. They are establishing separate l
-10 criter ia for electric motor-actuated valves separate from 11 the other types of valves, and ' they considered reducing the j 12 allowable magnitude of the change in the MOV stroke time.
l
(} 13 In fact, they have reduced it to 15 pe rcen t , plus or minus 14 15 percent change in stroke time as the criteria for action, 15 which has changed from 25 percent.
16 Whether that is any more realistic or not is l 17 another issue. It seems to me that there are inadequacies l 18 in relying on stroke time to determine MOV operational l
l 19 readiness. Those inadequacies could be addressed in OM 8, 20 but the s ta tement is still a valid statement. And such I
l 21 things as dif ferences between DC and AC motors with regard l
l 22 to torque speed relationship, as was pointed out to me by l 23 Joel, there are very few DC motors in plant safe ty systems.
l l 24 However, the criteria are very dif ferent.
25 MR. MICHELSON: There are quite a number of them O -
l l ACE-FEDERAL REPORTERS, INC.
l 202-347 3700 Nationwide Coverase 800 336-6646
26378.0 126 REE 1 in boiling water- reactors. A large number in boiling water 2 reactors. Because a lot of the isolation valves, one is AC 3 and one is DC. I assume that is true on Limerick as well.
4 MR. CURRY: Yoc.
5 MR. MICHELSON: So look carefully. There are a 6 lot of them on boilers.
7 What do you do about the variation on the AC 8 motors? I guess Limitorque has kind of said that there 9 isn't much. ,
10 MR. EISSENBERG: We ll , we think it may have to 11 do with these two (indica ting) . S troke time can be taken 12 to mean the time between the one lig ht , the close light
(} 13 going off and the open light going of f.
14 MR. MICHELSON: That depends on how you adjust !
15 the limit switches.
16 MR. EISSENBERG: That is right. You are not 17 telling the full stroke but the intermedia te part, between 18 say 95 percent to 5 percent open. That is one way.
19 Other utilities might define it as f rom the time 20 that the switch is thrown to when the light indica te s 21 closed. But that also isn' t the full stroke. I don't know 22 that you can at this point measure the stroke time going ;
23 all the way to seating because there is no indication of 24 seating in the control room. There is just an indication 25 of 95 percent of stroke or 90 percent or whatever.
(
i l ACE-FEDERAL REPORTERS, INC.
202-347-3700 Nationwide Coverase 800 336-6646
- - . - =- - .. ~ . . -. _. ..
f
.i 26378.0 127 ,.
2 REE 1 MR. MICHELSON: Are there extra contacts on a 2 torque switch?
3 MR. WARSING: Not normally, no.
4 HR. MICHEL3ON: You could use it, if you usad :
5 power input as the initial and opening of the torque as the
! 6 final, that is probably about as close as you would get 7 today.
i 8 MR. EISSENBERG: The ambiguities associated with -
, 9 that may be a contributor, because if a maintenance action I 10 involves resetting of the limit switch and then you take 4
11 stroke time , you might find a different stroke time .
I 12 MR. MICHELSON: If you are out of tech specs, l,
(} 13 you go to limit switch to get back in, which isn't quite 14 the spirit of the idea.
- i 15 MR. EISSENBERG: I think that the definition 1
j 16 needs to be addressed more carefully. Also the te s t :
, 17 conditions. The stroke time will vary with the load. And l 18 if the , for example, the packing tightnesses change or the
, 19 lubrication on the stem nut, it could affect these.
20 (Slide.)
21 I am getting near the end, I think. Briefly on ;
j 22 OM 8, it is progressing slowly. There are some i
, 23 unsatisfactory issues involved. They are currently
[ 24 preparing a draf t which incorporates changes to accommodate 25 Subcommittee comments and new information. But our feeling ACE FEDERAL REPORTERS, INC.
202 347 3700 Nationwide Coverage 8f6346646
26378.0 128 REE 1 is that the proposed monitoring methods in OM 8 -- I have a 2 draft copy of that standard for the Committee if they are 3 interested. They are not suf ficiently identified or 4 defined, the drates that wc havc se c .i . Oc there in m"ch 5 work to be done on OM 8. ,
6 MR. MICHELSON: Do you have a draft of OM 10? :
7 MR. EISSENBERG: I don't have it, but Joel --
8 MR. PAGE: I do not have it with me, no.
9 MR. MICHELSON: Could you send one copy to our 10 staff?
11 MR. PAGE: Yes. One copy?
12 MR. MICHELSON: Yes. !!e will make extras. Are
(} 13 14 there questions from the Committee? If it is agreeable with you, we could take a question or two from the audience.
15 I am sure there are perhaps some and then from the Staff as ,
16 well. If you wish.
17 MR. EISSENBERG: That is fine.
18 MR. MICilELSON: Are there comments or questions?
19 MR. EISSENBERG: The second presentation which I 20 have on check valves shouldn't take as long. So we can 21 move into tha t --
22 MR. MICHELSON: I would like to cover- this first 23 because this other is a little different. Are there 24 questions or comments that people might like to make.
25 MR. WARSING: You said earlier that in using ACE FEDERAL REPORTERS, INC.
202 347 3700 Nationwide Coversee MS3366646
26378.0 129 REE 1 motor current as a diagnostic tool, that you didn' t need 1to .
2 know'what the motor power was. Can you explain how you are 3 using it as a diagnostic tool in a little more detail?
4 MR. EISSENBERG: I would like to, but let me say 5 that it has to do with the time variability within the 6 stroke. There fore , the absolute magnitude , unless the 7 voltage is changing during the stroke , the voltage is not 8 some thing that tells us anything about the health of the 9 valve.
I j 10 MR. WARSING: But if you compare one run to the f
11 next, say you tested last month and this month and the l 12 voltage changes from test run to test run, that will'have ,
l
(} 13 14 an ef fect on the current.
j MR. EISSENBERG Yes. And let me say that we 1
l 15 need to incorporate voltage into a diagnostic. It is not a 16 sensitive diagnostic in itself. It is a normalizing
! 17 pa rame te r , if you will. In that regard, yes, I agree with 18 you. Not even just voltage but voltage times phase angle.
19 You need watts. I agree with that. But as a diagnostic, 20 again, to tell us the mechanical behavior of the valve 21 during the stroke, motor current is sufficient.
22 MR. MICl!ELSON: llave you looked at power input, t 23 time-dependent-power input?
i 24 MR. EISSENBERG: No.
25 MR. MICllELSON: The form of power, have you O
1 ACE FEDERAL REPORTERS, INC.
l
! 202 347 Nationoide Coverage 2 346646 I __________________._.___.._.______________________________-3700_ . _ _ _ _ _ _ _ _
.- = . . . . - - _. .-.- , ...- .
26378.0 130 i REE ,
I watched the wave form of power?
2 MR. EISSENBERG: W'e haven' t measured power. ;
3 What I am saying is, we can and I accept the comment.
'4 MR. MICHELSON: It is a'very cimple measurement, ,
i S MR. EISSENBERG: We have dono signal r
6 conditioning on motor current so.that it tells us, - we have .
)
i 7 all the wave forms of motor current.
i 8 MR. MICHELSON: Yes.
i 9 MR. BROOKS: I am going back to one of your !
, 10 pages. It is indicated as diagnostic approaches dependent .
11 on the type of abnormality to be detected. You indicate on 12 one of those approaches, criteria based on desired 13 adjustment. I am thinking about a torque switch adjustment.
[}
14 One of the problems that we face in our work is how do you 15 really know what end product you are getting when you
{
16 adjust a torque switch to a given cetting? .In other I i
) 17 words -- e i 18 MR. EISSENBERG: What do you mean by "ond ,
19 product"?
f i
20 MR. BROOKS: Stem thrust or sten tension?
! 21 MR. EISSENBERG Well, you would measure it. I I
i 22 have in mind here the MOVATS-type device. And NOVATS-type
)
4 23 dev ice involves calibration of the stem load with 6 load
. 24 cell and you relato the two.
j 25 MR. BROOKS: But are you doing anything that i
1 i
i ACE-FEDERAL REPORTEPS, INC. '
I 202 347-3700 Ndonwide Coverage 800.'36 4646
.-____j
- - - .- - . . . - ~ _ - -
. ,. .- -- ~.
f i ,
, 26378.0 131 i- REE 2
1 would indicate that the measured stem load is adequata for i
- 2 the performance that you want to occur? In othee Words, if 3 it takes 10,000 pounds of thrust to upen the valve, you
- 4 would deter,ine this by some calculationel means; you roay
}
5 confirm that triat hr.s actually bebn produeno, bu't it you i 6 e don' t do the first stop, you will never know wr:sthar the
- , second step is right or not.
4 4 i ?
8 MR. SIGGENBERG: You are absoli,'tely'right. 1 ]
i
- 9 think the IEEE bulletin addresses thct. It has heen j 10 addressed by othorr . TFp criteria e. gainst rehich t.he 11 diagnostic in neaaured needs to be established. And one i 12 way of establishing the criteria is by' calculation. It 16 13 not the only way. But I agree. That wasn't -- we are 14 l talkirig about the hardware , the diagnostic raethods and t,he '
1 13 I trending. -And that is a scparate subject, calculating what f 16 the criteria is at th'is point.
l :17 It is part cf the overall problem, but w,s ara
-[
- 18 davelopicg diagnostics. If we develop *- it wa are i
- 19 successful, then somebody cun calculate-it. Someonc elso 20 can calculate it and we could use that number. I agree l
l 21 with you.
j 22 MR. BROOKS: One other thing. On the next page i
j 23 where you had indicated list of diagnostic monitoring 24 par 6neters --
]
I 25 *1R. EISSENDERG: Yest
!O i
i i i.
i ACE FEDERAL REPORTERS, INC.
u 202 447-3700 Natjorwide Covciase - 80(L3%-6646 l
u__________.__
25378.0 132 REF. ;
1 MR. QROOKS: -- on the list that we have como up 2 with, we do have a display th .t shows what the valve's 3 ctroke time is. !
4l tiR, f.ISSENBERG You are right.
5 MR. BROOKS: I recall now, l
6 f tR. BROOKS: That wasn't a question, tha t was
~
7 just a s ta tement .
8 MR. MICHELSON: Does Staf f have any questions?
9 okay, then let's proceed with yottr next -- the 10 same speaker I guess. Noxt subject? I 11 t1R. VORA: Would it be okay if Frank Cherny 12 would give you a p3rspective on valves before we . discuss 13 that?
14 MR. MICliELSON: Sure.
15 MR. CHERNY: We have in NRR several I i valve-related generic issues which wo are spending a fair 16 17 amount of effort on. I .think the issoe t. hat probably comes ,
18 ,
closest to addressing most of the things that we have heard [
i 19 ' this morning and I think in'several other Subcommittee j 20 meetings of this type, we have the generic issue 2E61 which (
21 I know in at least in responses to questions at previous
'22 meetings -- the title is "In Situ Testing of Valves." This !
23 generic issue was prioritized quite some time ago. It has 1 s 24 ) had a number of concerns fed into it ovar the last couple 25 of years which possibly, if it was reprioritized today, it O
ACE FEDERAL REPORTERS, INC. .
l Io214t$he Nationwide comase answ.s46 j
26378.0 133 REE 1 might come out high.
2 The idea of this issue is to in a comprehensive 3 way, take a look at the various valve types in the plant, 4 tako a look at the types of design basis transient and
+
5 accident conditions that they have to perform under, and 6 take into account the types of currently performed in situ 7 tests, which in most cases are your ASME section XI te s ts .
- 8 With some exceptions, there are some special tests with
! 9 some of the plant test specs also.
i
- 10 We are going to be covering juat about all valve l 11 type s in there with the exception of safety and relief i
12 valves. They will not be covered in there. There are-a l
(} 13 14 number of other ways that of covering those. There is either generic issue or there is ASME code work involved,
, 15 which I think I would rather end on.
i
- i 16 2E61 is going to take into account numerous I
- 17 interfaces in other work that is going on within the agency, 18 in IE, research, for example, you have the IE bulletin 8503 a
19 which was recently sent out af ter the Davis-Besse event l
! 20 that addressed for a certain class of valves, the torque 21 switch / limit switch concerns.
1 j 22 We are going to be taking into account all of 23 the work that has been talked about today with regard to l 24 any kind of anything that anybody is doing to evaluate
- 25 dif ferent ways of signature tracing measurements. There is ACE. FEDERAL REPORTERS, INC.
202 147 3700 Natkueide Coverage IWXM M 6646
26378.0 134 REE J
1 the Oak Ridge work that was talked about, wQrk going on at 2 ETEC under a different contract. The industry is doing 3 work on that. We intend to look at all of that and see l 4 whether any of 'that should be required in addition to wha t
~
5 little is already in section XI of the code. What kind of 4
6 valves it should be applied to, we are going to be looking 7 at that rather vigorously.
8 As a result of the San Onofra 1 event, an IE-9 task group was established to look at every aspect of i 10 design and testing and maintenance and whatever else you 11 can think of in terms of check valves. For the resolutien
~;
12 of generic issue 2C61, we intend to pretty much take the ir i 13 recommendations into account, which I believe are duo
[}
14 some time this coming May. Whatever they might recommend l 15 from their very extensive effort in terms of what 16 additional in situ testing should be done on check valves,
- 17 we intend to fully utilize whenever they recommend.
- 18 There is a number of other issues, generic 19 issues that cover certain particular valve types. For 20 example, we have an interface with generic issue 105 which .
21 is the issue that covers the pressure isolation valves for 22 both PWR and BWR. Those are the two valves that form the 23 barrier. We are not quite sure what we will find in that i
24 issue, but we are closely tracking that.
25 We have an interface with, a minor interest face i
ACE. FEDERAL REPORTERS, INC. I 202-347-3700 Nationwide Coverage M10-336-M46
. I
26378.0 135 i REE j
- 1 with generic issue B 55 which is a safety relief valve i
1 2 issue that directly addresses the reliability of target '
3 route for safety relief valves on BWRs.
4 As far as the safety valves on PWRs, there 5 recently was an ASME operations and maintenance ef fort in l 1
6 that area which we had extensive participation on. The 7 standard'OM 1, in-service testing of nuclear power plant i
8 pressure relief devices which is a substantial improvement 9 over ASME section XI of the code. This standard was 10 recently adopted into section XI of the code. That goes a 11 long way in our mind in resolving any concerns about in 1
- 12 situ testing of safety valves that are required.
}
(} 13 As far as -- there has been some discussion
- 14 regarding generic issue 70 which is reliability of PORVs i
- 15 and mock valves. It a separate issue for those components.
16 Anything that is deemed necessary to improve in situ 4
17 testing for those components will be handled as a separate i 18 item and will be covered under 2E61. I should mention that l 19 there is an OM standard that is quite far alrig that 20 addresses in-service testing of PORVs. That is OM 13 by l 21 number. We can get the Subcommittee copies of those , too, 22 if you so desire.
23 Let me just check my notes here. I think that 24 pretty much is covers the -- I think the issue of testing 25 of valves is mostly covered from what I have just said.
(
I
- ACE FEDERAL REPORTERS, INC. ,
i 202-347 3700 - Nationwide Coverage 800 336-6646
26378.0 136 REE O)
(_ 1 There are a few other generic issues that I think I would 2 characterize as more system oriented in nature , which I 3 don't think I will be into myself. I don't know that much 4 about them. Anybody have any questions on any of that?
5 MR. MICHELSON: Let me ask you a couple. Which 6 organization are you af filiated with?
7 MR. CHERNY: I am in the engineering issue s 8 branch of NRR. W'e have the lead on the 2E61 and most of 9 the things that I just talked about.
10 MR. MICHELSON: So you are as close to the one 11 in charge as identifiable at this time, is that it?
12 HR. CHERNY: On things that I mentioned.
(} 13 14 MR. MICHELSON:
the operating facilities.
IE, of course, is the closest to The regional of fices I guess are I 15 close to the operating facilities. Their inspectors walk 16 through the plants on a daily basis. Do they have their 17 own program that you are aware of or how is the valve issue 18 being handled at IE? I guess you wouldn't want to talk for 19 IE.
20 MR. CHERNY: I wouldn't want to do that.
21 MR. MICHELSON . You can tell us what you know.
22 MR. CHERNY: I guess the closest inte r f ace we 23 have had with IE would be in the development of that 1
24 bulletin that I mentioned.
25 MR. MICHELSON: I am getting down to --
1 ACE-FEDERAL REPORTERS, INC.
202 347-3700 Nationwide Coverage 900 336 6646
c-26378.0 ,
137 REE ,,'
t r *4 k_%) 1 MR. CHERNY: I don't think that I should talk 2 about their overall valve program.
3 MR. MICHELSON: Le t me ask you , from your 4 vantage point, where are you getting your information? Do 5 you get it from -- do you go and talk to the IE inspec tors?
6 Do they tell you what is going on out in the field, what 7 the sta te , the condition of valve operability at the plants 8 is? How do you find this out?
9 MR. CHERNY: I guess in our own case , we have 10 been involved in reviewing, at least prior to the 11 reorganization that occurred last November, we were 12 involved in reviewing the in-service testing programs for
(} 13 valves that had been submitted for the last five years.
14 MR. MICHELSON: Programs are paper.
15 MR. CHERNY: As a result of those reviews, we 9
16 have had . meetings with 'Atilities and regional people and 17 the residents, discussing those.
18 MR. MICHELSON: IE is responsible for the sa fe ty 19 of plants. They arc the responsible organization for 20 granting and withdrawing licenses, aren' t the y?
21 MR. CHERNY: I don' t think that is -- did I say 22 NL? Did,I say IE? I meant NRR. You are responsible for 23 granting licenses and if necessary, revoking licenses. So 24 the safety of the plant. ultimately comes back to you for 25 decision-making purposes. Is the plant adequately safe to (C)
ACE FEDERAL REPORTERS, INC.
202 347 3700 ' Nationwide Coverage 800 3 4 6646 '
26378.0 138 REE 1 continue operation. So you have to have certain inputs in 2 making those judgments. One of the inputs is the 3 operability of the equipment. If it isn ' t ope rable , you 4 better start thinking about it. Research tries to do work 5 to tell you how to make such judgments. Gives you -- does 6 tests and experiments and tells you what ways it might be 7 done. But you ultimately have to make the decision.
8 IE will tell you how things are going on in the
, 9 plants. That is fed back to you. It is in that regard 10 that I wondered, are you getting periodically or in some 11 organized manner information flow from the field relative
+
12 to valve operability?.
(} 13 MR. BOSNAK: Bob Bosnak. Wha t we had -- prior.
14 to the reorganization, we had been getting inputs from the 15 regions, from the residents, from people going out to 16 inspect the programs that Frank Cherny has been talking 17 about. So that type of input is the kind of thing we were j 18 getting back in the way of feedback. But we are not 19 getting anything -- we are not getting this thing right now 20 because all of the in-service testing programs have been 21 spun of f to the vendor divisions. In other words, the 22 Westinghouse group, the B&W, CE and the GE are now the ones 23 that are handling on a day-to-bay basis the in-service 24 testing program.
25 MR. MICHELSON: So the flow of information is.
(~)
v ACE-FEDERAL REPORTERS, INC.
202-347-3700 Nationwide Coverage 800-336-6646
1 l
l 1
26378.0 139 l REE l s l k_) 1 back to them and not to you?
2 MR. BOSNAK: We are now the safety review and 3 oversight. In other words, we handle the generic problems.
4 If there are day-to-day problems, they should flow back 5 through the vendor divisions and back to us. We are in the 6 process of setting up our oversight review. We intend to 7 have several review groups and one of these would get into 8 this because we know valves are a problem.
9 MR. MICHELSON: The reason I asked is we got a 10 letter at my request f rom one of the inspectors in Region 3.
11 I think they can get copies of it. The le tte r indica te s 12 the results of an inspection of, I think at Byron, as I
(~) 13 recall in the form of part of the inspection report. I
\-/
14 just wondered, do those interesting parts of inspe c tion 15 reports somehow get back to you indicating the kinds of 16 problems in the field, or apparently they get back to the 17 vendor divisions and then they are supposed to bring it to 18 your attention? Is that the process?
19 MR. BOSNAK: Prior to the reorganization, we 20 used to get these kinds of things. Right now we are not 21 getting these things. Whe ther they are actually getting 22 inspection reports, basically many of those go directly to 23 IE headquarters and then IE calls NRR if they feel there is 24 a problem.
25 MR. MICHELSON: But if I were in NRR and I were O
ACE-FEDERAL REPORTERS, INC.
202-347-3700 Nationwide Coverage 800-336-6646
26378.0 140 REE
/~N J
(._) 1 responsible for valve situations, I just wondered how I 2 would get my grist for the mill.
3 MR. BOSNAK: Part of the contract that we have 4 started on 2E61, we have a contractor and we intend to have 5 staff oversight of that. It is going to be going out to 6 visit the plants and talking to people right at the sites, 7 including our inspe c tors in the regions. But particularly 8 the utilities.
9 MR. MICHELSON: Who is the contractor?
10 MR. BOSNAK: The contractor in that is 11 Brookhaven.
12 MR. MICHELSON: And they will monitor --
(} 13 MR. BOSNAK: They will go out and see just what 14 -- how do people in-service test valves? What are the y 15 doing now? Are they actually doing what they are saying 16 they are doing in the programs? See if they can find any 17 inadequacies. This is part of what is going to start.
18 MR. MICHELSON: If you had told me that you had 19 a contractoc that was going to go out and watch for valve 20 troubles -- but if he is going out to worry about 21 in-service inspections, he has already got blinders on.
22 MR. BOSNAK: He is not supposed to have blinders 23 on. He is supposed to have an open mind.
24 MR. MICHELSON: He certainly has a focused view 25 of the problem.
O ACE-FEDERAL REPORTERS, INC.
202-347-3700 Nationwide Coverage 800-336-6646
26378.0 141 REE
(_) 1 MR. BOSNAK: He knows this is how valves are 2 monitored today and how we can do a better job.
3 MR. MICHELSON: He has just started on this.
4 MR. BOSNAK: Just getting started.
5 MR. ROTHBERG: We are going to have a kickoff 6 meeting.
7 MR. MICHELSON: So it will be some time before 8 you will be in a position to --
9 MR. BOSNAK: That is correct. And we would like 10 to have feedback from your committee on the progress. We 11 e xpe c t to get back to you to tell you how things are going.
i 12 MR. MICHELSON: I think Glenn has a question.
(} 13 MR. REED: I think what you are talking about is 14 very interesting. I.just read in the last 30 minutes this 15 report from the reactor inspector, a reactor inspector in 16 Region 3. I read the first page and in - f act I should have 17 read that and I could have addressed that or just read it 18 off and it is all the opinions that I have. It is 19 bea u t i f ul .
20 (Laughter.)
21 Here is a man-out in the field, close in the 22 field, not as close as perhaps the work is, but he is out 23 in the field and he comes out with almost the same opinions 24 I do.
25 MR. MICHELSON: I read some of this from a O
ACE-FEDERAL REPORTERS, INC.
202-347-3700 Nationwide Coverage 800-336-6646
26378.0 142 REE
(_) 1 couple of others in the field. They are not unalike. It 2 disturbs me a little bit. I would hope that NRR is very 3 close to it, even though it comes through IE.
4 MR. HERNAN: Ron Hernan with the Sta f f. The 5 length -- well, one of the philosophies of the new 6 organization is to make the assigned project manager for 7 each project more involved in the day-to-day stuff, more 8 involved with resident in s pe c tor s , with the region. I can 9 tell you for a fact that the project managers do get each 10 and every inspection report that is made on their plant.
11 That may or may not trip a trigger in terms of is this a 12 problem with plants other than mine, but maybe within this
(} 13 vendor or maybe transcends vendor lines also. That is the 14 first step.
15 The other part, I think very important part, is
, 16 the fact that once a week we have a weekly meeting with all 17 of NCC management on operating reactor events over the past 18 week. The screening for those events is not only-19 personnel-problem oriented but equipment-problem oriented.
20 And Mr. Den ton , in the course of those meetings, 21 does pursue -- maybe it is a problem that is brought to 22 this meeting for that plant -- but he does pursue: Does 23 this have generic implications? We have a group that 24 worries about just that - type of thing.
25 The next step in that process, once every two
.3 ACE-FEDERAL REPORTERS, INC.
202-347-3700 Nationwide Coverage 800-336-6646
26378.0 143 REE f~s
(_/ 1 months we select problems which we will bring to the ACRS 2 Subcommittee on Reactor Operations. We are sensitive to 3 valve problems and equipment problems and in our screening 4 to what we bring to you, we try to determine if there is a 5 generic application. So there is --
6 MR. MICHELSON: It is always triggered by an 7 event. I am looking for things that inspectors are seeing i
8 that would give an earlier indication of difficulty other 9 than the event itself. I just wondered how they are 10 screened by other than inspectors? Some times inspec tors 11 are great reporters and somebody else has to kind of piece 12 together each piece to begin- to see a pattern or problem.
(} 13 Sometimes I wonder whether that is being done. I assume it 14 is done on the regional level.
15 MR. HERNAN: It is not a surefire process.
16 There are some problems that --
17 MR. MICHELSON: The first problem that I hava is, 18 do we have an adequately informed and trained inspectors on 19 the valve-related problems so they can look for them 20 intelligently? Have we been tuned to the problem? -
21 MR. HERNAN: I they we probably rely. pretty 22 heavily on the IE bulle tins.
23 MR. MICHELSON: I haven' t seen .the kind that I 24 think -- they are, again, generally event oriented. After 25 a few events, someone starts tel'_ing inspectors to look.
O ACE-FEDERAL REPORTERS, INC.
202-347-3700 Nationwide Coverage 80 4 336 4 646
26378.0 144 REE
/~'T
(_/ ~1 MR. HERNAN: But events can range from a fairly 2 minor event at one plant that maybe perturbs the operation 3 schedule for that plant and keeps them shut down. That 4 gets people's attention.
5 MR. MICHELSON: I would sincerely hope that 6 inspectors are given a one or two-day indoctrination so 7 that they can now inspect these from time to time and give 8 you much be tter information.
9 MR. HERNAN: We ll , like Mr. Bosnak, I will not 10 speak for IE. I agree that this is an important step.
11 MR. MICHELSON: We will have to ask IE how they 12 do this.
13 MR. HERNAN: I want to add,'though, that we are
{~'}
14 relying more heavily on our project managers becoming 15 involved and spending more time at the plan t site . And he 16 is really the link between what is going on in the field 17 and what ultimately gets identified as a generic issue 18 within the NRR organization.
19 Other questions or comments? If not, then I 20 guess we better get on to the check valves, which are also 21 a subject of some recent interest.
22 MR. EISSENBERG: I am going to discuss the 23 project on check valve aging assessment. I want to make 24 two points. W'e are following the same overall strategy in 25 the aging program as we did with motor-operated valves. I O
ACE-FEDERAL REPORTERS, INC.
202-347-3700 Nationwide Coverage 800 336-6646
~
26378.0 145 REE
/~s
(_) 1 want to note again that there have been , since we have 2 started the project, there have been the se indications of 3 problems associated with check valves at San onofre and at 4 Turkey Point. They were different problems, but they were 5 both pertaining to check valves.
6 MR. MICHELSON: You might point out two or three 7 years ago, I guess it was, at Dresden they lost cooling 8 water to the diesel engines because of multiple' check valve 9 failures. I expect if there had been an IIT team thing at 10 that time, it would have prompted one. So there have been 11 a number of past events just as worrisome as these two. We 12 are just starting to tune into them a little be t te r ,
e 13 perhaps.
}
14 MR. EISSENBERG: I have a listing of -- from the 15 report that we have quoted, which lists a lot of events on 16 a lot of systems.
i 17 MR. MICHELSON: Check valves have been falling 18 apart for some years, maybe not getting quite the attention 19 that is warran ted.
20 MR. EISSENBERG: This is just a repetition. A 21 definition of operational readiness and diagnostics which 22 we would be applying to check valves.
23 (Slide.)
24 Also the scope of the program. We are at this 25 stage in the program. We have comple ted the. report which O v ACE-FEDERAL REPORTERS, INC.
202-347-3700 Nationwide Coverage 800-336-6646
26378.0 146 REE s) 1 characterizes the aging-related failure modes and defects s
2 which cause them and we are -- we have issued the report.
3 We are now in the process of identifying what we need to do 4 in phase 2 in this next area.
5 MR. MICHELSON: Are you anticipating any kind of 6 flow testing of check valves?
7 MR. EISSENBERG: Yes.
8 MR. MICHELSON: There is this story about at 9 cer ta in flows, the check valve is fine. Are you going to 10 do some actual flow work?
11 MR. EISSENBERG: Yes, we will. You have to in 12 the case of a check valve. In fact, the Turkey Point event
~T 13 was a small leak through the check valve which didn' t get (V
14 the check open , and it kind of just vibrated itself.
15 Just to put perspective on check valves, this is 16 typical numbers of check valves in BWRs and PWRs and sa fe ty 17 systems and balance of plant. They range -- a very wide 18 variation in the size of check valves. And there are 19 reasonably large number of check valves, sa fe ty systems.
20 (Slide.)
21 The frequently used ~ type of check valves, swing 22 check valves are the most popular type . There are many 23 piston-lif t check valves and there are a number of 24 so-called stop check valves. Now, a stop check valve is 25 kind of a hybrid between a piston-lif t check valve. It is O
v ACE-FEDERAL REPORTERS, INC.
202-347-3700 Nationwide Coverage 80433M446
26378.0 147 REE
/'s
() 1 a piston-1 .aeck valve where you can also mechanically 2 stop this piston from rising. So it is a stop check valve.
3 It has a slip stem connection so that one can close , but 4 one can' t open the check valve. It was that valve that 5 failed at Turkey Point, a stop check valve.
6 The valves that failed at San Onofre were swing 7 check valves. By and large they don't have anything 8 sticking out that one can hang a monitoring device on.
9 They don't have much hardware inside, so that compared to 10 the motor-opera ted valve , the diagnostic monitoring 11 presents a d4fferent kind of problem. There is just not 12 much to monitor.
{} 13 MR. MICHELSON: Are you including testable check 14 valves in the category of swing checks?
15 MR. EISSENBERG: Ye s . But those are te s table ,
16 correct.
17 MR. MICHELSON: They have operators and 18 indicators and so forth on them?
19 MR. EISSENBERG: Well, they can be actuated.
20 They in general don't have indicators?
21 MR. MICHELSON: They can be lifted?
22 MR. EISSENBERG: Yes. They can be actuated.
23 MR. MICHELSON: They can only be exercised.
24 MR. EISSENBERG: I think I will be referring to 25 that.
()
ACE-FEDERAL' REPORTERS, INC.
202-347-3700 Nationwide Coverage 800-336-6646
26378.0 148 REE
) 1 (Slide.)
2 MR. MICHELSON: It is important to understand 3 the dif ferences between behaviors of, say, stainless steel 4 check valves and carbon steel check valves. Yet your 5 report seems to focus on stainless steel when some of the 6 real problems are with carbon steel valves.
7 MR. EISSENBERG: We did not address that issue.
8 Problems in terms of what?
9 MR. MICHELSON: Well, in terms of corrosion 10 behavior, a number of ef fects in boiling water reactors i
11 which use a large number of carbon steel check valves, as 12 do water systems also, safety-related water . cooling systems
(} 13 14 use a lot of check valves. And the corrosion, the whole bit is dif ferent on the carbon steel variety than on the 15 stainless steel.
16 MR. EISSENBERG: We better go back and check 17 into the issue.
18 MR. MICHELSON: I notice in your report you just 19 limited yourself. I think the real problems might very 20 well be with --
21 MR. REED: To underscore what Carlyle is saying, 22 you have corrosion, erosion and cavitation problems which 23 are much worse on carbon steel valves and even some brass 24 valves.
25 MR. MICHELSON: Yes.
s ACE-FEDERAL REPORTERS, INC.
202-347 3700 Nationwide Coverage 80 4 336-6646
26378.0 149 REE g
k- 1 MR. EISSENBERG: I think that in our review of 2 LER and the da ta base , we didn' t make the distinction as to
, 3 the materials. They contain the failures of carbon steel 4 check valves so that the failures and the failure modes and 5 the causes which come from check valves are incorporated.
6 We should separate them out, though.
7 Just briefly, we reviewed, as we did in the MOVs, 8 the tech spec requirements and the tech spec requirements 9 in turn refer to the section XI, subsection IWV, which 10 indica te s that you exercise the valve to verify travel, 11 tha t the disk moves.
12 MR. MICHELSON: You made a statement a moment 4
'T 13 ago that really does need clarification. You said you (O
14 should have dif ferentiated between stainless and carbon.
15 Your report says: Only stainless steel valves are 16 considered in this report. So I am~1ed -- I am confused.
17 MR. EISSENBERG: I am certain there the LER and 18 NPRDS data bases included carbon steel.
19 MR. MICHELSON: But the report, does.it include 20 only stainless valves?
21 MR. EISSENBERG: We looked at only stainless 22 valves when we discussed the issue of ' stressors and impact
- 23. stressors.-
24 MR. MICHELSON: This report contains a lot of 25 information about valve behavior. Several times I wondered n
U ACE-FEDERAL REPORTERS, INC.
202-347-3700 Nationwide Coverage 800-336-6646 4
._.._______________.____.__.______________.__.___________-_.___.___.___.___._____2_
26378.0 150 REE
.fh
\._) 1 did you include carbon steel in this or not. And the 2 statement on page 9 clearly says you considered only 3 stainless steel. I am wondering if you really believe that.
4 MR. EISSENBERG: At this point, I would say that 5 it only -- I know that it only considered stainless steel 1
6 when we did an analysis of the impact of the enviro :meint on 7 the valves in that section. I don't believe that stitemen t 8 is true when we look at operating experience. And when we 9 discuss monitoring methods and failure causes, I know that 10 it includes both, 11 MR. MICHELSON: It is a very important one to 12 consider internal to the valves because stainless steel
{} 13 14 valves have considerably different internals than the carbon steel valves.
15 MR. EISSENBERG: Well, that is a point which is 16 taken.
4 17 MR. MICHELSON: Okay. Go ahead.
18 MR. EISSENBERG: With regard to the section XI, 19 one can use such items as visual observation, position 20 indicator for those valves that do have position indicators, 21 or indications of pressure and flow to determine that the 22 check valve has actuated, becau se there are no other ways, 23 other than this type of measurement.
24 There is also leak testing of check valves. One 25- uses air to determine leakage rate. And that is a subject
(~
i ACE-FEDERAL REPORTERS, INC.
I 202-347-3700 Nationwide Coverage 80 4 336-6646
26378.0 151 REE
(_) 1 of some work being done at ETEC, determining how 'one goes 2 about measuring leakage through a check valve.
3 MR. REED: I am sure you are going to get to 4 noise analysis?
5 MR. EISSENBERG: Yes. Ag a in., the purpose of 1
6 this report is not to lead anywhere but to provide the 7 basis. So this is from LERs from 1969 to 1982, 1983. This 8 is some analysis of what the LERs told us. Major mode of 9 failure was leakage and internal problems with the valve 10 in te rva ls . In fact, the problem at San Onofre and the 11 problem at Turkey Point.
12 Mode of detection, by and large , you find out
/~T 13 during tests; also during normal operation. You don't find V
14 it out from demand because there are very few emergency 15 conditions in accidents. So it is the normal group.
16 Not really much information here . By and large, 17 maintenance activity is repaired and the causes generally 18 distributed, bt it looks like wear, crud,. corrosion, 19 erosion, failure to seat -- that doesn't tell us much --
20 installation problems. A lot of the problems are just 21 unknown, that the valve is repaired and the LER doesn' t 22 contain the information as to what was wrong.
23 (Slide.)
24 A look at the NPRDS system and we find.a similar 25 set of information. Seat leakage and external leakage are O
ACE-FEDERAL REPORTERS, INC.
202-347-3700 Nationwide Coverage 800-336-6646
26378.0 152 REE
( .
\ l large causes of failure for the valves listed in NPRDS.
2 In-se rv ice inspection and surveillance covers most of them,
- 3 and repair is listed as the major activity. And the 4 failure cause is listed as normal and/or abnormal wear 5 sticks out as a major i te m.
6 It says that aging and cycle fa tigue is low, but 7 if you lump aging and service wear as the aging issues or 8 the time-dependent degradation, then that is a major i tem .
9 MR. REED: That normal / abnormal wear in 50 10 percent of the failure cause , that bothers me because - who 11 is to say what normal / abnormal is? Is it design or 12 application; right? That is 50 percent?
13 MR. EISSENBERG: Yes. The information that you
(}
14 get from NPRDS or from LERs is limited. You need to 15 in terpre t it. They are only what the utilities report. If 16 their reporting procedure doesn' t dif ferentiate, then NPRDS 17 doesn' t dif ferentiate whether it is a design problem that 18 led to accelerated wear or was it a normal wear.
19 MR. REED: Or an application that shouldn't have
. 20 been in the first place?
4 21 MR. EISSENBERG: Ye s .
22 MR. REED: I could wear the heck out of a valve 23 simply by having a carbon steel valve in a borated water 24 stream. I could tear it up in seven days.
25 MR. EISSENBERG: That again points to me to the O
ACE-FEDERAL REPORTERS, INC.
202-347-3700 Nationwide Coverage 800 336-6646
26378.0 153 REE
,)
_, 1 importance of a monitoring method which can trend whether 2 it is wearing. Be cause it may wear for reasons that you 3 can't sus pe c t . But if you know the rate at which it is 4 changing, then you know something and you can do something 5 about it. The ignorance factor that you are implying can 6 be handled by monitoring program, if we can develop one.
7 (Slide.)
8 This gets back to the comment made earlier that 9 in addition to the most recent two events, in the report 10 'that was issued we listed these IE publications, bulle tin s ,
11 circulars and notices, historically, which covers all the 12 systems. Diesel generator system is included. I didn't
{} 13 14 list the utilities, but the one you mentioned is in here.
The automatic depressurization system, LPSI, HPSI, they 15 have all suf fered reportable events and the se have been 16 documen ted in IE publications.
17 In addition, these reports have been issued 18 which cover check valves. They are not -- they were not i
19 written about check valves necessarily, but they do cover 20 operating experience with check valves. And we reviewed 21 all of these reports in preparing our understanding of how 22 they fa iled.
23 (Slide.)
24 Some of the results then, the failure modes and 25 causes, the failure modes are pretty straightforward. It O
ACE-FEDERAL REPORTERS, INC.
202-347-3700 Nationwide Coverage 800-336-6646
J 26378.0 154 REE 1 can only fail in one of these ways. So there they are.
2 MR. MICHELSON: Have you found cases where the y 1
i 3 literally weld shut? Failure to ope n mean s , to me , I apply .
4 my maximum differential and nothing happens.
5 MR. EISSENBERG: It doesn' t have to weld shut.
[ 6 If the hinge pin breaks off and plugs -- well, plugging is 7 an external foreign material.
- l. 8 MR. MICHELSON: I thought that was the valve
- f 9 flapper going down the pipe .
10 MR. EISSENBERG: No, this is external.
11 MR. REED: Very key question. In any of the i 12 data or any of your reviews, was there an occasion where
(} 13 somebody had said that it had welded shut? That is,_high 3
14 back pressure, an interfacing check valve between a 15 low-pressure system and a primary system, and it sits there 16 for a year or two or five years, was there ever a case - in 17 any of the data that there was any metallurgical bonding?
- 18 I would like . to hope not.
1 19 MR. EISSENBERG: I don't know. Le t me make a 20 qualitative judgment that for the swing ' check valves, it 21 _w ould be hard for me to see that you can weld it -shut so 22 that a tension, a tensile force ' won't pull it apart. For a a
j 23 piston-lift valve, you may have that problem. But the I 24 direct answer to your question is, I don't know. Perhaps-25 Earl Brown or somebody from IE -- and incidentally, in the -
n i (_)
! ACE-FEDERAL REPORTERS,' INC.
202-347-3700 Nationwide Coverage - 1 800 336-6646
26378.0 155 REE
~
! k_)N 1 appendix to the report which is NUREG/CR-4302, there is a 2 summary of each of the notices and bulletins, and so you 3 could go in and read which utility and what happened. But 4 to my knowledge, I don't recall anything and I would say it 5 is improper for a swing check.
6 (Slide.)
7 Those were the failure modes. This Vugraph then 8 discusses failure causes. This is a shopping list of the 9 various ways that valves can fail. These are, again, the 10 failure modes and these are the causes that potentially or 11 actually have led to these failures. And the guide wear, i
12 the obturator guide wear is the Turkey Point problem, and
/~T 13 the hanger pin wear and the obturator fastener loosening V
14 are the issues associated with San Onofre plant. There are 15 other ways that check valves can fail and have failed.
16 (Slide.)
17 We get down to the hard question on what are we 18 going to monitor, beca use these are simple devices. And 19 this is what we have come up with. Visual examination, 20 inspection during maintenance, take it apart, inspection 21 and disassembly and inspe c tion . You might -- you- look 22 inside. Look at the fastener tightness; you determine 23 obturator movement indirectly as per section XI. Or l
24 alternately, as Mr. Ward pointed out, I guess, using 25 acoustic noise.
ACE-FEDERAL REPORTERS, INC.
202-347-3700 Nationwide Coverage 800-336-6646
26378.0 155 REE 1 appendix to the report which is NUREG/CR-4302, there is a 3
2 summary of each of the notices and bulletins, and so you 3 could go in and read which utility and what happened. But 4 to my knowledge, I don't recall anything and I would say it 5 is improper for a swing check.
6 (Slide.)
7 Those were the failure modes. This Vugraph then 8 discusses failure causes. This is a shopping list of the 9 various ways that valves can fail. These are, again, the 10 failure modes and the se a re the causes that potentially or 11 actually have led to these f a ilure s . And the guide wear, 12 the obturator guide wear is the Turkey Point problem, and
{} 13 the hanger pin wear and the obturator fastener loosening 14 are the issues associated with San Onofre plant. There are 15 other ways that check valves can fail and have fa ile d .
16 (Slide.)
17 We get down to .the hard question on what are we 18 going to monitor, because these are simple devices. And 19 this is what we have come up with. Visual examination, 20 inspection during maintenance, take it apart, inspection 21 and disassembly and in spe c tion . You might -- you look 22 inside. Look at the fastener tightness; you determine 23 obturator movement indirectly as per section XI. Or 24 alternately, as Mr. Ward pointed out, I guess, using 25 acoustic noise.
O V
ACE-FEDERAL REPORTERS, INC.
202-347-3700 Nationwide Coverage 800-336-6646
26378.0 156 REE
/~T :
l
(-) 1 MR. MICHELSON: Under section XI, can I really 2 tell how far the valve opens or can I just merely verify 3 the fact that it cracked open?
4 MR. EISSENBERG: You can verify that the flow 5 has reached the value which is acceptable.
6 MR. MICHELSON: Does section XI require full 7 flow as the check valve test? ,
8 MR. ROTHBERG: Yes. Anything less than full ,
9 flow is considered prartial.
r
- 10 MR. MICHSLSON
- How often do you have to do that?
! 11 MR. ROTil BERG: Supposedly every 90 days. But 12 there is a --
3
{} 13 14 MR. MICHELSON: How do you do full flow on check ,
valves used as containment isolation valves, naybe backed i '
l 15 by a high concen ?. ration boron tank where you never want '
16 full flow? You never want any flow at all. ,
17 MR. ROTHBERG: You mean during cold shutdown or ,
18 at refueling outage s? .
19 MR. MICHELSON: At full flow.
20 MR. ROTHBERG: No. A good example is the 21 accumula tor checks. You can never really get full flow <
22 without a check. So what we have been doing in the past t 23 essentially is asking people to disassemble the valve on a :
24 rotating basis.
- 25 MR. MICHELSON
- How of ten? g O-s l
't i
h ACE-FEDERAL REPORTERS, INC.
202 347-3700 Natronwide Coverage 900 336 4646 *
. ._. _ . _ _m ._ _ . _ _ . . _ .. _, - .. _ .._ _ _ . _ . . _ , , . .._
e 26378.0 157 REE i <^
- k. 1 MR. ROTHBERG: Every refuel ng. They take on-a I a
f 2 valve apart.
3 M'R . MICHSLSON: llow long does ^it take to get N I
. 4 around a given valve?
5 MR. ROTHBERG: Eight to 16 years. '
MR. MICHELSONs That doesn' t leave you very warm '
f 6
'7 that it is going to work in 16 years, does it?
8 MR. ROTHBERG: If you take one valve apart --
9 MR. MICHELSOt{: But are they all identical l
i 10 applications? .
11 MR. ROTHBERG: I believe they are.
12 MR. MICHELSON: You rotate through just the
]! f
(}
13 accumulator checks.
i 14 l MR. ROTHBERG: Precisely. A refueling ou'tage is !
- 15 every 18 months. f ,
, 16 i MR. MICHELSON: That applicat' ion wouldn' t --
17 MR. ROTHBERG: You have a valve that is ;
18 essentially closed over time. A feed system valve , you 19 have a valve that is open all the time and is sitting in '
20 the strean all the time. When you call it on it to close, i
4 21 it nay not.
- 22 MR. MICHELSON: That is why I asked the question.
23 ! Is there any instance where a valve isn't fully shut? ,
i
, 24 Because accumulators are in that ca tegory.
25 MR. KIESSEL: Are you talking about being welded?
O ACE FEDERAL REPORTERS, INC.
202-347-3700 Nationwide Coverage 600-336-6646
l '
26378.0 ^158 ,
KEE I
1 MR. MIC!!ELSON: Corrosion or whatever causes it.
O 2 r Either in the hinge pin area or in the --
3 3 MR. KIESSCL: I believe Palisades had an event ,
4 where their accumulator check va].ves failed to open under 5 full accumulator pressure. I think the problem was that >
s 6 they had been over-tightened.
7 MR. EISSENBERG: Piston lift with a --
8 MR. KIESSEL: They had been manually tightened 9 tod much. Then they backed of f. And later when they went i
) JO to try and lif t them, they didn' t. This was written up and 11 vould have been or.e cf the information notices documented 12 in that Oak Ridge report. ,
t ,
I 13 MR. MICHELSON ,: That type of,valvaww-ill he
} .
14 ( tested depending on the --
15 MR. ROTHLERG: It could be . But, again, this [
i 16 vas ir. the IST bulle tin. They found the problem with a 17 particula.r valve and yoy had to disassemble and look at all
'18 of them.
I 19 MR. REED: I would like to -- in your methods of 20 monitoring, periodic tecting is us6d all the time, waokly 21 testing, monthly testing, on systenM and a lot .of 6 heck. -
22 valvos in these syst, ems, Certainly that is a way of i 23 monitoring --
}
24 MR. EISS$NBERGi Tnat would be cavored under .
25 section XI, periodic testing, ir.-service lospection, i
a ACE FEDERAL NyPORT]ikS, JNC.
202-347-3700 Nationwide Co-ery BOO-Lo 6646
_ _. .,_ -_- . . _ _ , , , . _ ..._ _ ~ - . . . -.
I 26378.0 159 REE 1 MR. REED: I didn't understand what that crazy 2 word means. You have got to uso simple words for us.
3 The other thing I was noticing is you don't put 4 up X-ray.
5 MR. EISSENBERG That is -- it appears later on t
, 6 as another. It is a potential one , but we haven' t te sted s' 7 it.
8 These are some of the methods that look f
9 I straightforward. As I noted, the program that. is funded ;
10 from, I guess through Jerry Weidenhamer at ETEC, is looking 11 at acoustic noise due to leakage. We have suggested to 1
12 them that they look at the acoustic noise during actuation 4 i
~
13 MR. REED: The best way to follev check- valves, 14 those that have no stem out through them to a lift device, l 15 is noise. Just stethoscope all the time. Opera ting ,
a 16 shutting down, starting up, closing under back flow. I 17 MR. EISSENBERG: I am glad you said that,.
18 because there isn' t many other things, but that does look 4
19 like a very good way to understand whether you are 20 l stif fening up. It will tell you that something inside is j 21 stif fening up. It is not thumping as hard as it should be .
22 MR. REED: Or it is rattle, rattle, rattle.
23 MR. EISSENBERG: Yes. It is not the best 24 trendable diagnostic, but thero aren't very many other l
[ 25 choices 6 1
O ACE-FEDERAL REPORTERS, INC.
' 202 347 3700 Neionwide Cmcrge 800 336 4646
26378.0 160 REE
, /'5
< Al 1 (Slide.)
2 l The next Vugraph is really related to this. It 3 is kind of looking at it from the device rather than from 4 I the monitoring method. This gets back to wha t Owen sa id ,
5 you can disassemble and inspect the -- get the dimension, 6 look around for erosion, corrosion wear, look at roughness, 7 cracks, deposition of foreign material. That is the hard 8 way, but that is a good way. If this is done periodically, 9 you can observe wear by dimension checks and you can trend 10 the wear with time. it is, again, a hard way to go.
i 11 Acoustig noise intensity for leakage , you can, 12 if the valve has an external actuator device. that can
/'N 13 actuate check. valve, when there is no flow through it, you i
(_/
14 can measure the forco necessary to flap the flapper when it 15 has this external device. ,
l 16 That is another thing you can monitor. If it 17 reaches some limit, you can do maintenance actions..
18 of course, you look at the .facteners ~ that hold i
19 the flapper in place and the other devices. Under other, I i
20 would use X-ray, I would use -- I would think of eddy 21 current. If you wrapped sone coils around the check-valve, 22 you might be able to see in the eddy current change that l
23 occurs during actuation. If you could identify that with
- 24 the motion of the flapper, ,you could get a correlation.
25 Those are more far out. X-ray is an expansive way to go. .
ACE FEDERAL REPORTERS, INC. ;
202-347-3700 Nationwide Coverage M3E%% >
_ _ . ~. _ _ _ . _ _ ., , _ _ _ . _
i i
26378.0 161 REE 1 MR. REED: That is only a follow-up. Once you 2 have determined where your problems might be, then you want 3 to make certain that they are and you can get good follow-up 1
4 monitoring from an X-ray. ,
5 MR. EISSENBERG: Yes.
't 6 MR. MICHELSON: Again, though, only a sample is 7 done at each section? You don' t disassemble every check 8 valve in the plant.
1 9 MR. EISSENBERG: You can't do that.
10 MR. MICHELSON: So on the average, how many 11 years are we talking about that we have to be comfortable 12 with this operability before we have a visual verification?
13 MR. ROTHBERG: Includingy--
}
14 MR. MICHELSON: It could be eight, 10, 12 years.
15 MR. ROTHBERG: The San Onofre check valves that j 16 f a iled , the last time they were looked at be fore the y
- 17 failed was 1980. In 1984 they tested them.
18 MR. MICHELSON: When would they have been looked l 19 at again if things had not happened? l 20 MR. ROTHBERG: Maybe never. ,
4 21 MR. MICHELSON: Never is a very long time. ;
22 MR. ROTHBERG Essentially noises were ignored P
23 in the system because there was no standard to judge where 24 the noise came from. Somebody has to make a rational ,
l 25 decision about when the noise is too loud. l i
r ACE FEDERAL REPORTERS, INC.
202 347-3700 NationwMe Cover. ige 80(L33MM6
_ . , ___ _ _ ~ . . . _ . . _ , , , . . . , _ . . _ -,_ .. . --. - . - _ . _ _ _ -
i I i 26378.0 162 l REE f
1 MR. MICHELSON: So this, really the monitoring i
2 --
there are monitoring methods that could be used, there
! 3 are some that are required to be used , but they don' t leave j 4 you with a lot of comfort?
5 MR. REED: On'that San Onofre thing, I thought i 6 they had made serious repairs on those valves two years in 7 a row and then they just forgot about it.
l 8 MR. ROTHBERG: What they had was a problem with 9 the -- that the obturator was rotating. They put these 10 lugs on them and when they put the lugs on, they considered f
i 11 .that the problem was fixed. So that was when they stopped 1
, 12 looking at them.
{} 13 MR. REED: I might point out another thing about 14 check valves. There are many check valves out there, 15 particularly stainless steel, that if you try to i
16 disassemble them, you have just finished the valve and you l
i 17 are going to cut it out and put in a new one anyway. So 18 you have to watch monitoring that calls for internal 4
19 measurement and taking them apart. Some you just can't i 20 take apart because of the way they are built. They 1
21 self-destruct.
l 22 MR. EISSENBERG: This is a catalog. Phase 2 of I 23 this program will get into the doing and I don' t -- I don't
- 24 think that that is a desirable way. It may be one of the
[ 25 only ways. It is at this point one of the only ways.
j 1
1 ACE-FEDERAL REPORTERS, INC.
202-347-3700 Nationwide Coverage 800-33H646
26378.0 163 REE
-('%
(_) 1 MR. REED: If you get the noise analysis here --
2 MR. EISSENBERG: That is what we intend to do.
3 MR. MICHELSON: The thing that bothers me though,
'. 4 we have a non -- what you normally think of as nonsa fe ty -
5 systems that nave some little safety features in them. For 6 instance, nonsa fe ty air supplies that -- that little sa fe ty 7 feature of those two check valves --
8 MR. ROTHBERG: If it is not a sa fe ty-rela ted 9 valv9.
10 MR. MICHELSON: This is. That is why it has the 11 check sign. It has an air accumulator right on the valve 12 because it has to have central air for one cycle of
{} 13 14 operation. So it is connected to a - nonsa fe ty system, but you say, I got two check valves here.
15 MR. PAGE: Those are safety-related valves.
i 16 MR. MICHELSON: But do they get the section XI 17 te s ting , all the good stuf f ?
18 MR. PAGE: This is kind of a long, detailed 4 19 history. They used Reg Guide 126 to, I guess, establish 20 the code class of each one of their check valven. Reg 21 Guide 126 gets a little hazy in the area of nonradioactive 22 nonwater-containing systems. Since the NRC did not review 23 the older plants when they did that reclassification, some 24 plants have chosen not to put those in.the program. We do 25 try to convince them to rat them in the program by whatever
(~'\
v ACE-FEDERAL REPORTERS, INC.
202-347-3700 Nationwide Coverage 800-336-6646
26378.0 164 REE s
1 means we have availablo-.
2 MR. MICHELSON: Unfortunately, they have just a 3 little water in them. r 4 MR. PAGE: It is not a water designed system .
5 MR. MICHELSON: But there are systems that have
~
6 ple n ty o f -- \
7 MR.' PAGE: They are indeed safety related.
8 MR. MICHELSON: Yes.
9 MR. PAGE: Some of them are tested because they
' 10 contain isolation valves also.
4 11 MR. MICHELSON: I am thinking of the ones on the 12 wall outside of containment. Those . are j ust as important.
{} 13 MR. PAGE: I agree.
14 MR. MICHELSON: How do we te st those? .Many of j 15 those can' t be disassembled either. They are the kind that 16 ,
you just tear them out and put a new one in.
17 MR. PAGE: Disassembly is not an option for full 18 stroke in the closed position. We really discourage people 19 from trying to use disassembly -- do you require -- you
! 20 don't require -- , , , -
1 21 MR. PAGE:, Where they are included in the 22 program, the sa fe ty position -- the exact' details of how 23 the te s t is performed we did not get into.
24 MR. MICHELSON: Every two years.
25 MR. PAGE: At least every two years. It A
v ACE-FEDERAL REPORTERS, INC.
202 347-3700 Nationwide Coverage 800-336-6646
26378.0 165 REE c^ :
g
(_) 1 basically boils down to each refueling.
2 MR. EISSENBERG: This, again, is a shopping list.
3 I would like to tell you what we are up to. We know of the 4 program at ETEC which is looking at the use of acoustic 5 ' leak detection systems. In fact, in conjunction with ETEC, 6 we supplied ETEC with acoustic monitoring device, one of 7 which we got from Philadelphia Electric and another from 8 the Navy, and they are being evaluated. This is primarily j 9 for leak detection, but also for actuation monitoring i
10 insofar as they can do it.
11 We hope to get some worn check valves and my 12 program monitor tells me maybe we could get the valves from 13 San Onofre to inspect and look at and try to understand the
(~}
14 aging aspects of it. W'e would like to get worn check 15 valves to see how they have worn.
16 We also intend to set up a flow - test stand to j 17 validate the diagnostic options that we will be looking.at. .
18 And at this point the options are acoustic leak detection, 19 acoustic actuation monitor, and other actuation monitors.
20 This would be perhaps eddy current and perhaps X-ray and 21 perhaps sonicator. We are going to exotic things because 22 we don't have much to work with. We need that breakthrough.
23 We need to get the motor current equivalent for a check 24 valva. We don't have it.
25 But we are going to be looking at them as much
(-)s s_ .
ACE-FEDERAL REPORTERS, INC.
202 347-3700 Nationwide Coverage 800-336-6646
26378.0 166 REE
- rm I
w) 1 as we can. The only other option is disassembly and 2 inspection.
3 MR. MICHELSON: Are you going to look at water 4 system check valves?
- 5 MR. EISSENBERG
- Yes, we will use water and we 6 will use two-phase water / air.
7 MR. MICHELSON: I am talking about raw water.
8 MR. EISSENBERG: Tap water?
9 MR. MICHELSON: Water out of the river. Water 10 systems that cool diesel engines and so forth use lots of 11 important check valves. They also are subject to 12 infestation by clams and silt and you name it, and be fore
{} 13 long those check valves -- you look at it, you will find 14 they don't work too well.
15 MR. REED: You get a mushy noise signal.
16 MR. MICHELSON: If any. Most likely the clams 17 have already got it frozen wide open. But.do you think 18 about these a little bit? Some of those check valves are 19 very importan4 ' i ke where water pumps are running in 20 parallel.
21 MR. EISSENBERG: We had not thought about raw 22 water from main cooling water lines going into the plant.
23 That is a good point. We should be worried about dirty 24 wa te r .
25 MR. MICHELSON: Because a lot of -- in some (v
ACE-FEDERAL REPORTERS, INC.
202-347-3700 Nationwide Coverage - 800-33H646
26378.0 167 REE n
() 1 plants, most of the cooling functions are raw water. Other 2 plants use closed-cycle cooling water and then the water is 3 -- the situation is not so bad.
4 MR. EISSENBERG: W'e have talked about -- again, 5 we haven't done this ye t. We have talked about using 6 air / water to simulate two-pha se flow. How does a check 7 valve which is supposed to open under single-phase flow 8 respond when you have two-pha se flow? We think we can get 9 answers with air / water rather than going to a steam water 10 system.
11 The other area that we would like to ge t is we 12 would like to get the periodic disassembly and inspection 13 data base from nuclear plants, what have they found in
{}
14 terms of wear. This involves a cooperative program with 15 the utilities and we hope that we can work with utilities 16 in the same way that we are now working with them with-17 regard to MOVs, to get a data base.
18 MR. MICHELSON: The real problem is you can' t 19 get a handle on a check valve to tell what it is doing. Is 20 there any way -- well, with stainless, I would think you 21 could put a hunk of iron inside somehow, but that doesn't 22 work so well with the magnetic body. But it is hard to get 23 a handle on it without --
24 MR. EISSENBERG: Which isLan equipment 25 modification.
5 x-ACE-FEDERAL REPORTERS, INC.
202-347-3700 Nationwide Coverage 80 4 336-6646'
26378.0 168 REE
('T
(_/ 1 MR. MICHELSON: It is not provided for on many 2 of the valves. Most of them.
3 MR. EISSENBERG: It is an important piece of 4 equipment because they have caused some severe headaches.
5 MR. MICHELSON: It is something to think about 6 for future design as to whether or not we need to indica te 7 tne position.
8 MR. EISSENBERG: Ye s , it may be advisable -in 9 future plants to have safety-related valves have a better 10 diagnostic.
11 MR. REED: Don't bring a shaft out, because then a
i 12 you have got to pack and then you got doling and all kinds
(')
\_/
13 of problems, and more shaft-out check valves fail than no-14 shafts out.
15 MR. EISSENBERG: That'is a check valve designer 16 problem. I am not going to_ redesign one. That may be 17 called for in the long te rm , some -- in fact, some pump 18 manufacturers are going to pumps which contain their own 19 diagnostics. You buy it and I will have temperature 20 indicators and vibration indicators as part of the pump 1
21 that you buy. And that may happen in the case of MOVs and 22 it may happen in the case of check valves in the future, if 23 we could encourage the manufacturers. It is a lot easier 24 to install during the design phase to incorpora te the 25 diagnostics. That may not be an option.
f'\
\. )
ACE-FEDERAL REPORTERS, INC.
202-347-3700 Nationwide Coverage 800 33M6M
i 1
26378.0 169 REE i 4 (")T N_ 1 That is all I have to say.
2 MR. MICHELSON: Okay. Any questions from the 3 Committee?
4 MR. WEIDENHAMER: My name is Gerry Weidenhamer.
5 You got to hear about the NRC effort in the valve activity.
6 You didn' t get to hear what is going on.in valve activity 7 in research during the equipment qualification program. I 8 thought maybe just to summarize.
! 9 MR. MICHELSON: Sure, we wanted to get as much 10 of the picture as we can.
11 MR. WEIDENHAMER: I am in the mechanical 12 construction engineers branch. Under this-branch we have i
i
(} 13 14 some valve work that is going on. And so this is all of that. I have already talked to you in previous 15 qualifications discussions, program plans and so on, about 16 the containment isolation valve test program. This is 17 primarily related to obtaining information on the leakage.
! 18 characteristics of certain kinds of containment isolation 19 valves once subjected to accident environment. And that 20 program is going on right now, by the way. I - don' t have 21 any results to report to you.
22 . This part right here is, this established for 23 the Heissdampfreaktor in West Germany. We are installing 24 an 8-inch gate valve that we got from the Shippingport 25 plant. We refurbished it and are installing it in a pipe
()
ACE-FEDERAL REPORTERS, INC.
202-347-3700 Nationwide Coverage 800-336-6646
26378.0 170 REE r
(_)/ 1 system over there. The piping system. We are attempting 2 to determine how those kinds of forces, when generated down 3 through the containment, will affect the operability of 4 this valve. We intend to close and open this ga te valve ,
5 escape valve, against characteristic flows to get some of 6 these characteristics. How do these forces combine? What 7 are these forces that you are concerned about when we 8 throttle this valve way down, close it, open it? Trying to 9~ get at least a first handle on this information.
, 10 Frank Cherny talked on this a little bit ago, 11 work that is going on at ETEC. That- is handled by another 12 fellow. I think this is related to acoustics, how can we f'}
v 13 detect degradation of a valve by using acoustical 14 techniques.
15 This work here is nothing more than reviewing 16 some of the standards that are used in qualifying valves to 17 see if there is something we can do in the equipment 18 qualification program that can clarify some of the 19 uncertainties in the standards. So although it is not 20 anything like these valve tests, it nevertheless is valve 21 re la te d . We are looking at the valve qualification 22 standards. So that is what I mean by related work under 23 this valve operability.
24 So this one here is the one I want to 25 concentrate on. That is the one you expressed in te re s t in (3
, 's_/
l ACE-FEDERAL REPORTERS, INC.
202-347-3700 Nationwide Coverage 800-336-6646
26378.0 171 REE O)
(_ 1 about a year ago when we were here explaining the equipment 2 qualification program. We were on record at that time as 3 telling the ACRS that we would do something in this area.
4 This is just to summarize. This concern is the 5 capability of isolation valves to shut off a high' flow 6 system in the event of a downstream break outside 7 containment that could af fect the operation of other safety 8 systems. And since you talked to us, we have received a 9 user need letter from NRR. That is what I passed out to 10 you today from the Division of Safety Review and Oversight 11 and it essentially addresses the same concerns that ycu 12 have. So what we have done, then, is taken what the user 13 needs letter has indicated and on the back of it, we have fN d
14 identified some work that we are planning to do, starting l 15 right now. So instead of 1987 when we had told you a year i
16 ago, we are now beginning this work now. W'e are just 17 starting. So this is hot off the press. And hope fully we 18 will address some of your concerns in this capability of
! 19 these valves to isolate the system.
20 We don' t know what the next phase will be. This 21 first phase is really what we are going to accomplish this 22 year, is really an attempt to decide what is out there.
23 There are all kinds of valves in these kinds of systems, 24 the se kind of piping systems. We understand that.a-valve 4
25 operator combination on one plant might- be dif ferent from a O
ACE-FEDERAL REPORTERS, INC.
202-347 3700 Nationwide Coverage 80 4 336-6646 -
26378.0 172 REE
(^'t Nm/ 1 sister plant. There are just all kinds of combinations 2 that we don' t know anything about yet. So our. attempt, 3 first attempt will be to -- what is that population.
4 MR. MICHELSON: One valve , of course , of 5 particular in te re s t is the reactor water cleanup on boiling 6 water reactors. Some of those are nonseismic qualified.
7 Some plants don' t even have OA on the reactor water cleanup i
8 system. It is big pipes, high energy, normally open valves.
9 So if a failure of any sort occurs in that system, it is 10 important to i sola te , and then you want to be sure the 11 valves don't misbehave at that po in t . So that is included-12 in your study?
(~ 13 MR. WEIDENHAMER: Yes. I am hoping that you V}
i 14 will read that and give me some comments.
15 MR. MICHELSON: I will read it when I get a 16 chance.
17 MR. WEIDENHAMER: So also, what are the 18 conditions that we have to look at? There are at least 19 three dif ferent lines coming off of there. Like you say, 20 t.he reactor wa ter cleanup, high-pressure core injection 21 line that NRR refers to as -- there are at least another 22 one or two lines. So our attempt then will be to identify 23 what are the valves, what is the population, what are the 24 conditions that are in these lines that we have to 25 essentially _close against? What are the qualifications and O
ACE FEDERAL REPORTERS, INC.
202-347-3700 Nationwide Coverage 800-336 6646
l l
I 26378.0 173 REE 1
/'T
\m) 1 conditions that these valves may have already been 2 subjected to. We want to know al. Sat informa tion be fore 3 we decide what it is we might have to do next year.
4 MR. MICHELSON: You may have to deal with flow 5 conditions in the water systems that are expanding as a 6 result of a break.
7 MR. WEIDENHAMER: We don' t know what all that 8 will entail yet. We hope to at the end of this year --
9 MR. MICHELSON: Is there a size of program in 10 mind at the moment in terms of budget?
11 MR. WEIDENHAMER: Right now we are iden tif ying 12 something like 30 grand to do the work this year.
/~N 13 MR. MICHELSON: But all that will do is hire
(.)
14 someone to think about it. One person.
15 MR. WEIDENHAMER: You are right.
16 MR. MICHELSON: Le ss than a person.
17 MR. WEIDENHAMER: He is going out and actually 18 performing the survey to see what this population is.
19 MR. MICHELSON: So this year you are going to 20 think about it. Next year you will start --
21 MR. WEIDENHAMER: At the end of this year we are 22 going to use that information to identify what the te st 23 program might have to be, what valves we should be using.
24 That type of thing.
25 MR. MICHELSON: That 30K for FY-86 or '7 --
O ACE-FEDERAL REPORTERS, INC.
202-347-3700 Nationwide Coverage 800-336-6646
26378.0 174 REE
,]
(- 1 MR. WEIDEdHAMER: The rest of 1986.
2 MR. MICHELSON: -- that is not so bad.
- 3 MR. WEIDENHAMER
- So I wanted you to hear that.
4 I knew you would like to. That is really all I have to say.
5 We are doing a lot of valve work in the equipment 6 qualification program. Particularly this new one that you 7 identified a year ago which has now been supplemented by 8 NRR. That is all I have to say.
9 MR. MICHELSON: Thank you very much. Any 10 questions from the Committee?
11 All right. We appreciate your presentation.
12 One of the general questions -- I don't know if
(} 13 you are the man who wishes to answer it -- but I am 14 wondering as a result of' the check valve problems of late 15 -- de fe r to NRR , I know they have been doing some general 16 following. Is there anything more elaborate than that in 17 mind at the moment on check valves as a result of the --
3 18 MR. BOSNAK: I think as a result of the San 19 Onofre issue, IE, as I think Frank mentioned, has a program 20 and is supposed to come back with a report of the results 21 sometime in May.
22 HM R. MICHELSON: That is the sort of inquiry.
23 MR. BOSNAK: It is to look at check valves, not 24 necessarily geared to only San Onofre but stemming from the 25 San Onofre problems.
ACE-FEDERAL REPORTERS, INC.
202-347-3700 Nationwide Coverage 800-336-6646
26378.0 175 REE I
\_/ 1 MR. MICHELSON: It is looking at the experience 2 records and so forth as opposed to doing te sting ; is that 3 right?
4 MR. BOSNAK: I don't think they are going to do 5 any te sting.
6 MR. CHERNY: One of the things they are doing is 7 they are visiting a number of plants, three or four plants 8 and they are looking at, in extreme detail, at the exact 9 ways that the in-service te s ting is performed on the check 10 valves in the plants. They have been instructed by the EDO 11 to make recommendations -- I think it is by middle to la te 12 May, something like that -- as to whe ther NRR should do
{} 13 14 anything f a ster in terms of recommending any additional te s ting , periodic testing or surveillance on check valves, 15 anything earlier than what would come out of our planned 16 resolution schedule for 2E61, which is now something like
! 17 May of 1988. Something like that. In other words, should 18 something be done quicker on check valves.
19 MR. MICHELSON: What bothers me a little bit is 20 that there is some indication perhaps that the rate of flow 21 through the check valve has some bearing on how quickly it 22 might wear or whatever. But I suspected that is mostly
- 23 folklore and not necessarily hard scientific fact. Is 24 there anything being done to get some testing done to 25 determine whether or not your in-service testing even means
()
ACE-FEDERAL REPORTERS, INC.
202 347-3700 Nationwide Coverage 800 336-6646
i 26378.0 176 REE 1 anything from that kind of a viewpoint?
2 MR. BOSNAK: I think this was a problem of 3 application of the valve. In other words, if you had --
4 and its location.
5 MR. MICHELSON: But the in-service inspe ction 6 has to address somehow the application.
7 MR. BOSNAK: First of all, you have to have the 8 valve properly designed. It has to be located and you 9 can ' t have improper conditions of flow, for instance.
10 MR. MICHELSON: Do we know what improper 11 conditions of flow might be?
12 MR. BOSNAK: Well, that varies from -- that can d
/~ 13 vary from location to location.
G}
14 MR. MICHELSON: How is that being established? -
15 MR. BOSNAK: I think it is one of the conditions 16 that they mr.y or may not recommend, that this would be i
17 something tha t --
18 MR. MICHELSON: But nothing is being done yet to 19 de termine what the flow regime e f fect might be , if any?
- 20 MR. BOSNAK: I haven' t heard of anything.
i 21 MR. MICHELSON: But that needs to be done by 22 either the utility industry, the valve industry or somebody.
23 Beca use the valve people, manufacturers, say, well, you
, 24 have a problem with certain flow ranges. Is that based on 25 pretty firm evidence? If not, somebody has to firm it up
(~)
U ACE-FEDERAL REPORTERS, INC.
202 347 3700 Nationwide Coverage 800 3 4 6646
26378.0 177 REE ,
1 so we know whe ther in-service inspection is even meaningful 2 until af ter the fact. Once the valve breaks, I guess we 3 know it .is broken.
4 MR. BOSNAK: If'you misapply the valve, you can 5 wear it out in a very short period of time.
6 MR. MICHELSON: Perhaps well within the seven or 7 eight years that we might have between inspe c tions .
1
{ 8 MR. CHERNY: I don' t know that we know that the 9 IE test group is addressing this specific question. I l 10 guess one of them would be better suited to answer that 11 than we would.
12 MR. MICHELSON: That is what I wondered.
(} 13 14 Questions?
If there aren' t any other questions, or any i 15 other comments, then what I will propose to do is adjourn 16 the transcript of the meeting and the Subcommittee members 17 will discuss what we would like to do next. Anything else i
! 18 that needs to be on the record?
- 19 If not, then this mee ting is finished.
- 20 (Whereupon, at 12
- 50 p.m., the meeting was 21 adjourned.)
22 4
23 24 l 25 t
(2) i ACE-FEDERAL REPORTERS, INC,
! 202-347-3700 Nationwide Coverage 800 336 6646
CERTIFICATE OF OFFICIAL REPORTER O
This is to certify that the attached proceedings before the UNITED STATES NUCLEAR REGULATORY COMMISSION in the matter of:
NAME OF PROCEEDING: ADVISORY COMMITTEE ON REACTOR SAFEGUARDS SUBCOMMITTEE ON RELIABILITY ASSURANCE (VALVES)
DOCKET NO.:
PLACE: WASIIINGTON, D. C.
DATE: TUESDAY, APRIL 1, 1986 were held as herein appears, and that this is the original transcript thereof for the file of the United States Nuclear Regulatory Commission.
i l
(sigt) / rIn (TYPED)
REBECCA E. EYSTER Official Reporter ACE-FEDERAL REPORTERS, TNC.
Reporter's Affiliation ,
1 O
i
/
l lO 4
i j VALVE MOTOR i
i i
OPERATOR IMPROVEMENTS ;
1 i
j EPRI PROJECT RP2233-2 1
1 4
- PROJECT STATUS REPORT -
- PRESENTED TO O
THE ACRS SUBCOMMITTEE 4
i ON 1
1
! RELIABILITY ASSURANCE (VALVES) i
! WASHINGTON, D.C.
I i
2 ,
l APRIL 1, 1986 '
i 1
I i EPRI PROJECT MANAGER BOYD BROOKS l
4 O
l 3041E0ses
e PROJECT OBJECTIVES:
IDENTIFY ROOT CAUSES OF IMPROPER PERFORMANCE
- 0F VALVE MOTOR OPERATORS .
DEVELOP MEANS TO EFFECTIVELY CORRECT THE PERFORMANCE DEFICIENCIES PROVIDE FOR THE TECHNOLOGY TRANSFER AND COMMERCIALIZATION OF THE DEVELOPED CORRECTIVE ACTIONS g
i e
gE0cxBPB
Q MOV FAILURE ANALYSIS 1
INPO REPORT 83-037 MECHANICAL 22%
FAILURE TO OPERATE BENT STEMS i DAMAGED VALVE SEATS
- GEAR DAMAGE i
ELECTRO MECHANICAL 32%
4 TOROUE SWITCH FAILURE O -
TOROUE SWITCH ADJUSTMENT i
LIMIT SWITCH ADJUSTEMENT
! ELECTRICAL 27%
I 1
! MOTOR !
! I i CONTACTS
- MCC I
ALL OTHER CAUSES 19%
!O l
3041E0BPB
. . - ---- ---_-. __ - . - - -- _. A
Y e
PROJECT RESULTS:
MAJOR PERFORMANCE DEFICIENCIES IDENTIFIED AS:
FAILURE OF ELECTRO-MECHANICAL SWITCHES DIFFICULTY IN ACHIEVING CORRECT ADJUSTMENT OF TOROUE AND LIMIT SWITCHES INADEQUATE THERMAL OVER-LOAD PROTECTION MICROPROCESSOR SYSTEM DEVELOPED TO PERFORM THE e
FUNCTIONS OF THE DEFICIENT COMPONENTS AND ADDITIONALLY TO: -
PROVIDE DIAGNOSTIC INFORMATION BE READILY RETROFITTABLE e
gE0CKBPB
I
!O I -
! INFORMATION PROVIDED
} TO THE 4
i MICROPROCESSOR SYSTEM i
i MOTOR VOLTAGE :
I j -
MOTOR CURRENT
!O l
PHASE ANGLE BETWEEN l
CURRENT AND VOLTAGE j WAVE FORMS
, VALVE STEM POSITION ,
I !
1
- i l VALVE STEM LOAD i
1 i
i j i
!O l
l 30lilE0ses
/
D O
CLOSE O'EM o
l L
IS II YES
$ TEM THRUST POSIT!CN YES .
700 HIGH7 FULLY CPEN7 ~
h3 ..
K0 15 FUTURE YES 15 STEM THRUST 309 yg3 HAT.ER BLOW 7 FUTURE POSITith TOO HIGH7 (Ulli CPEN7 CRIVE NO yEs to MOTCR q
_ NO CR foo" g '
LCAD TOO HIGH7 NG YE5
!$ F.; TURI TES raq
'v N
INRUSH CURRENT 7 0 11
? C
,0 e
04"F5 N = <s =
HIGH?
. P:0 fl5 INRUSH CURREN!? 'N -
NO U
CONTiiOLALGORITHM FOR VALVE fiOTOR OPERATOR MICR0 PROCESSOR SYSTEM O
3057E0cxBPB l/96
i VALVE MOTOR OPERATOR 7
^"
O MICROPROCESSOR CONTROLLER SYSTEM SET-UP -
d DETERMINE VALVE PERFORMANCE REQUIREMENTS VERIFY MOTOR AND CONTROL POWER SUPPLY VERIFY MECHANICAL AND ELECTRICAL INSTALLATION AND l
INTEGRITY CALCULATE REQUIRED VALVE STEM LOAD O -
CONFIRM VM0 PERFORMANCE CAPABILITY .
CALCULATE VALVE STEM LOAD STRUCTURAL LIMIT CORRELATE TOROVE SWITCH SPRING PACK DISPLACEMENT
! WITH STEM LOAD CORRELATE VALVE STEM POSITION SENSOR WITH VALVE ;
TRAVEL LIMITS UTILIZE THE ABOVE INFORMATION TO ESTABLISH MICROPROCESSOR CONTROL SET POINTS t APPLY APPROPRIATE LEVEL OF PREDICTIVE CONTROL TO AVOID SET POINT OVERSHOOT O
3041E0BPB
- - - - - . - , - , - - - , , - . , , , - . -r , . . ,., , . - .,- - - - . - , - . - , - , . .__ -- ~ - - - . ~ . ,
a- .
P O
CVFUTURE (ACTUAL WITHOUT PREDICTIVE C0flTROL)
OVERSHOOT C0flTROL n OF SET POIflT VARIABLE DUE TO CV (CV) ROTATICflAL IflERTIA 7 FUTURE (PREDICTED) 7
/l CV SET P0lflT :-
CVFUTURE (ACTUAL WITH CV PRESEtiT T PREDICTIVE C0flTROL)
LAST g
~
/
PRESEilT FUTURE
_ _ TIME TIME ,
SAMPLIflG INTERVAL PREDICTIVE CONTROL 9
l l
3057E0cxBPB I/9G
O PROJECT RESULTS PENDING IN-PLANT DEMONSTRATIONS LICENSING AGREEMENTS FOR C0f1MERCIAllZATION Q -
EPRI FINAL REPORT (4086)
VALVE TECHNOLOGY WORKSHOP / SEMINAR AT MEAC EPRI NDE CENTER, CHARLOTTE, NORTH CAROLINA DEC 3 & 4, 1986 I
l l
t O :
3041E0BPB l
/0 0
AVAlLABLE INFORMATION EPRI INTERIM REPORT NP-4254 "lMPROVEMENTS IN MOTOR-OPERATED VALVES' (NOV. 1985)
EPRI JOURNAL - R80 STATUS REPORT (DEC. 1985) 0 ASME PAPER NO. 85-JPGC-PWR-23
" MICROPROCESSOR BASED CONTROL AND DIAGNOSTIC SYSTEM FOR VALVE MOTOR OPERATORS" t-O l
3041E0BPB l
l i
MICRCPROCESSOR-BASED CONTROL AND DIAGNOSTIC SYSTEM FOR VALVE MOTOR OPERATORS Robert A.S. Lee Foster-Miller, Inc.
350 Second Avenue Waltham, MA 02254 ,
4 ABSTRACT As part of a research and development effort initiated by the Electric Power Research Institute (EPRI), a newly developed microprocessor-based control and diagnostics system has been designed to alleviate many of the historical problems related to the performance of conventional valve motor operators (VMos).
The principal advantages of the new system are that:
- 1. It eliminates dependency on conventional electromechanical '
torque and limit switches.
- 2. It provides easy, accurate set point adjustments for position control, stem load control, and motor overload protection. '
- 3. .It provides an easily interpreted graphical display of important +
s/ system parameters, and informs the operator of changing operational characteristics.
The last feature is considered to be of particular significance
' because it allows both instantaneous evaluation cf the condition of a valve motor operator and diagnosis of abnormal VMo performance.
INTRODUCTION In recent years, the issue of motor operated valve (MOV) reliabt1 Lty :
has become the subject of ir. creasing concern, particularly in the nuclear power industry, studies have been conducted by utilitiosl, architect-engineers 2, nuclear power organizations 3-5 as well as by ;
the manufacturers themselves.6 Multiple investigations have been conducted, expectedly, by the Nuclear Regulatory Commissich7 ~9, et al. ;
The most recent of the industry studies INPO's " Assessment of Motor operated valve Failures =3, investigated 644 different licensee event ,
reports (LERs) submitted to the NRC between 1974 and 1982. A major conclusion of this report was that the majority of the MOV failures analyzed were in some way attributable to the electromechanical switchen ,
within the conventional VMO. Discussions centered around the difficulty / uncertainty involved in establishing correct setpoints, especially for torque switches, and the overall reliability of the torque i and limit switches. Other discussions in the INPO report addressed the
('}
(_/
generally uncatisfactory method of providing motor thermal overload protection, and the lack of knowledge as to the mechanical /
i electromechanical condition of a given MOV. To address these concarns. a -
i l
.. . ._ - - . . . ~ - , . , .
.. \'
S f
C~ i l
research and development effort was initiated by the Electric Power .
Research Institute (EPRI) in 1984.
With the identification of the above VMO problem areas, sohn cf the ll) ,
objectives of the EPRI research project were pursued through the following activities:
- 1. Design hardware improvements which can minimize or e11mincte the historically documented problems associated with electromechanical components in valve motor actuators. This would include, but not be limited to, improvements in torque I switches, limit switches, and motor overload protection devices.
- 2. Identify and evaluate apprcpriate instruments, oc design new ones, which can provide diagnostic information on the electrical and mechanical condition of valve motor actuators.
IMPEOVED TCRQUE AND LIMIT SWiTCHE3 In addition to the shortcomings discussed above, a major disadvantage ,
inherent in the conventional torque and limit switchco is that between set points, during valve stroke, there is no knowledge as to what the valve / actuator is doing. Where exactly in mid-stroke is the valve stem? Is a set point going to be reached momentarily? Is a set politt being approached more rapidly than normal because of an overvoltage condition, or more slowly than normal because of incre&Jed stem Packing friction or gear train binding?
These types of questions point to the fact that ths presently 3 available equipment is relatively inflexible, only capable of reacting to a fixed set point valte regardless of the conditions under which the set W
point is reached. Obviously, a first improvement would be to provide
- continuous, rather than discrete, end point information regarding valve stem position. A soluticn to the problem of continuous yalve stem
- l position indication is already available commercially trom several l
rources. Both Limitorque and Bently Nevada offer precision totary potentiometer 3 to replace the standard dial position indicatcra on Limitorque actuators. It is easy to see how, with a constant oc voltage supply, a continuous tracking of the valve stem position could be easily realized.
l The measurement of instantaneous valve stem force (which is directly proportional to the torque exerted) is also quite attaightforward. It is a generally known fact that the thrust exerted on a valve stem by a VMO
, is directly related to the movement of the Belleville spring pack in a ,
(Rotork Series 1400 actuators oporate
( typical Limitorque actuator.
largely on tne same principle.) Dy installing a precision, spring loaded i slide potentiometer in continuous contact with the end nut of the Belleville spring pack, a continuous signal analogous to the dynamic valve stem load can be sensed. -
Ey incorporating the minor equipment revisions described above, available information on the instantaneous operating characteristics of the VMO is increased dramatically.
VMo control then becomes an electronics process whereby continuously changing analog signals, during valve stroke, are compared with fixed set
- point voltages. Moreover, the fixing of set points, by establishing l reference voltage levels, is far easier and more precise than the l conventional means involving the inaccurate positioning of set screws and
the tedious setting of mechanical rotation counters. In addition, by specifying the potentiometers appropriately, the problems of moisture O' intrusion, contact corrosion, and loss of electrical continuity, inherent in most torque and limit switches can be eliminated.
IMPROVE 3 THERMAL OVERLOAD PROTECTION Presently in the industry, thermal overload protection of a typical VMO is provided by a snap action bimetallic switch which is installed in the 115 Vac control circuit. The bimetallic switch is exposed to an electrical resistance heater through which motor winding current flows.
In cases where the current increases ir.ordinately and for a sufficient period of time, the additional neat ftpm the resistance heater causes the bimetallic switch to open, thereby opening the motor starter coil and shutting down the motor. Eoth the bimetallic switch and electric resistance heater are normally mounted in the motor control center (McC).
The technique described above is the subject of great dissatisfaction in the industry. The method is indirect, with resistance heating simulating thermal conditions within the motor winding itself, Simulated conditions are not actual conditions, so there is always doubt as to whether proper overload protection is being provided. This technique is
- nore normally used for continucus duty motors. Because of this, manuf acturers' recommendations for sizing of heaters and selection of bimetallic switches are not suitable for the VMO motors which are only used intermittently. If it has been determined that the thelm31 overload i
iO prote: tion for a particular VMo is not appropriate, then removal and replacement of component 3 ic requireds This situation cculd be improved significantly if a direct, rather thar. indirect, measurement of the notor current were emploVed. This sculd te more r:onsistent with typical VNO manufacturers' recennendations
- for operation which include limitation of motor stalled or locked rotor current to a certair. period of time in seconds, dependin3 on size, model number, applicaticn etc. Monitoring oE motor current, which increases as resistance to motor rotation increases, would also yield important trending data as the valve strokes. Observing motor current rise as the j
valve disc enters the valve seat, for instance, coJld provide an
- indication as to how energetically the two valve camponents are being
, driven against each other.
A potentially more useful control parameter would be motor lead, a complex functicn of motor current and motor voltage, which can be expressed as follows:
Motor Load = KVI cos 0 (1) wherc .
K = Constant V = Mctor Voltage I = Motor Current 0 - Phase angle between the current and voltage waveforms.
Use of motor load will allos tracking of supply voltage as well as instantaneous u.otor current. By continucus calculation of motor load, and comparison with other applicable control variables, volve stem position,
() and load, a mcre accurate de.ternination of the nature of an abnormal condition can Le made. For instance, if motor load it excessivoly high, W
but stem load is normal, it could be concluded that there is a mechanical binding in the drive gear train. On the other hand, if motor load is excessive, and there is a noticeable increase in stem load while the h
valve is in midstroke, then it could be concluded that there is increased stem friction impeding motion caused, for example, by excessive l tightening of valve stem packing or by a stem burr.
REVISED ACTUATOR CONTROL Improved actuator control can be realized by taking advantage of the increased performance data available from the additional measurements described above. Control of the actuator, i.e., switching the drive motor on or off, becomes simply a matter of devising an electronic circuit which measures instantaneous control signals from the above stem load, stem position, supply voltage and motor current sensors, compares these values with pre-established set points, and then energizes or de-energizes the motor starter coil.
There are a number of methods available for realizing the desired control function. A completely analog, hard-wired control system was first considered. In such a system, a series of comparators and operational amplifiers would be used to compare set point voltages with instantaneous sensor voltages. An initial control circuit was designed, but the whole concept of analog control was abandoned due to several considerations. The most significant of these was the fact that in order to alter the control logic, e.g., changing from torque priority to position priority on valve closure, as in the case of a large butterfly 3
W valve, rewiring of the circuit or even replacement of components would be required. This was seen to be a major disadvantage of the purely analog control scheme.
Other design alternatives were also considered, such as using speed control for the VMO, either lay providing variable drive frequency or by using a de or stepper drive motor. These schemes were also subsequently rejected because of the technical complexity and high costs involved.
l MICROPROCESSOR COlJTROL Af ter rejection of the aimlog control approach, it was felt that the proble:n of providing improved VMO control and performance 414 gnosis i capability would best be solved by using state-of-the-art microprocessor i technology. As illustrated in Figure 1, the microprocessor control would function by continuously reading applicable inputs,/such as instantaneous i
MEASURE 0 CALCULATE / EVALUATE CONTROL VARIABLES AND CONTROL QUAN!! TIES l MAKE CONTROL DEC15!ONS OUTPUTS MICROPROCESSOR ,
l Figure 1. The Microprocessor Soluticn l
l l
valve stem position or valve stem thrust, evaluating these measured quantities, and making control decisions based on the evaluations.
O- with typical microprocessor components available today, several hundred iterations per second through the above SENSE-CALCULATE / EVALUATE-DECIDE process could be made, if necessary, thereby providing a fine, responsive control function. Control programs ,
which allow the type of data manipulations described above would be very simple to write. As already mentioned, set points would be quite easy to define or change. The strongest advantage of the microprocessor solution, however, is its flexibility. To adapt an existing microprocessor control system to another VMO control application, with different set points and different control logic, would require minor
- subroutine rewriting, but no rewiring or replacement of components, as would be required with hardwired analog systems.
With microprocessor control system inputs identified. (stem load, stem position, motor load) and control functions understood, the system ;
architecture can now be outlined. The continuous tracking of multiple inputs requires multiplexing so that each input is associated with a discrete channel of data to be i processed. Microprocessor design requires digital data, therefore analog-to-digital
)j conversion must be provided.
Lastly, the requirement for
= _ "llE
_ ,y,, Z t,.o.
, flexibility and adaptability 4 l
i dictates that control programs be written and stored in an erasable, programmable, -aqar5 , ,
, read-only memory (EPROM) that ,,,y ,, ,x, o r , _ , , , , , , , , ,
i can be modified to accommodate == cat M.G' ""
- different control logic. '7,"2" Figure 2 shows the basic configuration of the Figure 2. Microprocessor Control microprocessor system. System Architecture i PREDICTIVE CONTROL i The most powerful capability of the microprocessor approach to control is that it allows an assessment of what is likely to occur at some point in the future. By sensing that a particular set point is about to be reached, the microprocessor control system may be arranged so as to trip the VMo before the set point is actually attained. This i scheme would be most effective in minimizing thrust overshoot caused by i rotational inertia, a problem of great concern particularly with fast closing valves. The principle of anticipatory control is illustrated in -
I Pigure 3.
i At the "present" time, the control variable being monitored has not ;
l exceeded its set point. An approximation of the first derivative of the l control variable with respect to time is calculated by subtracting the '
l' last measured value of the control value from the "present" value, and O aiviai e er tae 911 2 teterv 1. i= ece e - 1 tai =
instantaneous rate of rise of the control variable, e.g., stem thrust, is
- er - ,
l derived. Assuming that the driving force for the control variable were I i
l 1
,. .., , .,-,,-.---- - , ,~ - -,..-.-n- .,_..e-,,...,.. .,n- w , ,, -n-- .-~..,,-. -- , - . . , . - --- -- ~ . . .
1
.]
l Cury79,g Wu MEDICityt e c',vm ("<82cuni
%Es ",,,,,, /
l -----
,,,,P..,
7 Ci rcTunt WM WW CVPatgaT% PREDICTivt contact)
C'Last% ]
/
~
PRE $ TWT FUTual Il4 114 NM Figure 3. Predictive Control removed at the "present" time, the future value of the variable, say n sampling intervals past the present value, could be approximated by the expression:
Present +
resent - b ast x not (2)
CVfuture(n) = g where Cvfuture(n) = nvalue E control variable sampling intervals into the future CV present
= Present value of control variable Cv g ,,g = Last measured value of control variable at = Sampling interval i n = Number of sampling intervals into the future.
The actual terminal value will be somewhat less than that calculated above, because frictional forces will eventually bring system dynamics to a halt. Used properly, this phenomenon will tend to reduce the overshoot above desired control variable set points due to mechanical system inertia.
CONTROL ALGORITIDIS Control algorithms will, of course, depend on the specific control applications involved. For instance, the typical gate or globe valve will close on stem thrust priority, whereas a large diameter butterfly Q
7 A
valve in cold water service will be arranged to close and open on
! position priority.
Motor shutdown priorities will be normally " backed up" by secondary shutdown criteria for reliability. Although the control algorithms will vary from valve to valve, the control elements or tests used to assemble ,
the algorithms are quite simple to list. They are:
- 1. Maximum stem thrust
- 2. Stem position !
- 3. Motor load !
- 4. Future value of any of the '
above variables.
, For the gate valve example which l mu l l == l 1s arranged to close on maximum stem thrust set point and to open I on position limit set point, I (*" l ns ggsgg
- backed up by an open position , ,,,ig, i thrust set point, the operational ta a=.
' [ canu l control flow diagram would be as ,,, n, j, shown in Figure 4. This simple o i ns f,,3%u,, e control algorithm simulates the j s= """
l control functions of the conventional electromechanical torque and limit switches. Figure 4. Simple Control Algorithm In this example. any one of three conditions will shut down the valve:
excessive stem load at either end of travel, or attainment of full open O tet -
In order to provide a finer, more capable control, combinations of the other available control elements given above can be added to the rudimentary control scheme, so that a broader spectrum of possible g ,, g g m, g situations can be recognized and 1
controlled. Figure 5 shows how the l sample control scheme has been !
/ modified. On valve closure, future l EU'\ "'
5% =j' EU value of stem load is predicted and
" " used as a' motor shutdown criterion in order to minimize thrust
,#!00, ni n,wr%m. ns overshoot in the valve seat. In addition, motor load has been added l gg l . . so that abnormally high running g" loads or mechanical interferences in 5%','*g5,'
the actuator gear train can be sensed. Detection of motor load has
, been added to the valve open algorithm for the same reasons. The j l gg l, l g predictive control element has been l
gja added to the position test to !
prevent or minimize backseating of the valve in the open position. l There is no predictive element added I '" I to the stem load test on valve Figure 5. Improved Control opening since this test is already Algorithm a back-up to the position testing.
9 The improved control algorithm depicted in Figure 5 is still not in usable form, even though it does represent a vast improvement over the previous diagram. For instance, the improved control scheme can be h fooled into thinking that motor inrush current is a condition which requires shutdown. An actuator hammer blow, needed to unseat a valve disc, may be interpreted as an excessive stem load, and would therefore incorrectly call for a motor shutdown.
Figure 6 shows a further modified control algorithm which eliminates these difficulties. In it, additional tests have been placed to enable discrimination between normal, acceptable transients and real conditions warranting shutdown of the valve. The hammer blow test is designed to recognize the sharp, short duration spike in stem load associated with this event. The inrush current test disables the motor load protection function for approximately 200 msec after the control system has received a close or open command, analogous to the torque switch bypass on conventional valve actuators. The foregoing descriptions have illustrated just one way in which the four basic control elements can be combined to provide comprehensive VMO control and protection functions.
Obviously, there are many other possible control programs which can be constructed by combining the basic control elements in modular fashion or by introducing simple modifications to control elements. For instance, motor overload protection can be modified to allow maximum motor load equivalent to stalled rotor current for a predetermined interval of time before motor trip, rather than the simple peak detection presented above. The specific control program details will be dictated by the specific control / protection needs of each individual valve actuator.
one.
l ctosa l l l
,,,,s "5
,_os,n ..
, "5 s.n ._
. o sg'j"i "5
. .to.,
rudhrgo.
i
. j.,,
NIGN,
>=b' M1 s
'h'!h/ " " " ' " '
l EE l N_,,, .,
=.- . - !
fil leeUSN CuesER 5T0, Figure 6. Completed Control Algorithm
! MICROPROCESSOR DIAGNOSTICS Besides VMO hardware improvements, the other principal objective of this investigation was to develop and/or identify systems or components '
which would be capable of informing the plant operator of the electrical and mechanical condition of VMOs. There are several devices available on
{ the market which can perform this important function. However, their effectiveness can be severely hampered by the administrative controls in j
a nuclear power facility which govern access to, and temporary j
modification of, plant equipment. Another primary disadvantage of current VMO diagnostic equipment is the time and labor required for attachment and removal of the analysis equipment at each valve being
! tested.
l A clear improvement to this. concept would be to have diagnostic capability in place, permanently installed, without the need for i
additional equipment to be attached whenever VMO performance evaluation is desired. Benefits from such an approach would include decreased time j
required for VMO diagnosis, reduced radiological exposures to personnel, and elimination of the documentation required by plant administrative procedure governing temporary equipment modifications.
During the development of the microprocessor control system, it became obvious that its design was well suited for the generation of I
potentially useful diagnostic information. By using the dynamic VMO data l
being evaluated by the microprocessor control system, valve stem position i
and valve stem thrust, it would be a very straightforward exercise to graphically display this information in the form of a bank of light O emitting diodes (LEDs). As shown in the prototype design in Figure 7, one row of LEDs displays the instantaneous position of the valve, even in midstroke; a second row of LEDs could indicate force in the valve' stem, and furthermore, differentiates between compressive and tensile stem force.
Sine :he microprocessor control software functions by continually asking questions (e.g., Is position limit attained? Is stem thrust too large? Is motor load too high?), then when a condition requiring motor trip is detected, or predicted, the controller chip can be programmed to ,
provide an auxiliary output telling why the trip occurred.
This is seen to be a particularly powerful capability because the ;
operator would be able to discern changes in the performance of the VMO.
For instance, if a VMO is configured to trip on a stem thrust set point upon closure, and the diagnostic LEDs indicate that the last trip ensued :
i
' due to excessive motor load, then the operator could possibly conclude that a gear binding prevented full delivery of thrust to the valve stem and caused an undue increase in motor current caused by the mechanical interference in the actuator gearing. Or, if the valve motor trips out on attainment of. stem thrust set point, but the position LEDs indicate that the valve is in mid-stroke, then the operator could conclude that the stem packing is somehow exerting more frictional force than it should (caused perhaps by excessive tightening of gland follower nuts to stop a leakage). The above examples are indicative of the useful diagnostic i
possibilities which can result from implementation of the
() microprocessor-based control system, simply by analyzing the information provided by the diagnostic LEDs. As also shown in Figure 7, the diagnostic capabilities are further enhanced by the digital readout of l
- . _ - _ _ , _ _ . _ _ _ _ -._ , _ _ - - _ , , . . . _ . - _ , ,_.._,.m m . - -, , ...,- . . . . - ,,.,-_,.__,
lD
~ _..
? M 1 _ s m E p a i E :: Z L M B p r 77, 3 .. ,
Kausse nur i f
summan j-4 .
.p w .m i:
_ : : a:r. .
6( l '.1 E n !
p -
. itzb e p
i:,.y '
- :q :
.y
,p 3 in Figure 7. Control and Diagnostics Panel stroke time. Since the microprocessor samples discrete data points based on quartz crystal clock timing, it is a relatively easy task to count the number of clock pulses from receipt of input command, e.g., "CLOSE", to motor trip, thereby providing an accurate account of elapsed time between those two events. This feature will be of particular use for the periodic testing and documentation of critical safety related valve stroke times, as currently required by the Nuclear Regulatory Commission. .
Lastly, the diagnostic and control panel for the microprocessor '
system provides for a higher level of diagnosis capability, through the standard BNC connector shown on Figure 7 to a storage oscilloscope, or to a computer, for further detailed analysis. Traces of stem position, stem thrust, or motor load, as functions of time, can be selected and stored to allow visual comparison for detection of subtle changes in trace profiles which may be the precursors of more serious mechanical or j electromechanical problems.
TESTING Comprehensive testing of the microprocessor control and diagnostic functions was conducted in a laboratory environment using a WKM 2-1/2" POW-R-Seal gate valve mated to a Limitorque SB-00 actuator. Test results showed the control system to exhibit highly accurate control characteristics, capable of attaining a target stem load within 1105 lb (full scale stem load :: 4000 lb). Anticipatory shutdown through the use of the predictive control algorithm was seen to be capable of l
~
r l
//
reducing inherent thrust overshoot due to rotational inertia from 50 percent down to about 10 percent.
A Demonstration of "last trip" diagnoses was conducted by programming !
six LEDs to indicate the following trip conditior.s: l e Open cycle I Position set point exceeded Thrust set point exceeded Motor load set point exceeded e Close Cycle Thrust set point exceeded Thrust set point exceeded by the future value of stem thrust Motor load set point exceeded.
Several additional lines of control assembly language were written to allow generation of an auxiliary digital output to switch on each LED.
By using this arrangement of last trip LEDs, the various control elements of the program could be tested individually to demonstrate the comprehensive protection provided by the microprocessor.
It should be stressed that this particular arrangement of LEDs was selected for demonstration purposes only and that final LED configuration can be selected and programmed by the end user to satisfy diagnostic requirements unique to the valve and/or unique to the control application.
Testing of the diagnostic LED functions was conducted by first testing the normal valve trip criteria, namely, MOV shutdown on position limit in the opsn direction and future value of thrust in the close direction to minimize rotational inertia effects in the test valve.
After these tests were completed, the open position set point
() potentiometer was intentionally reset at a lower, unattainable resistance so that the valve would trip on valve stem (backseating) thrust.
Finally, the motor load set point potehtiometer was adjusted to simulate a mechanical interference (e.g., gear train binding). Again, simulation of this last situation was chosen in order to avoid possible damage to the demonstration valve / operator.
This last test, conducted in both directions of travel, was of particular value because it also served to demonstrate the effectiveness of the motor inrush current test feature of the control program.
As expected, the last trip LEDs operated flawlessly, each lighting only when the particular motor trip conditions associated with a particular diagnostic LED were prevalent.
FUTURE DEVELOPMENT The next steps of our research effort will involve the demonstration of the fully developed microprocessor control / diagnostics systems in an operatir.g nuclear power facility.
CONCLUSIONS The microprocessor-based control and diagnostics system for valve motor operators, developed for EPRI under this research effort, has
() satisfied all of the design objectives initially established for it.
Laboratory implementation and testing of the prototype has shown it to be an easily retrofitted system, capable of providing accurate, consistent
.,-r - ,. .- --, , -- -,
/ V I
control, while at the same time greatly facilitating the means for VMO i System features practically eliminate the possibility of set up.
mechanical damage due to rotational inertia or faulty torque switch ll ;
adjustments. ;
Most significantly, the microprocessor diagnostic system provides .
continuous, comprehensive information on the current mechanical and )
electrical condition of the valve, by graphically displaying critical l operational parameters such as instantaneous valve stem position and I load, and status of performance limits. Diagnostic capability is further enhanced by advising the operator of the' precise reason for the last trip of the VMO. The detailed information available from this diagnostic ,
display will enable the operator to recognize trends in actuator '
performance, and will help to identify and isolate problems before actuator failures can occur.
REFERENCES
- 1. NGD Operational Experience with Electrically Operated Valves on Emergency Coolant Iniection System. RMEP-lR-03600-32, Ontario Hydro.
Ontario, Canada, 1981.
- 2. Study of Valve Failure Problems in LWR Power Plants. ALO-73, Burns and Roe, Inc., Oradell, NJ, 1973.
- 3. Assessment of Motor Operated Valve Failures _._ INPO-83-037, Institute of Nuclear Power Operations. Atlanta, GA, October, 1983.
- 4. INPO Significant Event Reports (SERs) 26-80, 28-80, 28-80, 85-81 1 99-81, 19-82, 45-82, 49-82, and 63-82. Institute of Nuclear Power Operations, Atlanta, GA, 1982.
- 5. Assessment of Industry Valve Problems. EPRI NP-241, Electric Power Research Institute, Palo Alto, CA, 1976.
- 6. Fast Closing Valve Operators. FC-77, Limitorque Corporation.
Lynchburg, VA, updated.
- 7. Review of Licensee Event Reports (1976-1978). Nureg-0572 U.S.
j Nuclear Regulatory Ucrmission, Washington,-DC, 1977.
- 8. Survey of Valve Oporator Related Events Occurring during 1976, 1979, 1980. AEOD/C203 U.S. Nuclear Regulatory Commission, Office for Analysis l and Evaluation of Cperating Data, Washington, D.C., 1980.
- 9. Malfunction of Limitorque Valve Operators. NRC Circular 77-01, U.S.
Nuclear Regulatory Commission, Office of Inspection and Enforcement, Washington, DC, 1977.
O I
1 1
u 1 m
s ** l 4 -
Clt M I _
O}
p zgae ,
W mm-v
" ~
- 8 M-2T _ __ .g " -
iii N -
g r
3 3W 2
k ~ "
b ,
E o i
> l g >=
ame l
l N
w s W a se
-5 O
= 0. -
G. >=
i w 85o >
j a.
O '"
>W ,
q p
w
= $ --- N .,
l lc W ts
~
z m
j-5g .l
-n
.l
,____4 Im 1 -l
_ _ _ _ _ _ _ _. -j . ;
, i I l i
l 2 l 1 0
t____; a:
W C !
SS6 ..
{
i ,
i O l
PROGRESS ON CHECK VALVE AGING ASSESSMENT FOR THE NUCLEAR F'_ ANT AGING RESEARCH PROGRAM ,
- SAME STRATEGY AS FOR MOTOR-0PERATED VALVES
- RECENT INDICATIONS OF OPERATIONAL READINESS PROBLEMS SAN ONOFRE TURKEY POINT O
O
I O
THE OBJECTIVE OF THIS PROGRAM IS TO PROVIDE RECOMMENDATIONS FOR USE BY NRC REGARDING THE AVAILABILITY AND CAPABILITY OF DIAGNOSTIC METHODS FOR ASSURING THE OPERATIONAL READINESS OF NPP SELECTED SAFETY SYSTEMS AND COMPONENTS.
DEFINITIONS OPERATIONAL READINESS WILL IT FUNCTION WHEN REQUIRED AT A FUTURE TIME (INCLUDING EXTENDED LIFE)
UNDER THE EXTREME POSTULATED CONDITIONS (E. G., IEEE 382)
O DIAGNOSTICS WHAT ABNORMALITIES ARE PRESENT 4
HOW HAVE THEY CHANGED WITH TIME HOW DO THEY AFFECT FUNCTION PERFORMANCE
1 l
O :
THE SCOPE OF THE PROGRAM IS TO STUDY THE EFFECT OF AGING OF REPRESENTATIVE SAFETY SYSTEMS AND COMPONENTS. l l
ESTABLISH A GENERIC APPROACH TO INVESTIGATING AGING AND DIAGNOSTIC MONITORING FOR ASSURING OPERATIONAL READINESS FOR SPECIFIC COMPONENTS CHARACTERIZE AGING RELATED FAILURE MODES AND THE DEFECTS (DEGRADATION AND SERVICE WEAR) WHICH CAUSES THEM l
- IDENTIFY, CHARACTERIZE, AND VALIDATE DIAGNOSTIC MONITORING METHODS FOR DETECTING DIFFERENTIATING AND l TRENDING THE DEFECTS O
- DEFINE ACCEPTANCE CRITERIA BASED ON DIAGNOSTIC MONITORING .* .H WILL ENSURE OPERATIONAL READINESS (DETECT INCIPIENT FAILURES) ,
l PROVIDE RECOMMENDATIONS FOR METHODS OF DIAGNOSTIC MONITORING O
4 L
O CHECK VALVES ARE USED EXTENSIVELY IN BOTH SAFETY SYSTEMS AND BOP SYSTEMS VALVE NUMBER SIZE SYSTEM OF CVs (IN.)
BWR (TYPICAL)
LOW-PRESSURE 10-18 2-28 CORE SPRAY HIGH-PRESSURE 10-14 4-24
. COOLANT INJECTION (HPCI)
LOW-PRESSURE 10-21 4-24 COOLANT INJECTION (LPCI) [ INCLUDES RESIDUAL HEAT REMOVAL (RHR) AND CONTAINMENT O SPRAY]
BOP SYSTEMS 200-400 1/2-24 PWR (TYPICAL)
AUXILIARY FEEDWATER 4-23 4-8 CONTAINMENT SPRAY 4-14 6-14 HPCI 12-28 ~2-1/2-4 LPCI/RHR 5-14 8-10 BOP SYSTEMS 200-400 2-GO O
FREQUEllTLY USED TYPES OF CHECK VALVES INCLUDE SWIi1G, PIST0lb BALL AllD STOP CHECK O , ,.. .
l
- 7 mawM Fig. 2.1. Swing CV.
COVER-fffh A f A /
$W b$
pamTg
_II & IU ~
- !l
. .. . \
L .',
~%g * 'A C
-: ; h Fig. 2.3. Horizontal-piston-lift CV.
O STOP CHECK VALVES ARE GLOBE OR ANGLE VALVES MODIFIED BY MAKING A SLIP STEM CONNECTION TO THE VALVE OBTURATOR l
L i
O TYPICAL TECHNICAL SPECIFICATION REQUIREMENTS FOR CV PERIODIC SURVEILLANCE TESTS ASME B&PV CODE SECTION XI SUBSECTION IWV
- EXERCISE THE VALVE TO VERIFY OBTURATOR TRAVEL TO OR FROM FULL OPEN AND CLOSED CONDITION AS REQUIRED VISUAL OBSERVATION POSITION INDICATOR RELEVANT PRESSURES, ETC.
10 CFR 50 APPENDIX J O
- CONTAINMENT LEAKAGE TESTING AIR TEST (EXCEPT VALVES USING SEAL FLUIDS)
SHUTDOWNS (<2 YEARS)
- ~ - , - , , - , - - - - - - , - ,, ,, - . , , , . . - - - , . . w.
() Table B.1 Check valve failures reported in LERs for period 1969-1983 Percent I. Mode of Failure Leakage Seating surfaces 52 l Casket 4 l Seat-to-body 3 l Internals 32 !
Body 2 1 Slow response time 1 Operational error 2-Other/ unknown 4 II. Mode of Detection Surveillance testing 32 Leak rate testing 27 Normal operation 28 Maintenance 9 Demanda y Other/ unknown 3
() III. Maintenance Activity Repair 54 i Replace In-kind 8 Different 3 Modification 9 Other 1 Unknown 25 IV. Cause -
Wear 8 Crud 15 Corrosion / erosion 5 Failure to seat 4 Design errot 6 Crack / fatigue 2 Installation / fabrication 9 Binding 3 Other/ unknown 48 aResulting from emergency or acci-dent condition.
O
CHECK VALVE FAILURES REPORTED BY NPRDS SYSTEM O
Failure mode Percent Seat leakage 70 External leakage 16 j Failed to close 8 l Failed to open 2 Damaged internals 4 l
. . . . . i." <
Detection Percent In-service and surveillance 67 test Incidental observation 4 '
Routine observation 14 Operational abnormality 11 Maintenance 2 l
{
Special inspection 2 l O i Activity Percent Repair / replace 93 Modify / substitute 4 Temporary measure 3 Failure cause Percent Aging / cyclic fatigue 7 Normal / abnormal wear ,50 Binding / mechanical damage 6 Lubrication problem 2 Previous repair / installation 2 Corrosion 4 Weld related 2 s Dirty 14 Particulate contamination 1 i Out of adjustment 3 Foreign / incorrect material 3 l Unknown 1 Connection defect / loose part 3 Material defect 2
l l
O OTHER FAILURE INFORMATION REFERENCED IN CV REPORT I & E PUBLICATIONS IEB 83-03 DG/CW
- IEB 80-01 PNEUMATIC SUPPLY TO ADP IEB 79-04 VARIOUS IEC 78-15 LPSI IEC 77-08 HPSI IN 84-12 FW IN 84-06 AFW SYSTEM IN 83-06 PUMP DISCHARGE VALVES 1 IN 82-35 HPCI IN 82-26 RCIC, HPCI IN 82-20 LPCI IN 82-09 HPCI IN 82-08 DG/CW IN 81-35 HPCI IN 81-30 HPCI, AFW l IN 80-41 DECAY HEAT REMOVAL IN 79-08 SERVICE AIR REPORTS ALO-75 SAND 80-1887 EPRI NP-241
() NUREG/CR-0848 ID0-1570-TS u
SAND 80-0743 i w . - ,
i 3
} ,
4
\
1 1
- j. CV FAILURE MODES AND CAUSES
) i i
- FAILURE TO OPEN
- FAILURE TO CLOSE
- PLUGGED
!
- REVERSE LEAKAGE
- EXTERNAL LEAKAGE I
6 l
l' f
L .
3 1
i o
I i i !
i 1 ,
4 1 i s_.__,--_,__._--n _ _ _ . _._a_ - ..._._-._.,__;_.,__ __...__ . _ . - . _ . _ _ . . _ _ _ . _
r TABLE 6.2. VALVE FAILURE CAUSES RElg.D TO AGING AND SERVICE WEAR (j) W r'.
V SUB FAILURE FAILURE REVERSE EXTERNAL ,
COMPONENT FAILURE CAUSE f0 OPEN TO CLOSE ELUGGEQ LEAKAGE LEAKAGE CV TYPE: PISTON LIFT BODY OBTURATOR GUIDE WEAR, EROSION, X X X ASSEMBLY CORR 0S10N BODY WEAR, EROSION, CORROSION X X BODY RUPTURE X EQUALIZER PLUGGED X FASTENER LOOSENING BREAKAGE X INTERNALS OBTURATOR WEAR, EROSION, X X X CORRCSION SEAT CORROSION X SEAT WEAR, EROSION. CORROSION X X FOREIGN MATERIAL X X X SEALS CAP OR SONNET SEAL X DETERIORATION CV TYPE: SWING BODY BODY WEAR ERO3 ION, CORROSION X ASSEMBLY BODY EROSION, CORROSION X BODY RUPTURE X FASTENER LOOSENING, BREAKAGE X INTERNALS HANGER PIN WEAR EROSION, X X X X CORROSION, FRACTURE HANGER PIN FRACTURE X llANGER PlN BEARING WEAR, X X X FRACTURE, CORROSION OBTURATOR llANGER WEAR, X X FRACTURE OBTURATOR HANGER WEAR X OBTURATOR FASTENER LOOSENING, X X X TIGHTENING, BREAKAGE OBTURATOR WEAR. EROSION, X CORROSION SEAT WEAR, EROSION, CORROSION X FOREIGN MATERIAL X X X SEALS CAP OR BONNET SEAL X DETERIORATION HANGER PIN SEAL WEAR, X DFTFRTORATT0fJ
()
DIAGNOSTIC MONITORING OPTIONS FOR DETECTION, DIFFERENTIATION,
! AND TRENDING OF CV DEGRADATION METHODS OF MONITORING
- VISUAL EXAMINATION
- INSPECTION DURING MAINTENANCE
- FASTENER TIGHTNESS
- DISASSEMBLY AND INSPECTION (D/I)
- OBTURATOR MOVEMENT (SECTION XI)
- ACOUSTIC NOISE DUE TO LEAKAGE ,
- ACOUSTIC NOISE DURING ACTUATION O
i i
i 6
i l
4 k
i i
l
O DIAGNOSTIC MONITORING OPTIONS FOR DETECTION, DIFFERENTIATION, AND TRENDING OF CV DEGRADATION PARAMETERS TO MONITOR -
- DIMENSIONS OF ALL- INTERNALS SUBJECT TO EROSION, CORROSION, OR WEAR
- APPEARANCE ROUGHNESS CRACKS DEPOSITS OF FOREIGN MATERIAL ACOUSTIC NOISE INTENSITY OBTURATOR TOROUE/ THRUST DURING MOVEMENT
- FASTENER TORQUE TO LOOSEN
- OTHER O
CURRENT STATUS OF CV ACTIVITIES ,
- SCREENING TESTS FOR LEAKAGE MONITORING USING ACOUSTIC LEAK DETECTION (ETEC) ,
- . EXAMINATION OF WORN CVS
- CV FLOW TEST STAND TO VALIDATE DIAGNOSTIC MONITORING OPTIONS ,
ACOUSTIC LEAK DETECTION ACOUSTIC ACTUATION MONITOR 0THER ACTUATION MONITOR
- PERIODIC D/I DATA BASE FROM NUCLEAR PLANTS -[
C:)
(
t
($)
9
> % UNITED STATES l' " n NUCLEAR REGULATORY COMMISSION
- p, j c:ASHIN 3 TON, D. C. 2055S l
- *"* / MAR * ; '98E (3
V
, l MEMORANDUM FOR: Robert B. Minogue, Director Office of Nuclear Regulatory Research FROM: Harold R. Denton, Director Office of Nuclear Reactor Regulation
SUBJECT:
LICENSING NEED FOR SELECTED TASKS UNDER MECHANICAL d
EQUIPMENT QUALIFICATION RESEARCH PROGRAM (A-6322)
~
'~
A "HIGH" priority has recently been given to Generi Issue No. 87, " Failure of HPCI Steam Line Without Isolation ? The NRR Divisien of Safety Review and Oversight has been assigned to cocrdinate the actions necessary to resolve this licensing issue. .
The safety question is the ability of isolation valves in the High Pressure Coolant Injection (HPCI) System, the Reactor Core Isolation Cooling (RCIC)
System, and the Reactor Water Cleanup (RWCU) System, to close against the high flow conditions that could result after a break in those systems h- outside primary containment. Without isolation, such breaks have high potential consequences because other emergency equipment (both LPCI trains andthecorespraysystem)wouldbeexposedtoanenvironmentwhichcould result in comon-cause failures.
We understand that certain tasks in the RES Mechanical Equipment
! Qualification Research Program (A-6322) possibly could be modified and I
rescheduled so that we Could use their results in our program for this -
issue. These tasks and the proposed modifications are described in Enclosure 1. G. Weidenhamer and J. Richardson (RES) agree that these e programs are needed, and believe that the first phase (identification of ,
valves and their present cualifications) can be accorinodated within the existi g n program for'$2SK to $30K. The results of the first phase would be needed to determine whether or not the second phase could be accommodated within the existing program, and to estimate the second phase costs.
In sumary, we have a licersfrg need for the tasks described in Enclosure 1 to help resolve a HIGH priority gereric issue. To this end, we request a specific description of tasks that RES can perform as outlined in the Enclosure, and a schedule for coqoletion of those tasks. To best, suit our needs, the " valve-types-in-use" and " existing-qualifications" determinations should be completed by the end of Fb86, and the " critical flow tests of '
selected valves" (if it is determined that such a program is needed) should i
i L 1
. II E
. 1 O ~WC intend to use your description and be completed by the end of FY ,87.
- schedule in the NRR Task Action Plan currently being developed fcr this ,
isdue, whic,h is.to be submitted for NRR office approval during March 1986. ,
f -
n Harold R. Denton, Director Office of Nuclear Reactor Regulation 4
Enclosure:
Specific RES.-Sponsored Tasks Needed l
. - - cc: J. Richardson, RE M' '
! G. Weidenhamer, RES i
, V. Stello. EDO I i 1 ..
o :
m a
y 1
I k
- I 4
o
~
l l i i
. , , , . _ . _ - , . , , - - - ~ - , . . _ , . , . . . ,
4 .
4 m
(O Enclo~sure 1
~
Specific RES-Sponsored Tasks Needed -
The NRR Task Manacar for Generic Issue No. 87 (Roy Woods), has discussed with RES (Jim Richardson and Gerry Weidenhamer) the possibility of using some of the FY-86 funds from Tack 1 (Validity of Data Base) in a first phase program that-would survey BWRs to; 3) determine the types of isolation valves
. - used on BWRs on the HPCI, RCIC, and RWCU systems; and 2) determine tne present qualifications (if any) of tne existing valves to close under the critical flow conditions that would rebult from a downsteam break in these systems.
This phaie one effort would I;a needed in FY-86 to identify specific valves that should be used in any future industry or RES-sponsored valve testing program (tobedeterminedattheendofFY-86). Also discussed was the possibility of a second phase (. testing) program for these valves beginning in Q FY-87. We would be most interested in valve closure ability tests on the
. RWCU system isolation valves under critical water flow ccnditions, and we also have an interest in valve closure ability tests on the HPCI and RCIC systern
- isolation valvec under critical steam flow conditions (the isolation valves in
,f question on ths HPCI and RCIC systems are on the steam supply lines to the [
~
turbines that drive the high pressure injection pumps).
P di e
Y t
i.
~
s0
w ,, ca a.u
/
PROPOSED RESEARCH FOR RESOLUTION OF O ignaric Issus $7& ffailure of MPCI. Steam Line Without ' Isolation."
- a.
- Safety issue - Generic issue 87 addreaset the ability of isolatian valves in the High Pressure Ceolant Injectioh _
(HPCI) System, other systems that cens off the Main Steam Line (MEL), the Reastor Cers !scistion Ceoling (RCIC) i System, and the Reactor Water Cisacup (AntCfj) system, to close against the high flow conditions that could result after a break in those systems outside primary ec'nttiroent.
! Without isolation, such breaks have high potential
' consequencesescauseotheremergencyequipment(bothLPCI trains and the core spray system) would be exposed to ari j .
anvironment which could result in cosenon-cause failures.
- b. Resea*ch Ob.iectival O - m aa aat. .n the .,ecifi. ve ve as,embu .. ; ed in WR HPC1, other Mll., hCIC, and RWtu systems.
' - Determine the senditi' ens for which thebe v41ye assembitas have been qualif16d and identify valve assemblies that have adequate qualification to apsure isolation 67 a high energy line break, i !
- Roccanehd appropriate follow-up efforts required to assure adequate qualific4 tion sf gaastionable valve assemblies.
- c. Assaarch Descriation
- Through an assessmer.t of FSARs, the INtt Equipment l I
Qualificati6n Data Base, and potential'sburces such as the 1890 NPROS data tubmittals, determine the Valve and .
l actuator manufacturers, types, and times used in the ;
various HPC1, 6ther MSL, RCIC, and R'tCU systems. l f
l l
,- ) a --e_, n ,,o-,,-,- . - - - - L w-.s ,.,,w-- _J... - - - - ~- - -
_ . , _, - , e--,- r-----. ,__.--+e- ,- - - , - - ,,. - - - , - -- _ , - -, ~ --.m,
i- .4 ~~ +._ , ~ ~
' ~ ~
es a ==
4
$ )
. , l O - Determine the conditions for which each valve assembly was qualified and the method used (testing, analysis.
or a combination of both).
- Determine the approximate conditions anticipated as a result of a break ih the subject systems outside cf containment and identify thone cases where the vaive assemblies havs not been qu6lified for the anticipated ;
conditions.
' - Outline a qualification test research program, as appropriate, to resolve ths qualification deficiencies ,
identified above.
- Perform the test or research,pregram identified above,
- d. Products and Schedule
- Document task plan. (3/85) ,
l
- Report identifying speciff.c valves and actuators, and describing present qualification limits e a plant and system basis. (6/66)
- Report on approximate conditions for which valves and actuators should be qualified to encompass pipe break (flow interruptio'n) leads. A recommended program for ;
expanding the qualification of those valve assemblies '
l not already qualified for the pipe break condition will
' be included. (9/86) l0
, . 2 1
us s , . s. .-... .... ..
.8 9
O -
- e. ...ul. tor n .e
- '- Technical data en current states in SWA plants will be used by 1.icer.s*.ng in reselving eenaric !ssue 87. )
- The results of a program to expand the qualification of edstine valve enstablica, if implemer.ted, would be useful in assessing the adequacy of current valve qualificatten prtctices. Of specific interest would be
. the flow interruption qualification included in QV-4.
I t O
P 4
W e
O l 3
I l
l
% I ., *.e
._ NUCLEAR PLANT AGING RESEARCH PROGRAM ACTIVITIES *
.!,0.."o'.'.E.7.""L'.,T',,-
- =====::
l**". "E.T.'.,',"J." "' *
- i PRESENTATION FOR j ADVISORY COMMITTEE ON REACTOR SAFEGUARDS SUBCOMMITTEE ON RELIABILITY ASSURANCE (VALVES) a r
O DAVID M. EISSENBERG WASHINGTON, DC OAK RIDGE NATIONAL LABORATORY APRIL 1. 1986
'RESEARCH SPONSORED BY THE OFFICE OF NUCLEAR REGULATORY RESEARCH.
U.S. NUCLEAR REGULATORY COMMISSION UNDER INTERAGENCY AGREEMENT DOE 40-551-75 WITH THE U.S. DEPARTMENT OF ENERGY UNDER CONTRACT DE-AC05-840R21400 WITH THE MARTIN MARIETTA ENERGY SYSTEMS INC.
l l
l l
O NUCLEAR PLANT AGING RESEARCH (NPAR) PROGRAM (AGING CHARACTERIZATION AND DETECTION OF DEFECTS AND DEGRADATION MONITORING)
- PROGRAM OBJECTIVES AND SCOPE COMPONENTS TO BE STUDIED UPDATE ON MOTOR-0PERATED VALVE DIAGNOSTICS
- PROGRESS ON CHECK VALVE AGING ASSESSMENT THIS WORK IS BEING CARRIED OUT AT ORNL WITH FUNDING PROVIDED BY
) NRC 0FFICE OF NUCLEAR REGULATORY RESEARCH AS PART OF T PROGRAM.
l l
I
_ ,_ - , - ~ . , - - - - , -
. . i l
O THE OBJECTIVE OF THIS PROGRAM IS TO PROVIDE RECOMMENDATIONS FOR USE BY NRC REGARDING THE AVAILABILITY AND CAPABILITY OF DIAGNOSTIC METHODS FOR ASSURING THE OPERATIONAL READINESS OF NPP SELECTED SAFETY SYSTEMS AND COMPONENTS.
CEFINITIONS i
OPERATIONAL READINESS I WILL IT FUNCTION WHEN REQUIRED AT A FUTURE TIME (INCLUDING EXTENDED LIFE)
UNDER THE EXTREME POSTULATED CONDITIONS (E. G., IEEE 382)
O l DIAGNOSTICS WHAT ABNORMALITIES ARE PRESENT HOW HAVE THEY CHANGED WITH TIME HOW 00 THEY AFFECT FUNCTION PERFORMANCE O
l l
O THE SCOPE OF THE PROGRAM IS TO STUDY THE EFFECT OF AGING OF REPRESENTATIVE SAFETY SYSTEMS AND COMPONENTS.
- ESTABLISH A GENERIC APPROACH TO INVESTIGATING AGING AND DIAGNOSTIC MONITORING FOR ASSURING OPERATIONAL READINESS FOR SPECIFIC COMPONENTS
- CHARACTERIZE AGING RELATED FAILURE MODES AND THE DEFECTS (DEGRADATION AND SERVICE WEAR) WHICH CAUSES THEM
- IDENTIFY. CHARACTERIZE, AND VALIDATE DIAGNOSTIC MONITORING METHODS FOR DETECTING DIFFERENTIATING AND TRENDING THE DEFECTS
-)
- DEFINE ACCEPTANCE CRITERIA BASED ON DIAGNOSTIC MONITORING WHICH WILL ENSURE OPERATIONAL READINESS (DETECT INCIPIENT FAILURES)
- PROVIDE RECOMMENDATIONS FOR METHODS OF DIAGNOSTIC MONITORING O
(]) MOTOR-0PERATED VALVE ACTIVITIES
- IDENTIFY AND CHARACTERIZE MOV FAILURE MODES AND CAUSES,
~
AND POTENTIAL PARAMETERS TO MONITOR (NUREG/CR-4234 V1)
IDENTIFY AND CHARACTERIZE ABNORMALITIES FOUND IN A LIMITED NUMBER OF MOTOR-CPERATED VALVES INSTALLED IN FOUR NUCLEAR PLANTS (NUREGICR-4380)
EVALUATE MOVATS AS A METHOD FOR PROVIDING USEFUL DIAGNOSTICS INFORMATION (NUREGICR-4380)
IDENTIFY AND CHARACTERIZE DIAGNOSTICS MONITORING METHODS FOR ASSURING OPERATIONAL READINESS OF MOVS (IN PROGR.ESS)
O l
O l
l l
O TABLE 5.1.
SUMMARY
OF SIGNIFICANT MOV ABNORMALITIES IDENTIFIED BY MOVATS ABNORMALITY PERCENTA IMPROPERLY SET BYPASS SWITCH 75 INCORRECT TOROUE-SWITCH 50 CALIBRATION UNBALANCED TORCUE SWITCH 33 EXCESSIVE SPRING-PACK GAP 17 EXCESSIVE PACKING TIGHTNESS 8 EXCESSIVE INERTIA 8 LOOSE STEM-NUT LOCKNUT 8
)
VALVE BACKSEATING 8 STEM WEAR 8 GREASE HARDENING 8 GEAR WEAR 6 MOTOR DEGRADATION 3 MISCELLANEOUS ABNORMALITIES 35 l
-l A
THE PERCENTAGES SHOWN IN THIS TABLE ARE BASED ON A LIMITED SAMPLING OF VALVES AVAILABLE FOR USE IN THIS TEST PROGRAM. AS A RESULT, THE VALUES SHOULD BE CONSIDERED AS' INDICATORS OF FREQUENCY OF OCCURRENCE OF THE LISTED ABNORMALITIES RATHER THAN ABSOLUTE MEASURES APPLICABLE GENERALLY TO ALL MOVS.
O
i l
O TYPES OF ABNORMALITIES-FOUND IN OPERABLE MOTOR-0PERATED VALVES l TIME DEPENDENT DEGRADATION INCORRECT ADJUSTMENTS AND SERVICE WEAR l
v v DEFECTS WHICH PROGRESSIVELY DEFECTS WHICH DO NOT GET WORSE WITH TIME CHANGE WITH TIME O THE FOCUS OF THE AGING PROGRAM IS ON DEFECTS WHICH GET WORSE WITH TIME.
4 e
i f
I t
- O 4
- l i
l l
O TABLE 6.1.
SUMMARY
OF ABNORMALITIES DETECTABLE BY MOVATS CLASSIFIED BY TYPE TIME DEPENDENT DEGRADATION OTHER ABNORMALITIES BENT STEM EXCESSIVE INERTIA GEAR WEAR INADEQUATE STEM LUBRICATION MOTOR PINION BINDING IMPROPER SEATING STEM WEAR VALVE BACKSEATING GREASE HARDENING INCORRECT TOROUE-SWITCH CALIBRATIONA MOTOR DEGRADATION UNBALANCED TOROUE SWITCHA EXCESSIVE SPRING-PACK GAPA O EXCESSIVE PACKING TIGHTNESSA IMPROPERLY SET BYPASS SWITCHA LOOSE STEM-NUT LOCKNUTA A
ABNORMALITIES THAT CAN CAUSE VALVE FAILURE UNDER SOME
- ANTICIPATED OPERATING CONDITIONS.
- 1 l . .
I l
i o
DIAGNOSTIC APPROACHES DEPEND ON THE TYPE OF ABNORMALITY TO BE DETECTED ABNORMALITY' DIAGNOSTIC APPROACH
. INCORRECT ADJUSTMENTS - ONE SHOT TESTS SCHEDULED AFTER t
l INSTALLATION AND/OR MAINTENANCE i
- OPERABILITY ORIENTED
- CRITERIA BASED ON DESIRED I
ADJUSTMENT TIME DEPENDENT - PERIODIC TESTS TO DEVELOP TRENDS DEGRADATION AND BETWEEN MAINTENANCE PERIODS SERVICE WEAR - OPERATIONAL READINESS ORIENTED O - CRITERIA BASED ON EXTRAPOLATION IN TIME AND OPERATING CONDITIONS 4
O
O LIST OF DIAGNOSTICS MONITORING PARAMETERS EXAMINED FOR POTENTIAL USE IN DETECTING AND TRENDING TIME DEPENDENT DEGRADATION
- SPRING PACK DEFLECTIONIIII2) .
- TOROUE SWITCH ANGLE
- MOTOR CURRENT (1) '
- MOTOR LOADI2)
- LIMIT SWITCH / INDICATOR LIGHT POSITIONS (1)(2) l j
- VALVE STROKE TIME (3)
- MOTOR VIBRATION O
- GEAR BOX VIBRATION
- MOTOR TEMPERATURE l
- STEM POSITION / VELOCITY i
- STEM STRAIN ABOVE PACKING III MOVATS DIAGNOSTICS (U.S. PATENT NO. 4542649)
(2)EPRI/ FOSTER MILLER DIAGNOSTICS (EPRI NP 4254)
(3)ASME B&PV CODE SUBSECTION IWV (TECH SPECS) i O
, , , , _ , , - - _ - - , --- . , , ~ , -
-< r--,.. -- , ,- , , ,, ,, ., _
4
(]) CHARACTERISTICS OF A GOOD DIAGNOSTIC MONITORI.NG SYSTEM DETECT AND DIFFERENTIATE ALL SIGNIFICANT DEGRADATIONS LEADING TO FAILURES PROVIDE REPRODUCIBLE AND TRENDABLE DATA PROVIDE DATA DURING PLANT OPERATION PROVIDE CRITERIA FOR MAINTENANCE ACTIONS BASED ON OPERATIONAL READINESS UNDER WORST ANTICIPATED CONDITIONS E US BLE BY UTILI Y OPERATIONS D MAINTENANCE STAFF j
- BE COST EFFECTIVE IN TERMS OF RISK REDUCTION COMPARED TO ALTERNATE APPROACHES
()
i l
O C NCLUSIONS REGARDING THE M0 VATS DIAGNOSTICS MONITORING SYSTEM CAN DETECT AND DIFFERENTIATE INCORRECT ADJUSTMENTS (I.E., SWITCHES, SPRING PACK)
CAN DETECT AND TO A LIMITED EXTENT DIFFERENTIATE TIME DEPENDENT DEGRADATIONS UNDER SHUTDOWN CONDITIONS ABILITY TO TREND DEGRADATION?
ABILITY TO OBTAIN DATA DURING PLANT OPERATION?
CRITERIA FOR OPERATIONAL READINESS 7 NOT CONVENIENT TO INSTALL / REMOVE RELATIVELY EXPENSIVE I
j i
i O
O ORNL HAS BEEN INVESTIGATING POTENTIAL DIAGNOSTIC ME
! TWO OPERATING MOVS MOUNTED IN TEST STANDS i
- SMA-2 /
i 6" GLOBE VALVE (NON-NUCLEAR)
- /
- SMB-1 18" GATE VALVE (NUCLEAR QUALIFIED) i I
l 1
i
!O l
l l
4 l
!O i
i
i O e BASED ON THE INVESTIGATIONS. THE FOLLOWING PARAMETERS APPEAR 4
THE MOST USEFUL FOR DETECTING AND TRENDING TIME DEPENDENT
! MOV DEGRADATIONS i
! . MOTOR STATOR WINDING TEMPERATURE i i i MOTOR / GEAR BOX VIBRATION l
!O i
l I I
l
!O i
i l
- - - - - - , , - - , , , , , , , - - - , - - - , - ~ , - - - - . - . - - - , . - . , - - - , - - , . - - - - , , ,
- - ---.....?-.,,,----w., -n,n-n.,-----.,,.--.n,.. ,-~,,., -e-.
MOTOR CURRENT HAS BEEN FOUND TO BE P0TENTIALLY THE MOST VALUABLE O DIAGNOSTIC PARAMETER FOR MONITORING MOVS CAN BE USED BY ITSELF OR IN CONJUNCTION WITH OTHER DIAGNOSTIC PARAMETERS WILL ACCURATELY AND REPRODUCIBLY DETECT TIME DEPENDENT LOADS IMPOSED ON THE MOTOR ObRING VALVE ACTUATIONII)
DIAGNOSTIC EQUIPMENT CAN BE SIMPLY AND PERMANENTLY INSTALLED AT MOTOR CONTROL CENTER WITHOUT LIFTING I LEADS
!
- INEXPENSIVE CAN BE USED IN CONJUNCTION WITH OTHER DIAGNOSTIC j PARAMETERS TO DEFINE AND DIFFERENTIATE MANY TYPES
! 0F DEGRADATION AND SERVICE WEAR iO (II
) WHEN PROPER SIGNAL CONDITIONING IS UTILIZED 1
i
- O i
. , _ . . - .y .. , . _,- . . . . , _ _ _ .,.m._ m.. . ,,, , e ., -. , . , , . _ _ _. . . . _ . - _ . . . , , - , _ . - _ _ _ . , _ . . . - - _ . , . . . . , . - - . _
O CURRENT ACTIVITIES AT ORNL FOCUS ON EVALUATING THE DIAGNOSTIC METHODS IN FIELD ENVIRONMENTS I
- REPRODUCIBILITY (ABILITY TO DETECT USEFUL DIAGNOSTICS IN NOISY ENVIRONMENTS)
'
- SENSITIVITY ( ABILITY TO DETECT DEGRADATION AT AN EARLY STAGE)
- SELECTIVITY (ABILITY TO DIFFERENTIATE BETWEEN TYPES OF DEGRADATION)
- TRENDABILITY (ABILITY TO QUANTITATIVELY FOLLOW CHANGES IN DEGRADATION WITH TIME)
()
A DATA BASE OF DEGRADATION MEASUREMENTS OF A VARIETY OF MOVS IN DIFFERENT PLANTS WILL BE COLLECTED AND ANALYZED IN COOPERATION WITH UTILITIES AND OTHER INDUSTRIAL PLANTS
~
4 l
l l
(~)
1
O -
THE CRITERION FOR MOV OPERATIONAL READINESS IS OPERATION UNDER
" WORST CASE" CONDITIONS j
- MAXIMUM MECHANICAL LOADING OF VALVE ,
i
- MAXIMUM ENVIRONMENTAL TEMPERATURE
- REDUCED VOLTAGE TO MOTOR i
j TESTS AND ANALYSES WILL BE CARRIED OUT TO RELATE MEASURED
! DEGRADATION TO ABILITY TO OPERATE UNDER WORST CASE CONDITIONS 4
r O
l i
.)
1 l
I t
!O i
i i
I
i . .
1 4
lO l OTHER ACTIVITIES RELATED TO MOV OPERATIONAL READINESS i
t- .
I i
- UTILITY /ORNL INFORMAL WORKSHOP ON MOV DIAGNOSTICS -
1
- ASME 0&M COMMITTEE 1
i i
i i
I I
i i
a ,
!O i
i 1
l i
I
{
'l 1
} l i
l I
i O
i i
1 .
O AN INFORMAL MOV DIAGNOSTICS WORKSHOP WAS RECENTLY HELD AT 0AK RIDGE
- SEVEN UTILITIES PARTICIPATED IN AN EXCHANGE OF MOV OPERATING EXPERIENCE AND DISCUSSION OF DIAGNOSTIC MONITORING OPTIONS
- MOTOR OPERATOR MONITORING
- VALVE LEAKAGE MONITORING THE WORKSHOP PROVIDED THE BASIS FOR A FUTURE'C00 PERATIVE I APPROACH TO IMPROVING OPERATIONAL READINESS CAN PROVIDE A DATA BASE FROM NUCLEAR PLANTS FOR EVALUATING DIAGNOSTIC MONITORING OPTIONS CAN PROVIDE EXPERT JUDGMENT IN EVALUATING PRACTICAL DIAGNOSTIC DEVICES CAN PROVIDE RECOMMENDATIONS TO ASME O&M COMMITTEE ON NEW STANDARDS BEING DEVELOPED i
O
i
!O .
> CURRENT ACTIVITIES OF THE ASME OSM COMMITTEE 4
- ADMINISTRATIVE PROCEDURE 6200: STANDARDS FORMAT AND CONTENT
- OM 10: REQUIREMENTS FOR INSERVICE TESTING OF VALVES
! IN LIGHT WATER COOLED NUCLEAR POWER PLANTS
- OM 8: REQUIREMENTS FOR PRE-0PERATIONAL AND PERIODIC t PERFORMANCE TESTS OF MOTOR-0PERATED VALVE ASSEMBLIES 4
l t
I 1
i i
!O 1
1 1
1 1
O i l
ADMINISTRATIVE PROCEDURE 6200 WAS RECENTLY FORMALLY ISSUED BY THE
> l j ASME OGM COMMITTEE SUBCOMMITTEE ON PERFORMANCE TESTING j PROVIDES FOR CONSISTENT STANDARDS FORMAT .
i
- PROVIDES FOR CONSISTENT STANDARDS CONTENT j :
i
- PROVIDES GUIDANCE.IN PREPARING STANDARDS
?
! l 1
i PROCEDURE 6200 SHOULD IMPROVE THE QUALITY AND UTILITY OF ALL 0&M l
i STANDARDS ISSUED IN THE FUTURE :
I t ,
1 8
i I
i l .:
3 i
i 1
1 0 l
!~.._____... _ _ . _ _ _ . . , . _ _ . _ . _ _ _ _ . _ . . . . _ . _ _ , _ _ _ _ . . _ _ _ _ _ . _ . . _ . _ _ _ _ . . . . _ . . . .
- = _,
I
()
OM 10 WORKING GROUP IS CONSIDERING MODIFICATIONS TO THE STROKE TIME CRITERIA i
- SEPARATE CRITERIA FOR ELECTRIC MOTOR ACTUATED VALVES i
- REDUCE ALLOWABLE MAGNITUDE OF CHANGE IN MOV STROKE TIME l
7 THERE APPEAR TO BE INADEQUACIES IN RELYING ON STR0KE TIME TO DETERMINE M0V OPERATIONAL READINESS DIFFERENCES BETWEEN DC AND AC MOTORS WITH REGARD TO !
TOROUE/ SPEED RELATIONSHIP I
- DEFINITION OF STROKE TIME i
()
- TEST CONDITIONS t
t t
O
I i !
l !
O !
GM 8 STANDARD DEVELOPMENT IS PROGRESSING SLOWLY
- i CLERENTLY PREPARING DRAFT INCORPORATING CHANGES T0 ACCOMMODATE: ;
- SUBCOMMITTEE COMMENTS ,
- NEW INFORMATION :
- PROPOSED MONITORING METHODS NOT SUFFICIENTLY IDENTIFIED OR DEFINED IN FORMER DRAFTS i f
O i i
i i
O l
l
+
() ,
THE CRITERION FOR MOV OPERATIONAL READINESS IS OPERATION UNDER -
" WORST CASE" CONDITIONS. I.E.,
- MAXIMUM MECHANICAL LOADING OF VALVE
- MAXIMUM ENVIRONMENTAL TEMPERATURE
- REDUCED VOLTAGE T0 MOTOR I TESTS AND ANALYSES WILL BE CARRIED OUT TO RELATE MEASURED
~
DEGRADATION TO ABILITY TO OPERATE UNDER WORST CASE CONDITIONS s
b i
I
(
O !
ORNL PROGRAM IS CURRENTLY STUDYING ADDITIONAL COMP 0NENTS :
0 AUXILIARY FEE 0 WATER PUMPS 0
SOLEN 0ID-OPERATED VALVES [
o PORVS I
O CONCRETE STRUCTURES o
REACTOR INTERNALS l t
i h
i i
i O 1 e
t a
i t
l l
P I
i h
6 i
?
O- !
I t
_ _ . . _ , . _ - . _ . - . _ . - . _ . _ , _ . , _ . . . _ _ _ . . . . _ . , _ - , _ _ _ _ _ . _ _ _ _ . _ , _ _ , _ _ _ _ _ _ _ . - . . _ _ _ _ _ _ . , _ _ _ . _ _