ML20195D477
| ML20195D477 | |
| Person / Time | |
|---|---|
| Issue date: | 10/27/1988 |
| From: | Advisory Committee on Reactor Safeguards |
| To: | |
| References | |
| ACRS-T-1697, NUDOCS 8811070017 | |
| Download: ML20195D477 (323) | |
Text
--_^'-'"----m , , . _ _ _ _
---.,,_a "4 - - - ~ - - - , , ~ . .
'~
l UNITED STATES
' O NUCLEAR REGUIATORY COMMISSION m==menmemomme===== mememommmmmmme
.mm mmm ... meme=======
i l ADVISORY COMMITTEE ON REACTOR SAFEGUARDS
)
In the Matter oft
)
)
SUBCOMMITTEE ON MECHANICAL COMPONENTS l
L i
- I l !
t t
/
' i I
i i i l t
l 1
I f
DATE: October 27, 1988 t i
LOCATION: Dethesda, Maryland.
ll 260 throu h 4, ,,.. e
~ ' '
e).n}g PAGES:
m: , ;
Q...C.0 L(O?! .......
I j
! l I HERITAGE REPORTING 0$nelaiNeperent CORPORATION I
- !218 L Serest N.W, Salm det l .
WesMagnen,D.C.30005 j $r,
'-l' "
' M l' 310:
(181) 43N005 i
rn l - - - ____ - - .-. }
4 1 PUBLIC NOTICE BY THE o
'(,) ~ 2 UNITdD STATES NUCLEAR REGULATORY COMMISSION'S 3 ADVISORY COMMITTEE ON REACTOR SAFEGUARDS 4
5 i
) The contents of this stenographic transcript of the 8 proceedings of the United States Nuclear Regulatory 9 Commission's Avelsory Committee on Reactor Safeguards (ACRS),
10 as reported herein, is an uncorrected record of the discussions 11 recorded at the meeting held on the above date.
12 No member of tne ACRS Staff and no participant at 13 this meeting accepts any responsibility for errors or 14 inaccuracies of statement or data contained in this transcript.
() 15 16 17 18 19 20 21 22 23 24 25 Heritage Reoorting Corporation
() (202) 628-4888
260 t
1 UNITED STATES NUCLEAR REGULATORY COMMISSION o 2 ADVISORY COMMMITTEE ON REACTOR SAFEGUARDS ,
3 )
, In the Matter of: )
'. ) ,
)
I
'o SUBCOMMITTEE ON MECHANICAL ) :
COMPONENTS ) [
6 ) t 4 ;
r 7 Thursday, t October 27, 1988 !
8 I p Room 114 l 9 7920 Norfolk Avenue l Bethesda, Maryland l 10 i The above-entitled matter came on for hearing, ;
11 [
pursuant to notice, at approximately 8:30 a.m. ;
, 12 i bEFORE: HR. CARLYLE MICHELSON f 13 Retired Prncipal Nuclear Engineer L Tennessee Valley Authority O
14 Knoxville, Tennessee i and Retired Director, Office for i 15 Analysis and Evaluaton of Operational !
Data l l 16 U.S. Nuclear Regulatory Commission ;
Washington, D.C. L 17 ('
ACRS_11 EMBER _S_RRESE))I:
) 18 j j MR. CHARLES J. WYLIE '
- 19 Retired Chief Engineer j 0 Electrical Division j 20 Duke Power Company '
Charlotte, North Carolina 21 i
.i . MR. JAMES CARROLL !
l 22 Retired Manager, Nuclear Operations Support ;
. Pacific Gas & Electric Company !
- 23 San Francisco, California, i i I 24
! (:) 2s !
t HERITAGE REPORTING CORPORATION -- (202)628-4888 j
- . - - . . . . ~ - - - , - . _ . . . - - . . . . . . . - - - ..
F 4
261 1 CONSULT _ ANIS: ,
} g 2 P.R. Wohld t
3 ACRS CO'JGI,ZANT StAEP_REMBAR: i i
4 Al Igne f t
- S NRC STAFF PRF,SENTERS
6 H. Ornstein [
O. Rothberg i 7 i
< L 8 k i
L 9 ,
$ 4 10 l t
i 11 ,
i.
I, 12 l 13 :
i 14 1 i '
15 l i t i
! 16 l l
[ 17 :
i L
i t i 18 !
, ?
i 19 l l t
! 20 !
I i 21 l
t 22 !
23 j 24 i I e 2s !
l t
t 1
l l
HERITAGE REPORTING CORPORATION -- (202)628-4888 l 3
,..-,...,..~.4 , , . , .,,,,n.n,.,,,_,,-.n,..-,,,,,,,....,,, , , _ , , _ _ ..,.,r.
'/
262 1 g&QCEEDINQS.
(} 2 CHAIRMAN HICHELSON: The meeting will now reconvene.
3 This is the meeting of the Mechanical Components Subcommittee, I
4 dealing with motor-operated and air-operated valves and other i 5 related mattters.
6 At this meeting there will be further discussions o'. [
7' the air-operated valves, testing and operating experience.
8 Our first speaker is Pete Wohld, our consultant. For this 9 particular presentation, Pete is going to be giving us his i
10 views on work which he did at, in part at Davis-Besse. ;
11 Therefore, he is with us as an invited expert on this .
?
12 particular subject, not as our consultant, and so with that j 13 introduction, Pete, you can proceed.
14 Just so the members appreciate the difference 15 between an invited expert and a consultant, Pete has a 16 conflict because he works at u n.- Besse. Therefore, we can 17 bring him in as an invited expert, but certainly we can have 18 him as a consultant.
19 HR. WOHLD: I also don't get paid for this.
20 CHAIRMAN HICHELSON: You don't get paid for this .
21 hour.
22 HR. WOHLD: Okay. What I am going to discuss is the 23 air-operated valve diagnostic testing that we have been 24 developing at Davis-Besse. John Hayes had talked yesterday,
() 25 who is one of the principal people working on this.
HERITAGE REPORTING CORPORATION -- (202)628-4888
l
, 263 )
I 1 To give you a background on how CJ 4= started, i 2 Davis-Besse, as John explained, is interested in upgrading the 3 ' reliability of the air-operated valves. In 1 coking at the I i
4 valve problems, we recognized the need to have some tools to i i 5 diagnose what the problems were.
I 6 I need to give credit to Fisher Centrol Company who I 7 gave a paper at the EPRI stainar in Kansas City last summer. l 5
i 8 They-sort of gave us the seed material from which we started
! I
'9 off in developing a diagnostic system. We had originally [
10 intended to have Fisher Controls do the work and come in and e
1 11 show us their diagnostic system, but they were still in the ;
i
, 12 development phase. As it turned out, they were developing o 13 their software. We needed something right away at Besse.
J 14 There was some problems we wanted to look at, and in the time l
< 15 that Fisher was developing their software, we took some !
[
J 16 off-the-shelf diagnostic equipment, off-the-shelf software, l t
17 and the diagnostic system I am going to show you in the next l
[ [
4 18 hour cost about $300, ;
19 We had $195 of IBM data acquisition board that we [
t 20 installed in a PC portable computer. Of course, the $300 [
l 21 didn't include the computer, but we had $195 data acquisition
[ j 22 board, and a S99 software program from Unkle Software, which f I ;
l 23 is an MIT professor named Bill Unkle formed a corporation ,
24 called Unkle Software. He has got a diagnostic program not.'
O 2s ta t ie reir1r e11 rece2 ized ia twe i=du err er aice I i I
j HERITAGE REFORTING CORPORATION -- (202)628-4888
264 e' 1 transient analysis type of software in the data acquisition '
2 field. It is used in a number of university student programs. !
}
.3 CHAIRHAN HICHELSON: Is that working on HSDOS or .
4 something like that, using an IDH computer?
5 HR. WOHLD: Right. Sure.
6 CHAIRHAN HICHELSON: Is it programmed for MSDOS?
7 HR. WOHLD: It is programmed, right, for P.he HSDOS, 8 whole IBM structure, j 9 CHAIRMAN HICHELSON: Is this just a PC, or do you 10 use an AT? ,
11 HR. WOHLD: You can use a, I thin' you can use ,
12 the--is it programmed for HSDOS? The old, the original PC, it 4
13 will work on the XT. Works on the AT. Unkle Software l
() 14 accommodates all sorts of computer combinations, supports all ;
15 sorts of data acquisition hardware as well, so it is a very i
16 universal type of thing, but I want to give credit to Fisher l 17 Controls for getting started on this. They gave us the basic j 18 of the idea to get going. l 19 (Slide) !
l i
20 HR. WOHLD: They are pursuing a patent in this area. ,
21 Their system is much more comprehensive than what I am going f 22 to show you here, but this, what I am showing you today is the !
4 i
23 first time we tried it. ;
I :
! 24 We are also going to be developing a more !
L
() 25 comprehensive system. I don't know how the patent thing is I
i f
HERITAGE REPORTING CORPORATION -- (202)628-4888 l
265 1 going to turn out, so we ought to look on down the line.
i
(} 2 Everything I am going to talk about today, i 3 Davis-Bessa feels is not a patentable type of thing. It is l 4 something that you use off-the-shelf equipment. It is an 5 obvious application if you are interested in diagnosing an j i
6 air-operated valve operation. t 7 (Slide) f 8 HR. WOHLD: The slidos I have here, this slide is, i r
9 just I don't expect you to read this. 'rhis is the first page !
10 of the handout that I have given you. It is a summary of a ;
11 test we did on a valve. This valve is a--of course, :
s 12 Davis-Desse is a PWR. The valve is for throttling of seal !
13 injections to the reactor coolant pumps. It is very important O. 14 for operations. It has no particular position. It is 15 important to safety, so we had a lot of latitude in working !
I 16 with the thing. We didn't have to be concerned about a l l
17 hundred percent assurance of everything we did. ;
i 18 In looking at the history of this thing, it had an 19 ongoing history of sticking and then packing leaks, sticking, 20 packing leaks, and so on, so forth. When we got into it, it 21 turns out that if they adjusted the packing to stop the leak, 22 the valve would stick, so we couldn't work it, so they backed I
23 off en the packing the valve. It worked, but it would leak,
[
l
- 24 and when we got onto it, they had a bucket under the thing f i
() 25 with a funnel off of the tube going to the drain to carry off ,
{ !
l
- HERITAGE REPORTING CORPORATIOu -- (202)628-4888 ,
L
266 1 this fairly radioactive water.
{} 2 It is a highly contaminated area where this valve is 3 because of the leakage. Okay. When we got--my impression 4 when I got to the valve was how such a small device could l S cause so much of a problem--just a small operator, small 1 6 inch globe valve. What happened, they went to a live loading
{
7 on the thing. When they set the loading up to the desired 8 torque on the hand bolts, the valve would not fully open.
F 9 Okay. So that's where we came into it.
10 CHAIRMAN HICHELSON: Why wouldn't the valve fully [
11 open when they did that? !
i 12 HR. WOHLD: I'll go into that when we get to the f I
13 diagnostic. l
( 14 CHAIRHAN HICHELSON: All right.
I 15 (Slide) !
i 16 HR. WOHLD: Keep in mind this is a fail open spring !
and diaphragm, spring and diaphragm valve. The problem was 5 17 i
18 that the spring was not strong enough to overcome the packing i i
19 friction. It was that simple. 7t was a 60 square inch !
20 diaphragm, and the bench set that was mentioned briefly (
t 21 yesterday, the bench set is such that at 3 pounds, the valve ;
22 starts to move, so there was 180 pound spring force in the j f
23 fully extended position, and that 180 pounds was all that was 24 available to open the valve fully against the packing f I
() 25 friction, so if the packing friction was any greater than 180 l
l HERITAGE REPORTING CORPORATION -- (2021628-4888 l t
267 l
l' pounds, the valve would stick in mid-position. I 2 What we found was the dynamic pack friction load was 3 863 pounds, so there is no way the valve was going to go open l l
4 when the packing was tightened properly.
5 Talking to Bob Elfstrom who talked to the group 6 October 4th, about a thousand pounds would be the expected 7 packing load for this valve. It has to resist leakage at a 8 200 pound system presuure, so the packing has to be fairly !
9 tighc. So the conclusion that we came to was that the packing ;
l 10 was adjusted properly, that the spring was just not strong 11 enough to take the valve fully open. !
12 CHAIRMAN HICHELSON: Are the valves inside of 13 containment? l 14 HR. WOHLD: No. It is outside? It is outside 15 containment. So they, the proposal--this is the, these are j l
16 the results of the diagnostic. He didn't know that !
i 17 information to begin with. To start with, we had all these [
f 18 various groups that had their own ideas of how te handle thu l 19 problem. Of course, as John explained yesterday, that there 20 was the INC and mechanic interface. When the valve wouldn't [
I 21 open, the INC people were brought in. They disconnected frcm I f
22 the valve stem, and checked their bench ::et. The spring was l 23 adequate. They set it for, to start moving at 3 pounds. [
! 24 There it was at full extension at 15 pounds, so the INC people I l (
aa-O 25 ou14 v ta t the over *or ria - ta t ta =eca oic l;
f HERITAGE REPORTING CORPORATION -- (2021628-4888
_ ~-
268 1 problem.
(} 2 CHAIRMAN HICHELSON: Is there a regulator on that 3 for the 15 pounds? Or that is the actuating air pressure, or ;
4 did I misunderstand? i 5 HR. WOHLD: No. The full actuating air pressure is ,
6 80 pounds. For full closure--I am just talking about the !
f 7 spring. For the INC people to set up the actuator properly, 8 all they do is make sure that the spring starts compressions ;
. 9 at 3 pounds, and it has gone to full compression at 15 pounds.
10 CHA!RMAN HICHELSON: That is just a calibration? I 11 HR. WOHLD: That is just to calibrate the operator, l
12 and that is done without any loads on the valve. So the INC 13 people were determined that the operator was calibrated
( 14 properly, so the mechanics were faced with the problem how are
- 15 they going to get the valve to go fully open, so their t
16 solution was to 6 -k off the packing which gets into the same i i
17 old problem. It is going a sit there drifting throughout the l 18 next year and a half of this fueling cycle.
] !
1 19 The other obstacles and conflicts involved with i
[
20 solving this problem was the design of the valve. It appears !
i 21 the valve is properly designed, but no one understood how it !
i 22 was intended to operate. In this case, it takes the system l
?
23 pressure en the stem, and that's taken credit for in the 24 operability of the valve, so that there was no intent for this
() 25 valve to open fully without system pressure, so the valve i i HERITAGE REPORTING CORPORATION -- (202)628-4888 l
s ,- ;
269 i i design understanding was an obstacle that 'e had to overcome. )
2 The different people involved, the system engineer, 3 he said it has to go to full stroke or it is not operable. He .
I 4 says he won't tolerate a half stroke, and so forth, so his ,
t 5 position was that the packing had to be loose enough for an 6 operation.
I 7 The packing engineer who designed the packing, no ,
8 way was going to back off on his packing modes because he l
9 designed his packing right, so there was an interface there. [
10 Operations was not willing to readily give up the
[
11 system for testing under flow conditions, and then the ;
12 radiation control people, they wanted the paching tight so I i :
t i 13 they didn't have contamination to deal with, so we came in to l 14 the performance grou) at Davis-Besse and proposed that we do 15 this diagnostic testing. Very simple test--and eventually we f l
16 won out, and the solution was to leave the valve where it was, l
l 17 with the packing tightened, and the diagnostic that we did i i
i 18 shows that the valve would operate properly, Lut I'll get into a i l 19 the diagnostic curves here, f 20 (Slide) 21 HR. WOHhD: I show you--the test setup is der.cribed
- 22 on that, the page where the data is. I'll come back to that, f 23 Very simple test setup--we had two sensors, zero to a hundred l l 24 PSIG pressure transmitter. This was right off of the l
25 diaphragm of the valve operator. Zero to 5 volts output, and
- {
r i
, t HERITAGE REPORTING CORPORATIO!! -- (202)628-4888 !
t
270 )
1 a Lanyard pot was mentioned yesterday, the same thing as a ]
2 string st. Just a can with a spring-loaded spring, and you 3 pull it out and it runs a wiper arm across the potentiometer !
l 4 to get a signal out.
5 (Slide) j 6 HR. WOHLD: Okay. This shows the trace. I guess ,
7 you can't see it back there very well. You can come up closer 8 if you want.
I 9 This is the Unkle Scope trace. Basically Unkle l 10 Scope on the data acquisition equipment is looking at volta j
11 versus time. You can plot the data many different ways, but a .
12 this is the one, one way that was useful in presenting the i 13 data. This is the close--this is the first time we tried the
() 14 data acquisition system in the piant. We had played with it a l 15 little bit in the training center, found out that we could get j
- 16 a data on to the computer and get it back out. This is the 1
17 first time in the plant, i 18 As we go along, we are going to correct things like 19 making open in the up direction and closed down, but here tha i 20 open is down. To the right again is time. This is the signal l d
f f
21 off of the string pot, and this trace is the pressures off of r a 22 the diaphragm. The pressure was changed fairly slowly by l 1 l 23 toggling. Well, it is just a continuous holddown of a ramp
{
i l 24 generator in the main control room. This held the toggle :
L
() 25 switch in a continuous open position, and the pressure came f
i i
1 HERITAGE RFSO*. TIM 9 CORPORATION -- (202)628-4888 l
i 271 1 down.
2 You have a zero to 50 second time here, f i'f n times i'
1 '3 ten to the plus one, so there is a 50 second time scale here.
l 4 On this trace, you can see the problem fairly well from a 5 qualitative point of view. The pressure takes a dive elmost 6 to zero before the valve starts to move right here. I'm not ,
i 7 sure why we got this little bit of drifting here, but really 8 motion begins at this point.
9 CHAIRMAN MICHELSON: What was the spring attached f 10 to? (
11 MR. WOHLD: Had a vice grip an the stem with a--the j 12 spring pot, one end was attached and the spring pot was 13 attached to a support that was rigidly part of the valve body, f 14 CHAIRMAN HICHELSON: Good enough.
i 15 HR. WOHLD: Okay. So the valve was coming down. .
I 16 One of the things our system doesn't do at this point is zero :
l 17 out the position. When you set up this string pot, you set it 6
18 .'or mid-stroke and you gat some voltage, se fully open is !
19 somewhere along here some place. The valve failed to go fully [
i 20 open, and you can tell that. It is 1.05 inches per volts on l 21 the sprino pot, and on the vertical scale, trace 2 is the 22 string pot.
23 We are showing 1 volt full-scale from here to here f i
24 for trace 2, and so we have about .3 or .4 volts cha'.n, so it !
O 25 went about .4 inches stroke. That should have gone 1 is..h, so I
L HERITAGE REPORTING CORPORATION -- (202)628-4886 [
L 272 r i
i i the first thing we have, we have represented a condition that
{} 2 shows the, a possible problem 'sith the valve. It is a very 3 simple, very simple approach. There is nothing exotic about 4 it i
h 5 (Slido) !
6 HR. WOHLD: Don't let me get my slides all messed up i 7 here.
3 Here is the closure stroke. This shows, this shows 9 something interesting about the Unkle scope program, and this 10 is a cursor capability, but let me go through qualitatively L
11 what it shows here.
i 12 Again, we have a pressure signal on the diaphragm, [
13 50 second ramp-up, now the position signal. What is t
() 14 happening, on the--up is closed now. On the cloaed stroke, we i
15 are increasing the pressure on the diaphragm. Nothing happens i
16 at first. You have to overcome static friction in the i i
17 packing. You have overcome static friction, but you have more 18 pressure than you actually need to overcome sliding friction, {
19 so the thing takes a junp.
20 Fressure here you can see takes a little dip, okay. j 21 Valve stops a while, picks up some more pressures, appears to 22 take a little bit of a jt'" here, and finally goes on a ,
23 continuous ramp right through here. Then it reaches the fully t
[
24 closed position. The diaphragm, the volume between the can !
t
() 25 and the diaphragm is no longer expanding, so you have a fixed HERITAGE REPORTING CORPORATION -- (202)628-4888 i
u 273 1 volume, and the addition of air simply ramps the pressure up 2 on an expoential rato.
3 The thing that we saw in this was the jump, the ,
t 4 sudden jump. That's very undesirable f ro'a a control point of :
F 3
5 view. This does throttle flow to the seal injection, the !
I i
6 reactor coolant pump swal injection, and if you get a jump {
l 7 like this, you can have a valve that sits there and pops back j 8 and forth, so the one problem we found we need to look at I 9 under full flow and pressure conditions was a possible control i i
10 instability. [
t i
11 The cursor capability, it is a neat thing about
{
12 Unkle Scope. You can select cursors, puts two cut sors any f
13 place on any of the cur'res, and the readout on the left side 14 tel*.s you what the values of the cursors are.
i 15 The cursor number one, for instance, is at 5.07 f 16 volts timra ten to the minus one, so it is at .507 volts.
17 Horizontal, you have horizontal value point No. 80. Cursor 2 [
t 18 gives you the voltage, the time, horizontal value, the point ;
L 19 number, and you, right down here you have the ditference, so j 20 the difference in voltage between this point and this point is [
I 21 .14 volts. Since we know the volts per inch, we know how f
f 22 much, how man.y inches this thing jumped in one step is about j l
23 .15 .4 nches , so it is an extremely powerful diagnostic tool. [
24 You don't have to, you don't have to set a ruler and pick off f i
O 2s on tae xi if it w or ohio 1 tree or taine. !
HERITAGE REPORTING CORPORATION -- (202)628-4888
. ~ . - ._ . _ - - -.
I 274 [
1 In all of the data that we analyzed, it was on--the r
2 chart I just showed you was done with this user technique, so 3 .vou have accurate, accurate analysis values. !
4 CHAI7 MAN HICHELSON: The pressure you are measuring :
(
5 is the pressure right at the diaphragm? -
6 HR. WOHLD: Right.
f 7 CHAIRHAN HICHELSON: Right at the upper? [
8 HR, WOHLD: Right. (
)
9 CHAIRMAN HICHELSON: Did you also try to monitor f
10 supply pressure back of the airsets and that sort of thing? i T
L 11 HR. WOHLD: We will be, and Fisher Controls has that !
l 12 capabi3ity. We currently have the more powerful computer 13 because Fisher is writing their own program and they naven't i l
() 14 got all these features in it yet, but they have the more !
L i
f 15 sensor capability.
16 CHAIRHAN HICHELSON: What I am trying to fathom is ,
17 what is happening together. You know, there is a complex i
[
18 system now all the way back to a supply line. !
I 19 HR. WOHLD: Right. l 20 CHAIFNAN 'HICHELSON: And depending on the size of :
t 21 the supply line meeting the airsets and so forth, you can get l h
22 significantly different novement of the diaphragm simply j r
23 because it is a pressure sensitive device, and but I can't see f I
24 from this picture whether there is, the supply line is big [
l
() 25 enough or too, whether it is too small or things of that sort,
)
f t
l HERITAGE REPORTING CORPORATION -- (202)628-4888 I
, i
275 i
1 or whether the solenoids are sticking. That sort of thing is 2 harder to see from here. !
3 HR. WOHLD: Right. This, this is looking at one i,
4 component only. It is looking at the valve. You have sort of i i
5 a transfer function. You are looking at the input versus the l 6 output, and the input being air pressure. The output is l 7 position.
t 8 CHAIRMAN HICHELSON: Toggled the devices on !
[
9 actuacing the solenoid? Is that toegle-- :
10 HR. WOHLD: It is a throttling valve. It has a (
11 positioner, and the toggling device changes the input to l
12 the--well, there is an I/P converter and there is a positioner E
I 13 and then there is the valve. The toggling device actually l
j () 14 puts a ramp signal into the I/P converter which gives I f
15 believe 3 to 15 psi signal to the positioner, and then--
l 16 CHAIRMAN HICHELSON: It is the follow in the control I
~
! 17 room that started this? f
.\ >~
q 28 HR WOHLD: Right.
19 CHAIRHAN HICHELSON: But at the moment he turned his [
i i
20 switch, that toggled the tracers to start, is that correct? j l t i 21 HR. WOHLD: The tracers were toggled, and one of the j t
i l 22 features of Unkle Scope is you can set it to trigger on i
! 23 whatever you want, and in this case, we triggered it on f 1
I I
j 24 increasing diaphragm pressure, and I think it was-- i
() 25 CHAIRHAN HICHELSON: You actually waited until the l l HERITAGE REPORTING CORPORATION -- (202)628-4888 l
l 276 l 1 diaphragm pressure started to change before you started the
, l
( -
2 trace?
3 HR WOHLD: Right, and in this case, we started, it ;
4 triggered on increase. l i
5 CHAIRHAN HICHELSON: You can't tell what the airsets j 6 were doing and that sort of thing? f i
7 HR. WOHLD: No. This is strictly looking at f
8 ti.e--the valve was the problem. We didn't have any concern l 9 for any of the other equipment, so we wanted to look ;
I 10 strictly-- :
i l
11 CHAIRMAN HICHELSON: Eventually to diagnose a valve j 12 operation you have to look at the whole met? !
13 HR. WOHLD: Sure. There is a lot of--you can do, I i.
I 14 am going to Lhow you a hysteresis curve in a minute that vou 15 can do with any component. You look at the input versus the i 16 output through a cycle back and forth, and you can tell a let 17 about how it is working. We hope to do that with solenoids, t
[
18 the positioner, the I/P convert, lock at input versus output. i 19 You can tell if it is stable, repeatable, and all that sort of r 20 thing. You have to keep in mind this is the first time we had 24 this in the plant.
22 (Slide) >
I 23 HR. WOHLD: This is a hysteresis loop, and this i t
t 1
l 24 is--basically you can analyze the data off of those other two l l
l
() 25 curves, but this seems to be the most instructive. The i
l :
HERITAGE REPORTING CORPORATION -- (202)628-4888
r 277 l 1 advantage of that, the computer is that you can stroke the :
- 2 valve open, closed. You have got the data, and you walk away.
{}
3 You can present it any way you want. This is the exact same !
, 4 data that I just sho,c'? you on the other two curves. It is !
i 5 just presented--you take the time out of it and this is !
i 6 plotted pressure versus position, okay. ,
7 It is our intent in the future to plot position i
8 versus pressure, but we reversed the analog channel and got it !
I i'
9 it this way so it is--one thing we learned about these things, t i
- i i 10 you have got to pay attention to when you plug it in in the j 11 field. l
- f
- 12 This is showing--the open stroke is on this side, j 4
I
. 13 the closed stroke is over here. I guess the best thing to do j
( 14 is go through it on an imaginary situation if the operator was
! 15 on the bench, what kind of curve you would see, and in that f
! t
) 16 case, there would be no hysteresis discounting some minor 17 friction events effects that might be in the actuator bearing l 18 that guides the guide bushing for the actuator arm, but you !
! {
19 would basically have no hysteresis.
20 What you would have is the, as the pressure i
- 21 increases, you would hit the point of preload on the spring !
i '
(
I 22 and the spring would start to corpress. As the pressure !
- I 1
23 increased, the, the travel would go up until you got to the f l
! 24 full travel position of the device, and as you increase l 4 i
() 25 pressure, you would simply be against the stop. The i
i HERITAGE REPORTING CORPORATION -- (202)628-4888
it r
278 d
1 decrease--am I getting this right? Yes. Decrease pressure, j i
2 pressure will come down until you got to the point where the 3 spring was ready to relax. You come back on the slope, (
4 depending on the spring rate, and once you reach the full ;
5 travel and coming this way, you would stop, and any further 6 reduction in pressure wouldn't cause any further travel. ;
7 What actually happens when the velve is hooked up 8 and you have packing friction, you increase the pressure and i 9 you like to start compressing the spring, but you have to (
10 overcome the packing friction, so you go up from the center of 11 this, this line, you go up an amount equal to the packing l i
12 friction. Keep in mind we are looking at pressure, but there j r
13 is a cimple relationship of pressure times area. This was a [
() 14 60 square inch diaphragm, so each pound of pressure was. was 15 60 pounds of force, so it is directly related. j i
16 Overcoma the static friction, and fairly--remember 17 that quick jump--fairly wide spacing in the data points here, j t
18 and I don't have a good explanation of why the pressure dipped 19 right over here. I suspect it is something to do with the lag f 20 time of the pressure and my sensing line decaying off a little i
21 slower than the pressure in the diaphragm or whatever, but the l i
22 jump and the jerky notion was over right here, and then it !
23 went up on a continuous ramp according to the spring rate. [
24 What you have is a spring force plus the packing
() 25 friction, so this, the slope of this line right here is f,
(
I HERITAGE REPORTING CORPORATION -- (2021628-4888 (
i 279 i
I i representative of the spring rate, so we calculate the force ,
2 three times the difference in pressure from this point to 3 here. versus the stroke, and you have the spring rate for the i 4 actuator, so from the data on that, we were able to determine ,
l 5 that the spring was good. There was no, the valve doesn't ;
6 open when it is sitting there. You think the spring is weak.
7 We calculated 758 pounds por inch actuator spring rate, and l i
8 the design value was 780, so that's well, we felt well within l 9 the tolerance of our ability to measure that, {
f 10 (Slide) !
t i
11 HR, WOHLD: First you have the resolution problem.
12 Huch of this reso1'ation is the resolution of the graphics. I 13 You have to go sevatal steps of the actual data acquisition.
l
() 14 the least significant biddor, whatever they call it, before 15 you get a least significant graphical bit, so the data is not {
i 16 as jumpy as the graphics display shows, l
t 17 Okay. The otner things. I told you about the {
l 18 packing springs. When it is f ully in the closed positicit, !
19 backing off en the pressure of come-down, and you get mid-way 20 on this, on this line right here, you get mid-way, you get to i
2' the point where you want to let the spring b2 gin relaxing and {
I 22 opening the valve, but the spring also has to overcome the j i
23 packing friction, so you have to go an amount, en equal amount i
24 this way to overcome the packing friction in the opposite ,
() 25 direction, and then as the pressure decays, the valve came I i
I HERITAGE REPORTING CORPORATION -- (202)628-4888
r 280 1 open. Okay. Keep in mind it didn't come fully open. It [
2 would have been a hysteresis loop down here someplace.
3 CHAIRHAN HICHELSON: What do you do about the stem ;
4 force from pressure if this was an unpressurized valve.
i 5 HR. WOHLD: No flow, no pressure. [
6 CHAIRHAN HICHELSON: Okay.
7 HR. WOHLDt All right? So to calculate the friction j 8 force, all you have to do--we couldn't do it down here. There [
9 is some strange things going on here, and I will mention that [
[
10 in a second. t 11 We will take the--from this point to this point, f
12 that represents a certain force. Take half of that and that's ;
l 13 the friction load on the packing, straight out. There is no, t 14 we don't--you know, very active way of mesh measuring the I
15 friction load. j 16 What is the capability of the valve under flow? How j L
17 much is there to overcome the stem rejection load? l 18 Well, when we go closed with no pressure, no flow, 19 the margin we have in the system is represented by this, the l t
20 height of this set line segment right here, and that, that's 21 in the report I think. It was something like zero, let's see.
l f
- 22 seat load. I
! j 23 (Slide) l I
i 24 HR. WOHLD: That seat load scaled out to be 2675 :
i !
=a r e> r tie
O 25 re=#a o =v rore ta t - r aoi e t i l
l l HERITAGE REPORTING CORPORATION -- (202)628-4888 i l _.. _ ;
- 281 ,
1 os the valve, we have that much load available to handle it. l
[
l i l 2 Okay, It turns out I have got the calculation in there. The
{
l 3 three quarter inch stem, under 2400 pound conditions, the s*.em :
l
! 4 rejection force is something like 1300 pounds, so it turned .
I i
+
l 5 out to be almost exactly half of this, so what we would expect l;
6 to see in, with no flow the system, simply static pressure, is l 7 that this whole hysteresis loop would move up. This point l t
8 here would move up right here, and obviously the valve would l l i 9 go to full open, but you would have a wider hysteresis loop !
10 with the tail off on the end here, indicating it had margin to {
11 open. You would have about 1300 pounds of margin left on seat l I
i 12 load, and about 1300 pounds margin for opening the valve. I
! I 13 (Slide) [
14 HR. WOHLD: Of course, the stem travels on here, l
' i 15 too. The ster travel is just the, from this point to this l l
l 16 point. There is your stem travel. So from the data that we ;
17 got--oh, the, t.',e static f riction, the way I calculated the f I
18 static friction on this was I projected the slope down from ;
i 19 the, from the smooth motion of the operator, and I assumed [
l 20 that this component right here was a, the delta due to the ;
21 static friction, but that I guess the height of this line here [
I 22 was what we were calling static friction. !
23 (Slide) 24 MR. WOHLD: Some of these techniques ?e have to set l
O 2s o= so e sta dard a d we are sti11 trvi o te rioere e er t HERITAGE REPORTING CORPORATION -- (202)628-4888
282 i
'l these things out, but the static friction we es11ed by that :
l I
(} 2 line length was 1003 pounds, dynamic 863 pounds--types c, 1
3 values that you would expect to see. l
. 4 The packing design person--good packing design man, l t 5 but he didn't appreciate the numbers that ne was allowing for l
6 packing load. If this packing load was seen, 2492, the valve !
7 wouldn't have operated properly, so he was allowing for a 8 packing load that was just unacceptable for valve operation. ;
i t i
j 9 Fortunately it didn't come close to that value, and he indeed j t
10 knew that this was a very conservative number anyway, but if L 4
I 11 this were--the valve simply wouldn't work even under system 12 pressure conditions, so it kind of shows the, this diagnostic
{
- i. !
13 kind of brings together a lot of the disciplines that go into l
]
( 14 running one of these valves, and it shows everybody what, what
\
15 it takes to pull it all together, i 4
16 Of course the actuator spring rate we showed, that f t
! 17 wcs okay. Bench set, the bench set at closure. I'm not sure i i
18 if I explained that well enough. The bench set at closure is h
?
i 19 simply the pressure on the actuator without any friction load, j l
f 3
20 to get the full stroke of the valve. !
?
i
]- 21 Well, if this line segment from the beginn < af !
l 1
- 22 valve opening to this point up here, the full clocure, if f 23 that's twice the friction, snd my theory works on what it {
]
l 24 would be without the packing friction, then the mid-point of -
{
() 25 this segment should represent the bench set at closure, and I HERITAGE REPORTING CORPORATION -- (2021628-4888 ,
y i !
283 ;
1 i
i don't have the cursors. I did the cursor technique, 2 determined the difference here, split the two, and found out 3 what the middle was, and that turned out to be like 15.01 f l i i 4 pounds,-which there has got to be some good fortune in that !
l 5 because it was supposed to be 15 pounds, but it gives the I4ic j i
6 technicians a way of checking the calibration of the actuator j t
7 without taking it off the valve or disconnecting from it, so f 8 we could determine from this that the, the INC people had done f i
9 their job right. [
i 10 (Slide) i i
11 HR. WOHLD: Okay. Seat load I already discussed, f f
I 12 That's the excess beyond the full closure stem travel, and t 13 then you just look at the curve, find out what is .'unny about [
14 it. The hysteresis curve wasn't very good for, for
. i 15 determining jumps in the erratic motion, so this, this curve }
k 16 turned out to be valuable from that point of view wher< we saw I 17 the rapid jump of .15 inches per 1 inch, 1 inch valve stroke.
18 That is 15 percent, 15 percent tra$e1.
19 okay. Do you have any questions on thic, on those f 20 curves?
l 21 CHAIRMAN HICHELSON: Only question I have is you, of 22 course, were looking at this valve because you faund that it j I
23 didn't open the full stroke?
24 HR. WOHLD: Rignt.
l l
O 2s caxxxxx" "zcas' son: 28 exi== sieht = e==w t-t, HERITAGE REPORTING CORPORATION -- (202)628-4888 (
l
. t
_ m _ ._ ,g. .= _ - -- -- - -
.p. .
l 284.
1 ' keep it moving, that vas the problem. Yo ace.imu13ted the l
/~T 2 data. Now I guess it just.didn't come through to me real l V
3 clearly what you finally concluded was the reason why it 4 didn't want to stroka full stroke.
5 MR. WOHLD: Oh--
6 CHAIRMAN MICHELSON: From looking at the data 7 because it looks like everything came out like you thought it 8 ought to come out.
9 (Slide) 10 CHAIRMAN HICHELSON: Just the spring constant I 11 guess is not great enough on that spring?
12 HR. WOHLD: What--we didn't go back to the 13 manufacturer. What we concluded was that the manufacturer r
() 14 mu3t have taken crodit for the stora rejection force to operate
( 15 the valve.
- 15 CHAIRMAN MICHELSON
- All right.
17 HR. WOHLD: And we found out that it should be, it i 18 should not be unexpected that the velve won't st:oke in an 1S 'Inp'tcissurized condition.
20 CHAIAHAN MICHELSON: Are you saying then that it 21 really did open full stroke when you had pressure on the 22 system?
23 MR. WOHLD: From the diagnostic, we expect it to.
24 CHAIKMAN MICHELSON: I t.hought you concluded from t
() 25 earlier observations that it didn't seem to be opening full HERITAGE REPORTING CORPORATION -- (202)628-4888
+ , .. ._
285-il stroke, and I assume that was under full pressure condition.
You know, the fellow said well, if I tighten the packing f [}
2
!o 3 properly, then the valve won't open all the way. I assume 4 .that meant Gnder full press >tre.
5 HR. WOHLD: Okay. And in looking at, in 17oking at-6 the maintenance history, we didn't do a mode analysis to 7 determine what mode the valve was in.
8 CHAIRHAN HICHELSON: Hight have been it wasn't 9 opening full open because he didn't have pressure on it?
10 HR. WOHLD: It could have been during refueling 11 outages where they put maintenance on the valve.
r 12 CHAIRHAN HICHELSON: That part didn't come through 13 to me for sure.
14 HR. WOHLD: Ke didn't go into that detail.
15 CHAIRMAN HICHELSON: What you are saying, if y(u do 16 indeed operate it under operating conditions where you do have 17 full pressure, the thing really does stroke all the way opan?
18 HR. WOHLD. It should. We will find out in a couple 19 of weeks, and I don't have any doubt about it. It is just 20 straightforward physics. The pressure on the stem, three 21 quarter inch stem, it is going to give an extra 1300 pounds, t
22 It only takes, only takes a thousand pounds to overcoLe the 33 static friction. It is going to be 1300 pounds plus the 24 spring to overcome the packing. ,
() 25 CHAIRMAN HICHELSON: Is that valvo supposed to fail HERITAGE REPORTING CORPORATION -- (202)628-4888 !
286 1 open, or it is a non-safety valve? '
r~w 2 HR. WOHLD: it ii a fail open valve, but it has no--
b 3 CHAIRMAN HICHELSOti: Is there any safety reason?
4 HR. WOHLD: Has no safety reason; it is strictly an l
5 operating situation.
6 CHAIRMAN HICHELSON: It. fails open but depending on 7 what system pressure, it may not fail full?
8 HR. WOHLD: Right.
9 CHAIRMAN HICHELSON: If system pressure remains up, 10 then it will fail full open? IC the pressures drop, it may 11 not get all the way?
12 MR, WOHLD: The only reason it will fail open is 13 that if you lost the air for any reason, you could still
() 14 operate the plant. You have injection of the seals, you would 15 save the seals. You would have a little extra letdown to 16 compensate for the increased seal injection.
17 CHAIRMAN HICHELSON: But you are saying the spring ,
13 plus the stem thrust force of pressure is enough to push it 19 full open?
20 HR. WOHLD: Right. We are able to determine that 21 from the diagnostic. We had to overcome all these prejudices.
22 The system engineer said well, I'm not going to accept it 23 until it strokes full open, and we are not going operational 24 until it does, and the packing fellow refused to back off on
() 25 the pscking, and we finally had a strong discussion, and based HERITAGE REPORTING CORPORATION -- (202)628-4888
287 1 on the diagnostics, we said the valve is good. Leave it n -2 alone. 'So we expect to have a fully operational valve that 4
V(~~'T 3 doesn't leak and is not causing contamination.
4 CHAIRMAN HICHELSON: You tried to make that 5 determination without pressurizing the valve is what this 6 amounts to?
7 HR. WOHLD: Right. In the most convenient, 8 efficient way, we got the data, and we said hey, you are ready 9 to go, you don't have to wait until the plant starts up to 10 determine that there is a problem. The spring, the packing is 11 good. It is in the right range. The spring is good. The 12 actuator is set upright. The seat load has margin to handle 13 all the other forces during operation. The only problem is
( 14 the stem travel which that comes with the system pressure.
15 CHAIRMAN MICHELSON: I guess testing, it is kind of 16 difficult to test the air-operated valvas when you can't apply 17 full system pressure because the system pressure has a 18 considerable effect on the ability of the valve to open. You 19 are workiag against the spring essentially.
- 20 MR. WOHLD
- It has a large effect, but I don't know 21 that there is very many situations like this. I think that's 22 why this thing wasn't identified. It is as far as--
23 Cl! AIRMAN MICHELSON: 0.te of the circumstnnces I can 24 think of where it gives one great complication, and that is if l () 25 your valve is inside of containment, and you have to operate l
l HERITAGE REPORTING CORPORATION -- (202)628-4885 '
288 1 it during a design basis event where the containment pressure
{} 2 might be 60 pounds, you have to somehow figure out that wich 3 the back pressure of 60 pounds on the airset, you are still 4 going to bleed enough pressure off the diaphragm for it to do 5 the right thing, and you can't do that by any simple test. I 6 mean it is a, you have to be a little more elaborate to verify 7 that. .
8 HR. WOHLD: There is a lot. This is just a very 9 simple approach, and we are looking at a ramp change. We 10 could go to a step change in the position signal and look at, 11 look at changes in travel speed and so forth, trying to detect 12 adequately sized lines.
13 CHAIRMAN MICHELSON: Msnitored your airset; you
( 14 might have explained a little dimple you got in your curve 15 also. You never did tel] us what happened at the bottom of 16 the curve with--you have your littic dashed line.
17 HR. WOHLD: Okay. Let me--I don't know. Someone 18 who knows more about friction than I do might be able to
'9 explain this, but what it seems to me is happening here, the 20 pressure has almost gone to zero before the valve starts to 21 move, and I would have expected the vtive to go in the open 22 position a short distance and stop, but it didn't.
23 What it did was it kept slowing down and continued 24 to open ever more slowly, so the conclusion I would come to on
() 25 that is that there must be a velocity component of friction so f
HERITAGE REPORTING CORPORATION -- (202)628-4888
289 1 that as the valve slows down, the friction load is less, and T'T 2 it will continue to move, but in a continting slower velocity.
\~).
3 CHAIRMAN HICHELSON: Your dynamics friction is not a 1 4 constant?
5 MR. WOHLD: Right.
6 CHAIRMAN HICHELSON: Velocity sensitive.
7 MR. WOHLD: Right.
8 CHAIRMAN HICHELSON: In some manner.
9 MR. WOHLD: Right, some manner; so that was an 10 interesting thing.
11 CHAIRMAN '.iICHELSON: I guess 5at is not to be 12 unexpected. Just to predict what it should be might be 13 difficult.
( 14 HR. WOHLD: Right. I don't know tnat much about 15 friction. I probably shottid expect it.
16 Okay. We have goc a lot of plans for expanding the 17 capabiliti* of this. As you said, it would be nice to know the 18 supply pressure. We would like to be able to directly put a i
19 current or voltage in an I/P converter, monitor the output of r 20 that, the input and output of the positioner, which requires j 21 more transducers, more data acquisition capability, and we 23 want to make it very user friendly.
i 23 This system I showsd you right here--not this one, 24 but to get the data I have been showing you, required a power
[
() 25 supply to ba hooked up to the pressure transmitters that we HERITAGE REPORTING CORPORATION -- (202)628-4888
=
l l
290 l 1 brought down. In this case, it took us two hours 2 approximately from bringing in the equipment to exercising the 3 valves, setting up.the power supply, to leaving.
4 In the next valve that we tested that I did not, 5 really not discuss in here, it took two shifts to get the 6 power supply to work, so we want to remove as many of those ,
7' kind of variables as we can and have everything battery 8 powered like the pressure transmitters, and the, we are 9 looking-at a system. We are buying a system that I have put' 10 -on the board here.
11 CHAIR! FAN HICHELSON: How did you position your ,
12 pressure transmitters on the sys+.em?
13 HR. WOHLD: It was simply a, I believe about a 20
} 14 volt DC input to the pressure transmitter, and it put out zero 15 to 5 volta.
16 HR. HICHELSON; How did you got it into the air 17 system?
18 HR. WoliLD : Oh, you have to break the air tubing.
19 Put a T in the air tubing. Okay. That's another thing. We i 20 are looking at eventually a permanent modification so you can 1
21 have quick disconnects so you just take your battery-powered 22 equipment, walk up to the valve, hook into the quick 23 disconnects with an air sensing line that you brought with t
24 you, start t5e data ace,uisition, call up for a valve stroke,
() 25 open/ closed, unplug from the quick disconnects and walk out, HERITAGE REPORTING CORPCRATION -- (202)628-4888
. . - - . . _ - - - , . - - _ - - - - . - _ - - - - - _ - - - - - - - - - - - - . - - - - ._- _.-_ E
- e i
t 291 1 and it is all done. It is almost'as quickly as you can stroke
(~S "2 the valve.
%)
3' CHAIRMAN MICHELSON: One more thing--the leak 4 -safety-related valve.
5 HR. WOHLD: Well, I'm trying to sell the bicycle G tire type of valve obviously.
7 CHAIRMAN MICHELSON: They are pretty good.
r.
8 HR. WOHLD: I have got an old Buick that they worked ;
9 on for, since 1964, and I don' t 'aave any leaks, but there are l 10 hard inversions. If you look at the air conditioning system 11 on a car, they use the same valves that are obviously a lot 12 sturdier, but that kind of idea for a quick disconnect, and we 13 are talking largely non-safety-related equipment tying into a 4
4
() 14 non-cubed air line and so forth, so it should be fairly easy.
15 We have done a search of available data acquisition l
16 hardware. It is kind of an interesting area. I was trying to c
17 get into this, find out how you got scuff into a computer and 18 analyzed it when I was still in the Commission, but I couldn't i 19 find anybody close to me that knew anything more than DBASE or 20 LOTUS and that sort of thing, and then I found out the control .
21 industry is heaviP/ into this, but they are all talking much ;
22 more sophisticated r;scems and want to run an entire plant 23 with a PC with a thousand inputs or a thousand outputs, and 24 what I was looking for was something in between. '
() 25 I didn't want to write anything in TURBO PASCAL HERITAGE REPORTING CORPORATION -- (202)628-4888
g ._
292 1 because that would take me a year to learn TURBO PASCAL. I 2 vinted a-menu-driven system. We were looking for something
}
3 that was using Unkle Scope which we were used to, and it 4 appears to have good recognition in the industry, and we found 5 this basically boiled down to this box.
6 This is a small outfit in Alliance, Ohio, five 7 people. It had this box. They have been building it for a 8 year. It does all the things we want it to do. It does 9 data--well, we were accelerated in our effect. I was ,
10 analyzing this what, two weeks ago I guess, John? I had it on 11 my desk and I got a call from the Engineering. They wanted '
12 some data acquisition real quick on the pressurizer surge -
i 13 line. It is an item of interest right now, so this box would '
( 14 handle 13 inputs, actual valve, handle 16--they wanted 12 .
t 15 inputs r. hat would input pressurizer surge line position 16 information, and so we very hurriedly went ahead and ordered 17 one of these boxes. It is supposed to colao in tomorrow, and -
18 they are setting up the string pots, and I'm very quickly 19 trying to learn some of the additional software that goes 20 alono with it so we can have it operational. Expect to start k
- 21 the plant up possibly sometimo in the next week.
22 CHAIRMAN HICHELSON: That's all keyboard input? ,
23 MR. WOHLD: Yes. The nice thing about this, this 24 is, what we are buying is truly a black box. It is aluminum,
() 25 but it has got a black centerpiece that hides everything in -
HERITAGE REPORTING CORPORATION -- (202)628-4888
293 1 it. What you have'got is a. lap top computer. You can use r~ 2 almost any lap top computer, and that's normally bought D]
3 separately from the box. The box itself is from the center 4 point on down. It has got a black aluminum cover, and there 5 is nothing exciting about it. It has got 16 inputs on the 6 side, got some digital stuff here that we are not using, but
'7 you just plug in the lap top to the black box. The black box 8 is the data acquisition hardware.
9 CHAIRMAN HICHELSON: That lap top is a separate item 10 from the box? You set the lap top in there?
11 HR. WOHLD: Yes. You can use any computer. You can 12 use a table top PC, portable lap top, anything.
13 CHAIRMAN HICHELSON: Okay.
() 14 HR. WOHLD: What we are going to do is sample 15 pressurizer surge line position once a minute on 12 different i 16 channels. We have, this box has the capability of storing ,
17 data in that mode for 129 hours0.00149 days <br />0.0358 hours <br />2.132936e-4 weeks <br />4.90845e-5 months <br />. The nice thing about it is i 18 it is good for, was it 150 degrees, 140, 150 degrees :
t 19 temperature, so in the containment environment, it will 20 survive the temperature. You use the lap top to set up the !
t 4
21 data acquisition and you can unplug it, take the lap top out.
22 It is only good for--we are buying the best lap top in terms ,
23 of temperature capability that we could find that was good for 24 113 degrees. Most are good for like--
l
() 25 CHAIRMAN HICHELSON: You can put the Unkle program i
l l HERITAGE REPORTING CORPORATION -- (202)628-4888
294 1 in the lap top?
/T 2 HR. WOHLD: Unkle actually comes with the box.
k) 3 There is two programs. There is a data screen program that 4 these people have written themselves, and an Unkle Scope comes t
5 with it, but so we are looking, 1^< king at this for the 6 pressurizer data acquisition, but when we get done with that, 7 we will fold it into *'a air-operated valve diagnostic 8 testing. All you need, there is 8 inputs right here, minus 9 ten to plus ten volts, and 8 inputs down here minus two 150 10 millivolts to p'Jes 150 mil 11 volts, and it has got selectable -
, 11 gain of one ten in a hundred, depending on what your input 12 signal is. You program that with the lap top computer, walk 13 away. A couple days later, you come back and you have got a
( 14 continuous srt:trum of data.
15 CHAlRMAN HICHELSON: What does the box itself cost?
16 HR. WOHLD: The base price on the box is $7500, but i 17 with the, all the extras, it is about 9,000.
18 CHAIRMAN HICHELSON: Then you have to have a lap top 19 to go with it?
20 HR. WOHLD: Lap top goes with it, right. It is, the 21 nice thing about it is that we are using basically I 22 off-the-shelf equipment, and we can massage it and play with 23 it to suit whatever--
24 CHAIRMAN HICHELSON: Can you get display on the lap
() 25 top at all?
HERITAGE REPORTING CORPORATION -- (202)628-4888
295 1 MR. WOHLD: Sure. This has, you can got the Unkle 2 Scope displays just like I showed you.
(V~T 3 CHAIRMAN HICHELSON: They are instantaneous as you 4 are doing your work?
5 HR. WOHLD: The--if you want to watch it, as data 6 comes in, it scrolls across the screen as you are taking the 7 data, and you can present four curves at the same time, so in k l
8 the data acquisition, you can do one, one data collection per I 9 hour1.041667e-4 days <br />0.0025 hours <br />1.488095e-5 weeks <br />3.4245e-6 months <br />, or 20,000 per second, so you have a tremendous range.
10 Most data loggers are very slow. They take the shortest 11 intervals, normally run one second, where this has up to.
12 20,000 samples per second, so--
13 CHAIRMAN MICHELSON: That'; fast.
14 MR. WOHLD: And liko I say, we are planning en doing
! 15 all these other things, being able to test the positioner, tne <
16 supply pressure, look at the whole works. Okay.
17 CHAIRMAN HICHELSON: Any questions? .
I 18 HR. WCHLD: Oksy. Thank you.
19 CHAIRMAN HICHELSON: Thank you, Pete. I believe we i 20 are running a little ahead of time. ;
l 21 HP. WOHLD: I finished ahead? That's a first. I ;
l 22 usually take twice my time. l i
23 CHAIRMAN HICHELSON: We art keeping right on time. (
24 We want to do it today. We thank you very much, and the next L
() 25 speaker I guess is going to be Hal Ornstein? Are you speaking HERITAGE REPORTING CORPORATION -- (202)628-4888 ,
f
o 296 ,
1 for, on operating experience with air and hydraulic valves, ,
2 whatever? You will?
i 3 HR. ORNSTEIN: Good morning. I was speaking to Mr.
4 Michelson just a few days ago, and I guess I usually do best 5 on shortest notice, and told him that I can probably speak ,
6 from three minutes to three days depending upon what you would 7 want to bear with.
8 The subject that I have narrowed it down to is the t 9 area in which I am presently doing a case study on, and that i
t 10 is solenoid valves, and their experience, the experiences we !
11 have been having with them.
12 (Slide) 13 HR. ORNSTEIN: This basically constitutes total set
(/ 14 of vugraphs for this morning that I am presenting, 15 CHAIRMAN HICHELSON: To be sure now, you are going 16 to talk about solenoid valves being used as air control valves 17 to large diaphragm valves as opposed to solenoid valves which 18 can operate process directly?
19 HR. ORNSTEIN: Well, actually we have not really 20 made that distinction in what we are looking at. We have been 21 looking at solenoid valves in general, and we have been ;
22 narrowing down the areas of interest. We have seen them !
23 actually directly working on a stream and also piloting larger ,
1 24 components based upon the control air going, you know, in j t
!- () 25 different directions.
r
! i I HEPlTAGE REPORTING CORPORATION -- (202)628-4888
- 297 1 We have not basically cut off either from interest.
2 We have been seeing failures of many different valves in many 3 different systems, and what originally started off to be a !
4 look at one particular manufacturer's particular valve in a 5 particular application, has grown and spread, and effectively 6 the case study that I am working on is going to involve about 7 a dozen different systems, many, many different failures.
8 Now I have been acquiring data at a fast and furious 9 pace since the beginning of this year. I'm hoping to get a 10 copy of the case study in draft for peer review before 11 calendar '88 is over.
12 The kinds of failures that we have been seeing have i
13 been very interesting, and I mean I could probably give you a i
l
() 14 whole slew of them. I'm not sure if that is exactly where I i
l 15 ought to be going right now.
1 l The important thing is the big picture, and that is 16
- 17 we started to observe failures which were more and more l !
r 18 important than previous ones. We have been seeing that they i
(
19 are happening in many dif f erent saf ety systeras. We have seen l 20 that the pace has been picking up. At least it is our 31 observation from data acquisition, from examining operating l 22 events, whether they come in on the early morning information 23 from the NRC operations center, reviewing LERs and NPRDS 24 reports.
() 25 If one takes a look at what has been happening over HERITAGE REPORTING CORPORATION -- (202)628-4888
298 1 the past four years, we seem to be getting an upward curve of 2 the frequency of occurrence. I can't say that any one 3 pArticular system has got problems more so than any other.
4 They are really rather widespread.
5 The thing of importance--
6 CHAIRMAN HICHELSON: Is there some kind of an aging 7 phenomenon then?
8 HR. ORNSTEIN: There is a lot of phenomena. There 9 ere a lot of the--one of the questions one would ask, is it a 10 reporting phenomenon? We find t'at it is difficult to get out 11 the data from the available data bases without some 12 manipulation so to speak, and part of the reason is that 13 different utilities report the data under different headings.
() 14 For example, if you have a main steam isolation valve, you may 15 never see an entry for solenoid valve.
16 On the other hand, if you put in solenoid valves, 1
5 17 you may never get certain MSIV failut'es which have been caused 18 by them, and it is sonething where there is no uniformity in 19 the plant reporting of the events for the NPRDS data base as 20 Well as the LER data base, and we have also been observing 21 very carefully many of the inspection reports from the plants.
22 i basically have been going through all the data of all the 23 5072 reports, all the regional morning reports for every 24 single day as well as being channeled information from all
() 25 inspection reports that Acon reviews which have anything to do HERITAGE REPORTING CORPORATION -- (202)628-4888 m- - -_ . _ _ _ _ _ _ _ _ _ __ __________m_ _ _ _ _ _ _ _ _ _ _ _ _ _
299 1 with solenoid valve performance, and it has been something we 2 are--the pace has been picking up, and we are trying to try 3 and get some kind of order and understanding of the reasons 4 behind it, whether they be individual failures or a generic 5 set of failures. I mean just, just among the things that we 6 have been doing is I accompanied a Vender Inspection Branch 7 visit, team up to visit the ASCO manufacturing facility up in 8 New Jersey. We spent about a week there, and we found quite 9 quite a bit of interesting things.
10 In particular, I was going over the ASCO internal 11 reports indicating their follow-up and analysis of the 12 failures that had occurred where users had returned the valves 13 to them or they had also internal produe ion problems. In f~
(,) 14 fact, about a week or so ago, I guess spurred on by some 15 failures that had occurred out in the Kewaunee plant coupled 16 with some discussions that we had with ASCO when we were up 17 there, ASCO determined a failure mechanism, that there was a 18 failure mechanism that was introduced in the manufacturing 19 process on their 8314 valves, and they filed a Part 21 report 20 on that about a week or so ago where a particular lubricant 21 was used in the manufacturing process which should not have 22 been specifi..d in the manufacturing process, where the people 23 who had been doing the work sort of knew that according to the 24 vender information, this particular lubricant would dissipate
[) 25 in time, and ASCO was also cleaning cut the valve internals so Q]
HERITAGE REPORTING CORPORATION -- (202)628-4888
300
_ 1 there was a thought that this material that was introduced g') 2 would not cause any harm.
V 3 Well, subsequently it has been found that this 4 lubricant which I believed is called P-80 which was like a 5 water-based rubber lubricant remained within the crevices, 6 could not get cleaned out by the cleaning operations and 7 subsequently on high temperature wound up sticking or actually 8 becoming a sticky substance which prevented the 8314 valves 9 from operating properly.
10 And as I understand, I'm not sure whether this 11 information came from the Part 21 report or subsequent 12 discussions I have had with the head of valve engineering at 13 Asco, but this particular is, I think they have about 500
() 14 units, which is fortunate because of the fact there is a total 15 of about 75,000 ASCOs out there in use, so it reprcsents a 16 small part of their product line, and as I say, this kind of 17 gremlin creeping into the production is the kind of thing that 18 we had been looking at, and there are a couple of other areas 19 within their production that we have had some dialogue with 20 them that they have not totally resolved yet.
21 I plan to visit some of the other larger 22 manufacturers of these kinds of valves to get a better 23 understanding from them as well on the wherewithall of the 24 ualves, the margin they have, forces that they can tolerate,
() 25 ':he clearances they can tolerate, the forgivinginess or HERITAGE REPORTING CORPORATION -- (202)628-4888
301 1 unforgivingness of the design relative to the lubricants that 3 2 may be used and other contaminants which may be coming by.
3 This particular study, as I said, started out being very 4 small, and narrow, and it has been branching out.
5 We recognize the fact that a lot of work has been .
6 done previously by the NRC Research, Franklin Institute. They l 7 had published some reports, and there is a lot of good data l 8 there. I 9 However, what we are doing seems to be based on
- 10 that. It is an update on it. We have been getting many 11 events in just recent weeks that we are following. ,
12 Right now I guess a few days ago there was a failure ;
L 13 of a main steam isolation valve at Hillstone 3 because of a
() 14 solenoid valve on that malfunctioning. It is an interesting I
15 event. However, checking back, it does not appear to have a 16 widespread, it may not have a widespread significance within 17 the overall U.S. industry since it is a valve made by a 18 foreign manufacturer, and we do not believe it hss much 19 application states in it, but anyway, we would like to find 20 out the root cause and see if there are similarities. i 21 MR. CARROLL: One question--you indicated that NPRDS 22 is not a, at least a foolproof source of data on these l 23 failures. ,
1 i l
l 24 HR. ORNSTEIN: I don't think I used those words.
l
() 25 What I said is the data is there, but you have got to be b r HERITAGE REPORTING CORPORA? ION -- (202)628-4888
~_
302 1 creative to try and pull it out of it.
{} 2 HR. CARROLL: Okay. So in other words, with your 3 MSIV example, my impression was that you could go in and query 4 the NPRDS data base on solenoid valves, as opposed to HSIVs, 5 is that po+ true?
6 HR. ORNSTEIN: You could get stuff, but you won't 7 get it all, you know. For example--
8 HR. CARROLL: In principle, NPRD3 has every failure 9 that has occurred in the industry on solenoid valves, is that 10 right?
11 HR. ORNSTEIN: Well, I'm not sure about that, but 12 assuming that that is correct, okay, the thing that you will 1
' 13 find is that people doing the reporting for their division
( 14 will look at Atwood-Horrell 'ralve which has an AVCO soler.oid 15 actuator in there and they will report it. and the way you 16 will pick it out of it is Atwood-Horrell, you will not pick it
- 17 up on main steam isolation valve and you trill not pick it up u 18 under AVCO and you will not pick it up under solenoid valve.
j 19 The only way that a particular failure that I found was found 20 was just by the name, manufacturar of the big valve, not of 21 the small driving part of it.
22 HR. CARROLL: Somebody isn't reporting right then.
l 23 HR. ORNSTEIN: Well, don't say what is right. I 24 mean the thing is, there are numerous field widths and the
() 25 question is are you reporting about a system called an MSIV or HERITAGE REPORTING CORPORATION -- (202)623-4888
]
303 1 are you talking about a component within the system that is,
< 2 that has failed? In other words, if I have an Oldsmobile that N]J 3 fails, and it really failed because of a lead gasoline filter 4 in the line, it can be reported as a car failure, where it did 5 not have adequate gas. You may never find out that it was a 6 lead filter until you go ahead and reach me as opposed to what 7 I plugged into the data base.
8 HR. CARROLL: Well, that certainly wasn't the theory 9 of that data base when it started. Have you--
10 HR. ORNSTEIN: I can't profess to that. All I know 11 is I try to use it, and in order to use it, and get useful 12 information, it takes more than just running through it once.
13 In fact, very often what we will have to do is know of a
() 14 particular failure which was not able to get pulled out of the 15 aystem, and then do an in-depth study on what, the identifiers 16 on that system to once you actually get the failure. This is f
17 somathing that we have done over in AEOD for quite a while.
18 It gets tricky, and then after about three or four times, you 19 can very often get a strategy with several runs that you hope 20 will get you all the data, but then again, we are not 21 absolutely certain with regard to completeness.
22 HR. CARROLL: Have you discussed this with the ;
i 23 responsible people down at INPO, the problems you have j t f 24 encountered in this particular--
() 25 HR. ORNSTEIN: I personally have not. However, I HERITAGE REPORTING CORPORATION -- (202)628-4888
304 1 have--
2 HR. CARROLL: Would you do that? Because I think r' }
3 this is something they ought to know that. The data base 4 ought to be useful for the kind of case study you are doing, 5 and you are telling me it isn't. !
- 6 HR. ORNSTEIN
- Well, I am not telling you it is not ,
7 useful. I am telling you it is useful, but you havc got to 8 pressure and tweak the system in many different ways to figure 9 out how to get the stuff out of it.
10 HR. CARROLL: I will change my word useful to it t
]
, 11 should be foolproof, i
4 12 HR. ORNSTEIN: For every foolproof system, there is
! 13 always a fool that is going to be there. I can't agree.
14 CHAIRMAN HICHELSON: That is not the only problem f a
15 with the system. ;
- 16 HR. CARROLL
- I know. There is a lot of probleras, t
2
- 17 but it isn't going to get better unless peopf.e start feeding i
18 back to IN2O the frustration they are experiencing.
- F 19 HR. ORNSTEIN: Well, if we--I think this is l
20 digressing from the original intent here, but Int ne give a 21 couple of quick things to try and give you a better fool for 22 it.
I 23 HR. CARROLL: I understand. I have had the same one '
24 when I have tried to use the data base, but it isn't going to i i
() 25 get better unless users feed back to the people that run it !
HERITAGE REPORTING CORPORATION -- (202)628-4888
305 1 the problems they are having.
3 2 HR. ORNSTEIN: Earlier this year I was given a fast J
3 and dirty job to come up with some information on solenoid 4 valves, on failure rates, manufacturers, and whether or not 5 this particular manufacturer has a bigger problem than the 6 other guys. If you read the research report, it says that 7 they are all abouc equal. If you talk to people of Franklin 8 Institute who have done the work, they are saying no, the guy 6 9 who you think is the worst is really the best. That's not the !
10 point.
11 The point that I am trying to make is that when I 12 try to get the information which would be able to answer the 13 question on counts and numbers and this one versus that one, I
() 14 wound up with a problem that I have just described. I went 15 back to my management, who in turn went back to the other part 16 of AEOD which is the interfacing with the people who work the ,
1 i
{ 17 data bases, and one of the things that they say at that point 18 in time is dell, maybe you ought to go ahead and write a quick :
l 19 and dirty technical review, indicate that this is a problem.
20 Well, this is something that I said sure. However, !
i 21 Rome started to burn at different plants. We were finding i
22 ma'jor problems, and rather than worrying about putting 23 together something about this particular as aspect of the data 24 acquisition, I am chasing the data itself and the problems
() 25 that it is with. However, it is not something that ir
{
I 1
l HERITAGE REPORTING CORPORATION -- (202)628-4888 l l
s.
306 1 forgotten about. The people within our organization who do 2 work with INPO are aware of it, and I believe that this kind i 3 of thing, dialogue is.indeed going to happen. However, I am 4 not the person who does it. I spotted the problem. I have 5 found out how to work with it and try and get useful into from [
il 6 it, and that that's basically where that part is.
7 HR. CARROLL: Would you confirm that this dialogue 8 is going on with INPO and get back to Al? Because I would i 9 like to know'. I uean the industry is spending an awful lot of ,
10 time and effort on an NPRDS, and if we are not getting the j 11 results out of it, we ought to do something about the about 4
12 the problem instead of living with it.
13 MR. ORNSTEIN: I will pass the word to my 14 management, and I'm sure this 13 something-- !
15 CHAIRMAN HICHELSON: We h6ve a committee that deals f 16 with what they do over there, and we might approach the 17 subcommittee chairman and ask. There are several other, l
) 18 several other things that I could easily add to the agenda if 19 they ever had a meeting because NPRDS, of course we haven't
! 20 visited for a long time, and it was, it needs some revisiting i
i 21 to see if it is working because at the time it was set up as 22 an industry initiative in lieu of the regulatory requirement, 23 there was agreement within two years we were going to review l
! I 24 the situation, and if it wasn't working, then we would proceed j l
l l
() 25 to take reculatory steps. I don't know if that occurred or l
i I
HERITAGE REPORTING CORPORATION -- (202)628-4888
307 1 where it is at or whatever, but one of these days we ought to (3 2 revisit it--very simplistic, and you just plug it in.
\_)
3 CHAIRMAN HICHELSON: It is a vcluntary system.
4 HR. ORNSTEIN: It is too complicated. The more 5 sophisticated it becomes, the more difficult it is to work 6 with it and you have got to go ahead and use some ingenuity .
7 and understand it. If it was a go, no go gauge, you know, 8 there in very little doubt in your mind we think, but this is 9 a little bit trickier. You have to try and use the tools 10 rather than have the tools use you.
11 HR. CARROLL: I would have the same frustration you 12 are describing, and in trying to use the data base, and having 13 the people that are very knowledgeable about it to help me, ,
]
i
() 14 and every time we have an examp'e of something that isn't
- 15 worr'ng well, NPRDS, those guys down in the lab ought to know 16 about it and ought to be doing something to fix it.
17 UR, ORN3TEIN' I think it is an understEndable kind :
t 19 of situation, but anyway, let me go en to what I think it why l 19 I'm here today.
20 The thing that is most important, as I think I have 21 touched on a little bit, is the fact that we have widespread 22 use of these components, and on some of the failuree that we s 23 have seen or can see, we came close to having, you have the 24 very strong possibility of failing multiple components or
() 25 multiple trains of safety systems, and these particular kinds l HERITAGE REPORTING CORPORATION -- (202)628-4888 ;
4 308 i 1 of events are not taken care of in the plant safety analysis.
2, As far as what is causing the failures, here is just ,
3 a partial list. Each of those, each of those lines will ,
4 appear I guess as a section of the case study, widespread i 5 types of design deficiencies, maintenance deficiencies.
6 Hisapplications, we have different types of misapplications as 7 far as failing open, failing closed, kinds of environment that 8 you have. I have that on the next line as well, and it is 9 just, you know, keeps ongoing.,
10 One particular kind of failure I think Carl was t 11 alluding to earlier today, was the differential pressure that 12 these components have to operate with, and recently, the NRC 13 put out an information notice on such a problem that was 4
() 14 recognized at Kewauneo and at Calvert Cliffs where there were dozens of componento and different safety systems which could i 15 l 16 have malfunctioned because the air pressure regulator in which l
. i i 17 case Calvert Clif f s was not necessarily qualified for the l 18 environment where if it had tai'.of on the high side, it could [
- l 19 have wound up with a malfunction of the controllers, which in i
20 turn would cause a failure of safety equipment. !
l !
21 That is sort of coupled with the design deficiency.
22 How do you start to say that such a problem is not a design 23 deficiency from the standpoint of when the system was j
. 24 originally put in? Did the designer take into account the
() 25 regulator could fail and the pressure could be high? Should i
l
, HERITAGE REPORTING CORPORATION -- (202)628-4888
309 1 the solution have been a higher differential pressure ,
! E' 2 operating valve, or should it have been a safety grade i 3 regulator? I mean this is a kind of thing where you start to 4 go into an area where it is difficult to adequately define j 5 something to everyone's satisfaction, but these are the kinde ;
6 of things we are finding. ,
i 7 Contaminants are very interesting. They are, of l
course, are contaminants in the air stream which hopefully the i 8
i
- 9 work we have done in the area of air systems, and future l t
10 actions that we are going to be taking, a lot of the component ,
, 11 problems that have been observed can be negated. !
I 12 In fact, to my recollection, on a data sort of air l
13 system problems, about 70 percent of the reported licensee j I
i
() 14 evenc reports on air involved contaminants doing the thing. ;
l 15 In additton to contaminants in the hydraulic stream i t
16 or the pneumatic stream, we also have contaminants that might {
17 be introduced in the production stage. We are also finding a l l
18 certain licerceo has chosen certain lubricants which sort of i i 19 surpriso us in their capability of performing properly.
1 20 That's another contaminant type thing that we have been .
i 21 looking at, but as I cay. this particular assignment which [
22 started out small has been growing, and we are trying to put a (
23 big box around i- - 2 package it up with useful information. l 24 We are still executing and sorting the data, f
() 25 Very quickly, I ran some NPRDS runs back in January l
HERITAGE REPORTING CORPORATION -- (202)628-4888
310 1
1 1 of this year, which went back to times zero, and I think I had 2 oh, maybe about 500, 700 failures with the final strategies 3 that I used. I think we did an update between January and 4 June and came Jp with about 177 more. I'm now doing a final 5 update betwes 2 June and November 1 and that will be the time 6 line I think on thitt case st'dr, i unless there is an 7 extraordinary failure that is worthwhile mentioning, but what 8 I am trying to indicate is had it been failures as such or 9 reporting as such, the pace is rather high, and it is 10 something where we have to get something out to the people to 11 understand more about the kinds of problems that have been 12 occurring at other plants. There is a problem with feedback 13 of information on the experiences that people have had, and l
t~h
\_/ 14 hopefully we will be able to get the root cause of many of 15 these failures, and get people alerted to ic, and at that
! 16 point, someone will have to nahe a decision where to go from 17 there, but our job is to try and pull out the data. and nahe 14 heads or tails out of it.
- 19 CHAIRMAN MICHELSON
- In the case of regulated power i
I 20 supplies, which you did mention, there the problem, of course, 21 is I guess failure of the regulator to function properly or i
- 22 somebody adjusted the regulator improperly, but another 23 closely related situation is where you are operating directly i
24 off of the air system, but the air system pressures range over
]
() 25 quite a wide range.
HERITAGE REPORTING CORPORATION -- (202)628-4888
311 1 Have you looked, for instance, at the effect of very 2 high air pressure which you might < jet by failure of the q{ )
3 pressure contrcl syJtem on the comp'ressors or whatever?
4 HR. ORNSTEIN: We talked about that. You and I 5 talked about that privately. Since then, I really haven't 6 done much, i 7 I have seen some failures on the diesel generator
}
8 starting system which is about double the pressure of the
{
9 normal plant pressures. You have several in containment 10 independent ~ systems that are high also.
11 l' haven't really, you know, spotted many failures. .[
12 You can do a lot of conjecture, but the actual data that we l
13 have-- .
( 14 CHAIRMAN HICHELSON: I was really curious to see 15 what the vender might have thought would be the, how they 16 would affect operability if ,vou elevated the pressure on the !
17 solenoid set beyond its normai design. 1 i
18 HR. ORl1 STEIN: Basically the one area, as I said, is 19 the naximum operating pressure differential, and we have seen l 20 many cases, I mean the number doesn't have to be 200, 500, f f
21 600. It could be, there have been many examples going back to i
22 about 1981 or so I think. VEPCO plants have found that they j 23 had like a 30 nound differential solenoid in a stream where 24 the operating pressuve was more like 70, and they recognized l
() 25 it as a problam after they had a failure, and I mean that was I
l HERITAGE REPORTING CORPORATION -- (202)628-4898 l
312 1 in an LER, so I think I spotted about half a dozen cases where
, 2 they actually did have failures from small overpressures
{
3 across the valves. They didn't have to get up to three, five, 4
4 seven hundred pounds. We are talking a 30 pound differential 5 valve failing at 40 and 50, so--
6 CHAIRMAN HICHELSON: There have been a few LERs 7 indicating that this has been a problem.
8 HR. ORNSTEIN: Recently, yes, but what I am saying, 9 many of us were unaware that buried in the LER data base there 10 are s lot of other events, and wo are in the process now of 11 trying to build our own little data base of these events that 12 rave been found in the data base, and try to sort them 13 ourselves to get a better feel for the statistical 14 significance of the failures that we are finding.
15 I do want to also mention something else, which is 1 16 think helpful for Mr. Carroll. t 17 One of the things things that I have spotted was in f 18 the NPRDS data base. I found a lot of Fisher controllers that 19 had failed, and the numbers that had been ent> red in these 20 failure records were num'its that matched ASCO valves, and low 21 and behold, wnen talking to people at ASCO, I found out that i
23 Fisher controllers is oae of the largest purchasers of ASCO 23 valves, so now where do you call it? Fisher-ASCO failure?
24 CHAIRMAN MICHELSON: Have "ou looked at the airsets
() 25 that are used to control piston-driven oporators say on HERITAG. .tEPORTING CORPORATION -- (202)628-4888
gy ,
\'
313 i I containment isolation valves? :
2 HR. ORNSTEIN: We have been picking up a lot of 3 those, yes.
4 CHAIRMAN HICHELSON: One of the interesting things l i
5 that most people I think have to do on containment isolation, }
I 6 they--of course, it must close quickly, but not so quickly as 7 to damage the valves, so they have to meter off the air from 8 the airset in a controlled manner in order to keep the valve 9 from closing too quickly. They have to keep the air cushion l 10 on it, and to do that, many people use needle valves, which ;
- 11 have extremaly fine ports as you get toward th1 full closed l' 12 position, and they are very susceptible to ninute contaminants 13 in the air stream which do not affect the solenoid valves at ;
() 14 all, so in looking at solenoid valve, one must also see what k 15 kind of an oriface do you put downstream when you bleed off ;
r 16 the solenoid, the air off the solenoid, to see--I consider 1 i 17 that a part of the solenoid valve problem in a way, and so you 18 need to look at those needle valves which is not an uncommon f
f 19 oractice. l 5
1 9 i 20 HR. ORNSTUIN: Okay. Needle valves, I/P converters t l
21 you have got in many cases, the manufacturer wi),' recommend or t
i
- 22 actually provide some kind of filtration on the solenoid j
- 23 valves. We see a very-- l 24 CHAIRMAN HICHELSON
- Needle valves, they don't put
() 25 any filtration in. Those are right off the port of the air i
! l 95RITAGE REPORTING CORPORATION -- (202)628-4888
314 1 system.
f 2 MR. ORNSTF.IN: But you see, you have to go back to ;
)
3 yesterday's discussion. The air systems, we have had generic t
4 letter 8814 I believe, that said see the compatibility between !
i 5 your air system and the components that it is serving. If you i 6 have got a needle valve, and it can't tolerate anything more 7 than one micron, you sure as heck should be certain that when 8 you are done with your review of your air system, you have got 9 something up in that needle valve to prevent it from seeing ,
r 10 anything more than one micron.
11 You know, again, we can't solve all the problems in 12 any one area. liowever, the big picture of the air system and ,
l
- 12 its compatibility with the components that are on it, or [
, () 14 vice-versa, is crucial. When you get to an MSIV air pack, 15 where you may have four different solenoid valves of, maybe l
16 six or eight even of different manufacturers, different t i
17 tolerances, you find that because of kind of the thing you are 18 talking about, slow closure, fast closure, damping, the speed (
f 19 of closure, you get the different valves port to the other i
20 valves, and one of the problems is on the back stroke, you can (
- t 21 get air from the containment coming into the solenoid valves, f
22 and if the solenoid valve of interest is not able to tolerate f 4 2J the particulates that are coming in, you should have some kind I L
24 of a filter, you know, on the discharge of the valve which L
() 25 actually works as an intake. I mean it is very complicated, f
l HERITAGE REPORTING CORPORATION -- (202)628-4888 j
l-315 i
1 and in some cases, for equipment qualification, the ,
- -~' 2 manufacturers have indicated that you have to have.the exhaust ,
t 3 port pointed in a certain direction so you don't get incoming i 4 spray which can also do a number on some of these valves.
5 Part of the problem that I was trying to attack is i
6 to understand what the threshold if pain is for some of these 7 solenoid valves. Like when I was up at the second Perry, Ohio 8 where one of the solenoid valves had failed, we were not 9 surprised that we found there were three small pieces of an 0 10 ring which didn't even amount to like a 16th of an inch, had l
14 caused the malfuncton, and it just was a simple situation 12 there where the frictional load upon the plung'tr was so big I
l 13 because the tolerances were so small, and that gets you into l
() 14 another big question--do you want to have a good technique to 15 rebuild the valve when it has gotten to the end of its service
(
16 life, or do you want to throw it away and get a new one?
17 Based on the failure experiences that we have 18 observed there is good reason why you may wa;t to discontinue 19 purchasing rebuilt kits. In fact, when you go to the 20 manufacturer of the solenoid valves, you will find he has six 21 different tests and certain test rigs to check out each valve 22 as he assembles it to make sure it is working properly whereas 23 the utility will have one crite ia test. Does it closa or 24 doesn't it?
() 25 Now it may close, but he may be on the ragged edge HERITAGE REPORTING CORPORATION -- (202)628-4888
316 1 from other espects of the valve safety. This is the kind of 2 thing that we expect to address in the case study as well. I
[
3 mean it is too early for me to come up with some, you know,.
4 recommendations. I could come up with some gross 5 recommendations, but what we are trying to do is to get to the 6 root cause of the problems and come up with what we think will 7 be important from the safety significance standpoint, and that 8 is reliable operating valves.
9 CHAIRMAN MICHELSON: In the last few minutes that we 10 have, can you tell us what you are doing, if anything, in 11 other air-operated type valves, just what kind of case studies 12 or other things?
13 HR. ORNSTEIN: I myself am not. This is something, 14 tAe solenoid portion has been elevated.
That's the focus of your effort i
, 15 CHAIRMAN HICHELSON:
l l 16 at the present time? You haven'e looked at the piston j r
17 operators on the containment isolation valves? l t
i 18 HR. ORNSTEIN: Not to the extent that the people t
! 19 have thus far. We started out with a couple of events where I l 20 was at Perry on two occasions. I was over at the, I guess it [
t r
! 21 was Dresden which had a different interesting event, and t
- 22 Brunswick failures as well.
I 23 CHAIRMAN HICHELSON: One of the potential l i
~
24 shortcomings of some of AEOD's approaches, you are event !
f
() 25 driven, and some of these problems won't, ycu won't see events f I i HERITAGE REPORTING CORPORATION -- (202)628-4888
1 317 i 1 until the real thing, and it is a little late.
\
2 HR. ORNSTEIN: We just have so'many people and so 3 much work you can do. You can just bore in on so many at t 4 once.
5 CHAIRHAH HlfaSON: Dat euch of the work is event j 6 driven, which sure, it f'19 vLe ld be event driven, but ,
7 perhaps occasionally we end lock as some of these areas f 8 where we weren't expecti o ?,o , rent driven but may have 9 reason for concern, and I s a't kr < how they handle those. I 10 guess they don't get very high on the priority list.
11 HR. ORUSTEI'rt: Well, Carl, you are the guy who got t
12 me going on air systems problems back in 1980, '81, and the !
5 13 event that got us going on that was a San Onofre loss of salt
() 14 water cooling system, and it is that type of thing. You have' t
15 just kept on accumulating data, and until we had enough events [
16 and like we tried many times to show people how important and f i
17 significant it was, but until we had a real hard core event, I 18 we weren't getting too much sympathy and we weren't getting j
{
19 too much action. [
20 The same thing when you had the thing with Asiatic !
(
21 clamps. You knew that they could be a problem, but it wasn't (
22 until they showed up in Browns Ferry that you lot loose one of 23 the guys to put together something and it was able to do it. f i
24 Save thing with SCRAM discharge volume--if you don't have an ;
i
() 25 event, it is very difficult to explain to people that it is an t
i HERITAGE REPORTING CORPORATION -- (202)628-4888 f i
r 318 c
1 important thing outlthere i
- 2. CHAIRMAN HICHELSON: I just agonize over the fact 3 that apparently you don't always have to be even' driven.
L 4 Are there any other questions befote Hal leaves? We -
i 5 do expect to get him back later on and I in, I think January r
6 if you haven't been told, we are going to have an auxiliary :
?
7 system subcommittee meeting on air systems. It was supposed 8 to be a part of this originally, but there wasn't the time, so i.
. 9 we moved it to January. ,
10 MR. ORNSTEIN: Sure, f I !
.: 11 CHAIRHAN HICHCLSON: And that way we will handle the i
I 12 other end of the problem--hew do we control pressures and 4
13 quality and all that of the air system? What kind of problems ;
]
() 14 does it generate?
- . 15 MR. ORNSTEIN
- Hopefully by then we will have gotten 16 more feedback from what the utilities have actually done in i 17 the area that was highlighted in the generic letter that (
4 18 recently went out. l 1
19 CHAIRMAN HICHELSON: That is tentatively scheduisd i
20 for January--always subject to change, of course, j i
21 HR. WOHLD: I would just like to ask him a question.
I j 22 Are you seeing any need for periodic preventive or predictive {
{
i 23 maintenance or replacements that is popping out of your study? l
, i I
24 I read the NUREG they put out which I think was vary good, but
() 25 the conclusion was, the conclusion that you drew in that NUREG I ,
l i
HERITAGE REPORTING CORPORATION -- (202)628-4888 I
319 ;
1- was that there really didn't appear to be anything at that 2 time that perhaps as a failure occurred, analyzed that for 3 root cause and then proceed from thero.
e
'4 I just wondered if there is anything else that you 5 discovered since, in a periodic maintenance program that might 6 be worthwhile? :
7 HR. ORNSTEIN: I'm getting a little bit confused.
8 You are talking maintenance with regard to?
9 HR. WOHLD: The solenoids. !
10 Mk. ORNSTEIN: Now I was not involved with the !
11 solenoid NUREG. There may be a little bit of confusion here.
12 MR. WOHLD: Maybe I have the wrong paper. You have 13 ptnt out something that dincussed-- t
( 14 HR, ORNSTEIN: As far as solenoid valves go, we have [
l 15 seen many plants have taken actions to try and recover from 16 repef;itive failures and cne particular actually--not in one, l
17 in many plants, what has seemingly worked is combination of [
- l 18 the, combination of changing out certain components within the !
f 4
19 solenoids.
I 20 For example, if yotl have something that is attacked j i
21 by hydrocarbons in the air, of course it is wonderful to clean l
- t 22 up your air system. However, that doesn't get rid of i l i 23 everything that is already in there from previous crimes, but i
24 on the other hand, what they have found is to, a solution is j i
r
() 25 to use materials which may be less susceptible to the f l !
! HERITAGE REPORTING CORPORATION -- (202)628-4888 l i
320 ,
\
1 contaminant or irritant, and more important than that, perform j 2 more frequent cycling or operation to prevent the sticking
)
3 kind of problem that has occurred.
4 I have got problems with certain aspects of the ;
5 environmental testing or equipment qualification aspect of 6 solenoid valves. Where you look at them, they cycle them 7 every day as opposed to a real life plant where some of these 8 valves may go for weeks and months on end without--you know, t
)
9 the cycling of the valves is important. The recognition of (
i 10 what, the life expectancy of the valve is important, but even <
l 11 more important than that is how you go ahead and service it, F 1
I 12 and you know, sometimes you may be better eff leaving the }
13 thing there beyond what you thought was okay as opposed to ,
( 14 having some person rebuild the valve and make a mistake'on it.
4 i 15 There is a lot of stuff out there. It is too early to put ;
16 together any firm recommendations, but it is something we are j 17 looking at. !
I 18 HR. WOHLD: Okay. Thank you. !
19 CHAIRHAN MICHELSON: Any other questions? If not, ;
i I i 20 we will take a 15 minute break, about 12 minute break, end be ;
21 back at 12:30. l I
l 22 (A brief recess was taken.)
f 23 CHAIKHAN HICHELSON: The meeting will reconvene. We i
24 are now going to shift gears and proceed to the portion of the j 4
() 25 program dealing with the resolution of generic issue 2 E i
4 >
HEkITAGE REPORTING OL... ORATION -- (202)628-4888
l 321 1
1 whatever.
HR. ROTHBERG:
2 61.
3 MR. MICHELSON: I do want to make a couple of 4 remarks concerning this matter. Thc original planning for the ,
i 5 meeting, it was to review the staff's final generic letter and 6 the industry's comments concerning this letter.
7 It is now my understanding that the final generic 8 letter is not yet ready. Therefore, we will have to have ;
9 another meeting on this just before the December Full j i
10 Committee meeting, and--
l 11 MR. BAER: December or November?
12 CHAIRMAN HICHELSON: It will be now December. You !
13 didn't get it to is when you-- [
() 14 HR. BAER: At the last meeting you said he prepared r
15 to discuss the changes we intend to make, and we can discuss [
i 16 those changes.
17 HR. MICHELSON: We would like to raad them, too. I f
18 CHAIRHAN HICHELSON: Discuss means get it to us j 19 before the meeting, and then we discuss them at the meeting. !
i 20 HR. BAER: The earliest we could meet with the !
21 industry was last week, and since we got comments from CRGR, 22 when we get it edited to try and accommodate all the comments. [
l 23 we would like to give them a chance to review it, also make {
l 24 sure it accommodates their--
() 25 CHAIRMAN HICHELSON: The indartry hasn't seen it t
i HERITAGE REPORTING CORPORATION -- (202)628-4888 !
l _ -- --- - _ _ _ _ - - - - - - - - - - _ - _ _
l 322-t h
1 yet, either,'and their comments may be different, depending on
( 2 what this final document looks like as well as our comments 3 might be different, so the earliest we can write a letter on -
4 it is the December Full Committee meeting. And therefore, l 5 what we would like to do is on December 13th, we would like to l 6 have a short subcommittee meeting in whi;h you would come in L
7 just to tell us what, and to be prepared to answer questions i 0 on what the final version looks like. t 9 HR. BAER: Why can't that be done in November? j 10 CHAIRMAN HICHELSON: Because there is, there isn't i 11 any other date set up. It is too late to have another meeting l l
12 noticed and so forth. ;
r I
13 HR. HINNERS: In November? It is too late to notice 14 a meeting for November?
I 15 HR. IGNE: November is, it is coming up within a 16 couple of weeks, week after next. ,
t 17 HR. HINNERS: Are we talking about a Full Committee f
18 meeting or talking about subcommittee?
19 CHAIRMAN HICHELSON: We are talking about a f
[
20 subcommittee moeting. We have to have an opportunity to sit i 21 down with you and discuss whatever you are coming up with, !
i 22 which we hadn't seen, and which we won't see today, j 23 HR. MINNERS: Why can't that be done in November?
I 24 CHAIRMAN HICHELSON: Simply because there is no !
I i
4
() 25 other dates available. All the-- l,
! I i
. HERITAGE REPORTING CORPORATION -- (202)628-4888 !
323 1 MR. HINNERS: Is that a restriction on your schedule-2 or our schedule?
3 CHAIRMAN HICHELSON: Yes. You have to now get in 4 the next available slot. There in nothing--ABWR is two days, '
5 subcommittee before the Full Committee; that takes that week. .
6 It is only next week you are talking aheut. And you can't !
7 notice a meeting, and give everybody a fair chance in a couple i
8 of days. [
9 HR. BAER: Generic letters are not generally sent I i
10 out for public comment.
11 CHAIRMAN HICHELSON: No, no. We are not saying they r 12 are. .
f 13 HR. BAER: But that's exactly what has happened in {
) 14 this case, and which I have gotten a fair amount of ,
i 15 aggravation about. j i
16 CHAIRHAN HICHELSON: The subcommittee can ask for i
{
17 public comment on something if they wish, and that's what we {
13 are doing, and that's what we are going to get today, and when f
19 you give us the document-- !
f 20 HR. BAER: We believe we have gotten all the f i
21 comments, and the question is trying to rewrite it to !
f 22 accommodate the comments. [
23 CHAIRMAN HICHELSON: Yes. f I
24 HR. BAER: I can see an endless process. What f i
() 25 happens in December if someone has comments? Do we then [
HERITAGE REPORTING CORPORATION -- (202)628-4888 i
324 l 1 rewrite it and get back on like a February schedule? j 2 CHAIRMAN HICHELSON: As a genera) rule, we get the 3 documents that you wish to have us comment on before we do the !
- 4 commenting. . We cannot write a letter without seeing your 5 document first. };
E 6 HR. BAER: Well, we have a complete package. We got j i
7 comments from you, from industry, from CRGR. We are trying to !
8 accommodate-- I I
k i 9 CHAIRMAN HICHELSON: If you wish, we will comment on l
. i
^
10 what you have already given us and you may have fixed our l s.
I I comments or not, but this is also an endless process to send 11 l I I I
12 to t,,4 Commission a document which does not reflect what they i i !
- 13 finally see. l 1 i f
i
() 14 HR. MINNERS: The Commission is not going to seo t
! 15 this document anyway. :
i I
16 HR. HICHELSON: I don't know that. If they got our !
The industry has the I 17 letter, they may. I don't know.
18 prerogative of complaining to the commissioners also, and then [
I
- 19 they will certainly read your document. Well, I suspect it j I f 20 won't go unnoticed. j 21 HR. CARROLL
- I think the position is if they can I 22 get us the final verse, maybe we can have a conference call or ;
) 23 something. [
24 CHAIRMAN HICHELSON: Depends also on what it is. j
) () 25 HR. CARROLL: But if it is-- f l
! {
t-HERITAGE REPORTING CORPORATION -- (2021628-4888 i
1 i t 1 325 ,
1 CHAIRMAN HICHELSON: When is this going to take ;
2 place?
3 HR. CARROLL: I am saying, that could very likely I 4 happen before the November meeting. If they can get the-- !
5 HR. ROTHBERG When is the November meeting? !
6 HR. CARROLL: 17th through 19th.
7 HR. IGNE: What Carl is saying, his schedule-- ;
1 l 8 CHAIRHAN HICHELSON: We have got other things. -
l 9 HR. IGNE: Won't acccmmodate it. l i
10 HR. MINNERS: It is not our problem of not getting I
! I l 11 the material in. It is your scheduling problem? [
l l 12 CHAIRHAN HICHELSON: It is your problem. The i l
l l
13 agreement with the staff is you always get the material to us, (
( 14 you are supposed to get it to us three weeks before we write a l
15 letter, but obviously-- [
l !
l 16 HR. MINNERS: I guess my question is we are looking ;
17 for a--when is the December Full Committee? j 18 HR. ROTHBERG: Thirteenth of December.
19 CHAIRMAN HICHELSON: November Full Committee is I }
20 think 6, 7--17, 18, 19. !
l t I
21 HR. CARROLL: Fi:f teen through 17. l I
22 HR. BAER: When was December again?
! 23 HR. CARROLL: Fifteenth through 17th. (
I 24 CHAIRMAN HICHF.LSON: It is pretty hard for you to do ;
() 25 un .] tor November in terms of getting a document to us. i l l HERITAGE REPORTING CORPORATION -- (2021628-4838 :
o _ _ _ . - . _ . _ _ _ - _ _ _ _ _ . _ - - _ _ . _ -
326
- 1 HR MINNERS: If we get it to you by November 1st, I i
2 mean that is three weeks, so do we meet your rules if we do
, 3 that? Or do you still have a schedule problem?
4 4 CHAIRMAN HICHELSON: We have to schedule a ,
5 subcommittee, but the rule is you have to get the information ,
i 6 to us three weel:s before we schedule a subcommittee meeting. l 7- HR. MINNERS: I thought it was three weeks before '
. 8 you had a Full Co,'ittee meeting. [
- I 9 CHAIRMAN HICHELSON: I will certainly try to oblige. !
. i 10 It is a question of when you, when can we have a subcommittee j 11 meeting if we need one? And I don't know whether we will need f I
j 12 one or not. We haven't seen your document yet. Now if it is i 13 minor change, fine, but I gather it is so, sufficiently
(
f () 14 extensive you couldn't get it to us for this meeting.
15 HR. BAER: Well, there is some management review f
16 that has to bo done. We just met with industry last week, and f
17 as I said, I definitely want to have CRGR have an opportunity l 1 1 2
18 to, at least to, the chairman to look at it and see if it :
i l 19 would accommodate their comments, but the comments generally l S
20 were to clarify the letter to make sure that work that has :
21 been done either on 85-03 or voluntary extensions of 85-03 j i
j 22 don't have to be repeated, and we thought we said that !
\
! 23 originally, but we apparently didn't say it clearly enough, l I
1 i 24 and we are trying to, to make that absolutely clear. I don't i
() 25 see this as a big technical change, but it is extensive i
l HERITAGE REPORTING CORPORATION -- (202)628-4888
i 327 j i
i rewording, and Owen will discuss some of the things that we I
t 2 intend to do for, or are doing.
3 CHAIRMAN HICHELSON: It is up to the subcommittee i 4 whether or not you want to go without seeing the document or f 5 whether you want to see the-- l 6 HR. CARROLL: I want to see the document, but if .!
i j
7 they have got it out to us by November 1st, maybe in place of (
- i 1 8 a subcommittee meeting we could get together on a conference i 9 call and say yes, it looks like they accommodated our
{
r i
10 comments, and talk about what kind of a letter we ought to l f
11 write and get the letter out. i 12 HR. ROTHBERG: Do you need it to be formally [
I 13 transmitted to you? i 14 CHAIRMAN HICHELSON: It has got to be a document
) 15 that your management has been willing to stand behind, of i
16 cour5n, I don't want to see it before your management says it I a
F 17 is okay. [
?
18 HR. ROTHBERG: Of course not, i i !
- 19 CHAIRMAN HICHELSON
- We are talking about next week [
i 20 you realize, i 21 HR. HINNERS: I don't understand your comment. You !
4 :
I 22 already have seen it before the managenent said it was okay, j 23 That was your request, and we provided it to you.
24 CHAIRMAN HICHELSON: Yes.
l 25 HR. HINNERS: Now you have a different request?
1 HERITAGE REPORTING CORPORATION -- (202)628-4888
328 l
-1 CHAIRHAN HICHELSON: At this stage of, at this stage f f
2 of the game, we would like to see the final document. That's 3 all.
l 4 HR. HINNERS: I am only trying to get the ground l 5 rules straight, Carl. !
6 CHAIRMAN HICHELSON: How we would like to see the r
f.
7 final document before we write a letter from the Full 8 Committee. The subcommittee is quite willing to look at 9 anything. l 10 HR. HINNERS: Is there any use to meeting? Because ;
t 11 we didn't have a clear understanding that you wanted the final !
12 document. I mean it doesn't seem to have auch use of coming ;
i 13 down to this meeting if we don't have a final document. i
() 14 CHAIRHAN HICHELSON: The original thought was that [
15 you would be ready to talk today about any changes you were t I
16 going to make in the final document that we could read. That f L
17 was-- I t
t 18 HR. BAER: The first part was my understanding. At [
19 the time that we met with the subcommittee, you pointed out j r
20 you were going to have another subcommittee today with [
i 21 presentations by industry. NUMARC had asked to meet with us. [
I 22 Ke arranged that for what was it? A week ago Tuesday. ,
L 23 HR. ROTHBERG: The 19th. i t
24 MR. BAER: We got, had that neeting. Owen sat down We try to accommodate their comments, and
(:) 25 and did a rewrite.
HERITAGE REPORTING CORPORATIOP -- (2021628-4888
?
329 i
i 1 CRGR comments plus all subcommittee comments, and it still [
2 needs some more work.
3 CHAIRMAN HICHELSON: Now we are talking about ;
{
4 writing a Full Committee letter on a document and we won't, f
5 generally do not write them on preliminary documents, although i
6 we can, but not generally. l l
r 7 HR. HINNERS: I am not sure we are asking for a l 8 Committee letter. I don't think it is usual that the f 9 Committee writes letters on generic letters. I think we cane 10 down here because you asked us to redo this, and I don't think ;
11 we are asking for a Committee letter. l l
12 CHAIRMAN HICHELSON: We intend to write a letter on I I
13 the resolution of this generic issue. And you know, we would j
() 14 prefer to write it on what we thought was the final 15 resolution, but I haven't seen the final resolution. The !r
?
i 16 resolution is this letter. It's the resolution. It is this i j 17 letter that's solving the generic issue, but I haven't seen i l t 18 this letter yet in its final form. I don't know what it is l f
l 19 going to be. I have seen it in nreliminary and intermediate
! 20 form. We have seen two versions of the thing already, and you l
t 21 are always one version ahead. l L
! 5 l 22 MR. HINNERS: You will never see it in final form. i i t i
l 23 You always have comments on it. It gets changed. You have to !
l 24 repeat it, so you never see it in final form, okay, until it t
() 35 gets published for everybody. That's the way the system HERITAGE REPORTING CORPORATION -- (202)628-4888 (
1 !
_ = - _ - . _ - _ _ _ _ _ _ _ _ _ _ _ _
330 ;
1 works, or else we go around in circles forever, f
2 CHAIRMAN HICHELSON: It is up to the subcommittee to l O 3 decide whether to forward with it to the Full Committee. The v
4 Full Committee very often says you fellows haven't got the 5 document ready yet, we will wait until next month now so you !
6 can get it in place.
7 HR. HINNERS: Okay. In order to get a letter in. t i
I 8 December, when do you require a package? I 9 HR. BAER: December or November? [
10 HR. IGNE: We need it two weeks before the i
11 subcommittee. j 12 HR. ROTHBERG: You want it the 1st of December [
13 about?
l
() 14 HR. BAER: All depends on when the subcommittee 15 meeting is. l t
16 CHAIRHAN HICHELSON: We have already picked the time [
17 that we would be available for a meeting, and it is on the ,
I 18 13th of December, i 19 HR. IGNE: The 1st of December would be the latest. [
t 20 HR HINNERS: Okay. ,
21 HR. CARROLL: That presumes that a meeting is really
[
22 necessary.
23 CHAIRMAN HICHELSON? Yes, If it turns out after we h t
{
24 have see the document thrat no meeting is necessary, we will i j () 25 cancel the meeting.
(
HERITAGE REPORTING CORPORATION -- (202)628-4888 L f
?
L 331 1 HR. CARROLL: And possibly get you a letter in November.
(}- 2 l
3 CHAIRMAN HICHELSON: No, not if we get it on [
I 4 December ist. l q 5 HR. CARROLL: I am saying--did I hear that you felt f
j 6 you could get a quote, final version, to us by early November? l l
7 HR. HINNERS: No, not if you want management [
f 8 approval. :
j 9 HR. BAER: Including CRGR approval. !
r
) 10 HR. MINNERS: Impossible. !
11 CHAIRHAN HICHELSON: CAGR would sometimes go with or !
) I I
12 before or after. That's not critical. We feel that we want
! 13 to review only documents that your management is willing to :
i 14 support. In terms of writing a letter, I don't think you want i
15 tc write a letter on a document-- ,
i
(
, 16 HR. HINNERS: You just reviewed a document that [
I I don't know why you keep 17 didn't have management review. I 1
t i
18 saying that.
f i
19 HR. BAER: They looked at it. It had both NRR and (
l 20 research management concurrence. The changes we are making, f
21 want to make clear. is the response to the subcommittee !
, I
.2 conments, to the industry comments, and to the CRGR comments. (
- 23 And so now you are saying well, okay, do it, go through it I
24 again. We didn't send it down before it had current-- i 1
() 25 CHAIRMAN HICHELSON: Don't confuse che issue. We i !
\ !
HERITAGE REPORTING CORPORATION -- (202)628-4888
332
. 1 are talking now about when do we write a Full Committee l I
2 letter, not when do we have a subcommittee meeting.
]
3 Subcommittees can'roview material for months or years, and 4 generate it, and obviously don't even have management support 5 along the way, but the Full Committee generally does not write [
t 6 letters that the staff hasn't even endorsed themselves yet. !
] 7 It would be I think foolhardy. f a
l 8 Oh, we can, but it would not be well advised, i l
9 especially for resolution of a generic issue. The only reason (
]
l 10 we would write one ahead of time is we didn't like the !
l' i i 11 direction they are going, but if you like the direction they b i
4
- 12 are going, we just want to see what they finally decide to do.
13 Then we can write endorsement or criticism, but we generally f
14 do it on documents which are management approved.
15 HR. CARROLL: That seems to be the heart of the k
- i I
16 problem--your definition or, and their definition of when i
t I 17 their management has endorsed something. [
i 18 CHAIRHAN HICHELSON: If the Full Committee wishes to i
~
19 write a letter anyway, there is there is no problem there, i (
20 HR. CARROLL: We always reference the document we j t
1
.1 are talking about. :
22 HR. HICHELSON: And we have done it, the past on a i l !
l 23 couple off occasions certain letters to the staff, just :
l I 24 changed it--I don't know where this is going. I hope it is (
O 25 o tv imer cw4oce . !'
I
)
d HERITAGE REPORTING CORPORATION -- (202)628-4888 i
] '
l l 333 i
i HR. CARROLL: Where do you think you will be in the 2 process of quote, management approval, early in November?
3 HR. BAER: I think we will have--if you say the lat ;
i 4 of November, which is next Tuesday, we will have, you know, a l 5 clean draft to give to Warren, but if you are saying, and I I
6 would think it would be Warren and our division director, j 7 Wayne Hausman, then it would go to CRGR at least for !
8 their--they have an option whether they want another meeting j 9 or whether they think we have satisfied the requirements. I 10 CHAIRMAN HICHELSON: If they have amther meeting, i I
11 when will it be? Do they meet just whenever there is a need [
I 12 or-- j 13 HR. BAER: Every two weeks, and in fact, I'm sure we C
O 14 can meet with them u they--I guess my hepe wou1d be that they ;
15 wouldn't even want another meeting, but if they did, it would !
I 16 be usually two weeks after we get them the package. I f
17 CHAIRMAN HICHELSON: Clearly around the 1st of 18 December.
19 HR. BAER: No.
I 20 CHAIRHAN HICHELSON: You are talking two weeks from !
F l
21 the first weck in November. It is getting toward three weeks, !
I 22 then Thanksgiving week. I think you are talking about the 1st !
I 23 of December. :
i I
24 HR. BAER: Haybe, maybe I'm an optimiet. They may f 6
O 25 send us back a letter or just verbally hey, this looks fine, HERITAGE REPORTING CORPOPATION -- (202)628-4888 ;
' 1
334 ;
i . f i 1 you accommodated our' comments. f 2 CHAIRMAN HICHELSON: So what is the problem with our O 3 giving you a letter in our December meeting in terms of your 4 schedule? What difference is it going to make? CRGR isn't f t
5 going to see it much sooner than that and you are not going l 6 anywhere without it, without CRGR.
- 7 HR. HINNERS: I guess. '
I i
8 CHAIRMAN HICHELSON: What is your problem now that >
l 9 we are somehow holding you up? f i
10 HR. HINNERS: My problem, I don't--I wish to i 11 understand the purpose of this meeting. i i
12 CHAIRMAN HICHELSON: The purpose of--
(
l 13 HR. HINNERS: If you want to see a final product, we l
() 14 are not ready, but we, there is no way we could be ready.
t l 15 CHAIRMAN HICHELSON: You told us that a month ago 1
l l \
16 when this was scheduled. You didn't know that, either.
l i
17 HR. HINNERS: I thought the procedure that you were i f
18 going through was hey, we gave you something that we proposed, i l !
19 you made comments on it, okay; we come down here and tell you !
{
20 what, you know, the comments were, and then we go to a final l
(
i t 21 product. 1 l f 22 CHAIRMAN HICHELSON: Then the main--
l
- t
, 23 HR. HINNERS: Now you are telling me you only want l i
! 24 to look at the final product. i I !
() 25 HR. MICHELSON: The main purpose of the meeting, t f
l HERITAGE RSPORTING CORPORATION -- (202)628-4888 j
335 ,
1 aftwr we heard wherever you were at, we wanted to hear from I
2 industry on their views, and we had not had an opportunity to j 3 do that yet. That we are going to do today;.in fact, this l 4 afternoon. That's the purpose of putting it on this. meeting, j 5 HR. BAER: One of the reasons we met with industry i
6 last week is I guess I didn't understand how we were even 7 going to be able to dise.uss the changes we were going to make !
8 li : were, you know, simultaneously listening to what, i t
9 industry comments, so cao of the reasons we hurried up and I
10 they had requested a meeting and we arranged a meeting on !
?
i il f airly short order--it was I t.hink a very productive [
i 12 meeting--was to get their comments first-hand, and but I guess i 13 I didn't anticipate that we would have a rewritten letter, i 14 We met with the subcommittee what, early this month 15 I think it was. I didn't think that we would be meeting with l 16 industry and completely rewrite the letter and be back today [
i 17 with a rewritten letter. We are prepared to discuss the !
18 changes, j The December ACRS letter is no
~
19 CHAIRMAll HICHELSoli:
20 particular impediment to your schedule. You haven't convinced [
i 21 me yet that CRGR is going to look at it before we meet in l l
22 !!ovembe r. Clearly you are not going to do that. They may
(
look at it a week or two after that, so our December schedule h 23 ;
24 is certainly no great impediment. We will proceed unless 25 other members feel differently. [
f f
HERITAGE REPORTI!!G CORPORATIO!! -- (202)o28-4888 ;
I
t 336 1 HR. CARROLL: Just one point--I think Bob said 2 before maybe CRGR is coing to say that's fine, no need for a 3 meeting in advance of our November meeting.
4 CHAIRHAN HICHELSON: They can't do that until two 5 weeks after they get their packago I guess. Is that the way 6 it works? Even if they decide in that two weeks to have a 7 meeting or not--
8 HR. BAER: They generally, you know, tentatively 9 schedule one and then--
10 CHAIRMAN HICHELSON: They like two weeks to look at 11 stuff, too, just like we kind of have two weeks, and I see no 12 reason for us to rush into it if we are not holding their 13 schedule up to anything significant. Maybe two weeks--
() 14 HR. HINNERS: Well, should we then limit our 15 discussion today to the subcommittee comments and let industry 16 make their own comments?
17 CHAIRHAN HICHELSON: That was the intention, that we 18 hear whttever happened since the last time we met. We didn't 19 want to hear--
20 HR. HINNERS: You don't want to hear us paraphrase l
l 21 what industry told us? You are going to get a direct reading l
22 today?
I 23 CHAIRMAN HICHELSON: We would like to hear you, ehat 24 you thought industry told you certainly.
() 25 HR. CARROLL: At least what your response to it is.
HERITAGE REPORTING CORPORATION -- (202)628-4888
i I
337 !
1 HR. L*NNERS: I don't think we are prepared to do i' !
2 that. I mean the way this was presented to me, and maybe I 3 didn't hear it correctly, was that we are going to come out !
4 here and have a discussion session on the comments. What I !
i 5 :.ow hear is that you are not prepared to react unless you have i l
6 final proposals, and that's two different purposes.
f 7 Fine. Let's come down and t,alk--the Committee has j l
8 always asked to be involved earlier in the issues, and I have j t
[
- 9 no problem with that. !
I 10 CHAIRHAN HICHELSON: We don't write letters early in
- 11 the resolution unless we have problems with the direction in I
} 12 which the resolution is going. Otherwise we wait until a l 1
13 resolution is insuod, and then if we wish, we write a letter, 14 We don't always write letters on generic issues, either. f l f 15 HR. HINNERS: I don't think we are prepared to give f
(
16 you a final statement on what the letter is going to be. We i 17 certainly are willing to discuss the comments we got and i 18 probably give some direction as to how we might resolve those I i
connents, but I don't think-- f i 19 [
L i 20 CHAIRMAN HICHELSON: I think that would be ou.4te [
j j 21 satisf actory to the subcorr.mittee, and we may not even ask to j I
22 talk to you again. Depends on what the material looks like [
l I 23 that you send us, but we will go to the Full Committee which :
24 we would like you to support in December.
f 25 HR. HINNERS: Okay. I just want to understand the i
HERITAGE REPORTING CORPORATION -- (2021628-4888 f t
P n .
+ l 338 1 schedule. If we get'someth!.ng to you by December 1st, we ,
r 2 should be able to get a letter at the December meeting? )
A O 3 CHAIRHAN HICHELSON: That's right.
4 HR. HINNERS: If we can do better than that, I guess f i 5 we will try, but I agree with you, Carl. I don't think it is :
6 practical.
7 CHAIRHAN HICHELSON: And be prepared tentatively, if 8 we have a problem, we would like to talk to you about it on l t
? 9 the 13th of December. !
l i 10 HR. HINNERS: Sure, j
11 CHAIRHAN HICHELSON: We have the time already set up 12 for that. I e
I 13 HR. HINNERS: My concern is once we are into ,
i i
() 14 December. I mean things get all fouled up.
i 15 CHAIRMAN HICHELSON: That's right.
[
i 16 HR. HINNERS: It is very difficult to get anything j k 17 out until February. :
18 CHAIRMAM HICHELSON: We are prepared to meet that i
- t I i 19 December schedule. I thought that was reasonable, and then I ;
5 l
] 20 thought suddenly things were falling apart, but I don't think (
l i 21 they are. 1
[
22 HR. MINNERS: I guess I was misled because I was 23 still confused on what the purpose this meeting is. It we are I
f 24 going to coce bach-- {
t 25 CHAIRMA)1 HICHULSON: From you people, the first time 4
(])
i I
HERITAGE Rs?ORTING CORPORATION -- (202)628-4888
I a
339 (
1- for the industry.
l t
{} 2 HR. CARROLL: You may not come back in December.
] r 3 HR. MINNERS: I would rather have had it the other '
i !
- 4 way round.
5 HR. WYLIE: Whet is the schedulo on the December l i
6 meeting, the ACRS meeting > A.7ybody got that? I know what the l 7 days are, but I mean-- !
I 8 CHAIRMAN HICHELSON: I have no idea yet. We will i t
9 talk about that, you know, November Full Committee.
i !
10 HR. WYLIE: On the ?.5th and 16th. (
11 CHAIRMAN HICHELSON: Yes. I don't know which days. l 12 okay. Well, with ths.t cleared isp nicely now, let's proceed. !
. I 13 HR. ROTHBERG: Can you hear me?
{
14 CHAIRHAN HICHELSON: We may have stole all your 15 thunder now. !
I l 16 HR. ROTHBERG: I don't know tnai ! had that much to j 17 steal, f
I i
18 Let me tell you what CRGR told us-- L 1
f d
19 (Slide) I
! 20 HR. ROTHBERG: About the generic letter; and we met j 21 with them on the 13th I believe it was. The first bullet, t; f e
22 program is needed and the letter should be sent. They diQn't I i
4 23 say that go send out the letter. What they said was that !
I !
) 24 incorporate our comments, show us a letter again, and we will (
' i
() 25 lend you the final approval. but essentially they agreed that I
1
!;ERITAGE REPORTING CORPORATION -- (202)628-4888 i
?
340 1 if the program needed it, the letter should be ser.t.
(')
G' 2 The second comment was that th e environmental 3 qualification bases and design bases are, should be fixed.
4 We should not be questioning the existing design 5 bases. I'm told that you had some discussion yesterday with 6 the people who are concerned with Generic Issue 87. It ma" be 7 that the design bases change, but that is not what we are 8 concerned with with this generic letter. And again, it won't 9 affect that.
10 MR. WOHLD: Can I ask a question on that? .The 11 position change, whole valve thing, isn't that a design 12 change, a change in the design basis?
13 MR. ROTHBERG: I'm sorry. I didn't hear.
14 MR. WOHLD: Isn't the position changeable valve 15 issue, isn't that a change in the plant design basis? Is the 16 CRGR taking that part of the thing out of the generic letter 17 or--
18 MR. ROTHBERG: No. The position change of MOVs were 19 always part of Bulletin 85-03. As a matter of fact, the 20 supplement to Bulletin 85-03 added specific words that let 21 people, now that they were to include position changeable 22 motor operated valves. We know there is a safety concern 23 here, and the fact that you could disable the system, I think 24 that that is listed in our, in the rationale, in the safety
() 25 rationale position change of MOVs.
i HERITAGE REPORTING CCRTORATION -- (202)628-4888
i 341 l l
1 MR.'BAER: I think there was some words in there ]
about reviewing the design basis and had some words that we
(-
U 2
3 were, had been previously suggested by HRR that we are pretty, 4 were pretty broad and general and talked abcut environmental 5 qualification, and the CRGR said they want to make it clear, I 6 have to make it clear in the letter that you are not 7 questioning the environmenta) qualification that has already 8 been done, but as one of the other points that Owen will get 9 to, they did agree that the considoration of the degraded 10 voltage and power supplies should be, you know, specifically 11 cited in the letter, but they, they didn't, they thought our 12 words--in retrospect, I think they were right--were rather 13 unclear as to exactly what we wanted licensees to do in th'e
() 14 design basis.
15 HR. WOHLD: One other--on October 4th, you alluded l 16 to a more common sense approach to the, te the design basis I
l 17 evaluation for the DP and so forth.
18 Design basis is not a clearly defined thing. You 19 might talk about the design basis that the valve was built to, 20 the design basis for the system and so forth. It might be 21 beneficial to, to emphasize I think what you were trying to 22 say is that a common sense approach to what is the maximum 23 differential pressure, for instance, a valve might see under 24 limiting conditions, identified through a system engineer's
() 25 evaluations of operation as opposed to going through the HERITAGE REPORTING CORPORATION -- (202)628-4888
y 342 n
1 original design books to find out what the valve was purchased 2 to do and that Jort of thing.
'3 MR. ROTHBERG: The way we said it in our generic 4< letter was on the first page we defined the design basis.
. w.
5 These words are right out of 50.49 for electrical equipment.
6 In the subparngraph A, the first action item, we 7 said that the design basis is defined in the licensee's 8 commitments essentially. If he, if the licensee, the owner 9 has to go back and look through his existing records to find 10 his design basis, well, that's part of it. It is kind of 11 revealing that that design basis is not readily apparent. I 12 think that's a problem as far as I can see.
13 MR. WOHLD: I guess what I am saying is, is it
( 14 really necessary to go back into the books if you can pull out 15 a function curve, for instance, and say the max possible 16 differential pressure is based on the curve, therefore, the 17 design requirement of the system is met if it--
18 MR. ROTHBERG: That's fine. That's up to the owner 19 to do. The point is, there is something very interesting 20 about this. The owners spent a long time under Bulletin 85-03 21 defining the design basis for safety-related motor-operated 22 valves, and apparently the information on that design basis 23 was not readily available.
24 If the information is readily available to the
() 25 ownar, I don't have any problem with it. It is a matter of HERITAGE REPORTING CORPORATION -- (202)628-4888
343 1 his document. I think it is kind of revealing--if he has to go back and go through a bunch of gyrations to find out what 7N 2
3 the design basis is, you have to question how he is saying I 4 that his safety-related components are satisfying his design 5 basis when he doesn't know what the design basis is. To me, 6 that's kind of a revealing thing.
7 MR. BAER: One of the comments that sort of came 8 through I think somewhat from CRGR and somewhat from the 9 industry was the concern that work that had been done under 10 85-03 or voluntary extensions to 85-03 might have to be 11 repeated, and to the extent that we possibly can, we are going ,
12 to try and make the first three action items A, B and C 13 identical or as nearly identical as we can outside of--
() 14 HR. ROTHBERG: A, B, C and D.
15 MR. BAER: To those that were in the, well, it is 16 the continuing program, that I think was a littis different, 17 but certainly A, 3 and C we are going to try and make as close 18 as we possibly can, if not identical to what was in 85-03, to 19 help alleviate the concerns expressed that were there once 20 some work that is repeated, and so, you know, one of the l t
21 things when I was looking at it yesterday, Owen's rewrite was I t
22 to try and accommodato that request, but in, also in f
23 coordination with all the other comments we got which wasn't 24 in some cases to relax A, B and C, so-- l l
()
25 MR. ROTHBERG: Or increase it.
l t
HERITAGE REPORTING CORPORATION -- (202)628-4888 w _ _-_-------
344 1 CHAIRMAN HICHELSON. Well, before you--I guess this 2 is a good time. I assume that really all you are going to do
'3 is go through the two slides you have and then discuss any 4 questions we have on the letter.
5 It might be best to bring some of the questions en 6 the letter up witt; the bullets that you have on the slide, or 7 I can do it either way.
8 One of them, for instance, as an example is the 9 mispositioning question, which is the item A in the generic 10 le*ter.
11 One would worry about what is the design basis for 12 mispositioning? Do you assume that only one mispositioning 13 occurs at a time when you decide to look to see what the
( 14 pressure drop or delta P might be on the valve?
15 HR. ROTHBERG: The idea of mispositioning is there 16 are certain motor operators that don't have to function for
( 17 any design basis event. They may be convenience valves in a l
l 18 safety-related system, but if those valves are not locked 19 open, or have locked open or locked closed, and have power l
30 removed, or have power removed I should say, there is the 21 peasibility that they could become depositioned in position 22 and interfere with the safety function. That's a one-time 23 operation.
24 MR. MICHELSON: I think I understand. Let me give
() 25 the thrust of my concern and maybe you can give me a more HERITAGE REPORTING CORPORATION -- (202)628-4888
345 t
i explicit answer.
2 In the process of mispositioning,'if only one fixed 3 valve mispositions, then maybe the delta P to reopen and 4 properly position that valve is so much. If two valves happen i
5 to misposition at the same time, for whatever mispositioning ,
6 cause, then the delta P might be quite different on the valve.
i 7 If you go back and look carefully at ECCS systems, you will i
8 run into this occasionally wherein valves supposedly in, 9 norinally in certain position, when ECCS started, if they are ,
10 for any reason in the opposite position, the delta Ps are 11 quite different, and that mispositioning isn't addressed here 12 at all I don't think.
13 HR. ROTHBERG: Why not?
( 14 CHAIRMAN HICHELSON: It isn't clear, because I 15 thought the mispositionings were the ones that don't perform 16 safety functions to begin with. I am talking about valves 17 that do perform safety functions which are assumed always to 18 be in a particular alignment at the time the ECCS signal comes 19 through.
20 HR. BAER: I think we are talking about the same 21 valves, and the answer is that we are not, we have not l 22 required in the past and I don't think our intent is to ,
23 require multiple failures.
24 CHAIRMAN HICHELSON: That's the question. Are you f
() 25 assuming more than one?
1 i
HERITAGE REPORTING CORPORATION -- (2021628-4888
... - ~ . .- .
346 ,
1 HR. BAER: One of the comments, one group of--
2 CHAIRMAN HICHELSON: A single cause can cause 3 .several of these valves to misalign.
4 4 MR. KISSEL: I am Dick Kissel with o9.F. Carl, when 5 we were working on the bulletin and in response to licensees' ,
L 6 inquiries, and the, for the mispositioned valve, we wanted to 7 be able to-recover so that then the rest of the system can 8 function properly.
9 Now if mispositioning a valve, a position changeable 10 valve, means that the loading on a safety-related valve is 11 such that it can't function, the scenario is that we recover i 12 the mispositioned valve rendering the system so that it can ,
13 work. We are not requiring the safety-related valve to be
() 14 able to handle every conceivable configuration. ;
15 CHAIRMAN HICHELSON: The point is in doing such i 16 analysis, the ECCS valves you were referring to are supposed i i
- 17 to perhaps be already open. It happened at the time of this t
18 event to be closed for whatever reason. It now changes the L t
19 scenario quite a bit as to what the operability of the f
{
20 alignment is, and you can ignore that or you can include it in f 21 looking for what valves you worry about being in position, the '
- 22 non-safety valves that you worry about being in this position. t i
23 It is an analytical question only. Clearly in the practical !,
24 sense, sure--
() 25 HR. KISSEL: In reality, we have defined these i
l HERITAGE REPORTING CORPORATION -- (202)628-4888 t
347 1 valves as being any valve that can, because of its
, 2 mispositioning, render the, render the safety-related system 3 inoperable.
4 CHAIRMAN HICHELSON: One at a time.
5 HR. MINNERS: You are not answering his question.
6 CHAIRMAN HICHELSON: I think I am close enough. I 7 think you are saying about one, do it one at a time?
8 HR. BAER: One at a time, and the point I was trying 9 to make was that in Bulletin 85-03 we assume, licensees 10 asserted that they did not have to assume mispositioning, and 11 so the supplement was put out. It was always the intent of 12 the bulletin to assume mispositioning, and the supplement was 13 put out to clarify that intent.
() 14 CHAIRMAN HICHELSON: Single failure criterion kind 15 of required you to go back and make these assumptions also in 16 doing your analysis of the system to see if it works, assume l 17 that non-safety equipment or other valves not necessarily in l
l 18 the alignment Gre perhaps in the wrong position.
, 19 HR. BAER: The single failure analysis could also l
20 say I have already assumed this whole train, it doesn't work.
21 CHAIRMAN HICHELSON: I got tha answer. You intend 22 only one mispositioning to be analyzed at a time?
23 HR. BAEF: Yes, sir.
24 CHAIRMAN HICHELSON: Just want to understand; it
() 25 isn't clear in the document.
HERITAGE REPORTING CORPORATION -- (202)628-48PB
348 1 MR. MINNERS: Are you suggesting something else, or
^
)
2 is that just a clarification?
ss) <
3 CHAIRMAN MICHELSON: It is a clarification. I can 4 think of a number of cases where it doesn't work tco dell, but 5 it perhaps is good enough for, this will get most of the 6 situations.
7 Another question that comes up is this question of, 3 that comes in this item C where you talk about isolating 9 breaks, and the question, of course, is basically what kind of 10 breaks are you talking about?
11 Well, the answer that comes back is we are talking 12 about design basis breaks, but that has been a variable over 13 time, particularly when we got into the leak before break and
() 14 so forth, so as an example, if you were to qualify the HPCI 15 steam line downstream of the isolation valves to be a leak 16 before break line, which I could be doing some analysis and do 17 other things--
18 MR. MINNERS: I doubt that.
19 CHAIRMAN MICHELSON: Let's assume for argumentative 20 purposes I could. If I could, then does that mean that I 21 analyze only for the very small break, that I still am in that 22 line as far as the ability of the valves to isolate?
23 MR. ROTHBERG: Whatever it is in--
24 CHAIRMAN MICHELSON: That was not in the original
() 25 design basis.
HERITAGE REPORTING CORPORATION -- (202)628-4888
349 1 MR. ROTHBERG: In the design basis then.
2 CHAIRMAN MICHELSON: Design basis as modified at any 3 time? How that gives me a real pain because when we carefully 4 went through the Mechanical Branch with all of this, they 5 assured me that this would never be used for that kind of 6 purpose. Leak before break was only for restraints, et 7 cetera, not for compartmr.nt pressure, not for valve isolations 8 or things of that sort.
9 MR. DAER: I don't think that it was our intent to, 10 to not have people--we have some limitations in the original 11 bulletin which recognizes that it is pretty hard to do design r 12 basis testing on those valves, but the intent was that they, 13 you know, try and set switch settings so it could close under
() 14 those conditions.
15 CHAIRMAN MICHELSON: I am just trying to identify 16 whether or not you think these are still break flow isolatien 17 valves or nominal crack isolation valves? And that was, if 18 you change your story on leak before break, that it will, 19 reopens the leak before break issue.
20 MR. MINNERS: I don't think leak bef ore break 21 applies to lines. These are carbon steel lines.
22 MR. ROTHBERG: Wait a minute.
23 CHAIRMAN MICHELSON: Can be applied to any line 24 outside containment, if you want to go through the mechanica O 2s or it-3ERITAGE REPORTING CORPORATION -- (202)628-4888
350 1 MR. MINNERS: We have not. We have only applied 2- leak before break to the lines--
3 CHAIRMAN MICHELSON: I underJtand, but this document .
4 that I am trying to just identify what you mean by design:
~
5 . basis breaks, and I think the answer that came back is 6 whatever the latest acceptable design basis break is, not the 7 ones in the original-- [
8 HR. BAER: Warren is saying they haven't changed for~ !
t 9 these lines. I don't think the intent of the staff--my 10 understanding is the same as-Warren's, that there is not an 11 intent to change the design basis in that regard.
.12 CHAIRMAN HICHELSON: The last I asked, they had not 13 gone outside of containment with it except the penetration.
( 14 Fine, but there is nothing to prevent them from going out--if '
15 you look at ABWR, the current design document, they are 16 proposing there on some of it to go outside of containment and
- 17 now I have a problem with what kind of breaks am I isolating P
l 18 with chose valves? I 19 HR. MINNERS: I think that a more appropriate 20 question, the people want to extend the leak before break :
, 21 concept to other lines to say hey, does this, are you going to f 22 to change the valve design basis?
23 CHAIRMAN HICHELSON: It was questioned at the time 24 we accepted the leak before creak as a solution of a ;
() 25 particular generic issue, and the answer the staff gave us HERITAGE REPORTING CORPORATION -- (202)628-4888
351 1 then, they assured me it would not be used for this purpose.
-~ - 2 Not be used--
- i 3 HR. BAER: I know that's the intent of the staff. i 4 HR. ROTHBERG: And the letter doesn't, I don't think 5 the letter gives you any indication.
6 CHAIRMAN HICHELSON: The letter is unclear what you 7 mean. I am just pointing out how to--
8 HR. ROTHBERG: The letter is specific in this #
9 regard, that it does not question tho existing design basis 10 for the plant when we have words in there or we are going to t 11 make even in addition t.o that words, that say that the 12 licensee's, owner's commitment is the design basis.
13 Now that's in there in paragraph A now. We have
() 14 added words to that that make it even clearer. It is in 15 paragraph, right in the first part of paragraph A where it
! 16 says review and document the design basis for the operation of 17 each motor operator. Then it clarifies exactly that that ;
18 design basis is the owner's commitment.
19 CHAIRMAN HICHELSON: Yes, and that changes from time 20 to time is all I am saying. And it is changed for a different l
21 reason. You get rid of some restraints and also it could l 22 conceivably be used to back off. If I have a valve problem on 23 a particular line, I may take it and go back through leak ,
- 24 before break until I don't have-- ;
() 25 HR. BAER: You nesd to get the staff's, the staff's HERITAGE REPORTING CORPORATION -- (202)628-4888 [
i 352 j 1 approval on a revised design basis.
1 2 CHAIRMAN MICHELSON: You can see where this could be 3 I think easily misused in the future.
l l
4 The third question.has to do with the use of the l 5 applicability of this document to future plants, that there is 6 no, the document, it is a resolution of a generic issue, and 7 it remains silent on how it is to apply to future plants, and 8 that was our question.
9 HR. ROTilDERG: What we intend to do is to use this 10 in revised I guess standard review documents. The letter 11 itself obviously can't go to future plants. There is no 12 addressee.
13 CHAIRMAN MICHELSON: Part of it--what do you mean
( 14 there is no addressee? ABWR is an addressee today. You are 15 running through an FDA on ABWR. The agreement on ABWR was 16 that whatever generic issues were on the books in July of '86 17 would apply to ABWRs. Now the interpretation that has come 18 back is that okay, when you see the resolution, and it 19 mentions, does not mention future plants, then it doesn't 20 apply to future plants, so I read this as not applicable to 21 future plants. Fine.
22 What does apply to future plants in regard to this 23 particular generic issue? It clearly, clearly the issue 24 applies to future plants as well as present-day plants.
() 25 HR. MINNERS: In that what the advanced people are HERITAGE REPOP?ING CORPORATION -- (202)528-4888
//
353 ,
1 saying? Is that their interpretation of that?
2 CHAIRMAN HICHELSON: That's my understanding in
.O 3 asking them, yes. They Nill address all the generic issues 4 tht6t are identified as applying to future planta.because a lot 5 of generic issues don't aipply to future plants. Some do. And 6 this one I think does.
.i 7 HR. MINNERS: llot very many of them don't. ;
8 CHAIRMAN HICHELSON: Beg pardon?- There is quite a i
9 few that don't, but they are plant-specific more and core, 10 today's gene */ation.
i 11 HR. HINNERS: We weren't saying our intent was to 12 have this incorporated into the SRP. Okay. Exactly where
, 13 those words are going to be, I think we intended to have a i
() 14 letter to NRR whi<:h says hey, and by the way, this general ;
l 15 ought to be incorporated into the appropriate SRPs. (
l 16 HR. MICHELSON: See, the SRP doesn't apply to ABWR !
l !
j 17 except for the revision. !
18 HR. MINNERS: That means it is future plants,
} ,
i 19 but--that means it is future plants.
1 You are talking about future
! 20 CHAIRHAN HICHELSON:
l 21 plants now? )
22 HR. HINNERS: I am saying if you put something into l i !
j 23 the SRP, that means it applies to future plants. I i
24 CHAIRMAN HICHELSON: I am saying, though, you people f
I
() 25 agreed with GE that the revision of the of the SRP to be used i
, i HERITAGE REPORTING CORPORATION -- (202)628-4888 i
- i
.y .- - - __ _ _ - . -
' !)
In F 354 ,
1 and revision of the generic, the generic issue is to be i
g- 2 applied so it was all as of a certain date. That was to get 4, .
3 you out of this race of keeping inventing new issues forever.
4- They got out that problem by putting a limit on it, which was 5' a fine idea, except they agreed that it, a generic issue that 6 was on the books had to apply to future plants as well. Then 7 they would recognize it and do whatever the generic !
8 requirements might be, but it had to be identified in the 9 generic issue resolution as applying to future plants.
10 HR. BAER: There is another--
11 HR. MICHELSON: And this one does not say anything 12 about future plants.
13 HR. BAER: There is another activity going on, and
() 14 I'm not sure exactly how it will end up, but there is going to 15 be an attempt to revise the regulations dealing with Section 16 11 to break out valve and pump testing separate from ISI, you 17 know, inspection of welds and things like that, and as part of 18 that, there is some consideration of trying to, in one manner l
19 or another, have these sort of requirements as part of the 20 regulations as the supplement to OH 6 and 10, which will be, 21 I guess is now referenced by Section 11 I think of the last ,
22 addenda or will be coming out shortly.
23 I'm not sure that's an answer to your question. If 24 that does come about and it is in the regulations, then it !
i
() 25 would clearly apply to all plants, but I am not sure how that HERITAGE REPORTING CORPORATION -- (2021628-4888
o 355 i
1 would come out.
2 Our goal here was to fix problems on currently 3 operating plants. You know, that's what we set out to do.
4 Hadn't spent, given a lot of thought other than this, Warren i
5 mentioned getting this into the $RP.
6 HR. MICHELSON: This is not a unique response for ,
7 any particular issue. You can raise the same question on 8 several issues brought up recently for final resolution in
{
9 which no mention was made of future plants, yet they clearly ;
10 applied in certain ways, and the resciution is supposed to say ,
11 in what ways this will apply to future plants if it applies at 12 all, and this resolution remains silent, and the [
13 interpretation remains c'ilent. It doesn't apply.
() 14 HR. HINNERS: That's not my interpretation of the 15 words, of the interpretation that has been applied in the t
l I 16 past. Generic letters have always been applied by the staff ,
l 17 to new plants as voll as the past ones.
L 18 CHAIP.HAN HICHELSON: Even though it didn't say in 19 the resolution that it applied to future plants, that's an 20 interesting, that's a perfectly--if that vero your policy, I r l
21 would have no problem. ;
i 22 HR. HINNERS: In the old days, we were licensing L
23 plants, issuing our letters. That's my understanding of the 24 way things went. If you talked to a reviewer, I mean--and it ;
f
() 25 was, a generic letter out there, that wasn't explicitly, avon HURITAGE REPORTING CORPORATION <- (202)628-4888 1
P I
356 1 wasn't explicitly in the SRP, he would be applying it to new 2 plants.
O 3 HR. BAER: Did have a bulletin.
i 4 CHAIRMAN HICHELSON: You reached a different 5 agreement on future plants or you wouldn't have put in this 6 provision that your cutoff date was July of '86 on generic j 7- issues. Generic issues after July '86 don't apply. Those ,
8 before it do apply. If the, if the generic issue-- l 9 HR. HINNERS: I don't know where that statement is i
10 made.
11 CHAIRHAN HICHELSON: It is made in--what do they 12 call that thing? That Memorandum of Agreement? !
13 HR. WYLIE: I don't know. I remember the words. ;
() 14 HR. MINNERS: I'm not that familiar with it, but--
f 15 HR. MICHELSON: One has been written for each item. ;
i i 16 HR. CARROLL: Who should they talk to to find out?
s >
17 HR. HINNERS: You have got some information I don't !
18 have. .
i 19 CHAIRHAN HICHELSON: You are talking to whoever is ,
t 20 doing ABWR. Talk to Scalletti. Ask him what the agreement is !
21 on generic issues to be--how are you, they applicable to l
. l 22 future plants? E i l J 23 HR. HINNERS: I don't quite understand your j 24 question. Are you asking us whether we intend to take care of
() 25 future plants, or are you asking me as a representative of the (
HERITAGE REPORTING CORPORATION -- (20?)628-4888 f
357 1 staff as a whole whether that's the staff policy?
2 CHAIRMAN HICHELSON: No. You as the writer of the 3 resolution, I am asking does this resolution apply to future 4 plants?
1 .
5 HR. HINNERS: I'm not the future plant guy, okay.
6 CHAIRMAN HICHELSON: But you are the writer of the [
t 7 resolution.
8 HR. HINNERS: But that may be. I think we are not 9 the future plant guy. The future plant guy decides what goes 10 on in future plants.
11 CHAIRMAN HICHELSON: Why then in some resolutions do '[
i 12 you put down how it applies to future plants and sometimes you ,
13 don't? You ought to be consistent then and never mention
() 14 future plants and let somebody else figure it out, but you I
15 have, you have mentioned how this one applies to cps, how it i 16 applies OLs, and how it applies to future plants. [
1 t 4 17 HR. HINNERS: That is usually only done--well, our 18 surveillace is not the best, is that there just isn't a ;
l 19 mechanism for generic letters to do that. It has never been f i
i 20 done on generic letters. It just happens to be this form. I f
- 21 mean maybe that's wrong. The only place that that's ;
I i 22 applicable is in standard review plans ar d regulatory guides ,
f 23 where there is an inplementation section in which you have to [
24 address backfit and--
25 HR. HICHELSON: Ars you saying the resolution of I
(])
1 !
) HERITAGE REPORTING CORPORATION ~~ (202)628-4888 t
358 1 generic. issues in the past have never addressed future plants?
3 2 HR. BAER: No, I am not saying that.
3 HR. MINNERS: No. I am saying that if, only if the 1
4 resolution was in the form of an SRP or a reg guide. If the 5 resolution was in the form of a generic letter, you won't find 6 any specific statement on that.
7 CHAIRMAN HICHELSON: I would have to go back to 8 research that.
9 HR. MINNERS: That's my, you know, that's my 10 understanding, and my understanding was that a generic letter 11 applied to all plants.
12 CHAIRMAN HICHELSON: There are specific statoments 13 in here on how it applies to OLs and cps in this generic
() 14 letter. Why are you, why do you even put that in? Why didn't i 15 you--
1 16 HR. BAER: Only in terms of schedule; only in terms 17 of schedule, because you have to make a distinction between, 18 you know, you can't tell a CP who may not be operating for i 19 five years he has to finish all this testing in three.
1 20 We are were trying to recognize on cps that they f
i.
j 21 have a different schedule.
22 HR. HINNERS: I don't understand your statement 23 this doesn't address future plants. We have got statements in
[
24 there about cps.
() 25 CHAIRMAN HICHELSON: I didn't say it didn't address HERITAGE REPORTING CORPORATION -- (202)628-4888
F 359 1- it. I said it remained silent.
("] 2 HR. HINNERS: You just said that it talks about cps.
(/
3 MR. ROTHBERG: He is talking about plants that 4 don't even have a CP yet.
6 CHAIRMAN HICHELSON: It does address those. It does 7 not cddress plants that are. going under the--
8 HR. HINNERS: It doesn't address future cps.
9 CHAIRHAN MICHELSON: It remains silent. That's all 10 I said. I wondered if it, it applied. If so, how we do it, l
11 because the agreement on ABWR was indeed the resolutions would j 12 be applied, but if the resolutions said that you don't do 13 anything, or doesn't mention, remains silent, then the
( 14 question is well, how does it apply or was it even intended to ,
i 15 apply?
4 16 HR. MINNERS: That statement has never been fed back '
l 17 to the people that are doing the generic issue.
18 CHAIRHAN HICHELSON: It has been fed back at more ,
I
- 19 than one subcommittee. q 20 HR. HINNERS
- It has not been fed back to the people j 21 who are doing the generic issues. They may have told you [
i j 22 something, but they haven't told us anything. [
23 HR. DAER: Yes. t I
J 24 HR. HINNERS: I'm not sure they have to, but you !
i
() 25 conm to be asking us the question, and it seems to me that-- f r
f I HERITAGE REPORTING CORPORATION -- (202)628-4888
u 360 1 CHAIRMAN HICHELSON: I think I have got my answer.
2 HR. MINNERS: It is their decision on how they want.
3 to apply generic issues to future plants, not us. !
4 HR. HICHELSON: Why don't you proceed?
i 5 HR. ROTHBERG: Schedule.
- 6 HR. CARROLL: Before you have get there--
7 HR. BAER: Did you talk about the degraded voltage?
8 HR. ROTHBERG: I haven't gotten to that yet. I 9 think that the business of degraded voltage is on the next ;
10 slide. [
11 HR. CARROLL: No. It is in the paragraph above your I
12 schedule.
13 HR. ROTHBERG: I'm sorry. That's right. I lost
() 14 track.
15 There have been a lot of discussions about the i 16 business of degraded voltage, water line losses, power supply 17 problems, and basically they can disable the motor-operated .
I 18 valve. We want to make clear that that is part of the design i 19 basis, and to the extent that the licensee has to, has to 20 consider degraded voltage in his-- !
21 HR. CARROLL: Our problem la-t time was what you 22 meant by degraded voltage. If I have a HOV that is designed [
P 23 to operate at 80 percent voltage, is it satisfactory for me to !
s 24 demonstrate it there, or do I have to, have to postulate all l
(} 25 kinds of terrible things happening to my power supply where--
i HERITAGE REPORTING CORPORATION -- (202)628-4888 I
361 1 MR. ROTHBERG: What we intended was that you would 2 postulate what the original design basis commitment was, what
.O 3 your design basis commitment was.
4 HR.-CARROLL: That is going to be made' clear in the 5 letter?
L 6 MR. ROTHBERG: That will be made clear. We don't 7 want to impose anything that nobody has committed to. He are 8 not--
! 9 MR. BAER: Degraded voltage, and I think we are I 10 recognizing that you can't even do the testing. At degraded 11 voltage, you ought to--
12 MR. CARROLL: Design basic degraded voltage.
i 13 HR. BAER: So the valves will operate at design i q
() 14 basis voltage.
4 15 HR. CARROLL: That's all I wanted to know.
~
16 HR. ROTHBERG: Again, this will be put into the
- 17 generic letter. t
, i
! 18 There was a lot of discussion about the schedule, i 19 the recommendations that we made. As you pointed out, that it 20 is somewhat arbitrary, the schedule. What the scheduling 21 recornendations, what the industry wanted, what we will try to 22 do is to put in sono words, make a recommendation, about
[
23 testing based on an owner's schedule, perhaps based on his 1
24 assessments of the importance of the HOV. We would like to do 1
() 25 it that way if we can. We are not quite sure of the words f
l HERITACE REPORTING CORPORATION -- (202)628-4888 i i
362 1 'right now. This is one of the reasons, this is one of the '
2 things that we are jockeying with it right now.
3 It is kind of difficult to come up with words that 4 allow someone to prioritize his work. It is much easier, as
{
5 you can understand, to come up with nomething like this. ,
6 Wh'at we would like to do is allow the owners f
7 flexibility in their scheduling of their work. We recognize 8 their problems. We recognize their shortages in manpower and 9 time, and what we are trying to do is work uomething out right i
1 10 now that will do that, so you will probably cee something--in 11 fact, you surely will see something that will be either as an
)
12 alternative to this, well, what I would like to do is offer it 13 as an alternative to this sort of schedule.
() 14 CRGR said that the letter is safety significant, and 15 really what you should do is put new some deterministic 16 argements about its compliance with the 50.109 paragraphs that j 17 say essentially that it is safet; significant. We are going 18 to do that, too. That will, prooably won't go in the letter.
' 19 It will go into the cover memorandum.
20 (Slide) 21 HR. ROTHBERG: The list of degraded conditions--
l' 22 what we wanted to do, and we fully intend to do, is to make 23 that list of degraded conditions something that the licensee 24 should look at as guidance for setting up a program. We don't
() 25 want to make it a check-off list. We don't want to make the
{ HERITAGE REPORTING CORPORATION -- (202)628-4888
363 1 licensees make a commitment to saying that, that they have to 2 do a specific action for each model on a specific schedule.
3 We don't want to do that. What we are trying to do is to have 4 them put together a program that addresses these things and i
5 addresses these things rationally from their point of view.
6 This was a comment that I believe the subcommittee 7 had made about the balance of plant MOVs. What we would like 8 to do is put in at least the suggestion in the generic letter 9 that the balance of plant HOVs be addressed by some sort of 10 diagnostic system.
t 11 NUMARC comments concerning the proposal should be 12 considered. Believe me, they are.
13 I went through the maintenance policy, and the
() 14 maintenance rule, looking for problems that might arise from 15 this generic letter. ,
16 What we would like to do is add some words to the i 17 generic letter that make it clear that there is no intent to, ,
i !
l 18 there is no conflict intended between the things that we are 19 proposing and the maintenance rule, and if there is a 20 conflict, that the maintenance rule would govern. The 21 maintenance rule is kind of a general thing, as you know, and '
22 what things might arise, it is kind of problematical to see 23 whether, whether something that we have said might be in 24 conflict wi',h that or not.
l
() 25 As of now, after voading it, I don't see any i
HERITAGE REPORTING CORPORATION -- (202)628-1888
i 364 1- particular conflicts, but maybe some will come up.
2 Again, this boils into, this comes into the existing 3 design basis. It is parallel to that. The owner's QA program 4 and his recordkeeping commitments, we are not trying to make a 5 conflict with that. We are not trying to make something that 6 is redundant and'that is going to cause him to set up a wholo 7 new program. What we would like'to do is to make it clear 8 that, that this thing should fit in with his existing 9 programs.
10 CHAIRHAN HICHELSON: How can that be . lone?
11 HR. ROTHBERG: I think the first *hlng to do is to 12 say so in the generic letter.
13 CHAIRMAN HICHELSON: But I am wondering in a
) 14 practical sense, do you think that can be accomplished or--I t
15 think you are sort of suggesting that you can use whatever '
16 your present system is.
f
] 17 HR. ROTHBERG: For quality assurance requirements? [
i f j' 18 Well, let's take-- [
] f 19 CHAIRHAN HICHELSON: I am thinking of 85-0?,. I (
t 20 assuue you might want to build on it. f i
21 Is that what you had in mind? i 1
22 HR. ROTHBERG: We are building on 85-03 as far as :
23 the procedures-- I l 24 CHAIRHAN HICHELSON: The recordkeeping was only a i l
() 25 minor aspect and a sheet-range aspect. It was a reporting i
HERITAGE REPORTING CORPORATION -- (202)628-4888 !
I 365 L 1 requirement which is no longer required.
l lt
' < 2 HR. ROTHBERG: Yes, sir. i i l 3 HR. BAER: In the previous draft, we had words like i
4 permanent records, and the CRGR said make sure that this is, l l
5 you know, permanent records has 6 legal meaning, you tnow. f
, l 6 Suppose someone eventually. throws away the valve or replaces
,7 it entirely. Does he now have to keep the records? Hake sure (
.. 8 it is consistent with, you kr.ow, existing requirements for ,
J 9 record keeping. j 10 CHAIRMAN HICHELSON: You want a living document, not 11 all the history. You can, if you throw the component away, [
l l 12 get a new valve, you can throw the documents away I guess you 4
f I
13 are saying?
I
() 14 HR. BAER: What they were saying, that was an f i
15 example that we shouldn't be, that all the licensees have j l
i ,
l 16 certain agreed-upon recordkeeping requirements, and we should i :
17 make sure that our requirements are consistent with that.
! 18 CHAIRMAN HICHELSON: This is admittedly a new batch j 4
J 19 of information that has to be recorded?
i l 20 HR. BAER: Yes. t i !
21 CHAIRMAN HICHELSON: It is just a question of format I
! I 23 then or the way in which you do it or what? j
> t 23 HR. BAER: I think that is one of the areas--I don't !
. 24 know if you have had a chance to talk to our recordkeeping -
8 d
() 25 experts. People had suggested words like permanent records, i
t j HERITAGE REPORTING CORPORATION -- (202)628-4888 i I
q l
366 I
1 and we had to admit to CRGR that we weren't prepared to tell
- 2 them what e:tactly a permanent record was, and so we may modify l 3 the words to say consistent with your commitments for keeping 4 maintenance records.
5 HR. ROTHBERG: I would like to keep it vague, but if.
6 we have to, we con refer to Reg Guide 1.28 and NQA 1. l 7 Unfortunately, some of the owners may not have that !
8 commitment, and that's one of the problems.
9 CliAIRHAN HICHELSON: One of the problems that I can 10 foresee from the, in the Agency's viewpoint is that it might .
11 be that we will want to look at an amount of this information 12 to look for trends or changes or, you know, do some .
13 industry-wide study, f
() 14 To do that I guess we will have to go to each plant 13 individually and look at the recoros, but if you, each record r
}
4 16 is kept in its own way, of course, the people that are going l 17 to do this mignt nase a lot of difficulty even in establishing l 4
i 4
18 comparability of what they view from one plant to another even t 19 though each record in itself might meet the requirements of a 20 generic letter. It may be very difficult to read them across 21 the board.
22 HR. ROTHBERG: Yes, indeed. Therein lies the 23 compromise which you have to make.
24 HR. DAER: I will point out when we met with NUHARC
() 25 and other industry representatives, they made a big point of HERITAGE REPORTING CORPORATION -- (2021628-4888
r i
367 1 programs I guess at least planned, maybe partially ongoing, 2 where they will be sharing data and so hopefully maybe they 3 can' agree on a consistent format.
4 CHAIRMAN HICHELSON: They certainly ought to 5 standardize, however this is done, and that standard ought to 6 meet whatever the requirements of thu generi letter are. It 7 is to their own benefit. It is a fine bunch of : ;ormation.
8 It is badly needed.
9 HR. BAER: And it it helpful to the utilities in 10 terms that I thiak overyone agrees that it is not practical to 11 pressure test or test that design basis delta'P, say it 12 correctly, all the valves or even a majority of the valves, l 13 and so--and it is hard, difficult for one plant, perhaps an l
() 14 individual plant to make a case that they have enough valves l 15 that are similar that they can make a case that some, you 16 know, zero delta P ot low delta P testing and some diagnostic 17 work can be extrapolated. The larger the data base, or the l 18 more similar the valves, the easier it is to make that case, I
i 19 and as I said, industry poinced out that they do have plans.
20 Maybe they will talk about that this afternoon.
21 HR. HICHELSOM: Your intention here is to try to 22 pake it as easy as possible for them?
23 HR. BAER: Yes.
24 CHAIRHAN HICHELSON: Okay.
() 25 HR. BAER: And CRGR's apecific comment was of the HERITAGE REPORTING CORPORATION -- (202)628-4888
368 -
1 -mechanism to be imposing new recordkeeping requirements per- l 3- se.
O p 3 HR. ROTHBERG: Excetly, so whatever we do, we will ,
4 be. careful to make it comply. . }
5 We want to also make it clear we tried in the first, 6 the draft that you have fer'the generic letter, to do this, i t
7 but we are also going to add some more words, that credit for 8 work done under 85-03 and any other additional surveillance ;
9 programs would be given. We want to-- t i
10' CHAIRMAN HICHELSON: In regard to recordkeeping, one '
11 of your old items and perhaps it is still there, I don't know, i
12 was the one requiring recording of trends and totals.
13 Is that going to be changed in wording or is that f (f 14 still--
l 15 HR. ROTHLERG: We want to k<ep the trending 16 recommendation. ,
17 CHAIRMAN HICHELSON: Now isn't that going to I i
18 constitute kind of something somewhat new? l f
19 MR. MINNERS: The whole thing is new.
CHAIP. MAN HICHELSON: I am thinking in terms of 20 f i
21 present recordkeeping commitments and so forth in the next to j 22 the last bullet. !
i 23 HR. BAER: It wasn't, my understanding of CRGR's :
t comment wasn't that we couldn't require them to keep I 24 i
() 25 additional records. It was tossing around phrases like I
HERITAGE REPORTING CORPORATION -- (202)628-4888
369 I
1 permanent records that was bothering them,.that as I said, 2 when faced with the precise question, exactly what is a 3 permanent record, none of us were expert in that area. ,
4 CHAIRMAN HICHELSON: .This particular item also 4 5 required permanent record in this. You are saying you still !
6 want it permanent, but you don't want to call it permanent? !
7 !!R . BAER: We want to keep it long enough to be' 8 trending, and we may end up using the word permanent record.
! 9 We just couldn't answer their question exactly what is meant t
10 by a permanent record and what is a requirement for permanent 11 records, et cetera.
r 12 HR. ROTHBEKG: Anyway, what we are going to try to l 4
] 13 do is avoid any recommendations that will result in having, l () 14 people having to go back to do retesting, do rework on their 15 design basis investigations if we can. i i
16 HR. WOHLD: I have:got a question on that last item i t
i There was some research information that was presented I 17 there.
> c 18 yesterday morning that indicates that the thrust requirements [
- l l i 19 identified during the actual DP and blowdown testing can be as 20 much as twice as high as the previously commonly assuned l
r 21 thrust calculations, and it is my understanding-- !
22 MR. ROTHBERG: That is pretty preliminary right'now, f l
23 though.
I
( 24 HR. WOHLD: I understand that, but I'm just I t
i j () 25 wondering about the, the extent that these calculations might !
l :
i l i HERITAGE REPORTING CORPORATION -- (202)628-48S8 i t
, 'i 370 l
1, be accepted in the old bulletin and continue to be accepted .!
2 under the extension here.
3 There has been I think enough indication that the 1
4 calculated, the previously calculated values ten years back [
5 and so forth, that are non-conservative often enough to set 7
r 6 the valves up now 'to the extent calculations of plus or minus 7 10 percent or so forth that wore used in the past co'ald result i 8 in common mode failures that are not currently being i
f 9 experienced because the, the old way of setting the torque j 4 !
, 10 switches has had a considerabic amount of margin in it.
4 i 11 MR, ROTHBERG: Pete, you are gettin7 out of the area l f
12 of in situ testing. In situ testing is, there is an implicit 13 assumption, and that implicit assumption is that the equipment i
() 14 as designed and insta11ad is pcoper, and the test statement ;
- 15 will verify that it is operable. ,
l-i 16 What we are talking about here is you are e i
i 17 questioning that the equipment ac designed and as installed is !
L i
! 18 not proper to the function. !
t i
19 I don't think that what we are doing here is going j 20 to address that. It is not going to help you.
21 HR. BAER: There is one footnote that Roy Woods, who .
22 was involved in that program, had us write in the previous i l 23 draft of the generic letter, and I don't see it changing, I t 24 pointing out that there is ongoing tests in that on these .
() 25 relatively limited nunber of valves, that there may be 1
l ;
HERITAGE REPORTING CORPORATION -- (202)628-4888
t l
- 371 I
K 1 supplemental' requirements coming. l N 2 MR. ROTHBERG: That is Appendix A.
('_) - : :
3 .HR. WOODS: Could I make a statement? This is Roy l 4 Woods.
P 5 Just this morning, we had some discussions. I L
(
6 haven't even had a chance to-discuss that with these people. ;
7 You got a presentation yesterday from Kevin DeWall or Bob 8 Steele, I have forgotten which,'and also Owen--this is Owen }
.t 9 Rothberg; also Jerry Weidenhammer. We had a discussion this ;
10 morning that we are going to make an effort over the next few !
1 11 weeks to analyze more data and get more input into trhat, what !
t 12 these requirements ought to be, and some way or another [
- 13 reflect maybe with a more extensive footnote or whatever than !
() 14 Owen's was just referring to we want to get more of those ,
15 results into this letter if we can, j t
, t i
16 HR. WOHLD: I think thc last item sort of indicates
- i 17 that you are going to have a licensee continue on in the same l t
18 vein that the bulletin valves were addressing.
i 19 HR. ROTHBERG: Again, what we would like to do is in i
20 the interest of efficiency, we don't want to have somebody, f L
21 and especially for norale--it is as much morale as anything r 22 else--we don't want to have somebody coming out with a new 'I 4
[
23 generic letter that says by the way, fellows, you have got to 1
1 24 go back and do everything that you did before and do it again, f
! i
() 25 only you have got to add this to it. We would like not to do f l
2 !
HERITAGE REPORTING CORPORATION -- (202)628-4888 h
I 372 L 1 that.
2 HR. WOHLD: I understand that. That's common sense.
3 All I am saying is that it may be prudent to revisit the 4 :aanner in which the valves were set up under the bulletin.
5 MR. ROTHBERG: No problem with that. All I am 6 saying to you is that this generic letter is not going to help 7 you up if you are saying that the equipment as installed 8 doesn't satisfy the design basis.
9 Again, we are talking about in situ testing. We are 10 talking about a plant that is built. We are talking about 11 operability testing to make sure that the equipment is, is 12 operable in accordance with the design basis.
13 If you, if you look at that design basis and say
() 14 gee, this particular valve as designed doesn't meet the design 15 basis, the in situ testing isn't going to help. You are going 16 to have to get rid of the equipment and get new equipment.
17 MR. WOHLD: Interpretation of design basis is it l 18 meets the manufacturer's thrust requirement.
19 HR. MINNERS: No. The.t is not an interpretation of 20 design basis. The design basis is what the licensee decides 21 the valve has to do, okay. And he looks at his transients and 22 his accidents and his degraded voltage and he decides what the 23 design basis is. Then he buys the valve to meet that design 24 basis. It isn't the other way around.
() 25 MR. WOHLD: Once you have that design basis, you HERITAGE REPORTING CORPORATION -- (202)62S-4888
4 373 1 calculate the thrust to make sure the operator delivers that 2' thrust. ,
3 HR. HINNERS: Sure.-
t 4 , HR. WOHLD: And the calculation has been shown i 5 invalid. t i
6 HR. HINNERS: I think that is too strong a- i 7 statement. '
8 HR. WOHLD: In many cases, it shows that it is a "
9 nominal type situation.
I 10 HR. HINNERS: They have only tested two valves, so 11 at best you can only make a statement about two valves. [
12 HR. CARROLL: There is also Marshall test. l t
13 HR. ROTHBERG EPRI Harsha11 test. :
i
() 14 HR. WOHLD: Have valves that were the cause--their t
15 coefficient friction was about .6 as I recall.
.t 16 HR. MINNERS: We don't know how to write a new 17 design equation for the valves yet. We don't have that 18 information yet. l l
19 CHAIRMAN HICHELSON: Hay just be a change in the
{
20 friction factor. ,
l f
, 1 l 21- MR. HINNERS: It may just be a change in the ;
22 friction factor. We don't know how to do that yet, so we .
f l 23 can't tell licensees to use a differont equation because you l 24 don't know what the right equation is. (
(], as CHAIRHAn HIcHELSON: This is a procedural l ;
t HERITAGE REPORTING CORPORATION -- (202)628-4888 l
f
? ;
~
,- 374 f 1 resolution, what you have here. ,
2 Hn. MINNERS: I think you have a good question of 3 whether te should be sending out this generic letter now 4 before we develop the design equation, but there are some 5 valves for which I don't think there is a problem.
6 MR. BAER: For the valves that don't involve a 7 broken line, you want to take the Davis-Besse valve hopefully, }
8 and you have a long presentation by Bob, and I have forgotten j i
9 his last nane, Elfstrom, about the Toledo Edison program both i a
] 10 at the previous subcommittee meeting, and at the, some 11 discussion on the presentation at the NUMARC meeting, and I t
12 think he is convinced and I guess convinced me that a :
) '
4 13 combination of testing at full delta P or partial delta P and ,
() 14 diagnostic techniques, you can have the high regree of 15 assurance that those valves will open and close against design ;
i 16 basis delta P for the case, at least for the cases where the !
j 17 pipe isn't broken. [
j 18 CHAIRMAN HICHELSON: If you know what the friction [
i t
i 19 factor is, i
i 20 MR. BAER: I don't have to, if I can do a design I
~
21 basis delta P, Carl, if I can do it.
22 CHAIRMAN HICHELSON: If you can, j f
23 HR. BAER: Then I don't have to know the friction [
i !
l 24 factor. i f () 25 CHAIRHAN HICHELSON: That's right.
{
f r
I I HERITAGE REPORTING CORPORATION -- (202)628-4888
375 1 MR. BAER:. For the smaller limited'aumber you have 2 valves-that involve isolating in a broken pipe, yes, there are j Gg -
3 these uncertainties. I guess-I feel strongly that we !
4 shouldn't hold up the whole generic letter for those valves. !
5 I think we can point out that there may be more information 6 coming and on those valves yes, licensees may have to do ;
I 7 something different. !
S CHAIRMAN HICIIELSON: A number of these valves are l 9 involved in ECCS injection, not necessarily it is because .
10 there is a broken pipe. It is not to isolate a pipe. It is l 11 to inject water after a break in the primary system, and those ,
i 12 see full delta P if the break is big enough to depressurize in [
13 a hurry, they will see pump head delta Ps right away, even
() 14 pump head delta Ps in some cases is--
15 HR. BAER: That, I would hope with a combination of {
1 16 testing and diagnostic techniques, you could determine that l l
17 those valves would open or close as needed against those delta l L
18 P. [
t
- 19. CHAIRMAN HICHELSON: Yes. !
?
i 20 MR. BAER: I guess I object to the idea of holding i
21 up the whole letter for, in terms of the fa:t that there may !
l 1 22 be a few valves where licensees later will have to do more. (
23 See, the other side of the coin was the strong )
24 statement by both representatives of Davis-Besse and San !
i i
() 25 Onofre who have done extensive programs that the letter as [
t !
l HERITAGE REPORTING CORPORATION -- (202)628-4888 ,
6
376 1 written would chuse them to redo everything, and that was 2 clearly not our intent, and that was, the CRGR reinforced 3 that, and we are trying to modify the words to make it clear i F 4 that someone that has gone out and done what appears to be a [
i !
5 pretty good program should have only a limited amount to redo, t 6 You can't have it both ways. :
i 4
7 HR. WOHLD: I don't want to hold up the letter.
8 This type of thing is an implementation concern, and I'm not t 9 concerned about the letter, the initial sending of it in this !
- 10 respect, and I am not concerned about Bob Elfstrom or 11 Davis-Besse's program, which is a good one, but you have 12 licensees out there that are going to take a diagnostic l
- 1.' machine with no meaningful differential pressure testing, and
() 14 they are going to set up to a thrust equation that has been l
. 15 used in the past. They are not going to have the type of i
I 16 comprehensive program thht Bob described. He seems to have a r 17 different program.
18 HR. BAER: We are asking them to demonstrate t
19 operability against design basis delta P, or to have a ;
20 combination of delta P testing at lower delta P and analytical [
?
t 21 work and diagnostic work that gives you the assurance that it i 22 would operate against delta P.
23 HR. WOHLD: You are saying the equation itself is :
. f t
24 inadequate. That's basically what I wanted to clarify. l
]
() 25 HR. MINNERS: All that has been demonstrated I I
I I i HERITAGE REPORTING CORPORATION -- (202)628-4888 i i
n
377 1 . understand from the tests I have seen, it may be inadequate
{} 2 for break flow conditions, and when it is saturated. ,
3 HR. WOHLD: It was proved inadequate for the i
4 Davis-Besse situation against the normal system. ,
5 HR. HINNERS: That wasn't my understanding. I'm not' 6 an expert. I think I thought it was just set up wrong.
7 HR. ROTHBERG: That's right. In. Davis-Besse it was 8 set up wrong. j
! 9 HR. HINNERS: I thought just the setup was wrong.
L 10 HR. WOHLD: The friction factor they determined from !
i 11 the diagnostic testing was twice what is popularly used. It t
12 was somewhere around .6 instead of .3. t i
1 13 MR. HINNERS: Okay. Well, that's why we want to [
14 have in situ testing at, with a delta P. (
15 HR. ROTHBERG: I don't think that you can make any, 16 you can make correlation when the, the Davis-Besse was a 7 a
17 failure to open the, as I understand the thrust of these f
18 tests, is failure to close.
19 HR. WOHLD: It is either way. l 2d CHAIRMAN HICHELSON: Either way, f i
l 21 HR. WOHLD: Davis-Besse valves, they had a high l 22 thrust opening and closing, both ways. They failed to open ;
[
, 23 and-- l i !
24 HR. ROTHBERG: The problem arose at Davis-Besse when {
l
() 25 those two valves were inadvertently closed, subjected to I HERITAGE REPORTING CORPORATION -- (202)628-4888
i
~
378 t
1 pressure, and failed to reopen.
2 HR. WOHLD: That's true, but had they been open and i 3 asked to close, it wouldn't have closed, either. Now the 4 safety argument is made that they would have closed enough to l 5 perform their function. However, the--
{
6' HR. ROTHBERG: That is problematic.
7 HR. WOHLD: The point is the friction factors are f
8 much higher than--
9 HR. MINNERS: I guess I don't understand the thrust 10 of your comment because--are we working at cross-purposes .
i 11 here? I thought the purpose of this issue was plain, is to l
~
12 try to get the test valves in place at design basis l 13 conditions. Okay. <
$ 14 Are you suggesting that that's not a useful purpose?
15 HR. WOHLD: I'm saying that some licensees are using l
\ t 16 an equation based on the design basis, and they are not doing i i
1 17 the DP testing and recognizing the need to address higher !
18 friction factors. f 19 HR. HINNERS: Okay. We are trying to get people to
{
20 do testing at the design basis conditions, which means in many :
l ;
h 21 cases a delta P. Or are you agreed that's a useful purpose? t
[
l 22 HR. WOHLD: Yes, sir, very definitely. {
! 23 HR. HINNERS: Okay, so I guess I don't understand. !
I i 24 Are you telling me that some licensees are doing it wrong? Or (
J
() 25 are you telling me that we have a wrong generic letter? I ;
i HERITAGE REPORTING CORPORATION -- (202)628-4888 ,
t
l' 379 i
1 don't understand.
i 2 HR. WOHLD: They are not doing the DP. testing that O 3 recognizes the need for higher thrust, and they are setting
[
4 the valves up to an equation not--
5 HR. ROTHBERG: That is too, very preliminary. -
t 6 HR. MINNERS: We are talking about what the generic (
7 letter wants licenses to do. ,
8 Is the generic letter being written wrong to ;
i 9 accomplish our purpose? [
I 10 HR. WOHLD: I think you can interpret it to set a i
11 valve up to an equation without the DP test.
12 HR. ROTHBERG: I think it is a little premature to !
13 start reworking that equation. I'm not suggesting that it
{
() 14 won't be reworked later on, but the state of the research f I
t 15 right now doesn't back it up right this minute, 16 HR. BAER: Let me ask Dick Kissel. He was the f 17 fellow that was reviewing the 85-03 results, which were for 18 the 85-03 valves, were licensees merely setting them up for l ?
I 19 friction factor equation? [
20 HR. KISSEL: The program said that we didn't accept ;
21 their program.
?
22 CHAIRHAN MICHELSON: What did they set them up to if [
l 23 they didn't set them up to their calculation using a friction f i
24 factor generally .37 l r
() 25 MR. KISSEL: Thi question that was asked was would I
HERITAGE REPORTING CORPORATION -- (202)628-4888 i t
380 1 we accept that as the simple, as the solution to the program?
2 That was what I understood the question to be.
3 CHAIRMAN HICHELSON: You don't accept .3 then if 4 that's what they use in their calculation?
5 HR. HINNERS: Why don't you say what we did do, 6 Dick, instead what we didn't do? What did we do in reviewing 7 their results?
8 HR. KISSEL: Okay. First of all, most of the 9 programs were not of sufficient detail, the submittal of most 10 of the programs were not of sufficient detail for me to be 11 able to tell you in detail what the licensee did to verify 12 that the valve would operate. Those programs that did that we 13 have accepted, did one of three things. They either tested an
() 14 identical valve, they had test results for an identical valve 15 and then were able to show that this valve that they didn't 16 test was set up so that it would function in the exact same 17 manner.
18 HR. BAER: That test was design basis delta P.
19 HR. KISSEL: It was done at design basis delta P.
20 They had four valves that were identical that they ustd to 21 establish a formula to demonstrate the effect of delta P.
22 They did not necessarily get up the full design basis delta P, 1
23 but they were able to establish in the testing of these four 24 valves, the relationship between delta P and required thrust. l
() 25 CHAIRMAN HICHELSON: These were not, were not break HERITAGE REPORTING CORPORATION -- (202)628-4888
st 381 1 flows, *<ere they?
1 2 HR. KISSEL: I'm sorry? ,
3 HR BAER: None of this was break flow. ,
4 CHAIRMAN HICHELSON: These were-- ;
5 HR. KISSEL: I said they didn't go to the design 6 basis necessarily. Okay. They could then use the result from l 7 those, from the testing of those four valves to make the '
8 setting on the valve that they did not test, or the? could use P 9 20 similar valves, the difference being going from four same, 10 same torques, 20 similar. t 11 Now we are starting to talk about variations in 12 size, and again, come up with an equation that predicted the
) .
l 13 effect of delta P on required thrust, and use that then to set ;
() 14 the valve. These were the three programs that when I have i
15 been asked I have approved. I have not approved blindly p 16 setting the thrust to the original manufacturer's setting. ,
17 HR. WOHLD: Okay. I think the problem is a lot of !
18 the responses were not clear, that you could determine how the ,
19 testing was to be done. There were a lot of caveats that were !
4 t
20 equivalent manner with the, would be used to demonstrate [
f 21 operability, and I have run across a number of plants that j l- )
22 would use the equation. I can't remember how I resolved that. !
(
23 I have left it as an open item, unresolved item, but that was j i
z 24 the question I had in inspecting against the bulletin. [
I
() 25 MR. CARROLL: How many of the programs for the r
i I HERITAGE REPORTING CORPORATION -- (202)628-4888 h l
382 1 original 85-03 have you accepted? I gather that you haven't 2 accepted a great number of them? l O 3 HR. KISSEL: Let me try and say it again, perhaps a
[
l 4 little clearer. l 5 We have accepted a great number of programs, but l 6 they have not gone into great specificity as to exactly what {
7 they were doing on every if, and, but and or. Those programs ;
i 8 that did go into detail, we have accepted when they have
[
l provided one of those three. If they simply said they were i 9
i i 10 going to use the OEM formula and set to that, we did not [
- i 11 accept it. l 12 HR. BAER
- If they were vague, what happened?
4 13 HR. KISSEL: It depended upon how they couched the r
. t
() 14 vagueness.
l 15 HR. BAER: How vague. l i 16 HR. CARROLL: It seems to me that one thing that f i
17 constructively could be done in the present situation is to, i
18 is to clarify the Roy Woods paragraph a little more, and (
l 19 because I'm sure some utilities may want to be pessimistic and {
- l' 20 take the INEL data as it is right now, if they have the )
f 21 capability in their operators, crank them up to whatever that 22 suggests. j l i l 23 HR. WOODS: Just this morning we agreed that there [
I 24 would be a top priority effort in the next week to do exactly [
t t
() 25 that, and I didn't even have a chance to tell these people.
I i
HERITAGE REPORTING CORPORATION -- (202)628-4888 l
383 i
1 That's why I came in late. That's what we were doing this ,
2 morning. I O 3 CHAIRMAli HICHELSoli: When you think about the P
4 problem that the IllEL data seems to indicate exists, you might l l 5 also think a little bit about the' case wherein ECCS pumps are :
6 at three, four, 500 pounds pressure waiting to deliver, and {
7 then you start to open the valve and already the downstream f f
8 side is atmospheric, and it is a very, it is a different kind i I
9 of problem. That was not what was tested at Wyle, but it, 10 because it is cold water now, but cold water released at high !
i i
11 pressure, it also undergoes certain kind of flooking [
t 12 phenomenon. {
q From what I have seen of it, what was t l
13 HR. WOODS:
O 14 aoee wow 1d save see# i ta t e.
I 15 CHAIRHAll HICHELSoti: It does, it indicates those 16 cases also require higher friction factors than night have l
s
(.
l 17 been traditionally used.
. l s
18 HR. WOODS: We have seen the least of the, we have !
< I 19 seen the margin disappear most clearly in the areas where you i l
r 20 are very close to saturation and subcooling region you just f f
21 described-- [
l 22 CHAIRMAli HICHELSO!!: You have co1d conditions at !
1 !
- 23 high pressure. you can get into the same kind of problem, and j i
24 I am just wondering, it would be nice if they used the correct l 25 friction factor for that. .
HERITAGE REPORTIliG CORPORATIOli -- (202)628-4888 l
\
l 384 I I
1' HR. WOODS: I agree.
l 2 HR. WOHLD: I think Dick clarified it, though. The O 3 OEM equation is not acceptable by itself. ;
l 4 HR. KISSEL: Simply making the thrust setting meet j f
5 the OEM indications would not be an acceptable solution to the r i
) ?
6 bulletin. That's why the generic letter was expanded to ,
7 include those three criteria, to provide guidance to the !
8 licensees as~to what we considered to be acceptable, i
9 HR. WOHLD: I think a generic letter was changed to ,
t l
i 10 add equally acceptable methods or something. [
I 11 HR. ROTHBERG: It was. j i !
12 HR. WOHLD: But the OEM equation is not an equally }
l l
q 13 acceptable method. !
- () 14 HR. BAER
- Explicitly state that somewhere. I l 15 HR. ROTHBERG: We could say that. This next slide [
l i 16 you don't have. What you do have is the memorandum and I attachments of my meeting notes with NUMAI.C, and this, these 17 18 bullets merely paraphrase the comments on the second page of 4
t 19 that memorandum.
20 Basically the industry reps talked about their f l
i 31 programmatic approach to solving problems of motor-operated i d
f 22 valves. The problem that we found with it was that-- !
23 HR. CARROLL: The commitment is spelled wrong. f 24 HR. ROTHBERG: Yes. Too many Hs or too many Ts--we
{
r
(} 25 don't have that right now. The industry has a lot of HERITAGE REPORTING CORPORATION -- (202)628-4888
l 385
~1 technical information being assembled right now, and that may ;
2 be useful, but I think that what we have seen is that the -
3 generic letter would probably be necessary to bring everybody {
4 along.
5 Again, the comment about the test schedules, they l
6 objected to this hard three-year, two refueling outage test 7 schedule that I was talking about. We are working on the 8 words. The NRC costs are too low. They were estimated [
9 somewhere in excess of $400 million as a total industry 10 effort. [
I 11 I might add that, that the, those cost estimates !
i 12 just about match what we considered to be the benefit without j 13 the safety, so it is a wash at worst. It is a wash between j
() 14 what it is going to cost and what their benefit is going to f i
15 be. l 16 There are going to be problems of personnel and }
t 17 spare parts, test equipment and vender support. There is a 18 competing problem in that we have the safety problem out j 19 there, and-- [
i 20 CHAIR!!AN HICHELSON: Which problem?
21 MR. ROTHBERG: Well, the industry has a problem in f f
22 that they are not going to be able to respond as quickly as we !
i 23 might want. We understand their problem, but we have a i 24 problem of our own, namely, that we want to get this thing i
() 25 done as fast as possible, so we are going to have to balance HERITAGE REPORTING CORPORATION -- (202)628-4888 l t
386 l o i 1 that. He suspect that there will be problems over the years 2 as they get into doing this thing. Trained personnel, spare I
, (2) ;
3 parts, vender support are going to be problems for them. '
i 4 There is no question about that. l 5 They didn't like the idea of pressure testing in l
[
6 situ. and we tried to point out that the generic letter did }
7 give them an out, that we weren't makir.g it an absolute solid
[
8 recommendation that you had to do in situ testing, We ;
i 9 recognized that there are situations in the plant where you !
t 10 can't do pressure in situ testing.
11 We are changing the words further to emphasize that, :
12 and make it absolutely clear that there are Otagnostic '
13 alternatives to full pressure testing of each HOV. We have an
() 14 appendi:t and we are going to work on that to make it again 15 clear that, that there are certain Hovs that they just will i 16 not be able to do at any time. We understand that, and we L
17 will adjust to it. l i
18 HR. CARROLL: Let me ask this queJtion. Is one (
i 19 demonstration of acceptable valve performance the fact that [
20 the valve works under some say spurious safety injection or 21 something like that? Did you count that, in other words?
You mean did it work at practice (
22 HR. ROTHBERG: -
23 rather than during a test?
24 HR. CARROLL: Yes. k
(} bec".use i 25 HR. ROTHBERG: There is a pro HERITAGE REPORTING CORPORATION -- .. . I i t
n ! .. )
387 t 1 you, because it is not instrumented, you may not know exsetly l 2 what happened. (
O- I J MR. CARROLL: Well, no. I've set the valve up. l l
4 I've looked at how the valve works, but a month earlier, I, j 5 you know, by gosh, it really did work as the combination of !
6 those two things. !
i 7 MR ROTHBERG. I don't see that is a problem. You ,
i
~'
8 are saying that could you substitute that for a particular i
9 test--
10 "R. CARROLL: Yes.
i 41 HR. ROTHBERG: Sequence?
12 HR. CARRCLL: Yes. ;
I 13 MR. ROTHBERG: I hadn't thought of that. At first i
() 14 glance, I would be reluctant to accept it again because t
l 15 although it might have performed the safety function, you i
t ;
i 16 don't know that it had particularly closed or opened. j l
17 The other thing is that there is the possibility of i t
18 internal damage that you don't know about. Now there are t
, 19 diagnostic systems that are coming that I understand that r
I 20 might give you on-line diagnostic information so that you !
?
t 21 would be able to see what, how the motor operator performed, l
t 22 motor-operated valve performed while the anomaly was h
[
f 23 occurring. I gueso that would be okay, but-- [
I j 24 MR. BAER: 85-03, I guess I thought there was some ,
t
() 25 plant that did purely pressure delta P testing and not, and no i !
HERITAGE REPORTING CORPORATION -- (2021628-4888
c 388 i
i diagnostic as such? ,
2 HR. ROTHBERG: That is true, there were, but he is
- 3 not talking about that. He is speaking of the situation where 4 the valve responds to anomaly. ;
5 HR. BAER: My answer would have been different. If [
6 you really knew the delta P, I would tend to be somewhat 7 satisfied, even though I am an analyst at heart.
< e 8 HR. KISSEL: I think the question is how sure are :
9 you that during this spurious action, or maybe even the l r
{ 10 planned action, that the valve really was challenged to the !
11 design basis that you are trying to demonstrate? And if you i 12 can show that it had been, then you have demonstrated that the ;
l j 13 valve is operable as set. !
4 i
() 14 HR. CARROLL: I still want to do diagnostics on it,
, 15 so I can cone back to it, j 16 HR. ROTHBERG: Right, and use it as a reference (
l
)
17 perhaps. r 18 HR. CARROLL: Right.
2 L 19 HR. ROTHBERG: Yes. It might be a good idea. Are i t
i 4
20 you suggesting that we put something in the letter about that? j i 21 HR. CARROLL: I happen to know of one utility that }
l 22 is at least I think response to 85-03-- !
t i !
! 23 CHAIRMAN HICHELSON: One utility, huh? j l i l 24 HR. CARROLL: That type of approach.
I f
() 25 HR. ROTHBERG: And we discussed trending. We, of f
i HERITAGE REPORTING CORPORATION -- (202)S28-4888
389 ;
1 course, have put ir, a, a paragraph that said that the owners 2 were to take intistmation which would allow them to do C) trending.
3 t 4 What HUMARC and the owners suggested is okay, allow l
5 us to base our test schedules on that trending, and that seems t 6 to be reasonable. And the, there was some discussion about !
. r 7 well, you know, on some of these quarter turn valves like j
- 8 balls and butterflies, of we just may not have enough test 9 information so that we can establish the design basis, !
10 At first I was kind of sympathetic to that, and j t
11 there was words mentioned that well, it is a small percentage l
] ;
4 12 of the population, less than 15 percent, but then I thought l
, I 13 about it and 1 said well, what did you--I started thinking f
() 14 about that and I said to myself well, what did they base their 15 design basis on if they didn't know the, the loads that a I
16 motor operator would be subjected to? And I have come to the i
17 conclusion that that information will just have to be gotten. ;
1 L 18 They will have to go and work at their motor operators, !
(
19 perhaps compare, start with the design equation, and if they [
l 20 have nothing else, and go from there, but I don't think that !
1 21 ignorance is an excuse, ,
22 HR. MICHELSON: I won't take much comfort in that 15 l i I 23 percent thought because some like essenti-al cooling water :
I i might be 90, 95 percent butterflies, and therefore if they are 24 l i
(} 25 set up wrong or not functioning as required, you lose both
+
r HERITAGE REPORTING CORPC> RATION -- (202)628-4888
390 l
[
1 trains of an itsolutely essential x sm, so the fact that l 2 they are really small in rel 4' s tal population is
(~) 3 immaterial.
4 HR. WOHLD: There is an interesting interrelation 5 between the interrelated valve testing and motor-operated 6 valve testing. I'm 1.ot sure what is out there. I know there 7 is sorae ef forts being done to have diagnostics on 90 degree I
, 8 turn valves, but the, the air-operated valve diagnostic that I 9 just discussed this morning, that can be done on butterfly j 10 valves, and get the forces through the entire stroke just by 11 monitoring air pressure.
12 HR. ROTHBERG: It doesn't have to be motor-operated, 13 doesn't have to be air or hydraulic operated? It can be a 4
() 14 motor operator? Is that what you are saying.
- 1 15 HR. WOHLD: You could get a force, a valve thrust 16 requirement from an air-operated valve testing, and do some 17 difficult type tests and then make sure that the motor
' 18 operator put out at least that kind of thrust.
19 MR. ROTHDERG: Again, that's fine. It remains for i
i 20 the owners to do that.
21 As far as comments about testing air and hydraulic 22 operators are concerned, I really don't have much to say about
, 23 them because I really don't understand the problems in the 24 hydraulic operator yet. I haven't really looked into them.
(} 25 but I had looked at LERs on air and hydraulic operators, and HERITAGE REPORTING CORPORATION -- (202)628-4888
391 i 1 it seems that they have the same or an analagous problem to' j 2 these motor operator, namely, that they are not tested when 3 flow and pressure conditions are on the, on the operator, and 4 it really, when you do subject them to flow and pressure, I l
5 think that you get a lot different answers about their 6 operability, and so it may be that we need to come up with l l
7 something generic about the air and hydraulic operators or l 8 parallel, I guess similar to what we have here, but as far as >
9 hard information about the problems of air and hydraulic l l
10 operator, I really don't have anything for you. [
11 That's about all I have. Are there any questions? !
t i
12 CHAIRMAN MICHELSON: Any questions? This afternoon, !
13 we will hear from NUMARC on some of their views, and other j j
O 14 iaa trv repre t tive - ze is er uoaer t aatao ao ta t tuer !
l 15 will probably require in the neighborhood of two hours for i 16 their presentation, but that can give or take a little :
f 17 depending on how much we may have to ask. [
i 6 18 This, though, realistically still puts ourselves (
t 19 somewhat ahead of the allotted time. We originally allotted ,
[
20 three hourt;. It was to start at 12:30. We will now start at i,
21 1:15 instead of 12:30, but I think we will still finish up at j 22 four o' clock, and you can plan accordingly. ;
1 i 23 With that, I think we will adjourn the meeting until l l 24 this afternoon. !
i 25 (Whereupon, at 1:15 p.m., the meeting was recessed, f f
i
! t t
HERITAGE REPORTING CORPORATION -- (202)628-4888 l
___ y
! -?- 'f I
f 392 i
! i 1 1
} 1 to reconver.e at 1:15 p.m. the same day.)
1
!e
)
3 l
4 1
l
} 4 i
l 5 i
j 6 7
1 8
i '
l 9 l
10 i
11 12 l 13 14 !
l 1
! 15 l
l 16
! 17 18 l ,
19 '
l l 20 I
i 21 l
! 22 '
s
! 23 i
24 l
o 2e HERITAGE REPORTI!iG CORPORATIO!( --
(202)628-4888
.-. . t
393- ,
P 1 AETE&HQOH EEE11QR 1:15_p_dit 1 2 CHAIRMAN HICHELSON: It is time to readjourn here, !
C) 3 and what we would like to do now is hear the views of the r 4 industry on the generic reJolution, and the generic letter, !
5 which is the resolution, and for that purpose, NUMARC was i
?
d asked to come and make some presentations. ;
7 We certainly do welcome you, and I believe Tom ;
8 Tipton of NUMARC is going to make the introductions and go ;
9 through, we would expect he will present to us some work on I
10 the INPO program, and then on the EPRI program, and that will i r
i 11 be followed by three different representatives from industry, l 12 one from Fouta.ern California, one from Toledo Edison, one from i
13 Philadelphia Electric, giving some of the insight on, on the j
() 14 generic letter as they see it, so with that short !
i 15 introduction, I will turn it over to Tom, and it is his l 16 program. [
t 17 HR. TIPTON: Thank you Hr. Chairman. Hy name is l l 18 Tom Tipton. I am Director of Operations Management and the 1
i 19 Support Services Division of NUMARC. !
l (
20 We really appreciate the invitation to come befora l l l !
21 the ACRS subcommittee this afternoon, and give you an overview [
l of several of the industry programs, as well as comments that i l 22 I
l I
23 we have on the draft gersric 1etter. f l t l 24 With me today are Ed Moore from INPO, Jim Lang fron l
() 25 EPRI, erian Curry from Ph11ade1 hi. E1.ct3:ic, .oh r1fstrom i HERITAGE REPORTING CORPORATION -- (202)628-4888 I i
l 394 i
i from Toledo Edison, and Bob McPherson from Southern California l 2 Edison and Warren Hall and Mr. Calloway of my staff. r O 3 As reflected in the letter from Warren Owen,
[
I 4 chairman of our prior NUHARC Steering Committee when NUMARC l
5 was a csamittee, in the Victor Stello letter dated February 6 17, 1987, we agreed the industry shouldn't initiate any j i
7 additional action to improve HOV performance. The industry t 8 programs are aimed at ieproving the operability of valves at :
9 nuclear plants and are not intended to address current or
[
10 potential regulatory requirements. [
F 11 Although wo are very satisfied with our interactions !
12 with the staff, and most of the time we are not concerned in i
13 terms of our requirements and their requirements, but the
{
( )' 14 industry has recognized the need to improve an HOVs, enT we l
15 have taken steps to make those improvements. You will hear 16 from both INPO and EPRI today on thas matter. !
(
17 We have had meetings with the staff periodically to 18 update them on the status of these activities, and in [
t 19 addition, we briefed the ACRS Subcommittee on Reliability i
[
20 Assurance on March 7th of thip year. We would like to first 21 brief you on the INPO and EPRI activities, as you indicated, j r
22 and then we would follow that with comments on the generic !
i i
! 23 Intter, fI 24 As Owen Rothberg mentioned this morning, we met with f
() 25 the staff on Wednesday, October the 19th, to discuss the
)
l HERITAGE REPORTING CORPORATION -- (202*628-4888 }
t
395 1 generic draft letter. We felt the meeting was Nery productive
> 2 in providing the staff with our perspective based on the
~
3 experience of responding to Bulletin 85-03 and obtaining 4 additional insight or the staff on the letter. I want to 5 enphas'.ze we felt : a very good interchange of 6 information lasting about three and a half hourJ.
7 With that, and if there are no objections, what I 3 would like to do is to introduce Ed Moore tron INP3 to give 9 you an overview of the IMPO technologies associated with 10 valves.
11 (A discussion was held off the record.)
12 MR. MOORE: Gentleren, ry name is Ed Mc re. I'n the 13 Director of the Plant Support Division at INPO, and Vice
' > 14 President of INPO, and I'n pleased to be here this af ter'. con 15 to tell you the status of the INPO industry valve progrars 16 let you know, give you a current status.
17 If you will look at the handouts that hace beer.
18 prcvided. ry intention this afternoon is first to give you an 19 overview of the previous discussicns we had with this group, 20 and then to bring you up to date on rotor-operated valves, 21 check valves, and instrunent a'r air-operated valves failures.
22 This is a prescntation we rade back on March the 7th. Gary 23 Federsen and I care and gave you an update on otor-crerated 24 valves and check valves told you wnere we were at at that r'
()T 25 particular tire.
HERITAGE EEFORTING CORPCEATION -- (202)f;2-4Ss3
I; 396 l' We went through kind of a history of that thing. I
'2 pointed out that there are a lot of industry (events associated o 3 with motor-operated check valve failures versus something like 4 20 significant event reports for motor-operated valves, and
" 5 about 24, 25-for check valves, and we talked about'the NRC 6 responses to those events. about the different, the bulletins i 7 they put out, the reviews of motor-operated valve performance, 8 and I think the last one from AEOD was about December 1986. :
L 9 Following that, you probably recall that Mr. Stello i 10 wrote a letter to NUMARC and asked NUMARC to take on doirs 11 something about'new valves, and INPO responded in support of 1? NUMARC by developing an action plan to address motor-operated 13 valve failures.
() 14 This plant consisted of various things. First of 15 all, I have already mentioned the experience dissemination,
. 16 the SERs, significant operating event reports, and operations 17 and maintenance reminders. We have had considerable 18 interaction and coordination with both SPRI ar.d MUMARC in the 19 area of, particularly in motor-operated valve areas, and we ,
20 gave you an update on that motor-operated valve initiative, 21 and I will talk a bit mote about ti.at . We also talked about 1
22 the check valve activities at that time .sich were just barely 23 starting at INPO.
24 The Motor-operated valve initiative first consisted i
{} 25 of training or maintenance personnel so they wou14 be better HERITAGE REPORTIFG CORPORATION -- (202)628-4888
397 1 qualified to go_on' and address the motor-operated valves with t es 2 the industry. To do that, we did a couple of things. First, k_)
3 San Onofre brought in their expert. We brought their valves, 4 and INPO sat down and trained our maintenance evaluators, also 5 a number of the other people on the staff, to give them a 6 first-hand knowledge that had the hands on, see what it looks 7 like and so forth.
8 We also had already two legitimate exper':s in the 9 motor-operated valves on our staff. One of them is Jim Tills, 10 who some of you know, and the other is--geeze--David Gruber.
11 Both of those gentlemen we have been using very productively 12 to assess industry performance. I will give you some nore on 13 that.
() 14 We decided to look at motor-operated valves in the 15 industry in the area of first, during the, at the time you 16 will recall we had some maintenance assistance review teams 17 that were poing out in industry and looking in detail at some 18 of the maintenance programs, and we looked at
- operated i
19 "alves in detail on those visite, I think identified-certain i '
2C problems to help those people have better programs, and also l
21 decided to look at motor-operated valves during a regular l
22 evaluation. I will talk some more about that as we get on 23 down to the next page.
24 And we talked about how we would promulgate what we i
I
(} 25 found in those things. We did that by providing follow-up l
l HERITAGE REPORTING CORPORATION -- (202)628-4888
398 1 operating experience out to the industry, put out that
. , 2- information during maintenance, during our various workshops
~
3 at INPO, write letters to the executive points of contact and 4 so forth.
~
5' Let me ta1k a little bit about where we are now in ;
6 motot-operated valves.
7 (Slide) 8 MR. HOORE: This is an event we started in 1987 and 9 are still continuing it. Some of the results that were seen-10 coming out of this performance at this particular time is that 11 these two Limiterque models account for nearly 60 percent of 12 all the-inside motor-operated valves in models and have about 13 37 percent higher failure rate than the other others.
() 14 CHAIRt!AN MICHELSON: The other Limitorque operators?
15 MR. MOORE: Other motor-operated valves; the system 16 is not restrained just to Limitorque.
E l 17 CHAIRMAN MICHELSON: Ninet/ percent, thereabouts, i 18 are Limitorques?
19 MR. MOORE: Sixty percent are Limitorques.
20 CHAIRMAN MICHELSON: I thought 90 percent are 21 Limitorque and 13 percent are--
22 MR. HOORE: I may have a typo.
23 MR. MCPHERSON: Sixty percent is just--
1 24 CHAIRMAN MICHELSON: Total Limitorque population, 60
{} 25 percent of that.
HERITAGE REPORTING CORPORATION -- (202)628-4888
=nr -
399 k
- 1. MR. HOORE: We have a higher failure rate on these
{} 2 than we do some of the others. I think more of that answer, 3 there is some other reasons for that particular answer.
4 We also. find that torque switches and limit switches 5 and motors account for approximately 7 percent, 70 percent of 6 all the MOV failures.
l- 7 our data also shows--again we are talking primarily l 8 abotat NPRDS data--that DC motors don't crack. We are not sure 9 why that is. That could be, could be some other~ issue in 10 there that we havea't gotten around to looking at at the I 11 present time. ;
l l 12 CHAIRMAN !!ICHELSON: By don't crack, do you mean L
13 they are worse or better?
14 MR. HOORE: I don't have--
15 CHAIRMAN MICHELSON: Higher failure rates on on AC 16 tnan DC or lower failure rate?
t 17 MR. MOORE: I don't know the answer to that. Does 18 anybody?
19 CHAIRMAN MICHELSON: They are higher failure rates?
20 MR. ELFSTROM: For the same proportion. 4 21 HR. MOORE: We also found there are some attributes 22 that tend to cause increased failure rates--the number of 23 cycles in a particular application of the circuit, amount of 24 differential pressure, particularly high differential pressure
() 25 cuch as we are seoing in main steam feedwater systems.
HERITAGE REPORTING CORPORATION -- (202)628-4888
r 400 1
if ' 1 Forecasts of corrosion can get on the packing and cause binding, and #or material blocking valve seats which r)
~j 2
3 affect torque eetting, and you tend to find that in things 4 like service water systems where chemistry has caused a 5 problem.
6 (Slide) 7 MR. MOORE: We have put increased emphasis on 8 motor-operated valves during the plant evaluations and 9 acsistance visits. We made a commitment to follow up on this 10 significance recommendations from all of our SOERs. We went 11 back through the whole group of SOERs and identified any two 12 recommendations which we follow up on a three evaluation cycle 13 basis, but some of tham like the motor-operated valve
() 14 recommendations, we have opted to look at every time we do the i-15 evaluation.
4 16 We are currently reviewing the station's overall i 17 motor-operated valve program, evaluate how they do business.
18 Before we go on evaluation, we look at our NPRDS data, out of 19 that, the component failure analysis system identified as high i
20 failure raten, but for motor-operated valves, we look at all 21 the NPRDS failures before we do evaluation of every single 22 failure and try to draw some conclusions on what the problems 23 might be to help us do a better job, help them to try to 24 improve their programs.
() 15 Once in the cycle we review preventive maintenance HERITAGE REPORTING CORPORATION -- (202)628-4888
- - - - . - . . _ , - , - . - . . ~ - . . - - . - . . - _ - - _ - . . - ,
/
401 1 program, look at the. procedures, look at craft training and so 2 forth, post maintenance testing. !
(; s)
3 CHAIRMAN MICHELSON: Excuse me. When you say you 4 look at the MOV failure history, does that mean for that 5 particular site that you are getting ready to visit?
6 MR. HOORE: Yes, sir. Out data bunk we pull out 7 specifically for that site and we pull it out specifically on 8 the failure rate for those components for that site.
9 We have, I mentioned nefore I have two legitimate 10 really good guys in'the motor-operated valves. We are trying t
11 to use those as much .s we can, make them available as we do 12 rsgular evaluations to go out and take a more detailed look at 13 motor-operated valves. In the past 57 evaluations, we have
() 14 been able to get them on site 30 cf those times for a period 15 of two or three days. We also had them on site with the team :
16 for the maintenance assistance review team visits in which we 17 had about a dozen. i i
18 Result of that is that we have identified 19 recommendations for improving these programs at 24 to 30 sitors 20 that we have had these people visit. I would like to point i
21 out at this point in time that the reccmmendations vary from 4
22 maybe a single recommendation on some element of the program 23 that needs improvement to several recommendations, but it is .
24 certainly a wide variation in the kind of results we are
{} 25 saeing, everything from excellent programs to programs that i
i HERITAGE REPORTI!!G CORPORATION -- (202)628-4888
402 1 really need quite a bit of help.
-s g
2 Most of'the recommendations I indicate here are in
- J' 3 the areas of program control, physical condition of the 4 motor-operated valves, documentation, which deals with keeping 5 track of the results we .d with the new performed valve 6 maintenance so that we can look at it, see if it has changed 7 later on, and design control to make sure the right thing is 8 in the right application, procedures, training, testing, 9 preventive maintenance. We also, as you will recall, 10 committed to go out and look at some good sites where people 11 had'very low HOV failure rates and compile good records.
12 (Slide) 13 HR. !!OORE: I am not surprised I guess because what
() 14 we found when we looked at the good plants is nearly a myriad, 15 mirror image of the problems we identified when we sent out 16 experts out on site. Where we had the other vugraph, you 17 recall we had problems in programming and in design and in so 18 forth and so on. I'm just not surprised in that you find that 19 good programs have those elements in place.
?
20 We recently put together a summary of motor-operated 21 valve performance which identified the problems that we havo l
22 seen during these visits to date, and also laid out the 1 23 elements of a good motor-operated valve program, and we have 24 provided those, that information to the, all the plcnt t
(
() 25 managers, and all the maintenance managers in the industry at HERITAGE REPORTING CORPORATIOli -- (202)628-4888
403 1 our recent workshops. One of them was September. -The other 1 2 just finished in' October.
3 Also provided~information to the participants in the 4 recent EPRI valve symposium, and just to make sure it gets to r
5 the right level, as we do evaluations, we provide a copy down
! 6 at the working level to make sure people really have that ,
7 information available.
8 (Slide) 9 HR. HOORE: Just to remind you what those 10 recommendations are--excuse me. Does somebody have a ;
11 question?- These are, these are the recommendatio.ls that we 12 primarily focus in on. A lot of those recommendations come 13 out of a fairly old SOER, but I think unfortunately that SOER 14 839 was a green SOER, but in the early days they were coded by ,
15 color and green was the least urgent and so probably we made a 16 tactical irror at that point in time and that resulted perhaps i 17 in not having as effective action as early as we might have 18 had it otherwise.
19 I am going to just mention this subject only because 20 we are involved in the Nuclear Maintenance Assistance Center 21 along with EPRI, but that's going to be covered in more detail 22 in just a minute.
23 INPO has a member of the Technical Review Committer ,
24 here, Jim Tills. I'm a member of the Steering Committee, < .d
() 25 so we maintain close couple witn this evolution and are HERITAGE REPORTING CORPORATION -- (202)628-4888 [
.~ ,
c 404
- 1 working to get a good product out of that.
- 2 (Slide)
'(vr 3 HR. HOORE: As Jim will tell you a lot more about j
t 4 that.
5 We will take a look at our current NPRDS 6 information. It indicates that there is a decrease in the 7 rate of MOV plants that have been occurring s4. ace 1987, middle 8 of 1987. I will show you a draft. It would tend to indicate 9 at least to me that industry initiatives have been effective 10 in identifying areas where emphasize is needed to improve the 11 program, performance nnd reliability.
12- It also shows that long-term reliability is 13 dependent upon a comprehensive motor-operated valve program,
() 14 and as I indicated just a minute ago, we have talked a little 15 bit about, passed out infornation on the elements that would 16 constitute such a program, and of course, we believe that 17 future NHAC motor-operated technical repai: quidelines should 18 give assistance in both assessing and improving these
- 19 programs.
7 20 Let me show you the picture of that current vugraph, 21 and I am going to put it up here with a couple of caveats, but 22 it is, the dashed line across the middle is the average. It 23 shows--if you look at the failure rate, Limitorque failure i 24 rate, you will definitely see a downward trend that has
(} 25 occurred since 1985. Again, this is based on, this is based i HERITAGE REPORTING CORPORATION -- (202)628-4888
7 -.
405 1 on our data from NPRDS, and the caveats are that the results 2 are just as good as the data input. The data input has
}
3 certainly been getting better as time goes on, and but again, 4 .it is too early, and the other caveat is it is just too early 5 to really hang my hat and.say we have got all the problems in 6 the world solved. We have a little ways to go yet.
7 Are there any questions on motor-operated valves 8 before I proceed?
9 HR. WOHLD: If I could make a comment if I could on 10 the training, one of the problems that--I was inspector for a 11 while in Region 3, as you may know, on the bulletin, and one 12 of the problems I saw at the sites was a misconception in the 13 maintenance departments about what constituted appropriate 14 activities. Since the training on site is also put in effect 15 by the same people that are in the Maintenance Department and 16 so forth, these misconceptions never get addressed on site.
! 17 And I was, I had the thought that maybe t training center 18 outside of the local utility that had the correct and 19 identified proper training program might be the answer to some f 20 of these things.
21 To give you an example, one utility I looked at had 22 a policy of shining the trends on the valve stem, which the 23 valve just doesn't, the operator doesn't produce it or the 24 training program can be great. It can meet all kinds of
() 25 requirements, but if you have these misconceptions it can 1
i HERITAGE REPORTING CORPORATION -- (202)628-4888
'I 406 1 defeat the entire program, and I nad a' thought thas perhaps
~
2 INPO or whatever could set up a ; raining center that could O- 3 have~a quality training program that would have these type'of 4 things in it or would have a better tendency to be identified.
5 That's just a comment.
6 MR. HOORE: You know, it is a good comment. If the 7 training isn't any good, it isn't going to help anybody, but I 8 think Jim may have a comment?
9 MR. LANG: There are commercial organizations 10 available that provide good training courses. Power Safety is 11 one of them. Other organizations, other organizations provide 12- training courses.
13 MR. HOORE: I was really thinking in terms of your
() 14 repair guides that are coming out, they are going to be very 15 detailed on how you disassemble and assemble valves.
16 MR. LANG: Detailed guidance, but we won't be 17 offering training. We will be providing detailed guidance on 18 how to maintain.
19 MR. WOHbD: The training we talk about, one would be 20 the technician setting the switches, and changing the 21 lubrication. The other training for the management type 22 people that would address, for instance, the attachment B in 23 this generic letter the low voltage issues and all that sort 24 of thing, We are advocating a pretty detailed
{} 25 HR. HOORE:
HERITAGE REPORTING CORPORATION -- (202)628-4888
ll' ~> ,
_4 407 1 training program that gets into the theory of operation, and 2 so all the mistakes you can make.
3 CHAIRMAN HICHELSON: You are advocating that as 4 being a training program to be run by the particular utility?
.5 HR. HOORE: By a utility.
6 CHAIRMAN HICHELSON: I think the point is that--I 7 believe it is a valid one--that perhaps the home grown i
1 8 teamsters are not the best in this ca9e since they have 9 developed certain kinds of folk lore and whatever which would f
10 just be reflected in the training, and you know, you won't 11 make--these instructors might be very, you know, not like the 12 new approach. They might be teaching what they thought was j 13 the best practice of the last 20 years.
() '14 How do you get out of that syndrome?
15 HR. HOORE: We certainly will admit that's a 16 possibility, but we can also cite many examples of excellent 17 training for those kind of people, 18 HR. MICHELSON: I am sure there are.
19 HR. HOORE: Also cite what is happening in the area 20 of simulator training for operators by virtue of really making 21 sure the instructors know what they are talking about.
22 CHAIRMAN HICHELSON: Chances are the people who have l 23 got real good instructors probably, on site probably don't 24 nt d this instruction program anyway. It is the utility that
() 25 is weak in this area to begin with, and the instructors on i
! HERITAGE REPORTING CORPORATION -- (202)628-4888
(
J!
408 I
-l 1 site might'be the reason for-those, j 2 HR. MOORE: I'll take your comment', and I think it s 3 is a good come, and we will take a look at it as we~ work our 4 way further. ,
5 .Any other questions on HOVs? ,
6 CHAIRMAN MICHELSON: Hight even be one can consider 7 having a small group of instructors that you can kind of send !
8 sround to the utilities to completely instruet. That would 9 even cure this home grown problem.
10 HR. HOORE: Let me make one more comment on 11 training, and that is the training addresses both maintenance -
12 training because as you point out, maintenance -people can make ,
13 a lot of mistakes and the people that used to be able to tweak
() 14 and adjust the torque and limit switches are probably local 15 heroes because they keep the plant on line without realizing 16 they might be impacting some design factors, and so that
i' 17 changing.
18 The ether aspect is the operators don't realize they 19 are changing what happens when they went around and tightened 20 the packing, for example, so we have to train both the 4 21 operators and the maintenance crafts, and so our training i
22 program addresses both.
23 Third is that the accreditation program, we have l
i 24 hupes--this is really getting going good in the maintenance
(} 25 area now. We have hopes that is going to do a lot to help HERITAGE REPORTI!iG CORPORATION -- (202)628-4888 l
gu n
409-1 this kind of an area, the. maintenance thing. Maintenance has 2 been kind of a, nnt had the best training programs in the ,
3 country in the past. We are working to change that. 1 4 Any other questiens on MOVs?
5 MR. WOHLD: One more, a question on are you looking 6 at the, at all at the possible.qualificat' ion of mechanics?
~
7 MR. MOORE: We look at qualification of mechanics, 8 yes, sir. And what we are trying to do is make surc the 9 person is qualified to do the job that he is assigned to, and
.10 that can be done in many ways. It can be done by not ;
11 necessarily just training. It can be done with excellent 12 supervision. If somebody is qualified, make sure he does the 13 right job. It can be done with a gcad on-the-job training
( 14 program. ,
15 Our bottom line check is to see if on the job he L
16 really Wnows what he is talking about, knows how to do the 17 business, and that's where we are really checking the process.
18 If we find he is not qualified, not doing the job right, then 19 we try to bring it back to that part of the organization that 20 is able to make those kind of changes if necessary, and it may [
t 21 be supervision or management. It may be the Training l 32 Department, but we are really focusing on the bottom line. [
l 23 HR. WOHLD: Thank you. ;
24 CHAIRMAN MICHELSON: From your exposure to tne MOV
{
() 25 problem and so forth, there is something I have kind of HERITAGE REPORTING CORPORATION -- (202)628-4888
, u .v - --
O l 410 n
l 51 ' puzzled about and try to get some advice on from time to time, 2 and that is this problem of over-torquing the packing to the 3 point where it doesn't function as well as it has to.
f 4 What kind of controls presently exist out in the l-
$ industry on what should be the proper packing torque? Is 6 there a valve manaual that says must not apply more than so 7 many pounds? What prevents the, you know, what gives the, t
t 8 what guidance is given to the mechanic as to how tight to make 9 it?
10 Clearly if it is dripping a little bit, he tends to ;
11 tighten it. That stops the dripping, but that might be the 12 very thing he shouldn't be doing, so how do you decide at what 13 point does he replace the packing or enother problem you run
() 14 into, if you replace the packing with a different kind of 15 packing or brand new packing, you may not be able to pack to !
16 the same spec initially. I have not found a spec on how, how J
l 17 tight to tighten packings in the valve manual, but maybe l [
18 you--my experience is getting pretty old now and maybe you j i i 19 have got better for that.
20 How do you prevent something as simple as i 21 over-torquing the packing? How do you know you have [
22 over-torqued it short of it doesn't work at al1 because well, 23 it is a nominal test to begin with. The fact it does still 24 work doesn't lend itself to perform the safety function, that l c:) 2s it w111 .erk. ;
l r
HERITAGE REPORTING CORPORATION -- (202)628-4888 l f
r l
411 1 MR. MOORE: I expect--
2 CHAIRMAN MICHELSON: This is a very practical f')
-(/
3 question.
4 MR. MOORE: If you need to--let me say from my 5 perspective, there is two different questions. One is we see 6 a need to do diagnostic testing that you adjust the valve.to 7 make sure the valve still has sufficient thrust to operate 8 under the design configuration.
9 CHAIRMAN HICHELSON: You have to come back in with 10 the MOVATS or something and redetermine that everything is all 11 right after you have tightened the packing?
12 MR. MOORE: That's all we can do.
13 CHAIRMAN MICHELSON: That would be one way.
() 14 MR. MOORE: We are looking for easier ways to do 15 that. That is time consuming from a practical point of view.
16 If you change the packing, it might be a design change.
17 CHAIRMAN HICHELSON: Let's take the practice 18 practical case. It is weeping a little bit of water. The 19 mechanic walks up and tightens it.
20 What kind of instruction do you give the mechanic?
21 You give him a quick torque or what?
22 MR. ELFSTROM: In our experience, first of all, to 23 adjust the packing on these valves requires in our case 24 maintenance work order. I think we are not unusual in that
() 25 respect, and maintenance work order in and of itself requires HERITAGE REPORTING CORPORATION -- (202)628-4888
r- ,
412 1 the use of applicable procedures, and that at that point also
- 2 applicable retest requirements.
v 3 How in our particular site, they are not necessarily-4 given torque values for standard packed valves or live-loaded f 5 valves and torque values that the packing will torque to.
6 However, retest requirements would require us in the 7 case of motor-operated valves specifically, to verify that we 8 have not decreased the amount of thrust capable, available for [
9 overcoming delta P. We do that by taking the signature 10 analysis from the motre control center.
11 CHAIRMAN M/4-:LSON: You go back in and do another--
l 12 HR. ELFSTROM: No. It is a very simple minding I 13 test. It is not the full baseline test. It is really fairly i
() 14 non-intrusive. We go into the motor control center and you 15 can do it in some case using the permit signature. We use 16 essentially motor power, and stroke the valves, and we already 17 have a baseline signature from whenever we did our baseline 18 test, so now we can compare what we see at the tightened 19 packing to our previous baseline and see if we have changed 20 the running load, i.e., increased it, as such decreased our 21 margin of thrust for delta P.
22 CHAIRMAN HICHELSON: I thought I was kind of 23 convinced yesterday that that didn't work.
24 Pete, do you have a problem on, comment on that?
(} 25 HR. WOHLD: Well, there is problems with the HERITAGE REPORTING CORPORATION -- (202)E28-4888
413 1 saturation of the motor, and there is, I understand there has 2 been some problem with reading the motor control center. I -
3 don't know.
1 4 CHAIRMAN HICHELSON: That is the way you normally do 5 it?
6 HR. ELFSTROM: Yes, and we have, we have probably
[i; '
7 gathered data on over a hundred' occurrences, and for example, j y
- 8 this last, we start the live load, some valves from the !
s q.
i [
9 standpoint of essentially getting us away from after the, ,
10 having to adjust packing and to keep it from needing i
?
11 adjustment, and we did the full baseline test before and after ,
12 to determine what our increase in load would be, and we have i
13 done load signatures after that, and because we have run into ;
( 14 a few problems, you can indiscriminately go thrnugh life {
i 15 loading the valves, but it has held pretty true. You have to 16 consider that especially on the larger the valve, the smallor i
the portion of the total thrust that you are, your running
[ 17
., 18 load actually should be unless there is someching wrong with (
19 the valve. for example. On a large valve with two and a half .
i 1
20 inch diameter stem, normally you shouldn't be seeing more than ,
i 21 three or 4,000 pounds of load just to overcome packing [
I 22 frictien. !
i 23 Now if you wide load that valve, that is going to go ;
i 24 up considerably, but these loads are fairly easily quantified f
- () 25 and measured in using that tyre of a device.
I HERITAGE REPORTING CORPORATION -- (202)628-4888 f
414 i
1 Current sensing, the only real difficulty you have
- 2 there is it depends on the, what you bought your actuators to.
l 3 In our case, we bought a lot of ours to 70 percent voltage )
4 4 requirements because of the saturation the motor. It really
, i 5 doesn't change very much for a significant change in thrust, ;
6 so motor load which takes changes in voltage and that into 7 consideration, base angle versus current, and it comes out to 8 be--it is very sensiti'.e. As a matter of fact, you can tell 9 quito a bit of difference just f ron, changes in the lubrication 10 on the puching on the stem gauge and that sort of' thing.'
11 CHAIRMAN MICHELSON: That's--is there generally in 3 12 youe practice and so forth, is that generally what you have [
13 found then, that a utility dces go back, any time they
() 14 readjust a packing or chango it, they go back and do at least 2 15 a partial re-examination of the basic input parameters to make ,
t
^
16 sure that the, it is still the, what they thought was the i 17 correct margin? Because they can't do a real test. That's 18 the kind of test you have to do. ;
19 Is that even dono, though, in most utilities?
1 I
20 MR. HCPHERSON: With the utility at San Onofre, l
21 there has got to be a maintenance order to go out and adjust ,
l
! 22 the packing. It has got to met the procedure requiring us to l l
l 23 go into instrunented test, ,
l !
24 CHAIRMAN HICHELSON: Is that the type of test you {
() 25 do, or do you do a different typo?
i i l
, 4 l HERITAGE REPORTING CORPORATION -- (202)628-4888 f
415 l 1 HR. MCPHERSON: We do current because ours are 2 bought to a different degraded voltage edition so we see f-)
- (._)
.9 changes in current. We just finished life loading a lot of 4 shutdown cooling valves and did complete diagnostics before 5 and after, knew exactly what the changes were, were alle to 6 comparc it to the current. :
t 7 You had asked about torque specs. Besides for the 8 life-loaded valve, there is a very specific X number of pounds P
9 to torque the valves to that. The other, the standard--
10 CHAIRMAN HICHELSON: Said valve springs.
r I
11 MR. MCPHERSON: On life loading, to life load the 12 packing, you use the springs and that's a very specific life 13 load value, but for standard packed valves, there may or may
() 14 not be a torque spec depending on the specific valve. There 15 is guidance in the precedures that is called out in the work 16 order that says obviously one thing is if you bottom out the 17 nut, you would repack it. ;
18 CHAIRMAN HICHELSON: Yes.
19 MR. MCPHERSON: But we instrument it after we test i 20 it.
21 CHAIRMAN MICHELSON: This was my impression. You 22 didn't have the torque values to use, the practice of whatever f 23 they are, and I just wondered how you checked after the good 24 practice to see if you still have the required margin in the
() 25 thrust? l HEi!TAGE REPORTING CORPORATION -- (2021628-4888
416
, t 1 HR. CURRY: Carl, I am from Philadelphia Electric.
2 At Peach Bottom we are going to the Chester packing of the
) l 3- gravity packings. We have established torque values for--in 4 addition, after they have been torqued, we run a most current 5 check on them to see if there is any change also.
6 CHAIRMAN HICHELSON: If they start to leak later on, 7 do you change, do you tighten then some more or do you--
8 HR. CURRY: They would be done by--the first effort 9 would be tightening. It would be using a torque wrench, 10 current stroking following that, again using-a maintenance 11 request form.
12 CHAIRMAN HICHELSON: Do you have a certain maximum 13 torque that you are allowed to go to before you replace--
() 14 MR. CURRY: It is not a maximum. What it is is a I 15 torque value with a band width on it on either side, and what I
16 they found I believe is they have solution dumped through the
]
17 consolidation of packing, something like that, so it is, when i I 18 it loosens up, it takas increased torque for you to return the f
, 19 torques to the earlier valua, and it wants to--consolidation
,' 20 is complete; then that torque value would remain relatively !
- i 21 constant.
! I j 22 CHAIRMAN HICHELSON: You do a current check? l t
23 !!R . CURRY: That is correct. f 24 HR. WYLIE: Can I ask a question? All these good l
(} 25 practices we are hearing, are these part of the EPRI r
I '
HERITAGE REPORTING CORPORATION -- (202)628-4888
o 417 1 guidelines that have been developed, for the valves as well as
,i 2 operators?
3 HR. LANG: It depends. If they are related to the 4 maintenance of'the valve, they are reflected in the valve S guidelines. If they are related to--
l 6 HR. WYLIE: Related to operation.
7 HR. LANG: If they are related to training of 8 people, et cetera, that's an issue through INPO in their 9 normal way of controlling that and issue a good practice to 10 the--
l 11 HR. WYLIE: I understand that. But the guidelines 12 describe all these good techniques for valves, and operators.
i l 13 HR. LANG: Yes.
l l
() 14 HR. WYLIE: The policy.
15 HR. LANG: It really concentrates on operators and 16 the interface.
j 17 MRs WYLIE: I know. I mean raost of what has been
- 18 aaid in today's presentation talked about just the operator, 19 not the valve, but you are doing this for the valves?
20 HR. LANG: The principal focus is on the operator 21 and where it interfaces with the valve, which is on the stem 22 and the sten threads, but the technical repair gui6e which I 23 am going to talk about, really doesn't address details of 24 valve maintenance.
() 25 HR. WYLIE: What does it address?
HERITAGE REPORTING CORFORATION -- (202)628-4888
- c. _
418 1 MR. LANG: There are, there are today procedures in 2 the plaats for valve maintenance.
u' O 3 MR. WYLIE: I understand that.
4 MR. LANG: We are at EPRI working on a valve 5 handbook, and in fact there is the application guide being 6 prepared for motor-operated valves that addresses matching the 7 system to the valve, to the operator, et cetera.
8 MR. WYLIE: The development of that, is the intent 9 of the industry to then train the industry on those?
( 10 MR. LANG: Those are resource documents that would l 11 be used in training.
12 CHAIRMAN MICHELSON: Could I ask one more--
l 13 MR. WYLIE: Let me pursue that.
() 14 CHAIRMAN MICHELSON: Go ahead.
I 15 MR. WYLIE: How, how is that information l
l 16 disseminated across the industry and uniformity achieved such 17 that you don't have the outliers?
18 MR. LANG: Well, if you would like me to address 19 that, or you do you want to take it?
20 MR. TIPTON: One question--I have to, we plan to do 21 it the same way we did with the check valve guideline, sending 22 it to--that's my understanding.
23 MR. LANG: It will be sent to all nuclear utilities 24 in the U.S.; the guidelines will be sent, and as you
() 25 unde rs t arid , INPO also monitors the performance of the HERITAGE REPORTING CORPORATION -- (202)628-4888
/i 419 l .'
1 utilities, nuclear utilities in the U.S.
r 2 Our approach is to provide the utility with 3 information, and our judgment was that they have the incentive 4 to use those to improve safety and productivity. Then we have 5 the industry organization INPO that really goes in and looks 6 at how the utilities are performing and provides them feedback 7 on their program, but that's the extent of what we are doing 8 from a programmatic standpoint.
9 MR. CARROLL: I have kind of a related question that 10 might help.
11 You indicated that at the managers and maintenance 12 superintendent workshop you sensitized people to MOV 13 performance problems.
() 14 I guess we have been hearing a lot of discussion on 15 the NRC staff that the responses that they got from 85-03 were 16 pretty much a mixed bag. Some of them, some of the programs 17 have not been approved because they couldn't make heads nor 18 tails of what the utility really was going to do.
19 Do you feel the status today is that plant 20 management is pretty much sensitized to this issue that has to 21 be dealt with and deal with that in a timely manner?
22 MR. MOORE: I think plant management is sensitized 23 to the fact they have got to make their critical MOVs work, 24 and we have done that in every way we can think of, but the
]
() 25 biggest thirg that we are doing is that we are following up to HF".ITAGE REPORTING CORPORATION -- (202)628-4888
420 1' see that something is really being done, and'we are obtaining i 4
-2 -commitments from the utility when we go there for the problems l O 3 we see that they are going to fix those in a certain amount of i 4 time..
5 MR. CARROLL: What are you finding in the last few I
6 months on evaluation visits? Are most utilities moving ahead l 7 with this program?
8 MR. MOORE: In my opinion, yes, sir, but they are at 9 a different place in the program. I wouldn't argue that.
l 10 Some of them are well along and some of them have got a long 11 ways to go, but I think they are all moving in the right 12 direction. r i
1 13 MR. CARROLL: Have you found it necessary in any
- i
() 14 cases to disagree with plant management, take it to a higher ;
i
. 15 level of management and say to a CEO, hey, VP Nuclear, your 16 guys at the plant just haven't got the message MOVs are 17 important?
! 18 MR. MOORE: We also sensitized the CEOs at our CEO 19 conference last fall. We didn't leave anybody out.
I
! 21 MR. MOORE: Yes, sir. And we also--you know, ever i
i 22 time we do an evaluation we give a report to the CEO. We sit l
23 across the table from him, tell him what his problems are, if l
I l
{ 24 HOV is one of the more significant problems, we get him and I
HERITAGE REPORTIllG CORPORATION -- (202)628-4888
1 421 l l
1 CHAIRMAN HICHELSON: small clarification--on the )
2 current monitoring that you do after you tighten the packing, 3 'is this a similar comparison of previous shapes of the ;
\
4 signature in magnitude, or is this frequency response i
~
5 analysis? To what extant are you monitoring?
6 HR. MCPHERSON: In our case, it is comparing it to j 7 che previous data. l 1
8 CHAIRHAN HICHELSON: By that you mean-- l 9 HR. MCPHERSON: In the amplitude and in the 10 characteristics, shape.
11 CHAIRHAN HICHELSON: You don't do a frequency i
12 response examination of the signature and that sort of thing? !
I i
1 13 HR. MCPHERSON: I am not sure what you mean b7 that, i 14 CHAIRHAN HICHELSON: There are techniques.
0 !
l 15 HR. MCPHERSON: Therefore I guess we don't do it. j 1
16 CHAIRMAN HICHELSON: You can understand what is .
17 happening in the valve from doing a detailed examination.
i 18 HR. ELFSTROM: You are talking about the in-depth
)
I 2 19 transformer, et cetera? i 20 CHAIRMAN HICHELSON: Right. That is not what you l 2
i 1
L 21 neant you do?
]
22 HR. ELFSTROM: No. What we are doing, and I don't t l
Y 23 know, that there are some utilities who have tried this with l
[
24 the help of outside venders, and it is being looked at, but !
l
() 25 from the standpoint that I think we arv all talking f
f HERITAGE REPORTING CORPORATION -- (202)628-4888 ,
. 422 l'
1 essentially magnitude, ovecall shape,.a standard of ;
2 overlapping training type of thing--
, g
)
3 CHAIRMAN MICHELSON: I wanted--it was nothing more !
4 sophisticated than that? l t
5 MR. ELFSTROM: No; fairly simpleminded.
t 6 CHAIRMAN MICHELSON: Yes. $
7 MR. MOORE: Going back to your previous ques'.icu, 8 Mr. Carroll, we don't find any problem with the people wanting '
9 to do the valves right. We do have a misunderstanding of what l 10 the proper standard is sometimes. It is kind of an iterative ,
r i
11 process to bring the standards into agreement. Some of the ;
i 12 gentlemen sitting on th's side of the table have had !
i 13 significant MOV problems in the past, put in considerable .
I 14 effort to upgrado their programs, so they can tell you what is
[
15 entailed in that process. It is a pretty extensive process.
16 Any other questions on that? Let me shift to check ,
17 valves then. ,
r 18 (Slide) 19 MR. HOORE: The last time when you were here I I 20 believe we told you that we, wo were going to go out and take !
t 21 a look at ten, at our ten evaluations and see how people were {
[
22 doing in the check valve area. [
[
23 Before I get into that and tell you what happened f i
I 24 with that, lot me tell you some of the problems we have seen
() 25 and some of the conclusions we are drawing looking from our k
HERITAGE REPORTING CORPORAT. ION -- (202)628-4888
- - . -_= . . . , .
r 423
! 1 ~ data to'the larger size, multiple problems.- l 2 The interesting thing is we find more problems 3 showing up in the first few years'and also showing up at about 4 the 10 to 13 year point. Again, we feel the first ones are l U
5 'probably design errors, and the other ones down the line are l t
6 probably due to the wearing out type problems.
7 We find that the service water main feed and feed l
8 system valves are particularly prone to failure because there !
9 is a lot of, some of those systems have ,ntributed a great l
10 deal to the transients and SCRAMS that we t. ave seen in the f l 11 industry, and the most frequent failure is the stuck open [
L <
12 valve where.you have a gross leakage situation.
f L
l 13 Of course, we are very concerned about check valve [
j '
() 14 failure also because when it fails, systems perform in strange I t l 15 ways, and operators really haven't been trained to handle that j Y
l 16 kind of, or even recognize it sometimes. i b
I 17 We went back and took a look, wo went back and took f
i 18 a look at the ten utilities, and maybe I am getting ahead of 19 myecit, but our recommendation focus has been on, focus has l
20 been on those recommendations in SOER 83-3, which is the l
21 establishment of a prevent!.ve maintenance program and 22 performance of design reviews, that we find those--and if you 23 ~1ook at that SOER, specifically identified about nine systems 24 that wo thought people ought to take a hard look at.
() 25 Those included main steam, servica water, action HERITAGE REPORTING CORFORATION -- ( 20 2 ) 6 2 8 - 4 8 n' L __ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ . _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ . . _ _
4 424 [
4 1 - foedwater and main feedwater and diesel air start, and a few 2 - other systems, frca the design point of view in terms of .
- ~s I 3 proper valve location, proper size, proper type, proper !
I 4 orientation of the problems, and we find those design "eviews i 5 are in general started in the industry, and in ganeral, moving f I
6 toward, to getting those completed next year. (
l 7 Preventive maintenance, and we in addition to those !
8 systems that I mentioned, we also say that if, also if you [
9 have recurring failures in any other system, you should also 10 include that in design review, and for those systems we [
11 recommended a preventive maintenance program, and we have Leen i 12 going out since the 1st of April to look at check valve [
i 13 performance, and we have been finding problems not so much in [
t
() 14 the design review side of the house, but in the preventive [
15 maintenance side of the house, j t
16 We really feel you don't have to wait to start the i
17 preventive maintenance side of the house, so we found at least 18 a concern nearly in every plant we visited since April. I I
T 19 In response to finding that kind of problems, we I i i I
20 initiated sore other action, and on this last bullet here I !
I i
21 mentioned that, that we made personal contact until we called i i
22 up every executive point of contact in the industry in July, j
[
t 23 wen
- through the whole sequence from San Onofre down to he f r
24 present point, poir.ted out the program is of signific..
l t
i
() 25 importance and he ought to get involved in it and get it f e
L n
HERITAGE REPORTING CORPORATION -- (202)628-4888 !
- - - - * . - .vmr,--,w-.%.,,---e.-.,---._,,-,-m.o.-,.w-~ _ . _ , _m % %~,m - we, w, - e w .,--,,w-_, em .. . . _ . -w---'
425
~
1 moving, and we are seeing people moving in this program at 7~ 2 this point'in time, so--
' V) 3 (Slide) 4 HR. HOORE: Got it from a gambler early in the 5 business, but our technical support. depar'. ment at INPO has 6 responsibilities for this area, the rest of us supporting him. ;
7 Again, I maationed we looked at preventive'
[
8 maintenance, the design issues. We take a look-and see hcw !
9 they are handling the problems they do have on aite, how they 10 are monitoring whether the valves are performing right. Many i i,
11 times, walking through the plant, you hear a check valve 12 making noise. Maybe that's as good tn indication as a lot of ;
13 ways. ,
t
( 14 We also take a look at the overall reliability.
15 Again, we use our data system to look at the component failure ;
16 analysis and the failures'before we go to the site to see what 17 is going on. I will mention here again since mid-year '87, 18 our data is showing a decrease in check valve failures ,
is industry-wide.
20 We also, as I mentioned earlier, asked the utilities ,
7 21 all to use the EPRI guidelines, check valve guidelines in 22 doing their design reviews, and we are enphasizing that as we [
t 23 go around and we find that utility all are doing that and when [
[
j 24 they have a failure, they are using those guidelines. ;
' t
() 25 Just quickly taking a look at the graph that comes f l l t
HERITAGE REPORTING CORPORATION -- (202)628-4888 [
426 1 out of that--
2 (Slide)
-O 3 HR. HOORE: It is not a significant decrease, and 4 again, I would caution you to not to draw any conclusions on 5 that, but I would rather have the curve moving in this 6 direction than moving in the other direction.
7 HR. HOORE: Let's see if I missed anything. We are l 8 seeing progress in this area, gentlemen, and I have--are there !
9 any questions about check valves and what we are doing at j 10 INPO?
i 11 We are just starting to get into air systems and F
12 air-operated valves As you know, we have issued several 13 SOERs, our latest one being 88-1 that came out in May of this
() 14 year, and in that particular SOER there is a number of 15 recommendations which I will come to in a minute that tell you l 16 how we look at that problem.
17 CHAIRMAN HICHELSON: Was that 88-1 dealing with air t
18 systems or air-operated valves, or both?
19 HR. HOORE: It is really dealing with both. It f'
20 identifies failures in air-operated valves and then goes back 21 and points to the system problems that are causing a lot of [
22 those failures. ;
23 (Slide) f
?
24 HR. HOORE: Some of the more common problems that we
(} 25 are seeing is contamination of the air systems by dirt and i
HERITAGE REPORTING CORPORATION -- (202)628-4888
427
1 wear products and moisture and hydrocarbons that end up ,
2 fouling up.the valves and causing them to malfuncton. Some o' those' problems are pretty basic, dome of those probluas are !
3 4 caused by using the wrong kind of filter, wrong size of filter i
5 rather than not changing the filters often enough, not keeping i 6 the air systems dry, pretty simple maintenance type things, 7 but again, some of those haven't been done as well as th.ey 8 could be. We also are finding a few design problems as f 9 mentioned on this piece of paper.
10 (Slide) ,
i 11 HR. HOORE: Our approach to the problem is by r
12 following up cn the key recommendations in SOER 88-1, and t
13 again, it looks like the check valves. When air valves fail,
( 14 we really probably haven't written good procedures for a, a 15 trained operator how to handle it or recognize it, and so we
- 16 are focusing on trying to get, from a training point of view, l L
17 making sure the operators can handle the problem when the f 18 problem comes up, and from a procedure point of viewing, j
\
19 giving them some tools to work with; from a maintenance point 20 of view, set up a program to monitor the air quality, ensure 21 the preventie maintenance is in place and is being donc, and f i
22 again, from a design point of view, make sure that we have l 23 sufficient air available to make, to operate the valves when l i
24 they are needed.
()
25 Again, additional training is pretty key, and I I
l HERITAGE REPORTING CORPORATION -- (202)628-4888 !
I
y_----___ _ _ _ _ . - - _ . _ _ - _
't FN -
. -i 1
428 l I haven't really. emphasized that too strongly, but we really l
2 expect again that training, again possibly with accreditation l
^
3' programs providing support for that, is going 'o Le.a key part 4 in showing improvement.in this area. Again, our NPROS data l S shows a decrease in failures based on the valves we are ;
6 looking at, and again, as you gentlemen know, the NPRDS
- 7 doesn't
- over everything, fionetheless, we feel that there is 8 a correlation betyeen the results of the things that NPRDS !
9 does cover and the valven that aren't covered by the system.
10 ($lide) ;
i 11 MR. MGJRE: What we have found is all these three 12 val.e problems that we have been talking about are part of our
. r l 13 key equ'.pment problem list, but we have been taking them on as ,
t
() 14 we have enough resources to cover them and we are just getting i
15 started looking at the area at the present time. i i
! 16 CHAIRMAN MICHELSON: Could I get a clerification on f
?
17 your previous slide? When you talk about pneumatic valve i 18 failure rates, you are including any failure at the airsets j i 19 associated with that particular pneumatic valve? Failure in 20 either the airset or the valve itself? !
21 MR. MOORE: In the air what? !
22 CHAIRMAN MICHELSON: The airset, the control valves i I i
- 23 that control the air to the diaphragm operator, if it is a ;
i 24 diaphragm; is that included in your failure rates, or just the [
t ;
l j
(} 25 operator on the valve itself and the valve? f i
i i
HERITAGE REPORTING CORPORATION -- (202)628-4888 [
_ , - , _ . _-__ ,_ _ . _, _ , _ _ _ , _ - , _ _ _ _ _ . _ _ _ _ _ . - _ _ _ _._ _._i
0 '
429 5
'? 1 HR. HOORE: You know--
s 7
2 CHAIRMAN HICHELSON: Clearly if'the airset quits, 3 tne salve is done.
1 4 HR. HOORE: I would think it does, but do you know 4 -
5 the answer, Rick? If the vaive doesn't function, wouldn't I
6 function, and the control valve wasn't working? !
7 CHAIRMAN HICHELSON: Looking at your failure rates, t
8 it would seem to be perhaps a little different than we heard l
9 earlier today, and I was just trying to puzzle maybe you are 10 just talking about the valve and its operating and not the [
l 11 control airsets operator.
{
t 12 HR. HOORE: We will take the question and get back j l' !
- . 13 to you.
() 14 Yes, sir? l I
15 HR. FITZGERALD: bill Fitzgerald with Fisher i i
16 Controls--I would be interested in air operators problems in 3
17 particular. My experience over the last two and a half years l 18 has been that there is a lot of failures not necessarily in [
f 19 safety-related valves, but just in control valves in general !
i 20- with the thing being over-torqued. We have already touched on i i
21 that a little bit. l 1
h 22 Did you see that? Were there failures, a lot of 23 failures related to that phenomenon?
t i 24 CHAIRMAN HICHELSON: Ordinarily we don't let the r
! (
() 25 audience ask the speakers questions. In this case, I will
- {
- HERITAGE REPORTING CORPORATION -- (202)628-4888 f
430 -
i 1 . invite to you ask. .
2 HR. FITZGERALD: I'm sorry. !
3 MR. HOORE: I'm sorry, tool This set of data really ;
4 doesn't cover the balance of plant valves, so I don't really ,
L 5 know the answer to that. question, but I, I don't see this kind l 6 of failure as that kind of failure that you are talking about ,
i 7 in the SOER. It is not'part that SOER.
8 CHAIRMAN HICHELSON: Do you know if this valve !
i r
9 failure rate data was put together in some kind of INPO 10 report? Maybe we could read the report if you can tell us }
1 11 where it came from. !
b i- 12 HR. HOORE: I don't think so. It was put together j 13 just recently by the our Technier* Support Department in f
() 14 preparation for a presentation tc e valve symposium, and 15 that's the only place it shows at .
, resent time. j i
7 16 CHAIRHAN HICHELSON: If these are questions on that, j t
L 17 we will ask the fellott from the staff who made his !
i 18 presentation why this seems different from what he got on the ;
19 HPRDS. Who does he call to try to reconcile the difference so 20 he can tell me which way it is? ,
21 HR. CARROLL: He was talking this morning. Carl, l t
22 about solenoid valves, f 23 CHAIRMAN HICHELSON: But if the solenoid valve f
24 fails, then the valve that is controlling the air to it has !
25 essentially failed. These are valve failures. l
(]}
HERITAGE REPORTING CORPORATION -- (202)628-4888
431 1 1 HR, CARROLL: That is only a small part of the 2 population of solenoid valves in the power plant.
3 CHAIRtiAN HICHELSON:- No, no. These are 4' predominantly the airsets, the big population. That's what I i 5 asked him earlier. You remember _you are talking about l
6 solenoid valves where the valve is controlling the process 'I 7 directly as opposed to where it is controlling the air, big i 8 air operators, and pre-dominant population-is big air j I
9 operators, although there is a big population for heating, !
10 ventilating and several other things, too. This is very ,
11 interesting data because it seems to be quite contrary to what ,
12 I thought I heard earlier and I would like to get it ;
13 reconciled. l
( 14 Who should he call at INPO just to--is Till the i
15 fellow to call? l.
16 HR. HOORE: Not on air valves, you ought to caAl-- !
1 I l 17 CHAIRHAN MICHELSON: I will have him call you, but I [
don't thing you want to be bothered with details, b I 18 1
19 MR. HOORE: Gentlemen named John Broth or Mike f l
20 Scott. [
l 21 CHAIRMAN HICHELSON: Mike Scott? !
)
22 HR. HOORE: John Broth, B-R-0-T-H. !
l 23 CHAIRMAN HICHELSON: I think that probably, there l l
24 probably is a good explanation. I just want to get it f t
() 25 resolved.
i l HERITAGE REPORTING CORPORATION -- (202)628-4888 [
l
432 1 Thank you.
l 2 MR. MOORE: I guess in conclusion, at least from our
- 3 perspective, we see valve performance improving. I don't mean 4 to say that we are out of the woods, but we have got a lot of 5 work to do yet, but we certainly think it is moving in the !
6 right direction.
7 The initiatives we have'in place are moving to 8 further, move us in the right direction. We have found the r
9 industry has been very positive in their responce in trying to 1 10 solve the kind of problems that we have identified here. They [
c 11 have a vested interest in making it right also, and again I !
12 point out our evaluations continue to focus on valve f 13 maintenance design and operability issues. Certainly ;
() 14 reliability issues, we are just as concerned about anybody
~
15 else that is having reliable safely operated plant.
16 CHAIRMAN HICHELSON: What kind of thoughts perhaps [
, 17 have you given to this question that we kicked around a little (
18 earlier about perhaps, you know, suitability of site doing its [
1 t 19 own training versus the INPO providing the training and so I
4 20 forth? -
i i
21 Anything really in the cards for INPO to du that I
, 22 sort of thing?
i l 23 MR. MOORE: No, sir, not right now. We would be [
!i 24 more likely to go out and accredit their training prog.am, to l
() 25 sit in and observe people teach the program, but that's what t
4 HERITAGE REPORTING CORPORATION -- (202)628-4888 i - ,. - _ .
433' 1 we do the, accreditation program, and as'I said before, to 2 check the bottom line and see if the guys are doing the job O 3 right. I think there is.some advantage'to having'an on-site 4 training program. t 5 CHAIRMAN HICHELSON: Oh, yes. f 6 HR. HOORE: There is significant credibility gained 7 by having the people that are knowledgeable and belong to the 8 site giving that presentation.
i 9 HR. WYLIE: It was your intent to accredit the j 10 program, audit and accredit it? f 11 HR. HOORE: Yes. It is part of our maintenance 12 training effort.
l:
13 CHAIRMAN HICHELSON: You mean you will, you do have i
() 14 underway thoughts of accrediting valves?
15 HR. HOORE: Not specifically, not specifically for [
i 16 valves, but in terms of people that do valves, training those people, accrediting the programs for training those people. !
17 l
18 CHAIRMAN HICHELSON: Would be one of the modules I 19 maybe? ;
20 HR. HOORE: It might be one of the modules. I don't 21 know the details of that. .
22 HR. WOHLD: I may say there is so many ways to set l 23 up these valves wrong, and just one pieci of misinformation is ;
i !
24 difficult. The recent case at Fermi, the supervisor in charge l
r
() 25 felt millrights were qualified to install torque switches and i
! HERITAGE REPORTING CORPORATION -- (202)628-4888 i
, 434 1 it was a misconception on his part.
2 HR. HOORE: I grant you there is a lot'of history.
3 HR. WOHLD: I guess I would, qualification was the f
4 wrong word I used before. I guess I think more in terms of 5 certification of, and that's all I would say.
6 HR. HOORE: All right.
- 7. MR. TIPTON: Mr. Chairman, Jim Lang will now cover i
8 the EPRI activities associated with valves.
9 HR. LANG: I was just going to summarize EPRI
- 10 activities on motor-operated valves, and they really fall into 11 three categories--the things that we have done in the past, ,
?
12 which really were related to control improvements in 13 motor-operated valves; the work currently going on, which is 14 going on under the Nuclear Haintenance Assistance Center; and 15 future activities, activities to start early next year that :
r 16 concentrate on motor-operated valve testing.
I 17 (Slide) i i
'18 HR. LANG: The motor-operated valve improvement work 6 19 that we did focused on a control and diagnostic device that l 20 was a permanent installation, and it was dubbed Intel 11 torque. j 21 It could be mounted on, can be mounted on the operator or 22 attached to the operator, and used to control the operator and i
23 also be used to provide diagnostics. l i
24 It is a fairly expensive device, cost us about O 2s sas.ooo. 1t costs mans times th t te ee the e eineerine te :
I l
l l
HERITAGE REPORTING CORPORATION -- (202)S28-4888 1 l
l 435 -
1 install one, so it is obviously not the answer to a maiden's 2 prayer.
3 It is commercially available. It is being, it is ,
l 4 installed in two plants for demonstration right now and a !
i 5 report is available on it as noted here, and a final report ]
6 documenting the experience with it is due for publication 7 latter this year.
1 8 CHAIRMAN HICHELSON: Which units is it installed in?' !
9 HR. LANG: It is installed at Comanche Peak, and at j i
10 Point Beach.
[
i 11 CHAIRMAN HICHELSON: I assume this is a modified }
i 12 version of what we heard from you people about a year and a .
13 half, two years ago or so? !
i
() 14 HR. LANG: I assume it is. [
15 CHAIRMAN HICHELSON: Microprocessor and the whole i i
16 thing; at the time it was thought it was going to be at Duke. [
17 HR. LANG: It was originally going to go into Duke, !
i 18 but we could never get it in there because of the schedule, so 19 we changed.
[
20 CHAIRMAN HICHELSON: But it is the same one? [
21 HR. LANG: Same device; it actually has more appeal
[
22 on a new installation, new desigi. plant, than it does as a f l
23 backfit to an existing one. [
i 24 Now the Nuclear Maintenance Assistance Center, as j i
() 25 mentioned by Ed, has really been the focus of our work on j l
i L
HERITAGE REPORTING CORPORATION -- (202)628-4888 j
a 436 i
1 motor-operated valves right now.
2 The work of the Nuclear haintenance Assistance 3 center was undertaken at the request of NUMARC, but it became 4 what it is today, the old news, based on that letter that 5 Stello wrote to Warren Owen.
6 One of the first, well, the thing that NUMARC did is 7 turn to INPO and NHAC and say has NHAC been back on work on 8 the technical part of it, INPO handle the programmatic part of 9 it, and you two keep yourselves coordinated and work together.
10 One of the first things we did at NHAC was to put together aa 11 technical advisory group to tell us what we really ought to be 12 doing, and this group is composed of utility experts, three of 13 whom are in the room right now sitting at the table, vender
( 14 representatives, including those from the venders, Limitorque, 15 some of the service organizations, and an INPO expert, Jim 16 Tills, and I don't use the word expert lightly because one of 17 the things that we found when we got into putting this 18 together is, is that the real expertise on application and 19 operation of these operators and on the valves resides in the 20 utilities, okay. Doesn't, venders don't have a corner on it.
21 Limitorque is smart about their product. Diagnostic venders 22 are smart about their products, but the guys who really know 23 how to put it all together are the guys who use it in the 24 plant, so it has turned out that not only have those been the
() 25 pr,ople that will guide us, those have been the people that HERITAGE REPORTING CORPORATION -- (202)628-4888
54 i
437 ,
k 1 provide the input word for word, and we have taken the }
2 document, gone over and over and'over it, and they have l
3 actually changed detailed words.in here to make it a better j 1, i 4 document. Our emphasis has been on accuracy and usefulness. t 5 (Slide) !
t 6 CHAIRHAN HICHELSON: How big--this technical i 7 advisory group, are they the ones that are putting together i 8 these words? -
i 9 HR. LANG: We have had a contractor put together the l r
10 words, sit with an operator in front of--put together the l l
- 1 words and so on. Then we have subjected it to the review of [
f 12 this group, and the review has been sitting down in a meeting l 13 together and go through it word by word.
() 14 CF\IRMAN HICHELSON: How large was that group?
t 15 HR. LANG: I think the total number is about 12. P f
16 CHAIRMAN HICHELSON: Thank you, t
17 HR. CARROLL: I am getting lost. The words we are ;
18 talking about are now what you have up there? l t
19 HR. LANG: I am going to put then up here. This is 20 the first document. Okay. This is the first document. The !
21 technical advisory group said okay, let's have a technical !
t 22 repair standard. It says once you have decided that you have j 23 to maintain the valve or that you have got the valve installed f l
24 or that you need to set it, tell me how to set it, tell me how [
() 25 you to set the torque and limit switches, give me the t, HERITAGE PEFORTING CORPORATION -- (202)628-4888 L t
438 1 guidelines for corrective, preventive, and predictive )
2' ' maintenance, post-maintenance testing, information on failure 3 modes, spare parts, everything I need to know to maintain ,
I 4 4 this. !
5 This is currently in final type setting. We 6 finished the review by the technical advisory group last -
7 month, and we are setting the type right now. f 8 CHAIRMAN HICHELSON: The packing would be a part of f 4 3 this? r l
, 3L the packing? [
t i
17 MR. LANG: The post-maintenance test requirecents :
() 18 are in there. The first edition of this addresses--the SHVII 15 operators are on rising stem valves, gater globe valves, but {
i 16 the guidelines for 1crger sizes will follow. We decided to f
I h 17 keep the scope small enough so we could handle it. Even as it !,
[
18 was, it turned out to be a horrendus job. Once we get this j i
19 out, the next month, the guidelines for the larger sises will (
L l
20 follow.
21 (Slide) {
d
}
22 HR. WOHLD: Let me ask a question. f 23 HR. LANG: Yes. l t
j 24 HR. WOHLD: Now if you take something liPe this and j 1
(} 25 had a training program around it that was essentially HERITAGE REPORTING CORPORATION -- (202)628-4888
439 1 competent, I think you might get the quality that's needed. ;
2 HR. LANG: I think you are right. I think that the )
)
3 utilities will take this document, will form their training !
4 program using it as one of the tools available for them. You t
5 will have an organization like INPO that accredits training, ,
6 and you are going to end up with e, with a quality training l 7 program on this as well as the myriad other subjects that have l 8 to be trained properly, so I think you are right.
9 HR. CARROLL: Now if I was a utility maintenance l r
10 manager and I had your technical repair guide in front of me, .
11 and I got a proposed generic letter that listed 32 problems, {
12 would I be able to say oh, I know how to deal with each and [
t 13 every one of these 32 now? {
14 HR. LANG: I think--I haven't gone through the list 4 15 of 32 problems to confirm that, but I would be able, if I were [
j 16 the maintenance manager and had this document, I would have in j i '
i 17 front of me information on the key problems that can affect my (
h 18 motor-operated valves, and I would have information on what ;
19 caused them, and what could be done about them, i
1 i
20 HR. CARROLL: I would be interested in your doing an !
$ f l 21 item-by-item check. {
22 HR. LANG: I gathered from your question that you [
I i 23 would, but I just haven't done that. (
t i
24 HR. WOHLD: From the aspect if your guide was
{
l
() 25 followed, you wouldn't cause one of these 32 problems?
I 1 !
l I
, (202)628-4888 l _ __.___ _ . _ _ _ _ HERITAGE REPORTING CORPORATION --
\
440 j l
l
> 1 HR._ LANG: Well, you know, it was.really attempting 2 to say yeah, that's what is going to happen, but we are all i 3 practical people, and we have been out and we have worked on ,
4 equipment and we know that isn't the way li*e is. Problems !
5 are going to continue to occur no matter what you do. All we ;
6 need to do is put together the information that will help 7 people manage those, and that's what we attempted to do.
F 8 (Slide) ,
9 HR. LANG: The second thing that the technical i
10 advisory group said after ceiling us to put together the l-11 technical repair guide is that we really needed an application f i
12 guide, that many of the things that are cited as operator l
- 13 problems or operator maintenance problems really have to do i r
() 14 with linking the whole systen together, and they really wanted
, 15 information in a compact form on how to evaluate the valve 16 operator combination for compatibility with itself and ;
I 17 compatibility with the system if you are v. sing the right valve l i '
j 16 in the right place, at cetera, and provide guidance in that P'
i 19 document on how to accommodste existing installations, so we l
-j. 20 putting that together. [
l j 21 Now that's farther behind. We have a first dra*t of [
1 i 22 that coming out within the next month that we will get out to i t
23 the technical advisory group for review. (
24 CHAIRMAN MICFELSON: When you say publication
(} 25 schedule, do you mean that's when it goes on the street?
HERITAGE REPORTING CORPORATION -- (202)628-4888
$'c
-4
- -;e
, 441
$V n l 1 HR. LANG: "That's when it goes on the street.
2 CHAIRHAN HICHELSON: Thank you.
.O t
3 HR. CARROLL: It will be, it will provide some :
.4 guidance to the user on, on the new problem? I 5 HR. LANG: On what kind of problem?
6 HR. CARROLL: Friction problem, a
7 HR, LANG: It will have that in it, but it K
'! 8 identifies what is availab?e today, cites the shortcom!ng, and 9 says be cautious, because the part that I am going to talk 10 about next is something that we are going to do to try to fill 11 the holes, but part of the problem is there are holes.
I-12 Information is available today, that I don't think anyone has 13 the last word, and so the first, what we wanted to make sure
() 14 we did is provide people things to use today, but put cautions 15 around it where we knew that there are questions, so that's 16 the way 17, is currently writton.
17 HR. HINNERS: Mr. Chairman, can I ask a question 18 about availability?
l 19 CHAIRHAN HICHELSON: Yes.
20 HR. HINNERS: When you say on the street, what does 21 that mean? If I send you a check--
22 HR. LANG: On the street means you will get a copy 23 of it for free. I would be a fool not to give it to you, 24 right?
() 25 HR. HINNERS: May be the other way around!
HERITAGE RE?ORTING CORPORATION -- (202)628-4888
h 442 1 CHAIRMAN HICHELSON: Early '89 is not far off.
2 That's what I want to make sure.
3 HR. LANG: No. You are right. I have been accused 4 of being everly-optimistic before, but that's why I put early.
5 Early is sometime before July.
6 CHAIRMAN HICHELSON: Oh; I think the first quarter 7 is about early.
8 HR. LANG. That's what I think, too. I would say 9- January, but in fact it is not going to happen in January.
10 CHAIRMAN !!ICHELSON: I just wondered when to put in 11 my request.
12 HR. LANG: I keep a file folder of the NHAC 13 requests, and I keep the requests th:t utilities have made for
() 14 the technical repair guide, for example. Are people 15 interested? Yes. I would guess I--you know, again, I 16 haven't done a count, but I wvuld guess I have requests from
- ' representatives of 40 utilities right now for the guide.
18 CHAIRHAN HICHELSON: That guide is on the street, 19 isn't it?
20 HR. LANG: No, it is not on the street. It is soon 21 to be on the street. It will be before the end of the year.
22 It should be out in some form early November, late October, 23 and out in a beautifully published form in a three-ring binder 24 November, December. Okay.
() 25 The final thing that EPRI--I just talked about what HERITAGE REPORTING CORPORATION -- (2021528-4888 n
443 i 1 NMAC is doing. NHAC is concentrating on tapping the f 2 information that is currently available, and making it 3 accessible and usable for all the utilities.
4 Now back to EPRI, which is an R&D organization.
5 Okay one of the things that we have noted is that there has 6 been a, a lot of trouble utilities have had in complying with 7 85-03 and doing the in situ testing. You just can't set up 8 some of the conditions that people would like to set up, and 9 moreover, it is, we are using the plants as a test bed as 10 opposed to operating them productively to produce electricity, 11 which was why they were built.
12 What we would really like to do then is find, pull 13 together the information that is available, package it, look
() 14 at the information on performance of motor-operated valves, or 15 perfornance of operators, find the holes, if they exist, and 16 then go do the testing to plug the holes.
17 Now I'm not convinced that there are myriad holes 18 that if we can just get all the data together. There has been 19 a tremendous amount of data produced. The NRC is producing 20 some now. Utilities have produced, have provided reports on 21 testing done in accordance with 85-03. The calloway submittal 22 cited some testing, test data MOVATS had which is all utility 23 data, so our first task is to pull all that together in one 24 place so that we can see how well the subject is covered, and
() 25 then to get into the testing.
HERITAGE REPORTING CORPORATION -- (202)628-4888
444 1 You were interested'in schedule. This will go move t 2 out into 1990, and that then' concludes what EPRI is doing..
t 3 MR. WYLIE: Let me ask a question. 'You mentioned :
4 things you are doing here in valves. You also mentioned there 5 was a lot of other component programs and guides and you 6 referred to perhaps guides. .
7 Have you got a list of all of those guides, !
t 8 publications pertaining to maintenance and operation?
9 HR. LANG: I can get it to you. I could almost do i
10 it off my head, but I will give you a copy.
11 MR. WYLIE: there are not that many?
12 HR. LANG: There are ' tot that many. The Nuclear [
13 Mair.tenance Assistance Center just got started this year, and ,
t
() 14 based on the input from the Steering Committee that Ed is on, 15 and a rumber of utility executives, and NUMARC representative, r
16 we have prioritized inuues and we have started working at the f' 17 highest priority issues to produce maintenance guides. l t
18 naintenance information on those topics. ;
i 19 MR. WYLIE: Do you have a list of that that we could i
?
20 have? l I
21 HR. LANG: Yes, I will sit down and write it up. f 22 HR. WYLIE: That would be great. I 23 MA. LANG: Yes, !
i !
,' 24 CHAIRMAN HICHELSON: In order to do PRAs, it is i
< r
( () 25 necessary to have numbers concerning valve reliability. From I
i t i HERITAGE REPORTING CORPORATION -- (2021628-4888 f
\_ I
445
?
1 .some of the things we have seen and heard of late, perhaps the !
r 2 numbers we are using are not too good. They may be too '
O 3 conservative, not realistic.
4 Does EPRI have-anything in mind trying to como up l 5 with better numbers for PRA purposes ccncerning valve i 6 reliability?
r
[
. 7 HR. LANG: No, I caution you that that's a
- l l 8 horrendus task because--
1 I i
9 CHAIRMAN HICHELSON: Yes. It is not small. [
a L
, 10 HR. LANG: The reliability varies with application, !
11 so you can't just test the valve at a given size. You have to :
12 worry about how it in used and where.
]
13 CHAIRMAN MICHELSON: I was thinking more in terms of
() 14 if the industry is going to be required to accumulate certain 15 kinds of information under this new generic resolution, it 16 might be nice if they would accumulate it in a. uniform i i
I 17 fashion, that the industry would figure out how to do, and 18 such that the data could be used by people who tcy to come up 19 with the reliability numbers which might 1*e EPRI, for [
l 20 ins.ance, or at some pcint it would be nice to think that that (;
21 was one of the useful outputs of going through all the generic j 22 letter work, i
l 23 MR. LA'N : Well, faced with adversity, the industry ;
t 24 usually pulls together and gets semebody to set standards and l
() 25 gets somebody to do things like that for them, so I suspect I
i t
i
! HERITAGE REPORTING CCRP0 RATION -- (202)628-488S ;
__ ___ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ . _ _ _______________.-______________________.___1
446
,1 that's not out of the question.
2 CHAIRMAN HICHELSON: I guess the prime mover will be O 3 at such time as your numbers are really challenged that you 4 are using in the PRAs, either come up with something to defend 5 them or come up with a new set of more realistic numbers if 6 that's the case.
7 There is nothing in the wind as far as you know?
8 HR. LANG: No.
9 MR. CARROLL: I think somebody ought to look at that 10 issue. If you believe the analysis we are getting, we have 11 got a very perious problem with MOVs at risk.
12 HR. LANG: I think somebody else will address that 13 here today, and I'm really not prepared to.
() 14 CHAIRMAN MICHELSON: No. I just wondered if EPRI 15 had in their research which was your last slide anything along 16 that line yet?
17 HR. LANG: No.. Not yet.
18 CHAIRMAN HICHELSON: Okay.
19 MR. TIPTON: Mr. Chairman, that concludes our review 20 of the principal industry activities from INPO and EPRI. What 21 we would like to do now is we would like to change our focus 22 to the draf generic letter.
23 As I mentioned, we met with the staff last week, 24 October 19th. It was a very good meeting, good exchange. We
(} 25 would like to focus our corrents in three arcas--cost, scope, HERITAGE REPORTING CORPORATION -- (202)628-4888
. 47 1 flexibility. And we would like to start the presentation with 2 the cost itself. Bob McPherson from Southern California 3 Edison will make that presentation.
4 HR. MCPHERSON: I am Bob McPherson from the San 5 Onofre plant. I want to provide you some information we 6 compiled in the last two weekr, on the cost associated with the 7 generic letter.
8 (Slide) 9 HR. MCPHERSON: The first number up there was a 10 result of 24 reactor units of experience based on 85-03,
+
11 actual money spent to implement it in the industry--over half 12 a billion dollars. The 35.9 million dollars there underneath 13 I believe is what is in the value impact assessment.
() 14 Actually our numbers showed that it was 63 and a 15 half million dollars to implement Bulletin 85-03 in the 16 bulletin. The periodic testing totaled a cost of $420 17 million. That's based on about 75 percent of the work we will 18 do is going to be a recurring cost. In other words, there is 19 only about a quarter of it that you will do one time and not 20 have to repeat.
21 CHAIRHAN HICHELSON: I gueas you did, went past me a 22 wee bit fast.
23 Could you conpare your numbers here now with what 24 you think the staff's value impact statement, what you thought
(} 25 the numbers sitould be? I see the staff has gone through and HERITAGE REPORTING CORPORATION -- (202)628-4888
448 1 made these estimates. ;
2 HR. MCPHERSON: We based it on approximately $20,000 0- 3 per HOV.
4 . CHAIRMAN HICHELSON: I just wonder how do these 5 numbers compare with what the staff had in there?
6 HR. ROTHBERG: The $35,900,000 is the, what the 7- staff has estimated.
8 CHAIRMAN HICHELSON: How about, what you, do you 9 think that--
10 HR. ROTHBERG: That was our, it was in the range.
11 That was our best estimate.
12 CHAIRHAN HICHELSON: What does the industry think 13 the nunber is?
() 14 HR. MCPHERSON: Half a billion dollars.
15 CHAIRHAN HICHELSON: That's the--I want to make sure 16 it is the--
17 HR. HCPHERSON: That is correct.
18 CHAIRHAN HICHELSON: Okay. I just want to make
, 19 sure. It seens like the 570 million and 420 million, they are 20 somewhat divergent.
21 HR. CURRY: The S570 million is for the first time 22 you would have--
l 23 HR. MCPH2RSON: That's right. That is to establish j
j 24 the baseline to get the data for the generic letter as i
() 25 written. The $420 million is what it will cost us every HERITAGE REPORTING CORPORATION -- (202)628-4888
449 1 periodic cycle. If you look at the rest, the number of 2 cycles, say you have got 20 percent of plants doing this 3 testing overy, every three years. What it is going to cost is 4 three and a third billion dollars for the industry to go meet 5 the periodic and the baseline requirement of the generic 6 letter.
7 HR. CARROLL: Is there a conparable BNL number to 8 that?
9 HR. MCPHERSON: Brookhaven National Lab's number, I 10 believe it was four hundred--I don't know, 11 MR. ROTHBERG: Best estimate was 440 million.
12 MR. CARROLL: Period?
13 MR. ROTHBERG Total.
() 14 CHAIRMAN HICHELSON: Bottom number here, the 15 periodic, what nunber did you come up with?
16 MR. ROTHBERG: I don't remember what it was 17 periodically. I think we summed it up, total number oper the 18 plant life.
19 CHAIRMAN HICHELSON: That was the recollection I had 20 was about that kind of number.
21 HR. MCPHERSON: Our number in relation to the 22 Brookhaven National Lab plant life number of $40 million?
23 CHAIRMAN MICHELSON: Yes.
24 MR. MCPHERSON: Isn't shown on this slide. It is
-() 25 three and a third billion dollars. That is taking--
HERITAGE REPORTING CORPORATION -- (202)608-4888
450 i
1 CHAIRHAli HICHELSON: I am just trying to get the two' l
2 related, i O 3 MR. MCPHERSON: One of the requests that we were f
4 asked after the last October 19th meeting, what way can we [
- f. shave money off of that value.to make it not such a !
F 6 significant cost?
7 You can take a percentage of the number away by not. h 8 requiring so much des,ign basis teuting. There is obvicusly i
'l some fluctuation in there. We don't know how exact it n4y (
10 be--either a low estimate or a high estimate, but since this !
11 number was cortpiled from some real world examples of at least i
12 the three of us sitting in the room and three others that were t
13 here last week, there was 24 units represented and we came up [
h 14 with an average that represents a these numbers. l t
15 A significant way to lower these by an order of (
i 16 magnitude would be to limit thw scope of the letter to just j t
17 significant contributors to core melt frequency. Then we have l f
18 on? PAA that shows that that would be one sixth again what the ;
i 19 IE bulletin required. [
l !
20 If we focused this to be just significant l y l 21 contributors to cor e :telt frequency, then this nutiber gets l l
22 lowered by an order of magnitu . In fact, I said it was 63 l l I 23 and a half million dollars for ndustry to trplement IE l
l 24 Bulletin 85-03. It would mean an 11, Sil million expenditure l l
l 25 to go add the rest of the significant contributors to core i !
( ,
a
{ HERITAGE REPORTING CORPORATION -- (2021628-4888 I
- 0 '
B '
y 451 ;
1 melt frequency. i i
2 HR. ROTHBERO: Can I ask a question?
O 3 CHAIRL", 0:HELSON: Yas.
I i
l E l 4 HR. ROTHBERG: What is meant by significant? [
5 HR. MCPHERSON: Actually that is going to be i
i 6 addressed when we discuss scope. !
t 7 HR. TIPTON: When we talk about scope. we will go j 8 over the valves themselves.
I Keeping in mind we plan on a 9 CHAIRMAN MICHELSON:
{
l 10 Full Committee consideration of this whole business in f t
11 December, if NUMARC wishes to come in to talk to the Full j 12 Committee on this matter, please bring in a slide that I i
j 13 compares your numbers with the cost / benefit study that has [
(
l< 14 already been done so it is easy to track through, including 15 the core melt frequency numbers as you think are applicablo !
.I k r
16 versus theirs and so forth, so we can see where you think the [
17 differences are, because this is a little hard to track. [
18 HR. TIPTON: Yes, sir. (
I 19 HR. HCPHP.RSON: That covers the basics of what we I r
20 wanted to show on the inc.ediate costs. and I think--
21 MR. TIPTON: What we would like to do now it. focus t
22 on the scope itself and Bob Elfstrom fron Toledo Edison will !
23 cover that. [
24 CHAIRMAN HICHELSOP': Let me check one other j i
f 2
() 25 question. In terns of total cost. not considering benefit but f e
I t
' HERITAGE REPORTING CORPORATION -- (20:1628-4888 }
_ _ _ - _ _ _ _ _ _ _ _ - _ _ _ _ -f
5 452 1 just total cost, it appears that you are about a factor of ten 2 higher than the BNL report.
(4Y 3 Is that about a correct ballpark?
4 HR.. MCPHERSON: Right.
5 CHAIRMAN HICHELSOll: That's what I thought.
6 MR. ELFSTROM: 'At least; the one difficulty with 7 that, though, is thht 35, 9 number out of the BNL analysis was 8 over plant life If you are looking at it versus the way we 4
^
9 figure it, you are looking at a factor of a hundred because we 10 are talking 3 billion over plant life rather than 35, 9. What 11 we did was the 420 million that was being discussed by 12 Brookhaven National Lab was their estimated benefits to the industry due to reduced maintenance and increased 13
() 14 availability, et cetera, and it can be fairly easily shown l 15 that tnat benefit is sort of vashed out and then some. ,
16 CHAIRMAll MICHELSON: Okay. I'm just trying to get a I 17 feel for how much, how much difference you have with the B!lL 18 report.
19 MR. CARROLL: Did your dialogue with the staff--I 20 assume BliL was there. vid they--
21 MR. MI!!!!ERS : No.
22 MR. /"RROLL: They were not?
23 MR. MIN!iERS : !!o , sir.
24 C H A I R!!A ll M I C H E L 9 0 N : Is the staff going to go back c l 25 to think about these numbers?
[}
HERITAGE RLPORTING CORPORATION -- (202)628-4888
-4 .<t.
453 j i
1 HR..ROTHBERG: Tie don't know where their name and l I
- 2 numbers came from.
(
3 HR. MICHELSON: I don't. either, right now. l l
4 MR. ROTHBERG: We asked for their, you know, 5 breakdown of their numbers. I don't know if I am going to get 6- them or not.
7 C H A I P H ,'.N M I C H E L S O N : You certainly can't do a 8 comparison unless you get far more detailed breakdown than we y 9 have seen here.
[ 10 MR. TIPTON: We will definitely give them a i
11 breakdown on how we came up with the 570 and 420.
12 CHAIRMAN HICHELSON: Intuitively I know at least one 13 Committee member 1.1 coing to ask how dii the industry
() 14 cost / benefit study come out? You havs jot good, firm numbers, 15 then it is only a question of why they are different than the 16 staff numbers. You do have to have some good, firm numbers.
17 MR. TIPTON: We will provide that.
18 MR. ELFSTROM: Good. My name is Bob Elfstrom with 19 the Toledo Edison. My discussion is going to center on the 20 scope of the generic letter, and when we discuss scope, we are L
21 talking in relation to say the previous 85-03 bulletin as this ;
22 is all the time being estegorized as an extension of 85-03. ,
2 *) (Slide)
L 24 -
MR. ELFSTROM: The first thing that we would like to !
() 25 discuss is static testing versus design basis testing.
HERITAGE REPORTING CORPORATION -- (202)628-4888
-- _ _ ~ _ _ . . . _ _ . _ _ _ _ _ _ _ ___ _ . _ _ _ _ . . _ _ _ , _ . _ _ _ . _ _ _ __.
454-1 I guess to try to keep things in a position where
_ 2 everybody understande terms, it would be easy to say up front O that when we discussed the testing of these valves, the static 3
4 condition, we were ref er r ing to a diagnostic-bssed system typs 5 of test with no delta P, no flow.
6 .The design basis testing is exactly what it 7 implies--hgain, a diagnostic instrumented test under design 8 basis conditions. There has be;n a lot of discussion on 9 valves and friction factors, and I guess one of the things 10 that concern us is everybody keeps discussing this generic 11 letter as though it applies only to gate valves, and out of 12 the 165 safety-related valves that we had at Davis-Besse, 103 13 of those were rising stem valves, and probably less than 90 of
() 14 those were ge , valves, so it was almost 50/50. Globe valves 15 don't really give anybody a hard problem as far as friction 16 factor for the most part because they are essentially a 17 cylinder of water as you are trying to raise steam or 18 whatever. 'ihere is no sliding friction factors to be, to be 19 greatly concerned about. At least there hasn't seemed to have 20 been any.
21 But previous discussions as to generic formulas 22 provided by the manufacturers essentially gets back to the 23 same questica of your oefinition of design basis because it' 24 you take into consideration the fact that on quarter turn
(} 25 valves we are not at this time reasuring torque on these HERITAGE REPORTING CORPORATION -- (202)628-4888
t, t
455 1 valves, we are measuring overall performance of them.
2 If you look at it quite literally, those valves are 3 essentially being set up pretty much to the calculations that 4
4 have been provided by the manufacturers when the valves were 5 bought, and review is being done to ensure that the valve was 6 not being applied fn another, in a way other than was 7 originally intended, but it is going to give us some very 8 difficult problems because the Appendix A, for example,,
9 Section C of the generic letter, the wording is almost 10 identical to 85-03 with I think something thrown in along the 11 lines of document smd keep ycur information, and that's fine.
12 I don't know the exact numbers, but Mr. Eissel says 13 that a number of utilities have approved 06-03 programs at
() 14 this point in time. It would seem to me then that Appendix A 15 really is not necessary. If we have already got approved J
16 85-03 pr jrams, then the words from 85-03, if that was 17 satisfactory, should be satisfactory and we can use 18 requirements we already have established.
19 The 20 and 4 valve comparison that keeps being 20 referred to, to the best of my knowledge and through 21 discussion with the people involved, that applied to a 22 submittal by one utility--Calleway--who also had a cavsat in 23 there that--and in fact they actually did representative delta 24 P test, did not use this standing data base and 20 and 4 type
() 25 of a situation.
HERITAGE REPORTING C)RPORATION -- (202)628-4888
ngc -
456 1 There is no data. base that contains information for -
2 butterfly valves, quarter turn plug-valves, et cetera, et 3 cetera, and it is not going to be that easy to get it, so from 4 the standpoint of where we are coming'from, we seem to be ,
5 focusing on, an awful lot on one specific thing, i.e., the 6 concern for friction tactor and rising stem gate valves not 7 being credible to the original designer who purchased these 8 valves, and I guess %e did testing.on a total of about 40 l
9 valves and 3 of
- nose were full closures, and we had two 10 valvos tiat you can say will come out to probably have I
11 different friction factors higher than the initial standard 12 formula.would indicate.
13 However, it is our belief that these are nort of
() 14 outside normal boundaries, as indicated by the results at 15 Wyle. Again, they were going into two phase flow conditions.
16 Ours are a little strange also from the standpoint of high ,
17 speed and geometries, but for the most part, all the rest of ,
18 the testing that we did indicated that that formula was fairly l l
19 conservative. It is fairly accurate.
t i
20 CHAIR!!A!! !!ICHELSOft: Your testing was not under pipe ,
i 21 break conditions, for instance? Those flows were not l t
t 22 indicative of pipe break flows. That is much loss, and so you 23 have to look at the full, what kind of testing you do because 24 you are talking about the in situ advertising under 85-03, !
(} 25 not--and some laboratory testing.
f HERITAGE REPORTI!!G CORPORATIOli -- {202)628-4888 [
457 1 MR. ELFSTROM: Exactly.
2 CHAIRMAN MICHELSON: So quite a'few of these 3- problems may not show up when you keep the valves at the more r
4 nominallyfloaded conditions.
5 MR. ELFSTROM: But I guess my concern there is if 6 there is an actual concern that these fermulas are in fact no 7 good and can't be used, we don't have anything to put in their 8 place. I can't test valves for pipe breaka.
9 CHAIRMAN MICHELSON: I think that is perhaps the 10 present situation, and that's why the concern, of course,
^
11 about trying to resolve as quickly as possible whether or not 12 things are okay or not okay.
13 .MR..ELFSTROM: But you are telling me that we can't
( 14 make that determination becauce we can't duplicate those 4
15 conditions, and I agree I can't duplicate pipe break 16 conditions in the plant.
}
17 CHAIRMAN MICHELSON: Yes.
I 18 MR. ROTHBERG: The generic letter has a specific
) 19 paragraph that specifically excludes testing at pipe break 20 conditions in the generic letter. That specific paragraph is 21 copied verbatim from Bulletin 85-03. There has never been any 22 discussion of testing these valves at design basis pipe break j 23 conditions.
24 CHAIRMAN MICHELSON: That paragraph says that you f
l
() 25 must do the testing and as best you can determine whether or l
HERITAGE REPORTING CORPORATION -- (202)628-4888
f 458 1 not the valve could indeed perform under those conditions.
2- MR'. ROTHBERG: No. That test--
v
(-)s
~3 -CHAIRMAN HICHELSON: 'Then I read ~the paragraph 1 4 wrong.
5 HR. ROTHBERG: The paragraph says that they don't 15 havo to do it. They have to consider it in their design and 7 in their analysis, but it is not a test requirement.
8 CHAIRMAN HICHELSON: No, no. I didn't say it was.
9 I said-that you must, in whatever you do--this is called in 10 situ testing. Whatever in situ testing you do is for the 11 purpose of verifying that if the valve were exposed to those
-12 conditions, it would perform its function properly.
13 HR. ROTHBERG: To the extent practical I would say.
-() 14 CHAIRMAN HICHELSON: Well, the testing.to the extent 15 practical. It must perform its function. There is no .
16 question of practicality. So it is only--then you h ve to
> l 17 know about friction factors and so forth, because o,u are 18 going to go, you can only do analysis. Clearly vou aren't .
19 going to do this--
20 HR. ROTHBERG: I don't want to quarrel with the I 21 analysis. What he is saying is that they had to test. t t
22 HR. ELFSTROM: I was not saying that. That we--no.
p 23 There has never been any understanding on our part that we 24 were to test the line break conditions, no. The only point we b
(} 25 were trying to make is that saying that the generic formulas I
i HERITAGE REPORTING CORPORATION -- (202)628-4888
~ . _ _ _ _ _ _ . _- _ . . _ _ . . _ -
-459-l 1 that these valves were orovided under are no longer to be 1
2 determined to be acceptable. I cannot duplicate the testing 3 such'as that.which is being done, that was done at Wyle or 4 other places, and there is no guidance as to what is 5 acceptable other than similar tests for which there isn't i
6 enough data to do on a lot of these, on a lot of different 7 types of valves, for example, i
8 It seems to me that in accordance with previous 9 comments, where the 85-03 work hopefuly was not to be 10 repeated, that the methodology used to answer 85-03 should be 11 acceptable to be used for the generic letter. -
12 The second item---
l 13 CHAIRMAN HICHELSON: Before you go on to that, just
() 14 to make sure we are still together in reading item C of the 15 generic letter, it says that in part break in the line i 16 containing the HOVs should be cons'idered and analyzed 17 described in items A and B, and A and B require that you do 18 the fundamental determination of what the design basis is, and 19 B determine what the settings are.
20 HR. ROTHPERG: The line above that says--
21 CHAIRMAN HICHELSON: I know what kind of friction 22 factor to expect under C, the pipe break, in order to do A and 23 B, so they must somehow get the right numbers in their 24 calculation, and I think they were saying I'm not sure what
() 25 the right number is, for instance, on friction factor under HERITAGE REPORTING CORPORATToti --
(202)628-4888
n >
460
.. 1 blowdown conditions.
-, 2 MR. ROTHBERG: The bone of contention in my mind at k_
3 .least from the discussion here was that-there was some 4 requirement for testing under pipe break conditions. We F
5 specifically excluded that.
6 CHAIRMAN MICHELSON: Not at all; obviously you can't 7 do that, but.there is the real problem of how do you calculate 8 it so that you can set up your percentage properly?
9 MR. ELFSTROM: The only contention really at this 10 point is that the words in paragraph C are essentially almost 11 identical to 85-03's original words, and the only point to be 12 made was that the original philosophies and methodologies for 13 approved 85-03 programs that have been approved and reviewed
() 14 by the NRC should be acceptable for the generic letter.
15 And the other area of concern is that we are looking 16 for design basis testing on a whole family and groups of I
17 different types of valves where none has previously been done, 18 and there is no data in the industry at this point in time, 19 and unfortunately, they are very difficult valves in a lot of i
20 cases to reproduce the conditions on.
21 For example, I have a lot of valves that are 22 containment isolation valves that are 1 and 2 inch ball valves 23 that really the condition, the accident design basis condition 24 is almoct containment overpressura. They close against our !
i
() 25 stand IRT pressure 38 pounds of air. There really isn't any HERITAGE REPORTING CORPORATION -- (202)628-4888 i
461 L .,
-l way that I can duplicate'that kind of a flow condition at all' 7- 2 whatsoever. I could duplicste doing an LORT, the pressure, b 3 and_open it against that, bui it is not the same type of 4 thing. It is, that's the kind of-a limitation I am discussing 5 'i n this area.
6 CHAIRMAN HICHELSON: There are some other really r
7 dif ficult problems when you star'; getting into the details, 8 which you would eventually, and that is, for instance, when 9 you have isolation valves that are only few pipe diameters in 10 separation, one downstream of the other by a few pipe I
11 diameters, when the upstream,-downstream is signaled to close, 12 to be upstream creates a terrific turbulence in the process of 13 trying to pinch down.
.() 14 Is that going to stop the downstream valve to trip 15 out and torque? Well, if it does, then you are lost, and the 16 single failures--you still have to assume the single failure 17 criterion, but if one valve causes the other to fail, that's a 18 part of the same thing. I don't know how you get into 19 determining what--but tiiat is some of the stuff. We discussed 7 20 this a little bit yesterday. ,
21 HR. WOHLD: I want to make one comment on that !
22 friction factor. The thi.ng I was bringing out before lunch 23 was the, the concern for the equation, and what I think is 24 needed is a rational approach which you have done in
(} 25 implementing the Davis-Besse program. I HERITAGE REPORTING CORPORATION -- (202)628-4883
~ _ -.__,
p 462 1 'Other utilities that I have seen hire new people to
. ,ess 2 do the HOV program, they set the, all the valves up to the OEM L
.(_).
3 equation and walk away and feel like everything is great.
4 That's the kind of thing that I was concerned about. I think 5 any type of rational-approach that makes sense--what you have 6 in the Davis-Besse program is a good one, and I don't. take 7 issue with'that--but another aspect of this that I wanted to 8 bring out was the expectation of NRC is that the OEM equation 9 is not the basis'for the testing, and that it was needed to 10 bring'that out so that you and we can discuss that, and bring 11 it out without having a problem on site six months from now 12 when utilities adopt a program using the OEM equations, feel 13 they are home free and then NRC comes in and says hey, that's
() 14 not what we wanted, so I just wanted to bring out a 15 clarification on that.
16 MR. ELFSTROM: Yes. I think it is important to note 17 that in this discussion right here, again we are talking about 18 an equation for gate valves. I have not heard the same kind 19 of, at least in any of these discussions, concerns about the 20 equations for setting up globe valves, for instance, or the 21 quarter turn valves.
22 Now if that is an error, I would hope that that 23 would be clarified in the letter somewhere along the line. I 24 think most of the people in the industry are aware that the
() 2o standard friction facect concerns only the rising stem gate HERITAGE REPORTING CORPORATION -- (202)628-4888
463-1 valves. There'is a foundation for that. Now what the extenti 2 of it is, and how it really affects the plant in responding to 3 a design' basis accident, I feel still remains to be seen 4 because.I still feel it is a fairly rmall overall problem as; 5 compared to.the valve flat refuses to move or doesn't change 6 position.
7 CHAIRMAN MICHELSON: But you do agree if I 8 understand you correctly, that all MOV, you do understand that 9 all MOV valves are under this generic letter requirement? ,
10 HR. ELFSTROM: Yes, sir. That's one of the points I 11 was trying to make. We keep talking about this generic 12 equation which in effect we are talking about rising stem gate 4
13 valves, where in actuality the generic letter is talking about i ) 14 all types of valves that are motor-operated.
f 15 The same comments don't held true for all types of i 16 valves. In AEOD we said the generic letter is a major 4
17 expansion, and some of the areas that it expands in, in scope 18 over 85-03, are essentially in design review, I have heard the i I i 19 comment a couple of times now how the industry doesn't know j i ,
~
20 raat its own design basis was for its valves, and I guess from {
21 my own perspective and viewpoint, 85-03 required the industry i l.
22 to go out and review their design basis for their valves.
23 The purpose of owners groups getting together was 24 mainly from the standpoint of assuring that we were going in f
() 25 the same direc*, ion in our review process as far as the depth f t
HERITAGE REPORTING CORPORATION -- (202)628-4888 !
- - - - ~ ,-we- yem w- - ,-----w,--m-- --
...y.--,-------w- - a g e , w n- -o ,e.m ww m , w w w~m m m- n n -n e -e- ,,,- m m -e~---.-m--m --s~*----
m 464 1 'that was necessary to satisfy 85-03, and to share:information
~ 2 and to apprise each other.
3 For example, in the owners group th' I was-a member 4 of, and our high pressure injaction system,.my,particular 5 plant has a lower head pump than another plant does in another 6 part of the country. Sure enough, we have got a. request for 7 additional information how come delta P for the valves is 8 lower than your sister plant? That's the type of thing that 9 the owners group getting together mainly was'about, but 85-03 -
10 required us to do this design review for delta P.
11 liow we are adding additional things in here along 12 the lines of flow, if required, control logic of cable sizing, I I 13 of power supply degradation. I guess all of these things to a '
() 14 certain degree or the other is a part of the design basis of 15 an individual power plant, but some of these things I think .
16 are more unique than others.
17 The cable sizing issue, for example, to my 18 knowledge, that has only been a problem with DC motor-cperated I
19 valves, mcst notably in recent history the Brunswick example l
J 20 where the design was applied improperly. They also had '
i 21 problems with their control logic as concerned, concerning DC 22 operated valves. I do not know of any problems with 23 undersized scaling on AC motor-operated valves. l l 24 There is a large confusion factor in my mind as to J
25 what a degraded power supply is. We discussed degraded
{~}
Y l i i HERITAGE REPORTING CORPORATIOli -- (202)628-4888 l
[
" ~
465 1 voltage, and I don't' feel'that there was any disagreement 2 between the staff or the industry as far as whatLthat, what-
.O. 3 that referred to,'and this is no problem.- I'm still.not'sure ,
4 what degraded power supplyLis or exactly how far.I can go to 5- obtain that power. supply. Am'I. going to the MCC where a ,
6 breaker feeds the motor operatcr? Am I going to the next bus 7 or where am I goir.g to on this type of thing? So that that 8 seems sort of vague and nebulous to us. <
9 And like I caid, I feel that a lot of these problems j i
)
10 from the standpoint of power s'pplies and wiring size really 11 apply mainly to DC motor-operated valves. I don't know of any 12 significant history in everything that I have examined that ;
13 would indicate AC valves are a problem. l (f 14 Additionally, under retest requirements, we touched i
15 on this somewhat earlier, in that the generic letter flat, f 16 make no bones about it, says that the Section 11 isn't 17 satisfactory, isn't enough. I don't feel that that's a f r
18 surprise to anyone in the industry, and I don't feel that any ;
19 of us in the industry disagree with that philosophy at all, 20 whatsoever. Stroke time testing motor-operated valves to 21 encure operability, is sort of an effort in futility, not to ;
22 mention that the .;also sense of safety it gives you for f
i 23 something that may not work properly.
24 The implication there, though, is quite interesting )
l
() 25 from the standpoint that we are talking about all l
i HERITAGE REPORTING CORPORATION -- (202)628-4888 i l
466 1 safety-related motor-operated valves, and requiring, for 2' . example, packing adjustments.
'6 3 Normally a utility, to meet the letter of the law so ,
4 to speak, will stroke time test the valve, go on about their 5 business, and no argument. That is not satisfactory.
6 However, ASHE Section 11 stroke time testing is on a specific 7 number of valves. It is not necessarily all safety-related 8 valves. The interesting portion about this is not only does t 9 it tell you that Section 11 testing isn't sufficient, or needs L
10 to be a better test method. It also is telling you now you i
11 have to apply the test method to all motor-operated valves 12 safety-related whether or not they are under Section 11 test 13 requirements or not.
() 14 The thing that is said but unsaid, and personally I ,
15 would just as soon see it said, is this generic letter cannot 16 be accomplished without some sort of diagnostic system. Now I 4
17 think that the retest requirements in the periodic retest that i
18 we havo talked about previously, we need to look r.t it in some 19 detail just as far as what we are actually going to look at on l
20 these types of rett't, but in a way, I really think that we 21 ought to just flat out tell people that they need some kind of <
22 instrumented diagnostics rather than keep circling around the 23 bush when we all know what we are talking about.
I 24 And one of the things that we did discuss with the ;
25 staff, and correct me if I'm wrong, is that on these retest
(
HERITAGE REPORTING CORPORATION -- (202)628-4838 -
467 1 requirements and including the periodic retest specifically, "s 2 that that was not intended to be design basis testing type 3 requirements. That was one of the agreements that we had come s: 4 to understand in our discussion with the staff.
5 In other words, if I am going to do a rotest at a 6 two-year interval or whatever it happens to be, it,would be 7 instrumented. diagnostic testing, but not necessarily design 8 basis flow or differential pressure testing. In other words, 9 you get your baseline or establish your switch setting, and 10 then you can do it on a static test basis and determine that 11 you haven't degraded the machine to the point where it won't 12 function properly.
i 13 Limited diagnostics--the only reason that's up there
( 14 is it gets back to one of my little pet things is that there i 15 are a whole lot of different kind of valves we are dealing i t 16 with there. There is one system I have seen that hasn't gone i
17 into a lot of plants yet or been used very much that gives us
, i
- 18 somewhat of the ability to look at quarter turn valves where ,
r 19 you are measuring torque, not thrust. l 5 l i 20 We don't care about thrust anymore for a butterfly
[
21 valve. We care about torque. None of the other diagnostics ;
22 systems that I have seen have that proven capability to 23 provide that infor.sation. At this particular point in time, i 24 we monitor, in my particular case, the spring pack
() 25 displacement cutrent motor load, et cetera, on these actuators i !
HERITAGE REPORTING CORPORATION -- (202)628-4888
468 1 and we l'ook at them for anomalies, but we can't quantify them 2 as to'this gives me X foot pounds torque output. We don't 3 have that capability, and at that particular point in time, 4 that makes it very difficult to get back into the question of J
5 not blindly accepting if you will the vender's calculations 6 because I don't have anything to measure anything different 7 with at this particular point in time.
8 And I don't mean to say that iLproperly. When I say l,
9 blindly, I don't mean blindly. I don't accept much of
. 10 anything blindly because we are responsible for the operation 11 at that plant, but I don't have anything better to put in its 12 place, and I have to use my engineering judgment based on my 13 failure rate at my plant to say that I really don't have that
]
() 14 kind of a problem and what seems to be there seems to be 15 aderuate because it is functioning properly and has been and 4
16 it does under delta P conditions wherever I can achieve them, 17 and that's about the best I can get at this point in time 18 until something down the road gives me some better information 19 than what I aircady have.
20 Last item is critical valves. We alluded to this 21 when we were talking about cost, that the only way to reduce 22 the cost impact on the industry is to reduce the scope, and 23 where we came to with this critical valves was based on 24 presontations that we have heard here, at the previous
() 25 subcommittee meeting, essentially the Brookhaven National HERITAGE REPORTING CORPORATION -- (202)628-4888
469 l', 1 Labs, indicating that the major area of concern is.the 2 increased core melt frequency down to the increased failure b'3 3 rate of critical HOVs.
4 We took that and I looked at it from our own plant 5 at Davis-Desse from a fairly preliminary standpoint because
- 6 the time was fairly short, but we looked at the Brookhaven 7 analysis, and versus our own Level 1 PRA, and what it comes O down to is I believe the question was earlier asked what was 9 the definition that we used for critical valves, and I don't 10 claim to be a PRA expert by a long stretch of the imagination, 11 but it just so happens the first thing I have, that a PRA 12 expert gave me some crib sheets, so we used, we used the 13 Wesley-Fussell importance measure technique referred to by
( 14 HUREG/CR 4377 to determine that contribution to overall core 15 relt frequercy, and what it boiled down to is out of, what 16 tell out of that is that the only significant contributors
.7 were a total of 12 motor-operated valves. Eight of those 12
- 18 motor-operated valves have already been addressed by 85-03.
i 19 Four others have not, but for me, that's approximately 8 or 9 i
20 porcent of the total of my safety-related motor-operated 21 valves, so all we are saying is to put forth for consideration 4
22 the costs in public safety f rora a standpoint of core melt l 23 frequency is a, is the concern, and if the speed at which we l 24 resolva this concern is a paramount issue, then we should
() 25 attack those valves that are major contributors to core melt i
. HERITAGE REPORTING CORPORATION -- (202)623-4888
_~. ,
470 1, 1 frequency.
2 Not every plant has a PRA, but if you can go to O' 3 Chapter 15, make a deterministic assessment--I know without a 4 PRA what my design basis events are and what valves have to 1
5 work. The only things that are going to give you, give you 6 more valves is that if you had a good Level 1 PRA, you would 5 7 be more conservative.
8 HR. TIPTON: In our discussions over the last week, 9 it came to my attention that you car do it either 10 probabalistically with a PRA or deterministically either 11 looking at your Chapter 15 design basis on the plant or you J- 12 can go and look at your emergency operating procedures and 13 those pieces of equipment that the operator needs once the
() 14 ir.itiation of the event starts, over an hour, two hours, three 15 hours1.736111e-4 days <br />0.00417 hours <br />2.480159e-5 weeks <br />5.7075e-6 months <br />, to determine which systems.he needs to maintain the 5
16 integrity of the core so you can either do it 17 deterministically or probabalistically.
, l' CHAIRMAN HICHELSON: The problem you get into, of 19 course, is that doing this, for instance, you have a boiling 30 water reactor and you have a reactor water clean-up system 21 which fails. Now there is a finite hopefully small 22 o ability of failing to isolate the break.
pro' If you fail to 23 isolate the break, however, you are almost assured of core 24 molt.
25 You have to be very careful with those two
(})
HERITAGE REPORTING CORPORATION -- (202)628-4888
- - - , = _ , - . . .
(~1 I 1 particular valves, recognizing that one, that they aru 'both l
- '2 failing. Then you can do it'deterministically, but generally, i 3 we don't say two valves in a series ever fail. Only one 4 fails. Therefore, a little steam comes out and that's the end
.t 5 of game. The only thing you.have te address then is loss of .i 6 liquid from the reactors resu1. ting in the time it takes to 7 close the valves. You are dealing now with the possibility 8 that neither valve fails. It is a small but. finite 9 probability, and it does lead to core' melt. It is one of the i
10 most serious breaks in fact.
11 HR. ELFSTROM: The way we played it when we did ,
12 this--and we did this same t'11ng back in '85 before '85-03 8
l 13 came out. We were in a different position. We were trying to
, () 14 restart, and restart always seemed to be just a couple of
, 15 uonths anay so you had to prioritize and make sure you got the i 16 important thing done first, and we did an assessment and we
- 17 essentially tested the snot out of tbone valves. [
I
- 18 For example, in the aux feedwater systeen, we had a j
, 19 total of ten valves, and we De tested eight of them in full ;
20 flow, tested six of them. There were only two valves that t ,
21 didn't get that kind of delta P test. The main reason was F j 22 they were sister valves on the first, off the steam generator, i i t i 23 a little tough to do from the standpoint of pressurizing the r
[ 24 generator.
l
() 25 CHAIRHAN HICHELSON: It certainly does improve the, h I
! l
- HERITAGE REPORTING CORPORATION -- (202)623-4888 f
s, i 472 1~ . your confidence in the . numbers that y6u are 'Ising, oof PRAs you l i
2 are using. l 3 MR. ELFSTROM: As a matter of fact, that's one of i
4 the things that we have done with our PRA is'we have, we are
.'S putting in our plant failure data into it.
6 CHAIRMAN !!ICHELSON: That's a little more helpful, 7 but the--you can appreciate, of course, that there might be 8 probability of one that reactor water cican-up can't close. '
9 That depends how the torque switches are set and so forth-- [
10 under break flow conditions, and if the friction factors are 11' going up under break flow and so forth, it could be the 1
12 probability is one, but we don't use one in a PRA. We use !
.13 about ten minus 3, ten minus 4 rate of failure, because if we
() 14 don't, we end up with extremely high core melt. It becomes j the contributor, so you have got to be careful in using PRA :
15 l' 16 that you are sure that when you are looking for contributors, f I !
4 17 that you understand that the, what effects inoperability of ;
18 valve might have on the contributors and that kind of analysis i i i 19 can be done, but it is not as often because we are using j
'O numbers about ten to the minus 3, ten to the minus 4 on the {
i
[
21 valves. l 22 HR. ELFSTROM: The additional point here is going l 4
23 into a little more detail, a lot of source data that was used !
l 24 as far as the failure rates for the Brookhaven study, without j 25 actually getting into the source data, I would have difficulty r
( }. l, i
HCRITAGE REPORTING CORPORATION -- (202)628-4388 l
{.
t 473 1 in believing it, and I uacd.my own expcrience at the last
- 2 subcommitcee meeting where Brookhaven told a,e that initial (3 s.
3 85-03 reports from Davis-Besse said a hundred percent of the 4 valves were inoperablu. I was the one who wrote all cf those 5 reports, and I never wrote that, so there is some concern as 6 to the validity _of some of the date.
7 CHAIRMAN MICHELSON: We want good reports.
8 MR. ROTHBERG: Mr. Chairman, at Davis-Besse, they 9 admitted to failure of 17 out of 165 valves. That'e a 10 10 percent failure rate. That's the only thing that I have ever 11 asked about or supported.
12 We think that the 10 percent failure rate on 13 safety-related motor-operated valves is low. If you want to
() 14 dispute the 10 percent is a good number, that's fine. I would 15 like to hear your argument, but you can't tell me from what I 16 have seen that it is less.
- 17 CHAIRtfAN !!ICHELSO!!
- It is going to have to be a lot i
18 loss. Some of these becoma significant contributors.
l j 19 MR. CARROLL: We are running out of time.
l 20 MR. ELFSTRO!!: That's all I have to say.
I 21 !!R . CARROLL: I don't think that is an important i
22 issue.
l 23 MR. TIPTON: Mr. Chairman, what we would like to do 24 now is cover the flexibility aspects of it, and Brian Curry ,
() 25 from Philadelphia Electric will cover that area.
HERITAGE REPORTING CORPORATION -- (202)628-4888
\ _
i
4
+ .
474
~'
1 One thing'I may mention, a couple of comments here, 2 when we met with'the staff last Wednesday, we had an earlier 3 version of the draft letter, and as I'm sure you are' going to 4 hear in just a few minutes, the staff I think is making some 5 changes to the draft letter to address these issues, so I 6 wanted to make sure that is clear.
7 HR. CURRY: When we talk about flexibility here, we 8 actually are talking about the opposite of rigidity scheduling 9 as it was the, issued the version of the letter that we have 10 seen at this time.
11 We.have heard that Section J which required a retest 12 of two refueling cycles for three years has been revised, and 13 we think it is a good idea because that it is fairly well
() 14 known that good maintenance practice based on environments c.
15 service, inspections and trending results is what will drive 16 you to have good equipment history if you look at it based on 17 that basis, and Section G of the generic letter even requires 18 us to trend the data that we see in the valve test results so 19 that the change in retest period from a fixed period to a 20 trending, on a trending basis, we think it is a good move.
21 The other thing we have up there is implementation 22 schedule. The problem with--let's see. Implementation 23 scheduling is that we went from approximately 25 valves at, 24 for doing on 85-03 to approximately six times as many valves
() 25 or about 150 valves per plant. That's the number of valves we HERITAGE REPORTING CORPORATION -- (202)628-4888
473 1
use for our cost estimate. The number of safaty-related j
2 valves in the plant varied from approximately 12 to there are l 3 plants that have over 400 safety-relatec valves, have to be 4
tested by the expansion of the bulletin.
5 (Slide) 6 MR. CURRY:
In addition, besides, as we have 7
discussed earlier, besides the analysis for delta P, we have 8
to look at the environment that the valve is in, the system 9
logic, power supply, and in some ways, the way it is written, 10 we have seen at this point it is almost an open-ended 11 question, that it requires a critical review of all pertinent 12 design and installation criteria that were used in choosing 13 that valve for tha application.
() 14 A lot of cases--we have two fairly old plants that 15 we have got to retrieve, review all the original reccrds. We 16 l
i also have to determine that design intent back in 1969, and it 17 is a difficult thing to do. The records are there, but 13 getting them out is not sonething that you have the design 19 basis in your plant on a shelf and just pull that out for 20 valves and look it up.
21 There is a lot of work involved in going back and digging out those reports and figuring out that i 22 design basis again.
23 In addition, there is restraints to the, to this 24 testing, that it--sone of which we have control over and sone
() 25 of which we don't.
HERITAGE REFOP.TI!iG CORPORAI'IOri --
- - -- (2n2iAS2 do "
C e 475 1 use'for our cost estimate. The number of safety-related
, - 2 valves in the' plant varied from_approximately 12 to.there are
' ~
3 plants that have over 400 safety-related valves, have to be -
4 tested by the expannien of the bulletin.
5 (Slide) ,
L -
6 HR. CURRY: In addition, besides, as we have 7 discussed earlier, besides the analysis for delta P,~we have e 8 to look at the environment that the valve is in, the system 9 logic, power supply, and in some ways, the way it is written, f 10 we have seen at this point it is almost an open-ended 11 question, that it requires a critical review of all pertinent i 12 design and installation criteria that were used in choosing 13 that valve for the application.
() 14 A lot of cases--we have two fairly old plants that 15 ue have got to retrieve, review all the original records. We i
16 also have to determine that design intent back in 1969, and it l 17 is a difficult thing to do. The records are there, but i 18 getting them out is not something that you have the design 19 basis in your plant on a shelf and just pull that out for j 20 valves and look it up. There is a lot of work involved in f
i 21 going back and digging out those reports and figuring out that j 22 design basis again. ;
i 23 In addition, there is restraints to the, to this ;
24 testing, that it--some of which we have control over and some ;
() 25 of which we don't.
?
t IIERITAGE REPORTING CORPORATION -- (202)628-4888
476 E 1 MR. WOHLD: Brian, talk a little bit about the 2 design requirement again. I visualize an easier approach to O 3 the design requirement. If you don't know exactly what the 4 valve is designed to do, if you can evaluate it, a limiting '
'S condition that it will see during an accident, that you can [
L 6 test to that condition, I would think that would fully satisfy ,
t 7 the design evaluation. That's my own personal feeling. I l
8 don't know if thnt is what the staff says, but if you have a -
9 problem and the worst conditi.n is the reactor depressurized j i
10 and the pump full head, test the valvo at that condition, that-11 would eliminate any need to go through the records barring !
i 12 something I am not thinking about here. l 13 MR. CURRY: Pete, you are only looking at delta P..
l
() 14 We also have to look at the environment that the valve is in.
15 We have got to go and look at the system logic. We have got ;
16 to look at the power supply, reduced voltage requirements. It i
l 17 is more than just delta P in this one item that we have got to j l
l 18 look at which basically was done for 85-03. I know that (
19 plants, we also looked at reduced voltage requirements. You [
20 know, EQ is addressed by EQ, but this is, in some way it is, could be interpreted we have got to review our environmental !
- 21 22 qualification reports for the valve.
23 MR. WOHLD: Determine limiting DPs ahead with the {
f n
24 testing and evaluate the environnent, you are not going to ;
? -
25 test it, any heat and humidity, all that sort of th.ng I l
({} l l HERITAGE REPORTING CORPORATION -- (202)628-4888 f
1,, . .m .- .-.
477 l 1 wouldn't expect.
2 MR. CURRY: You have'got to go through the
'3 evaluation'acain. That's what we are saying. There is a big 4 neope of engineering work being added in.
5 The other thing is that for boiling water reactors, !
- 6 and trying to establish forecasts for a lot of these valves is
?
l 7 I think more difficult for us than PWRs. I could be wrong on l, 8 that, but I am going to make some assumptions about PWRs that 9 you can drain down.the steam generator and then pump into it. (
t 10 For us to drain down, we have got to drain down the vessel and 11' then pump into it and not overfill it. ,
l 12 MR. WOHLD: It was just a suggestion. I thought
,i
! 13 there might be an easier way. I think Owen Rott. berg was
() '14 alluding to a simpler approach. I don't know if you dig I 15 through the files on the design. basis for everything. It is a l I l
)
16 ' horrendous section.
i 17 Am I right, Owen?
! i l 18 HR. ROTHBERG: I was going to change the words, f
) 19 Let's see if I can read from the draft. I was going to say [
20 that the review shouldn't be restricted to documenting f i 21 estimated maximun design basis pressure, but I recognize your (
) !
i problem with going back and into the environmental ;
22
{
l 23 qualification bases for the plant.
' f i
24 As I said this r.orning, we have no intention of
() 25 changing any of those bases. The way I have--I have some l t i
HERITAGE REPORTING CORPORATION -- (202)628-48f8
4 478 1 notes here. I say the design basis review should include an 2 examination of the documentation of pertinent design criteria 4
( 3 that we used in choosing a particular HOV. I still think 3 l 4 that's a valid--I don't think, i don't believe that you should.
5 it seems to me operate a safety-related motor-operated valva 6 without knowing what the design be,is for it is, including I
7 pressure and flow and the other things that go .4mto it. I 8 don't see how you can, you can test a motor-operated valve and i say that it is, that it is operable under' design basis 10 conditions without knowing what those design basis condicions 11 are. l
, 12 HR. CARROLL: Does design basis in your mind, Owen, {
i 13 include EQ issues?
{
() 14 HR. R3THBERG: Well, you certainly wouldn't have to f f
15 test for that. I think you should know what they are. i i
l 16 HR. CARROLL: And peop.'.J do. ;
17 MR. ROTHBERG: I went back through several of the !
I f 13 environmental qualifications on motor-operated valves and what h J !
j 19 they say is a discrete set of bench tests for very specific
{
} i
, 20 things. There is no place that I saw where the whole motor j
- 21 operator--and even the valve is put together as a. unit hooked ,
, l l 22 up to a pipe and tested, and I'm not really sure in my own (
, I i 23 mind what that environmental qualification tolla you. L 24 Sure, it meets the specific thing that is written in
! t 25 the tule, the specific rules that are written in 10 CFR, but l (} 1 1
HERITAC3 REPORTIllG CORPORATION -- (202)628-4838 i
_ m_
j
t i* 479 1 I'm not sure whether it tells you.that that particular item in 2 environmentally qualified.
. ..() What.I am thinking is if a pump has to 3 MR..WOHLD: ,
4 get through a valve, and the worst condition downstream is t
'5. Mepressurize, you start the pump, downstream if the valve t
6 depressurized, open the valve, that's proven that the system-7 can meet its design requirement. You don't have to even 1
8 consider--you know you can't get any more pressure against it 9 than that.
- 0
. HR. ROTHBERG: Opening against pressu:;e?
11 HR. WOHLD: Right. That would cover the opening.
12 Then close it if it is required to close, but I, what I am 13 trying to get away from is the time and expense related to the ;
() 14 long evaluation of the design basis conditions. I was quite .
?
l 15 surprised at the, myself at it taking that along. There is i
16 other holes, and Bob has said many times you cut Lt with a !
l 17 chain saw and measure it with a micrometer. I don't know how ;
t
?
l 18 much is gained by looking at the paper, spending a lot of time i
19 on paper, and start the pump, open the valve. Maybe you have '
\
i 20 HOVATS attached or not, but I *.hink that's, you are .ome free j i
-21 95 percent of the time. ;
t 22 CHAIRMAN MICHELSON. I think the whole point of the I l
23 argument was that you must first of all understand what the j 24 requirements en the valve zeally are, and then look to see how !
i i a
() 25 you might do the test, of course. That's what item A war, and
' HERITAGE REPORTINc, COR3 ORATION -- (202)628-4888 j t
~
.r 480 1 determine what are the requirements.-
2 MR. WGHLD: Right.
3 MR. CARROLL: Your second bellet this morning, Owen, 4 was to the effect that the letter should clearly indicate' that- !
5- the existing EQ bases and design bases was not being modified.
6 Doesn't sound like you agree with what you said this t
7 morning. '?
l i.
8 MR. ROTHBERG: 'I'm sorry? e i i 9 HR. CARROLL: Your second bullet this morning Said ;
10 that the generic letter should clearly indicate that the ;
i 11 existing environmental qualification bases and design bases l
~
12 are not being modified by this letter.
f i
13 MR. ROTHBERG: That is correct, we are not. We are '
() 14 asking them to review and document that design basis. We f I
l 15 never changed that. That's exactly what we are saying, sad it e l Y 16 is the first words--review and document the design basis for ;
17 the operation of each HOV. It is nothing more than that. It i i
18 is nothing more thar. saying you know what the design basis is [
19 for your motor-operated. I don't see how anybody could argue ;
i 20 against that you have got to know that in order to say your i
21 n.otor operator is operable. !
(
22 MR. WOHLD: I can still come up with several i E
23 approaches on that. The original design basis is going to be [;
24 the pump. The valve opened when the pump starts, and in l l
(} 25 implementing that design basis, then you buy a valve for a L
HERITAGE REPORTING CORPORATION -- (202)628-4888 c
481- ;
, 1 specific design. You might buy it for 500 pounds differential
_2 pressure highe? than you need, but the real design basis is 3 the valve opened when needed, and that would be against some !
.~
4 . limiting conditions, but which you can determine by a fairly-hv 5 simple means. I'm trying to get away from what I saw in the I other years and evaluate the design basie.
6 I don't see that q
m 7 that's necessary.
8 HR. ROTHBERG: This was the whole--not to get into 9 cross-talk, but--maybe you don't want this discussion, but the I 10 whole thing that we got into was we were absolutely surprised s
11 by the fact that the industry took two years to come up with 4
12 the design basis for these 35 valves, for plant, 30 valves per-13 plant.
) 14 HR. WOHLD: That is why I am thinking perhaps the 15 industry was more conservative than was required in evaluating 16 the design basis. That's all. That's a comment.
17 HR. CURRY: It is somewhat related to the item.
18 What doesn't show up on the vugraph here is that we, I believe e 19 we tried to be conservative because we want to avoid getting a 20 notice or a violation against us, and one of the potential L
21 problems the way the bulletin is written right now is that we 22 are to review and document it and put it away in our files, 23 and then the residents will review it a* a later date so we 24 could be several years down on our way, and then 'ind out that
() 25 we did something wrong because we didn't interpret it HERITAGE REPORTI' " CORPOPATION -- (202)628-4888
i 482 L
1 correctly. We will follow that for a very long time until it 2 is'xeviewed and it is kind of an open item now that some ways
~ O:
3 we tave a ticking time bomb in'our filing ccbinets. l t
? 4 MR. WOHLD I agree wi*h you, and I guess, and on ]
5 that comment, I would'say that more dialogue is needed between L
6 the NRC staff and industry to settle some of the things before [
7 the program is implemented. I know I inspected a lot of I 8 program 1 before, but it was--as you know, some are not 9 approved yei. A more active dialogue can resolve most of the !
I 10 questions in a few weeks, and I share your concern.
.i t 11 HR. CURRY: Going on with the presentation, in the t i
12 area of restraints, we have both oxternal restraints that we ,
i t
13 really can't do anything about as an industry, and we have
() 14 internal restraints we have some control over.
15 In the area of external restraints, obtaining valve 16 vender data, when we talk about valve vender data, we are not l E
- 17 talking about what it takes to seat that valve, but what the !
18 maximum seating thrust that the valve can tolerate, that 19 motor-operated valves are not single set point devices. They I
i i 20 are actually set within a band, and we need to know what that i i
21 maximum is to see what the maximum comparable forces are.
s 22 We were doing some checking, and we received a (
l
', 23 quoto--this is filled up--for 50 valves, that would take u? f
- 24 six months for one vender to provide us with that data, and f
) I I
() 25 you figure that you are going to have a 150 valves per plant HERITAGE REPORTING CORPORATION -- (202)628-4888 i 1
,- 483 11 and you have 127 plants asking the venders for this 2 information, that lead time is going to stretch out to way
.O
- 3. beyond there. ,
4 i
[
4 In actuator data, this is to find out what the l
5 . configuration is, what its capasilities are. In this case, we
[
s 6 don't have a number of venders. We are down to basically a I
7 single vender for the 90 percent of the v41ves'in the plants.
8 That vender is now seeing the effects of requests for data, 9 and has stated at the EPRI valve conference'that they won't be ;
I 10 able to support the 127 plants coming in looking for data from i 11 them. l 12 In the area of $tagnostics support, while there is a j l
13 .sumber of venders out there now, if you look at the number of ;
r f
() 14 plants, 11 weeks per outage, with no training, to get the i
15 people into the plant, in order to tail off their work, you
- i 16 would have, if the plants outages are spread throughout the i i ?
17 year, you could figure you would have 125 plants per peak ;
t 18 outage requiring support from diagnostic means. ;
1 :
, 19 Last but not least, the external area is spare l I
20 parts, and this is, primarily we are talking about 21 consumables here. We have gotten hung up looking for gaskets, 22 O rings from the vender. l i
. 23 In the area of internal-- l l I 24 MR. CARROLL: Going back to the Limitorque problem,
() 25 I'm not sure I see what it is. There is a finite number of l L
I HERITAGE REPORTING CCRPORATION -- (202)628-4888 l i
I 484 d
gj 1 actuators.
2 HR. ELFSTROM: There is an infinite number. You 3 have actuator--there is like a million, million combinations.
4 Each one is unique, and they have to go back to their build 5 sheet by order and serial number for each individual one to 6 como up with the inf6rmation you need.
7 CHAIP. MAN HICHELSON: I understand on one model there 8 is 37 different ways you can put things in that change the 9 characteristics of that particular model.
10 HR. ELFSTROM: Yes, sir.
- lou have motor sizes, gear 11 sizes, spring pack.
12 HR. CARROLL: Thank you. Moving on--
t, 13 HR. CURRY: Okay. Restraints that we can do
() 14 something about--internal restraints, valve availability is 15 ene of the biggest of them. Technical specifications require 16 certain minimum equipment available even when we are shut down 17 so we cannot shut do:in the plant, line up our testers, and 18 start testing and doing valves per day on 20 hours2.314815e-4 days <br />0.00556 hours <br />3.306878e-5 weeks <br />7.61e-6 months <br /> shift.
19 Certain valves only become available when other trains are put 20 back into service. There is, work is done on the power supply 21 so these valves in the motor control centers for the 22 switchgear, if I don't have electricity, I can't run that 23 valve either for diagnostic testing or for DP testing.
24 Tht other thing is that frequently these valves ar,
() 25 system boundary block points for other work. If I am working HERITAGE REPORTTNG CORPORATION -- (202)628-4888
485 1 on my RCIC turbine, I will block'out all these boundary valves-
~
2 for the RCIC turbine, and I don't work on them.
()-- 3 Now the other thing is the number of valves we can l
- 4 test per week. What we found based on' people that have tests j 5 for some period of time, and this is not--starting out, this i
6 is at the beginning of the learning curve, that we can get j 4 7 approximately five MOVs in a seven-day week with 20 hour2.314815e-4 days <br />0.00556 hours <br />3.306878e-5 weeks <br />7.61e-6 months <br /> days. ,
8 That is with a dedicated crew of utility people and vender 9 support, and it is with planning and scheduling support. This 1 10 doesn't say that we will only do five valves in any given {
11 week. We maybe do 15 valves in one week and we will sit there ,
t 12 for two more weeks not being able to test valves because of l
. l 13 other activities going on. [
, e 4
() 14 That completes my presentation. Are there any l
l 15 questions? I l
i 16 MR. WOHLD: The problem, do you have a solution in f
17 terms of the phased approach to the valves? j r
I 4
18 HR. CURRY: I think the best solution that we have I
19 seen so far is what Bob presented earlier. If you reduce the j 20 scope of valves that you have got to cast, to those which have l 21 prime importance, then it gets down to a reasonable number of f
5 4
22 valves that can be done in the timeframe we are talking about. [
i i 23 MR. WOHLD: Okay. What about on the lesser l 24 important valves? I know if there are qualified people doing 4
() 25 the work, you can get I would say at least a ten-fold increase 4 i HERITAGE REPORTING CORPORATION -- (202)628-4888 )
I
486 1 in reliability just by having qualified people doing 2 inspections or QC or whatever. And going back to Fermi, three i l 3 problems on one torque switch, none of the problems were such 4 that they couldn't be identified by a good inspector, so while 5 we have been talking about diagnostic testing, I think there 6 is a lot of other things that can be done to inprove the 7 reliability without the diagnostic test 3.ng.
8 MR. CURRY: We have already brought up we have got a 9 prograrmatic approach to deal with !!OV problems , but you (now, 10 we are primarily discussing the bulletin as er the extension 11 as it stands rather than trying to restructure it.
12 MR. WOHLD: I guess I don't think it is out of the 13 realn of possibility to have a phased approach based on the 14 importance and maybe have a lesser requirement on the 1.sser 15 important valves.
16 MR. CURRY: I don't think we would disagree with 17 that at all.
18 MR. WCHLD: Okay. That's all I have got, 19 CHAIR!!AN !!ICHELSON: I h&ve one kind of question I 20 would like to get your view on.
21 You have hind nf led me to believe that perhaps the 22 cost of doing this might be as much as ten tires what the 23 cost / benefit calculation indicated.
24 Have you actually done an estination yet? In other 25 words, is this thing really cost beneficial in your mind, or ll4 HERITAGE REFORTI?iG CORPORATICf; -- (202)628-48SS
487 1 is it way off, and if--you know, where are you coming down?
2 HR. TIPToli: We didn'.t look at the benefit side of l i
3 the equation. What we did was in the short timeframe for-last 3.
4 week based on actual experience of 85-03,'three utilities had 5 their numbers with them, and-that's what we' generated. l 6 CliAIRMAli HICHELSoli: I was wondering-even if it is l l
7 ten times what the staff might have thought was the cost, is j l
8 it still cost beneficial?
l 9 HR. TIPTOll: We didn't do that calculation.
10 CHAIRMA!i HICHELSO!!: You might want to think about l
i i
11 that at least because otherwise, so what if it is ten times? )
l' 12 If it is cost beneficial yet in your mind, you will do it I 13 would think. l f,) 14 HR. TIPTO!!: That phase of it we haven't looked at 15 yet. In the interest of time, what I did was I took my 16 conclusion and wrapped it into two paragraphs.
17 The industry is addressing the valve issue through l I
18 the offorts of I!iPO, EPRI and the owners groups. We plan to I
19 continue to pursue generic improvements in our overall valve 20 program. We want to work with the staff. I am very l l
21 interested in this phase of it. We want to work with the 22 staff on the focus of the generic letter, whatever generic 23 letter comes out, in its implementatien.
24 As was previously noted, we need, we really need 25 consistent interpretation of the letter at the region level to HERITAGE REPORTI!iG CORPORATIO!! -- (202)628-4888
488 1 avoid different understandings-of the same letter, spending 2 significant time on both the industry and the staff's part to 3 try and resolve what those differences are as you get into I 4 implementation of engineering, so with that additional point, 5 where'we would like to work with the staff on making sure that 6 in fact we have a smooth program, we would like to thank the 7 Subcommittee for your time.
8 CHAIRMAN HICHELSON: Any questions remaining? The 9 plan still is we are going to meet December 13th for the 10 Subcommittee meeting.
I~ 11 HR. CARROLL: If somebody brings us the rock.
12 CHAIRMAN HICHELSON: I will have a proposed draft i
13 letter by that time to the Subcommittee to think about, and if
() 14 NUMARC has a feeling that they believe thct their m,ssage nead 1 15 be brought to the full metabership, if you so indicate your l 16 desire, we will see what cars be done to includc it in the l 17 December meeting. I would think so. I don't th3nk we can--
1 f
18 MR. WYLIE: I think we have all we need.
19 HR. TIPTON: We will get back to you.
20 CHAIRMAN MICHELSON: At that time, please bring your l
21 best estimates of cost and benefits so that we can get a 22 feeling of what you are really saying is it is not cost 23 beneficial or what you are really saying is it is going to 24 cost a lot more but we still think it is worthwhile, or
() 25 something in between.
HERITAGE REPORTING CORPORATION -- (202)62b-4888
t i.
i
- 489 l' That finishes it.
1 1 1 2 (Whereupon, at 3:45 p.m., the meeting was l'.
i 3 adjourned.) ;
i-l'
- i. 4 i
! 5 ;
l I #
! 6 l
j 7 r
( 8
(
I 9 1
i 10 l
{ f l
11 t 12 ;
I i 13 l 14 .
15 '
16 17 '
l 18 ;
l 19 i l
20 .
I r
21 l 22 23 24 !
l g 2s l
! HERITAGE REPORTIliG CORPORATIOri -- (202)628-4888
1 CERTIFICATE
(~'l
'~'
3 This is to certify that the attached proceedings before the 4 United States Nuclear Regulatory Commission in the matter of:
S Names Advisory Committee on Reactor Safeguards, Subcommittee on 6
M hanical Components 7 Docket Number:
8 Place: Bethesda, Maryland 9 Date: October 27, 1988 10 were held as herein appears, and that this is the original 11 transcript thereof for the file of the United States Nuclear 12 Regulatory Commission taken stenographically by me and,
~
13 thereafter reduced to typewriting by me or under the direction 14 of the court reporting company, and that the transcript is a 15 true and accurate record of the foregoing pro eedings.
16 /S/ WbamY '
/
17 (Signature typed): Catherine S. Boyd 18 Official Reporter 19 Heritage Reporting CorpoTation 20 21 22 23 24 25
.( Heritage Reporting Corporation (202) 628-4888
.v . .
i O O O i ,
NUCLEAR MANAGEMENT AND RESOURCES COUNCIL MEETING j
WITH 4
) ACRS SUBCOMMITTEE ON MECHANICAL COMP 0NENTS i
l ON t i
l l INDUSTRY ACTIVITIES AND
! DRAFT GENERIC LETTER ON MOVs
! OCTOBER 27, 1988 l -
1 I
j l
4
_ . _ _ . _ _ _ _ . _ _ . _ ~ _ _ _ _ . _ - . - - _ _ . . _ _ _ _ _ . _ _ _ . . _ _. .-- --_ ._ - ._ _ _ _ _ - - _ - - - - . - -
O O O COST IMPACT o BASELINE l 5570,000,000 -
ESTIMATED COST BASED ON ACTUAL EXPENDITURES IMPLEMENTING IE BULLETIN 85-03 l
S 35,900,000 -
ESTIMATED COST BASED ON BRG0KHAVEN NATIONAL LAB ANALYSIS !
o PERIODIC TESTS 5420,000,000 -
PER PERIOD ;
l
t 0
_G_EXE1ICLETTELSC0PE.
o STATIC TESTING vs DESIGN BASIS TESTING o MAJOR EXPANSION DESIGN REVIEW RETEST REQUIREMENTS o LIMITED DIAGNOSTICS O o CRITICAL VALVES O
l o !
E_L_E11_B_I LITY i
l l
0 RETEST PERIODICITY ;
FIXED ITERATIVE o IMPLEMENTATION SCHEDULE O
- O i
I i
r l
O f
i j I M P L_E M E N T A T I O N S C H E_D_U_L_E_ l l h
, I f
- o NUMBER OF VALVES / PLANT
.l i l
- o RESTRAINTS t
I i
! i i
O I 4 i l l I l r
i j !
l
) t 1
l 1
i
)
!, < t i
I r
I' !
a t
}
)
l O l i !
i ;
___ _ . .___.__.____.J
O RE_STRAIJTS_
i o EXTERNAL !
i VALVE VENDOR DATA ACTUATOR DATA DIAGNOSTIC SUPPORT SPARE PARTS 1
o INTERNAL i
- - O -
AVAILABILITY 1
FIVE (5) M0V'S/ WEEK i i i i
l l l t
i f
i i
i O
4 I
O EPRLNPD EPRI ACTIVITIES ON MOTOR OPERATED VALVES Jim Lang l EPRI
. Motor Operated Velves O EPRLNPD EPRI ACTIVITIES ON MOTOR OPERATED VALVES
- Previous AdNtes: MOV lmprovements
. Nudear Maintenance Assistance Center (NWAC) Adivties:
Technical Repair Guide Applicaton Guide
. Future Activmes: MOV Testing Motw Operated Vatves O
4 l
EPRLNPD -
MOV lMPROVEMENTS
- Developed the 'Inte!!storque' System Control and Diagnostic Functons for MOVs Permanent Installation
- Demonstraton Unas installed
,
- Commerciah Available
- Interim Report Available: NP-4254 l
- Final Report due Late 1988 Motor Operated Valves 1
O EPRLNPD NMAC ACTIVITIES
- Undertaken at the Request of NUMARC
- Guided by a Technical Advisory Group Utilny Expens Vendor Representatives INPO Expert
- Emphasis on Accuracy and Usefu8 ness 1
Motor Operated Vatvee O
6 t
O EPMLNPD NMAC TECHNICAL REPAIR GUIDE FOR LIMITORQUE OPERATORS
. First Edaion Mdre:ses SMB-000 Operators
- Comprehenske Maintenance Guidelines:
Guidelines for Sattsg Torque and Lirnd Switches Corrective, Preventke, and Prodctive Maintenance
. Failure Modes
. Spare Parts
. is FinalType Setting
- Guidelines for Larger Sizes to Fofiow Motor Operated Valves n ,w ..
O EPRLNPD NMAC APPLICATION GUIDE FOR MOVs
. MOV Design and Functonal Requirements
. MatcNng Vanes and Operators
. Selecten and setup of Control and Protecten Devices
- PuMem. ten Scheduled for EarY 1989 Motor Operated Valves n ,w ..
O
l J
l .
l l
! O EPMLNPD ,
i' MOV TESTING i
e '
l v t
. ObjodNes:
[
. Predict Performance of MOVs I
. Simpffy in Situ Tests [
a
. Methodology:
. Accumutme and Evalume Existing MOV Test Data
. Conduct AdditionalTests as Required
. Up Date MOV Sizing and Setting Calculation Data 1 l l
Y mwm ,,
Motor Operated Ysivee l
l 1
d i
O :,
2 i 4
I f
t
?
l J i
. i a
1
' l I !
b i
O !
t !
r
1 J
l t
I MAJOR COMMENTS BY CROR REGARDINP i
Q PROPOSED PROGRAM FOR INCREASED IN j SITU TESTING OF SAFETY-REL4TED MOYS !
I -
< 1 I
' I
' PROCRAM IS NEEDED FOR SAFETY AND THE LETTER SHOULD DE SENT. ;
I .
} ' LETTER SH0rlC CIIAELY INDICATE TilAT Tile EXISTING DTIRONMENTAL '
QUALIFICATION BASES AND DESIGN BASES ARE NOT BEING MODIFIED.
1 i
j
- LFJTER SHOULD CIIARLY INDICATE TilAT DOCUMENTATION ;
l Of THE DE5tGN BASES INCLUDES CONSIDERATION OF i f DEGPADED YOLTAGE, WUUNG LINE LDSSES, AND POWER j SUPPLY, TO THE EXTENT INDICATED IN TIIE OTNER'S i f COMMITTMENT. THIS IS TO BE IN ADDITION TO THE l DELTA-P DESIGN BASIS INDICATED IN DULLETIN 85-03. j
- THE SCHEDULE RECONDfENDATIONS FOR INITIAL PROGP.AM i
CON 5TRUCTION (TWO YEARS OR ONE REFUELING OUTAGE) !
l AND INITL3 IMPLEMENTATION (THDEE YEARS OR TWO *
- REFUELING OUTAGES), AS WELL AS PERIODIC SUPJElllANCE ECHEDULE (EVERY THREE YEARS OR TVO REFUE1ING OUTAGES),
VAS QUESTIONED. t i
i f
,
- CEG3 AGREED TH\T THIS PROPOSAL IS SAFETY SIGNIFICANT AND SUGGESTED THAT THE FORMARDING MEMORANDUM THAT l
{ FETURNS Tile DKWr LETTER FOR 'lllElR FINAL APPROVAE l SHOULD CONTAIN A DISCUSSION OF THE PASES THAT j
, SATISEY 10 CFR 50.100taK4Hi) AND (ii).
Q .: .
4
l (
U 1 l
l '
4 THE NRC RECOMMENDATIONS WITII RESPECT TO TilB UST OF DECEADED !
CONihTIONS (CONTAINED IN ATTACHMENT B 0F THE PROPOSED GENERIC i
f LETTER) SHOULD BE CLARIFIED. THIS UST SHOULD NOT BE MADE !
A "CHECE-0FF" UST FOR SURVEH1ANCE OF EACH MOV.
THE LETTER SHOULD AT LEAST SUGGEST THAT OWNERS CONSIDER AUGMESTED i i SUINEILLANCE OF THE BAI.UCE-0F-PLANT M0YS. i j '
I NUMARC COMMENTS CONCERNING THE PROPOSAL SHOULD BE CONSIDERED.
THE PROPOSED LETTER SHOULD BE CHECKED FOR COMPATIBlUTY MTR SEC l
MAINTENANCE POUCY AND TIIE PROPOSED MAINTENANCE RULE IN ORDER TO AVOID CONFUCT OR DUPLICATION. WORDS SHOULD BE ADDED TO INDICATE O ' ~ Trils Wis CONsiDERtD IN CONSTRucriON 0F THE PR0nRiM. !
f i l '
THE PROPOSED PE0 GRAM SHOULD DE COMPATIBLE WITH OWNERS Q.A. AND
- RECORD-KEEPING COMMITIMENTS AND NOT BE IN CONFUG OR REDUNDANT, ,
1, '
OWNERS Sil0ULD DE GIVEN CREDIT FOR WORK DONE TO COMPLY WITH i
I l ; U B'llETIN 85-03 PLUS ANY OTHER ADDITIONAL SURVEllLGCE PROGRAMS l l FOR OTIIER MOYS. NRC S1100LD AV0ID DECOMMENDATIONS THAT WILL RESULT ,
l s
IN DUPUCATING WORK DONE PREVIOUSLY.
1 .i '
, i l i
, < .c
, i 1 i
i f l
I
gmew
+ %, UNITED STATES
[ o NUCLEAR REGULATORY COMMISSION s I wAsmuorou. o. c. mss OCT 16 E8 MEPORANDUlt FOR: Robert L. Baer, Chief Engineering Issues Branch Division of Safety Issue Resolution Oftice of Nuclear Regulatory Research THRU: Frank Cherny, Section Leader A i Engineering Issues Branch Division of Safety Issue Resolution Office of Nuclear Regulatory Research FROM: Owen Rothberg Engineering Issues Branch Division of Safety Issue Resolution Office of Huclear Regulatory Research
SUBJECT:
SUMMARY
OF MEETING BETWEEN NRC STAFF AND INDUSTRY REPRESENTATIVES ON 10/19/88 TO DISCUSS IN SITU TESTING OF MOTOR OPERATED VALVES (MOVS)
On October 19, 1988, a public meeting was held at the NRC Offices in Nicholson Lane to discuss in situ testing of MOVs. An attendance list is enclosed (Attachment 1).
NUMARC acted as the lead spokescen for the utility attendees. Copies of the slides from presentations >y Tom Tipton of NUMARC and Bob Kershaw of Arizona Public Service Company are provided as Attachments 2 and 3. Robert McPherson of Southern California Edison and Bill Ross of Texas Utilities Electric Company also made presentations. Attachment 4 was included at the request of Gerry Weidenhamer and indicates results of blowdown tests on several MOV> equipped with sevtral commercial diagnostic test systems.
The discussion focused on the NRC's proposed generic letter to all nuclear plant owners regarding in situ testing of MOVs. In the generic letter, NRC would recomend that owners perform additional testing, including flow and pressure testing, if practical, of safety-related MOVs in order to assure design-basis operability. That proposed letter was presented to ACRS on October 7,1988 and CRGR on October 12, 1988. The CRGR package dated 9/28/88 had also been placed in the NRC Public Document Room.
Most of the industry representatives appeared to agree with the need for a MOV testing program that went beyond the requirements of Section XI of the ASME Code and IE Bulletin 85-03. The industry coments centered around the cost and schedular aspects of implerenting the program as proposed in the C!'GR O
I 2
O package. In general, the industry representatives disagreed with the flRC estimates of cost i.e., stated they were underestimated, and indicated that the schedule did not allow sufficient time to set up the proposed program or conduct periodic testing.
Specific coments by the industry /NUliARC representatives can be characterizt.d as follows:
- a. The nuclear industry has an ongoing procramatic a >proach to MOV ,
surveillance and maintenance. Several INPO and E)Rl/ft!4AC documents and initiatives were discussed, it was noted that there is no ;
industry-wide commitn.ent to this program to date,
- b. The industry representatives suggested that testing schedules should be prioritized based on the importance of the specific tiOV's ;
contribution to avoiding core melt. Several suggestions were made about how this might be done, although the general impression was '
that prioritization would be a difficult task and there was no current consensus on details of how to accomplish it.
- c. The cost of implementing the plan would be about 10 times what i;RC l had estimated (Coment: Such a cost would be about equal to the benefit without considering improved safety),
- d. Problems in implementation of the I;RC prcgram due to shortages of O tr 4"ed verso"#ei. seere verts. test eavie eat ma ve"dor support were discussed.
- e. Differential pressure testing in place was considered by the ,
i industry representatives to be difficult and expensive. Periodic l l
l tes'.ing was also discussed along with trending, as well as industry's desire to adjust test frequency based trending results, t i
- f. Detailed test information on Butterfly valve operators and other i specialized operators may be difficult to obtai.. This will therefore male determination of design basis operability difficult for these corrponents. These components are less than 15%
of the i!0V population.}(Coment:
)
NRC representatives outlined ongoing stcff activities on Pt0Vs including several !
changes that are being made to the proposed generic letter, based en ACRS [
corrents, industry curtrents at the ACRS subcomittee meeting, and CRGR t coments . The concents from this meeting also provided useful information ;
to the staff, t
f O
3 NUMARC and the NRC staff plan to discuss the results of this meeting in forthcoming meetings with the ACRS Subcomittee on Mechanical Components to be held October 23, 1988 and the full ACRS Comittee in November,1988.
At 0
'vk N 'e t
Owen Rothberg Engineering issues Branch Division of Safety Issue Resolution Office of Nuclear Pegulatory Research Attachments:
As stated i
O
o i
, jttachment1
'" ATTENDANCE LIST '
10/19/88 MEETlHG ON MOV IN SITU TESTING NRC/RES NAME ORGANIZA,TJO,N PHONr Jerry Mazetis NRC/RES/RPS!B (301)492-3906 l
Owen Rothberg NRC/RES/ElB. (301)492-3924 John J. Lance EPRl/NMAC (415)855-2018 Richard J. Kiessel NRC/NRR/0GCB (301)492-1154 Earl J. Brown NRC/AE00/ROAB (301)492-4491 Frank C. Cherny NRC/RES/E1D (301)4923545 Warren Minners NRC/DSIR . (301)491-3980 Robert Baer NRC/RES/ElB (301)492-3930 Robert C. Elfstrom Toledo Edison (419) 249-5000 x7692 Robert Kirkwood NRC/RES/ElB (301)492-3928 Joel D. Page NRC/RES/ElB (301)492-3941 Eve Fotopoulos SERCH Licensing, Bechtel (301)250-3094 Philadelphia Electric Co.
. O Brian Curry Bob Kershaw AZ Pubiic Service /PVNGS (215)327-1200x4497 (602)371-4250 Ron Scherman Cleveland Elec. 111. (216)259-3737x5300 Bill Ross TV Electric /CPSES (817)897-5734 ,
Joseph Nadean MOVATS (404)424-6343 Roger CARR HOVATS (404)424-6343 Robert McPherc:7 SCE-SONGS (714)368-6987 R. Clive Callaway N9 MARC (202)872-1285 f Peter J. Kang NRC/ DEST /SELB (301) 492-0812 L David Lowry Liberty Technology I?15)834-0330 [
Frederick Burros NRR/ DEST /SELB t< ; 492-M33 l Gerald H. Weidenhater NRC/RES/EPEB % ! 492-3839 Roy Woods NRC/RES/RPSIB GS) 492-3908 Nick Konstantinov CECO (312)294-8557 Bill Farmer NRC/RES/DE (301)492-3858 O
1 % .3 44- .
af. .
- G ; ,, ;
l y',' 1 s pc-ry
- L' O /l<,
.NAME. ORGANIZATION PHONE llJ0hnHuang; NRC/RES/ DEST' (301)492-0921.
M ,z 3 9., ' E. L; Igne
-NRC/ACRS (301)492-1892 I
- j
' Warren-J. Hall NUMARC (202)872-1280 Tom Tipton. .NUMARC (202)872-1280-Randy YanLear' Babcock & Wilcox (804)-385-2.789
, .t E
4 e
? 'b . ,
J y
- -O h
h
{.
i i
I i
i j, -
i i
4 O
5.
- l .
- j. .
,-,c,-- .- n -,, , _ _
.t.
~
i .
0
! O O l
1
\3 t*
l NUCLEAR MANAGEMENT AND i
RESOURC.ES COUNCIL MEETING WITH U. S. NUCLEAR REGULATORY COMMISSION ON I
DRAFT GENERIC LETTER ON M0ys
.l l
i
)
i
l S l C) MOV GENERIC LETTER CONCERNS o STATIC TESTING vs DESIGN BASIS TESTING o PERIODICITY o IMPLEMENTATION SCHEDULE o GENERIC LETTER SCOPE o M0V PROGRAMMATIC APPROACH O
o CURRENT EPRI ACTIVITIEd o COST IMPACT 6
0 0
l O
l STATIC TESTING vs-DESIGN BASIS TESTING o CANNOT IMPLEMENT o APPENDIX A O
o ALTERNATIVES l
, o PLANT IMPACT i
O
O
)
l PERIODICITY o FIXED o ITERATIVE O
O 1 - - - - - -
j,.
O
. IMPLEMENTATION SCHEDULE o PRIORITIZATION o NUMBER OF VALVES / PLANT o RESTRAINTS O ~ .
l l
I II l
O I
L _ __--___.-_._ _ _ _ ..
O RESTRAINTS j o EXTERNAL ,
VALVE VEND 0R DATA ACTUATOR DATA DIAGNOSTIC SUPPORT SPARE PARTS o INTERNAL O -
AVAILABILITY FIVE (5) M0V'S/ WEEK ,
{
I
. l i
l i
l 0 l l
a
O GENERIC LETTER SCOPE 1
o MAJOR EXPANSION l i
DESIGN REVIEW l
RETEST REQUIREMENTS o LIMITED DIAGNOSTICS 1
o CRITICAL VALVES e
O
(
l i
i P
O
O MOV PROGRAMMATIC APPROACH l
. o CONTROL /0RGANIZATION o DESIGN AND EQUIPMENT REVIEW AND VERIFICATION 7
o DESIGN CONTROL o PROCEDURES TRAINING & QUALIFICATION
^
o O
o IMPLEMENTATION / MAINTENANCE AND TESTING l
o PREVENTIVE MAINTENANCE l
l O l
O O O EPRl/NPD -
/ CURRENT EPRI ACTIVITIES 1 RELATING TO
- MOTOR OPERATED VALVES (MOV5 . .
f ) RP2233-2 MOV IMPROVEMENTS l RP2814-2 TECHN! CAL REPAIR I
GUIDE-SMB-000 (NMAC) l RP2814-6 APPLICATION GUIDE-L!NES FOR MOV's (NMAC)
RFP2233-9 MOV TEST l
PROGRAM l
Motor Operated Valves (MOV)
Brooks-MOV 9388 1
l O O O l COST IMPACT o BASELINE
- $570,000,000 -
ESTIMATED COST BASED l ON ACTUAL EXPENDITURES l IMPLEMENTING IE BULLE. TIN l 85-03 i
i
! $ 39,500,000 -
ESTIMATED COST BASED ON BROOKHAVEN NATIONAL LAB ANALYSIS O PERIODIC TESTS
$420,000,000 -
PER PERIOD
O O O EPRl/NPD CURRENT EPRI ACTIVITIES s RELATING TO I MOTOR OPERATED VALVES JMOV) . .
RP2233-2 MOV IMPROVEMENTS RP2814-2 TECHNICAL REPAIR GUIDE-SMB-000 (NMAC) e RP2814-6 APPLICATION GUIDE-LINES FOR MOV's (NMAC)
RFP2233-9 MOV TEST PROGRAM f
Y emoks-MOV 9EB8 Motor Operated Valves (MOV)
)
i I
O O O EPRI/NPD l
RP2233-2 MOV IMPROVEMENTS DEVELOPMENT TO EPRI'S "INTELLITORQUE" SYSTEM g&s _
CONTROL & DIAGNOSTIC FUNCTIONS FOR MOV's PERMANENT INSTALLATION - mur pl od Jo/a A'J<e6 DEMONSTRATION UNITS INSTALLED /. .x n 4 -)
n_, c
- COMMERCIALLY AVAILABLE IdYD<d>&#-b' l INTERIM REPORT NP-4254 /c-4 y fgc g l FINAL REPORT NP-5696 PUBLICATION 401988 1 A Motor Operated Valves (MOV)
Brooks-MOV'wd,38
i o O O EPRI/NPD i
i RP2814-2 TECHNICAL REPAIR l GUIDE SMB-000 LIMITORQUE JNMACT COMPREHENSIVE MAINTENANCE DOCUMENT PREVENTIVE / PRE 01CTIVE FAILURE MODES l
SPARE PARTS j i IMPLEMENTATION AIDS PUBLICATION 40 1988 l l GUIDELINES FOR LARGER SIZE UNITS ANTICIPATED
)
i Motor Operated Valves (MOV) l Brooks-MOV 9/SB'3
~
i O O O EPRI/NPD =
RP2814-6 APPLICATION l GUIDELINES FOR MOV's (NMA.C?
sp w; a a 4d [ dem)u re&ur MOV DESIGN AND FUNCTIONAL REQUIREMENTS 4
j -
MATCHING OF VAL'/E AND THE MOTOR OPERATOR SELECTION AND SET-UP OF CONTROL AND PROTECTION DEVICES PUBLICATION 101989 Brooks-MOV 9/8/88
i o O O EPRl/NPD -
i RFP2233-9 MOV TEST PROGRAM OBJECTIVES l = PREDICT PERFORMANCE OF MOV's SIMPLIFY IN-SITU TESTS METHODOLOGY ACCUMLATE & EVALUATE EXISTING MOV TEST DATA CONDUCT ADDITIONAL TESTS AS REQUIRED '
~
UP-DATE MOV SIZING CALCULATIONS _
[knf(Scn EV =
Motor Operated Valves (MOV)
Brooks-MOV 9/EBS
s M(
ht EARLY RESULTS OF GATE VALVE
{) FLOW INTERRUPTION BLOWOOWN TE0TSa Kevin G. DeWa11 Idaho National Engineering Laboratory EG&G Idaho, Inc.
Idaho Falls, Idaho 83415 ABSTRACT The areliminary results of the USNRC/INEL bhigh-energy BWR line brea( flow interruption testing are presented. Two representative nuclear valyc assemblies were cycled under design basis Reactor Water Cleanup pipe break conditions to provide input for the technical basis for resolving the Nuclear Regulatory Comission's Generic Issue 87. The effects of the blowdown hydraulic loadings en valve operability, especially valve closure stem forces, were studied. The blowdown tests showed that, given'enough thrust, typical gate valves will close against the high flow resulting from a line break. The tests also showed that proper operator sizing depends on the correct identification of values for the sizing equation. Evidence exists that values used in the past may not be conservative for.all valve applications. The tests showed that improper operator lock ring installation following test or maintenance can invalidate in-situ test results and prevent the valve from performing its design function.
C Introduction As part of the Nuclear Regulatory Comission's (NRC) Equipment Operability research program, the Idaho National Engineering Laboratory (INEL) is performing research to provide input for the technical basis for resolving Generic Issue 87 (GI-87), "Failure of HPCI Steam Line Without Isolation." This INEL research program also provides iaformation applicable to two additional regulatory items: (1) Generic Issue II.E.6.1, "In Situ Testing of Valves" and (2) IE Bulletin 85 03, "Motor Operated Valve Connon
- Mode Failures During Plant Transients Oue To Improper Switch Settings."
The objective of the INEL research program is to determine wheth9r isolation valves in high-energy BWR piping systems will close against high flows in the event.of'a ' pipe break outside containment. GI-87 applies to ,
those process lines that comunicate with the primary system, pass through
- a. Work supported by the U.S. Nuclear Regulatory Comission, Office of Nuclear Regulatory Research, under DOE Contreet No. DE-AC07-76ID01570, under the direction of Dr. G. H. Weidenhamer, Technical Monitor.
- b. United States Nuclear Regulatory Comission/!daho National Engineering Laboratory.
O
containment, and contain normally open isolation valves. Three process lines meet these criteria: (1) the High Pressure Coolant Injection (HPCI) steam supply line, (2) the Reactor Core Isolation Cooling (RCIC) steam pd supply line, and (3) the Reactor Water Cleanup (RWCU) supply line. Of the three, an unisolated break in the RWCU supply line was determined to have the greatest safety impact. The design basis hot water blowdown testing was performed to provide information on valve operability questions associated with the RWCU environment (subcooled water flashing to steam). The majority of RWCU containment isolation valves are 6-inch, 900-pound, flexible-wadge gate valves with Limitorque electric motor operators.
Backaround The gate valve, Figure 1, is designed for use in a system where a positive shut-off is required with minimal pressure drop when the valve is open. It is ideally suited to those situations where isolation of one part of a system from another is required and control of the dynamic properties of the fluid (throttling) is unnecessary. With the disc (or gate) in the raised position, the run of the valve is free of any obstruction with ,
approximately the same head loss as in the adjacent piping. When the disc is lowered into the seat, the upstream pressure forces it against the seat, creating a seal and isolating the downstream system from the fluid.
Most of the valve and operator manufacturers use the same equation to determine valve closure forces, with only minor variations in coefficients.
In this equation, the required force to cycle the valve is equated to the 4 sum of the disc drag load due to differential pressure, the stem rejection load (stem end pressure load), and the packing drag load, as detailed below.
F p(Ad)(AP) A3 (P) + F p (1) t Where F t Total stem force p - Disc friction factor Ad = Disc area
- AP - Differential pressure A
3 Stem cross-sectional area P System pressure F
p
- Packing drag loao (a constant).
1he disc load is determined by multiplying the exposed disc area by the differential pressure and a disc factor to account for seating surface and guide sliding friction. The disc f actor normally used for wedge type gate valves in Equation (1) is 0.3. The stem rejection load is found by taking the 4 -
stem cross-sectional area and multiplying it by the system pressure. Note that in this equation the stem rejection load can be either positive or negative depending on whether the valve is closing or opening. This is i because the stem rejection load is always in a direction out of the valve '
body; the stem rejection load resists valve closure and assitts in valve :
! opening. The service conditions used in the force equation are supplied by
- each individual plant.
t O ,
t
U O
I \
' \
" w,,' ,
9 *= y Limitorque Operator L, ,1
+ 'w 'W
'1.ll:l I
l Yoke
! I d l: I f-ac t
- tA,N,t-. fl.4 4 ,en e mo ten Il i I I i 1 l} l.I lil lil* 'lil hl.
I i 1 8I N' '~
Seat Ring l ,,.
!- ge ,
M/ y, //)
(Closed Position)
Figure 1. Typical motor-operated gate valyo.
I l
O -
l
Test Desian In the blowdown test research program two full-scale, representative O nuclear valve assemblies were cycled under design basis RWCU pipe break conditions. Both valves were modified to incorporate extended yokes (4 inches longer than normal) and stems cut in half and threaded to allow installation of the special stem force instrumentation. Flanges were attached to both sides of each valve for mating with the test system piping.
The first test specimen, Valve A, was a 6-inch, 900 pound standard rated, cast steel, flexible-wedge gate valve with a pressure seal bonnet and butt
' weld ends. The valve seats were hard faced with Stellite and seal welded to the valve body. The one-piece flexible wedge (disc) was also hard faced with Stellite on the seatirg faces, and the disc guides were carbon steel. The valve was manufactured for this test program by Anchor / Darling Valve Company using a nuclear grade body casting, nuclear design and materials, without third party inspections. The valve was powered by a Limitorque SMB 2-40 electric motor operator. The valve design was representative of valves installed in early nuclear power plants.
The second test specimen, Valve B, was a 6-inch, 900 pound standard rated,
> forged steel, flexible-wedge gate valve with a bolted bonnet and butt weld ends. The valve seats were hard faced with Stellite and seal welded to the valve body. The one-piece flexible wedge (disc) was also hard faced with Stellite on the seating faces. The valve was manufactured for this test program by Velan incorporated using nuclear design and materials, without third-party inspections. The valve was powered by a Limitorque SMB-0-25
- electric motor operator. Representing one of the newer valve assemblies delivered since 1970, the Valve B design incorporated hard faced disc guide wear surfaces.
Q Both valves utilized 460 Vac, 3 phase, 60 Hz electric motor operators. To ensure valve closure and data collection at the anticipated greater-than-normal loadings, Valve A utilized a larger, greater-capacity motor operator than would normally be used. The motor operator used with Valve B was sized in accordance with curr nt practices to represent a typical valve used in nuclear power plants today. With their differences in internal and friction bearing surface design, the two valve assemblies represented a large number of motor-operated valves used in nuclear plants today.
The test system used for the subcooled water blowdown testing featured a large water tank, heated and pressurized so that various system water
' conditions.could be established and regulated, replicating actual BWR conditions. The water was propelled by high-pressure gaseous nitrogen. The water heating system consisted of a heating section and a high-pressure, high-temperature water pump. The heating section contained an electrical heater in an 8-inch pipe and heated the water as the pump recirculated water from the pressure vessel, through' the test section and heating section, and back to the pressure vessel. The test section was a 6 inch pipe with the test specimen mounting flanges and appropriate fittings for obtaining temperature and pressure measurements. The system is shown schematically in Figure 2.
O
N
, _'~ c;v ~
,9 O v O-T/g Test Specimen I,II,IltE,lE, 3
STILS,lLS,lTS
~
A i 'A 2 #As Bleed P4 ,
P3 o Bleed
- o i :le
- 4 e c H: l F-c &
Blowdown Mechanism " ^*#A s#As Pump u
. o Heater l[+
-i s
Water tank - -[ S c ; " c
- p Flowmeter Vent Valve BMGOO17F.
Figure 2. Test system schematic showing instrumentation locations.
. To accomplish the functional testing, the t,ystem contained bleed valves which provided the means to reduce system pressure on both sides (upstream and downstream) of the test specimen. In this manner, differential pressure
(].
conditions could be established across the test valve's disc. The test system also featured a fast-acting (approximately 300 msec opening stroke),
hydraulically operated valve, positioned so that when the valve was actuated, the system's fluid was abruptly dumped to the atmosphere, resulting in high-flow (blowdown) conditions through the test specimen.
The test system was instrumented to monitor flow, pressure, and temperature at various locations, including test valve upstream and downstream positions. Motor operator electrical characteristics were also recorded. Valve stem force was monitored using a special high-temperature load cell installed between the two halves of the specially designed valve stems. The test parameters measured are listed in Table 1.
A secondary objective of the blowdown test program was to determine how normal utility in-situ valve testing, using available diagnostic equipment, could be used to provide assurance of a valve's ability to isolate pipe break flows. By measuring actual stem forces and associated motor operator parameters, in-situ plant testing plans may be developed using available diagnostic equipment; the values of the functions measured by this equipment are then extrapolated to actual stem forces. Several motor-operated valve (MOV) diagnostic system manufacturers supported this objective by participating in the testing as listed in Table 2 in chronological order.
The manufacturer participation was not a competition but, rather, an attempt to determine what factors need to be considered to provide reasonable assurance of valve operability using each device.
O While general observations arc made in this paper concerning the use of this diagnostic equipment, in-depth analysis of the data taken and discussions of how the devices may best be utilized must await review of the digitized data, scheduled for early Fiscal Year-1989, as well as analysis coordination with the diagnostic equipment manufacturers.
Test Procedure Upon installation in the test system, each valve assembly was sebjected to a typical ANSI B16.41 g functional qualification test, including the valve leakage test (Annex A), cold cyclic test (Annex 8), and hot cyclic test (Annex C). These tests were accomplished to provide a baseline characterization of valve assembly operation for comparison with the later testing; The' valve leakage test established the mr.inseat. valve leakage rate and the packing leakage rate of the test valves, while the cold cyclic test demonstrated the capability of the test valve assembly to open and close under adverse combinations of motive power and system pressure with the assembly at room temperature not exceeding 100'F. The hot cyclic test sequence was performed to demonstrate the capability of the test valve 1
assemblies to open and close under adverse combinations of motive power and system pressure with the assembly at design temperature, in excess of 100'F. Annex G, flow isolation, was the subject of this test program and thus was not performed as part of the qualification series.
, O l
i
+
.k e
(Q._/
TABLE 1. TEST PARAMETERS MEASURED DURING BLOWDOWN TESTS Oscillo- Data X-Y Transducer Measurement FM Taoe araoh loaaer Plotter System Water Temp. X Tl T2 Test Valve Inlet Water Temp. X T3 Test Valve Body Temp. X T4 Hesting Section Water Temp. X T5 Load Cell Temp. X System Water Press. X X X X P1 P2 Test Valve Inlet Water Press. X X X P3 Test Valve Outlet Water Press. X X X P4 Discharge Section Water Press. X X X API Test Valve Differential Press. X X X AP2 Venturi Differential Press. X X' X AP3 Pump Differential Press. A X 11 Actuator Current X X 12 Actuator Current X 13 Actuator. Current X El Actuator Voltage X X ,
E2 Actuator Voltage, X l
i X
ST Valve Stroke - LVDT X X Open Limit Switch , X X
( )) LSI LS2 Close Limit Switch TS Close Torque Switch X X Valve Stem Force X X X F
l Al Actuator' Acceleration Y X X l A2 Actuator Acceleration X X A3 Actuator Acceleration Z X A4 Valve Body Acceleration Y X X A5 Valve Body Acceleration X X A6 Valve Body Acceleration Z X
__ l
' Control room light indicator only.
t
~ _ - - _ _ . _. . _ _ .
(n)
TABLE 2. VALVE DIAGNOSTIC E0VIPMENT USED FOR SUBC00 LED BLOWDOWN TESTS Test Diagnostic Valve Series Description _Eauipment A 1 Qualification Test MCSA A 3 Blowdown, 1000psig, 480'F MCSA A 2 Blowdown, 1000psig, 530'F None A 4 Blowdown, 1000psig, 400*F V-M005 A 6 Blowdown, 1400psig, 530*F V-MOOS A 5 Blowdown, 1400psig, 580*F MOVATS A 7 Blowdown', 1400psig, 450'F MOVATS A 9 Blowdown, 600psig, 430*F None A 8 Blowdown, 600psig, 480*F None A 10 Blowdown, 600psig, 350'F MAC, VOTES A 11 Blowdown, 1000psig, 530*F MCSA B 1 Qualification Test MCSA 2 Blowdown, 1000psig, 530*F HCSA, MOVATS B
B 3 Blowdown, 1400psig, SS0'F V-MODS B 4 Blowdown, 600psig, 480*F MAC B 5 Blowdown, 1000psig, 530'F V M005 MAC Limitorque Motor actuator Characterizer Os, MCSA ORNL Botor [urrent lignature Analysis M0 VATS MOVATS, Inc. (801 Analysis and Iest System)
V MODS WYLE Labs. Yalve Motor Operator Qiagnostic lystem VOTES Liberty Technology Yalve Qperator lest & Evaluation System o
O
Once baseline qualification testing of the test valve assemblies was completed, several test series were performed to address the questions of GI-87. Each test series included leakage tests, cyclic tests without flow, 3 cyclic tests at normal system flow, and cyclic tests at full blowdown (d conditions. A wide range of design upstream pressures and temperatures was maintained throughout the valve closures, with blowdown flow limited only by flashing and choked flow at the test valve or piping exit. Each test series was structured to optimize the amount of useful data obtained. Required non-flow data were collected during the preparation period for full-scale flow and post full-scale flow.
Fourteen blowdown tests were accomplished, ten on Valve A and four on Valve B. All tests were performed to evaluate the engineering parameters In addition, the required to calculate closure loads for a typical valve.
four tests with the Valve B and normally sized operator were performed to demonstrate expected in-service performance with the motive power closer to normal.
Test Results and Interoretation Valve A: A torque switch setting of 2.0 was selected for the motor operator so that the stem thrust capability was maximized without exceeding the valve and instrumentation capacity. (The torque switch was reset to 2.5 after test 10 to compensate for an observed torque-out anomaly, discussed later.)
The valve clused satisfactorily for all tests; however, '
higher-than predicted stem loads were observed during the blowdown flow isolation cycles.
Figure 3 is a reproduction of the stem force trace for the blowdown test with initial fluid conditions at 1000 psig and 530*F subcooled water.
O The figure shows the stem compression (negative values) increasing as the valve closes until it reaches a peak where the flow path is finally blocked and the valve disc is riding against the full seat ring. At this point the stem compression decreases sharply to a value representing the force required to slide the disc on the seat to the fully seated position. Then the force rises sharply through final seating and torque switch trip. The
) compressive force in the stem then continues to increase due to the time lag l
in the circuit dropout time and the momentum in the operator. The peak shown shortly before final seating of the valve is the force that must be j
overcome for successful closure. This peak is the flow isolation force referred to in comparisons and discussions throughout this paper.
. ~ Equation (1) was used to predict stem forces during closure at high flow. Following typical industry practices, the parameters chosen for use in the equation were conservatively chosen to be maximum design pressure (P
& AP) and assumed worst case packing drag force. The Valve A closure -
to calculate disc thrust was predicted using nominal valve size (6 in.)
area, maximum upstream pressure for P & AP, a packing drag force of 1500 pounds, and a 0.3 disc factor. The calculation was repeated using a 0.5 disc factor following recent industry recommendations for gate valves in high flow applications.
O
O O O Valve A,1000 psig, 530 *F 0, . ., . . g . . . .g . . . . g . . . .
- =
Closure stroke begins ,
_,o _
7 s..io ,o,c,__
g -
=2 _
8En o" -20~- -
IL 3 ~
E F' " '*d
,E F- _
p Torque ,
G _
switc83
- trip
-30 - ,
-40 O 10 20 30 40 Time (s)
BMG00181 Figure 3. Load cell mea:urements show the effect of blowdown flows on
. talve closure' stem loads.
4
Figure 4 shows a comparison between the predicted stem forces (c i a ti- ,
above assumptions) using 0.3 and 0.5 disc factors and the actual ste. .- <
measured during the blowdown isolation :teps. For Valve A testing, both ... '
l O a re s> # res e nd te a e te re s w ere v rie d t o c h e ns e t h e d e9 re e o f s ehc o oi tize and thus degree of flashing occurring at the valve during :losure. This figure shows that at the lower pressures the actual stem forces were closely predicted by the sizing calculation, as were the forces that were at higher pressures but greatly subcooled. The sizing caiculations did not bound stem forces for the tests at higher pressures with water closer to saturation temperatures. A disc factor of about 0.57 would be necesstry to reach the 19,400-pound stem force measured during the 1000 psig, 10*F subcooled water test.
In order to investigate how well Equation (1) models the actual behavior of the valve, the conservatism needs to be removed from the calculation. If the disc load is eliminated from Equation (1), such as would be the case without flow, what remains is a linear equation in slope-intercept form (y - mx + b), namely:
F t
=
As (P) F p (2)
Note that this equation has been written so that the stem rejection load is always negative (compression) while the packing load is either negative or positive depending on whether the valve is closing (compression) or c,pening (tension).
Figures 5 and 6 show the test data that cre expected to apply to the above equation. The data plotted are the stem force measured at mid-stroke (running load) for tests at varying temperatures and pressures but without flow. , The line fit through the data points has a slope equal to the stem crot.', sectional O area and provides an indication of the true packing load for each case. For Valve A, data show a packing load of 835 pounds for opening and 430 pounds for closing. Both values are well below the IE00 pounds used in the sizing equation, providing additional margin in the calculation. The difference between the t'wo values can be partially accounted for by the weight of the disc and lower half of the stem. This difference also provides evidence that the packing load is affected by direction of travel, possibly caured by water c;.rried with the shaft changing the lubrication of the packing / Item surfaaes or by other phenomena associated with stem travel through p h og.
Based on the packing load and stem rejection load characteris" Mo+ned from the running load evaluation, an evaleation of th 106. ', 'ce
. ..o (Equation (1)] was conducted. Here the conservatisms f a wir ; ion of the equation were eliminated by using measured parameLt ui
.- % sation. Figure 7 contains the results of this calculation for the bi t, .vn test shown in Figure 3 and compares it to the measm .d stem force.
As might be expected, the calculated force bounds actual values during the first part of closure, when little of the disc area is exposed to the flow and differential pressure across $he valvo is small. As the disc drops further into the flow stream and differential pressure increases, the calculation ceases to ' ound measured values, where the actual flow isolation stem force exct.eds t h e calculated.
j O i
l
\
! O .
O O i
i 1
25 . . . g . . i g ..ig .- i ,
g , , ,
~
a 10' Subcooled
',/-
X 50* subcooled '
20 - / 6 -
v 100* Subcooled A /,/ ,
- 0.3 Calc. f actor '
n
- --- 0.5 Calc. f actor
,s,/ _
j3
~q 15 X/ /,e -
-u _
s
,- y
@ c ,-
3 a -
3 - /, -
2 r =3 p
o
- /
/ V .
E s' j EF a-10 -
,/ .
us -
!l
~
w _
u .
't. -
1 5 -
1
. 1 . , , f , , . I , . . I . . .
0.5 0.7 0.9 1.1 1.3 1.5 I
(Thousandst i
Upstream Pressure ipsia) i
! sucooies l Figure 4. Valve A ineasured stem forces exceeded those calculated for tests at high pressure and slightly subcoolea conditions.
. , . - - - , , -u- m - - - . _ _m-w --w- , , - . __ y - e- ~ w - - - - . - , - , - - - - , . -, --
v-- - i ,
s t.e ,
i i
- . g .
i . g . g .
3
., - 0 Actuet -
- Case. rectes3Se s g _
L- .a .- . -
L
..t 8,
g h .
a3
.I$. .e -
8 1.4 -
g O -
O
. ..e -
g -
i i t i i i i i I . I
, g,, i i . i i
- e. .3
.4 .6 .s t4 1.2 1.4 16 (Thousands) .
System Pressure (psig)
Figure 5. Linear curve fit using industry equation closely approximates Valve A running stem forces for opening without flow.
l a w...
-- C r.. ..< se ~
O
.se -
. t.t - T1
,,,1 3 ..e 8 - i
' )g - -
l e .:.s - o -
- O
..e -
1
.at 4e -
. Q .
t ' ' ' ' ' ' ' ' ' ' ' ' ' ' ' 'U'
.6
- e. .4 e 1.2 te to a4 (Thousands)
Statem Pressure (psig)
Figure 6. Linear curve fit using industry equation closely approximates Valve A running stem forces for closing without flow.
o .
e . - - - -- -
D
~
O O O Valve A,1000 psig, 530 'F 4 0 . 1 . . g . . . ,, .. ,,...
_ G- _- . - - - . . -
1 W ~ ~ 2 ;- .3 ~ q ~~~ ,,,~- .. ...._..~. -
-s
- g _
. s
-10 -
- Calc - f actor =.3 's '
~
~
-- Calc. f actor =.5 9 i
2 ~
s\ '
~
g ,
- Actual ,
, y'- -
I o
u 8 -
i o"3 -20 -
~
~
u.
g E -
.t -
m -
l _
! -30 -
- ' ' ' I ' ' ' ' I ' ' ' ' ' ' ' ' '
-40 O 10 20 30 40
'ime (.')
BMG00197 Figure 7. Calculations using 0.3 and 0.5 disc factors and w ving .
typical conservatisms do not model ;;est data fe high-energy flow isolation.
The valve assembly was disassembled after testing to evaluate the condition of valve internals exposed to the high-flow testing. The disc (weige) was removed and examined to determine how the blowdow- forces had affected its condition. The disc was worn or, the guides and .ealing surfaces of the downstream face, indicating a cocking of the disc toward the downstream side. The upstream face was r.ot damaged. An examination of the assembled position of the disc valve body revealed large clearances of about 1/4 inch between the disc and disc guide rails mounted to the valve body.
This clearance plus the wear on the disc guides and guide rails allowed the disc to cock downstream far enough to come in contact with the seat ring during the last half of the closure stroke. This contact explains the wear pattern ;ound on the disc face, and may also be the cause of the high closure forces noted just prior to flow isolation during the higher-energy blowdown tests (see Figure 3).
Judging by the wear patterns observed, it is unlikely that the valve could have produced a tight seal using the downstream face alone. A tight seal on the downstream face would have been required to isolate flow if the valve torqued out prior to full travnl but with the disc on the seat. The valve, however, maintained a tight seal throughout testing, indicating a tight seal on the upstream face of the disc -a benefit of using an oversized operator with a higher-than necessary torque switch setting.
Valve C: A torque switch retting of 1.75 was selected for the Valve B assembly to provide the needed closure thrust for the given test conditions. This setting, however, resulted in delivered stem thrust (as determined by the stem mounted load cell) that was below the calculated magnitude. The torque switch setting was therefore raised to 2.0 prior to 1
the first system blowdown test.
The valve performed satisfactorily during the testing except for the 1400 psig test, where the operator torqued out before the disc reached the fully closed pc ition (1/4 inch travel reaaining). Even during this test,
> however, the valve had closed far enough to produce a tight seal, with no leakage observed. ,
In accordance with manufacturer procedures, the valve thrust was predicted using actual orifice area for disc area, upstream pressure for P &
AP, a disc factor of 0.3, and a packing drag force of 5000 pounds (the maximum expected for this packing configuration). Based on the initial results of the Valve A tests, all of the blowdown tests with Valve B were conducted at near saturation temperatures for the various pressuies. Figure l
J
- 8 shows a comparison bet,een the test measurements and the thrust calculations for the Valve B tests. The estimated stem forces, shown as the
- solid line, provide reasonable predictions of actual forces for I lower-pressure cases. Maximum flow isolation forces could not be measured for the 1400 psig test because the torr,ue switch tripped prior ta full l .
closure. Figure 8 therefore shows the extrapolated stem force ct 1300 psig (pressure at flow isolation) based on actual stem force characteristics prior to torque switch trip for two closure cycles. Here, as with Valve A,
' predicted stem forces are not conservative at higher pressures.
1 4
I W
1 l 0
. 8 1
0 0
G O M S
5
" 1
- - - - ~ _ _ - _
- ~ r o
f d
' e .
, ts an l o ui ct li 3 ad cn g 1 o g ec s
od he tl o
a do
)
ec g db eu i
s es p c 1 (
xy el g 1 t
)sre sh eg d u ci ns a s rl os f
' sar d s
o mn a
e u ce oP h t ea o se Tm O y f
e r
u 9
( a e
r t
du es t s ae r
e s s l r r
uh t c p op p
o d U ah l rg ce ' ti i
t - xh o a e to lo t t p Ba pa 3 cx mtr O es 7 vt 3 CE l s g
ae
- oa Vt 8
. O '
. . . , e r
/
u g
i
- F
- - - - - 5
- ~ - - - - -
8 4 4 r 0 a S 4 2 0 6 4 2 0 1 i 1 1 1 2 2 2 gE a:I 2O %:snr m O
I l t Illli ll'1 ll- I
Figures 9 and 10 show the Valve B test data that are expected to apply to Equation (2). Here again, the data plotted are the stem forces measured !
O at mid stroke for tests at varying temperatures and pressures but without V fl ow. The l'.ne fit through the data points has a slope equal to the stem cross-sectional area and provides an indication of the true packing load for both the opening and closing strokes. The increase in packing farces over .
those found for Valve A is believed to result from the different packing '
design and greater stem diameter of this valve.
The line fit through the data points for this valve shows a acking load
.of 1610 pounds opening and 1632 pounds closing. The force calcu ation for i this cise included a 5000 pound packing load, again providing conservatism.
The difference between the opening and closing values is less than expected, l given the weight of the disc and lwer stem half (over 50 pounds). This may j indicate a directional relationship for packing load, believed to be a '
I characteristic of the packing type used and its orientation.
Based on the packing load and stem rejection load characteristics determined from che running load evaluation, an evaluation of the total ;
force equation wai conducted. Here again the conservatisms found in the i 1
application of ths a Nation were eliminated by using measured parameters in the equation. For the Valve B assembly, stem force calculations ,
under predicted actual measurements. Figure 11 shows the maximum stem forces for closure during the blowdown testing at 1400 psig and 580*F. i As with the previous valve, the calculated force bounds actual values during most of the closure cycle. As the valve isolates flow and differential pressure approaches full system pressure, the stem force exceeds that .
1 calculated. !
O No wear patterns were noted on the Valve B disc after testing. The inspection of the valve internals noted tight clearances in the disc t
guide / guide rail interface and machined, hard faced disc guide surfaces.
Ocerator Toraue Switch Trio Anomaly: During the blowdown testing of Valve A, there were three incidences of anomalous operator torque switch tri) l behavior. It is believed these incidence's occurred in conjunction witi installation and removal of the HOV diagnostic test equipment. The valve !
stem forces associated with torque switch trip were normal in the numerous i tests performed with diagnostic devices installed. The anomaly appeared in ,
i the form of abnormally low valve torque-out stem forces during the tests immediately af ter removal of two types of diagnostic equipment. The ;
investigation that followed the discovery of the low stem forces showed that .
incorrect installation.of the motor operator spring pack lock ifna was the l
! i problem. The removal of the diagnostic test equipment and the sutiseqeent '
! incorrect installation of the lock ring invalidated the findings of tf e diagnostic test A recent problem investi atlan at Brunswick I (LER 87 023 01)2 has identified a similar ock ring installation probite .
and illustrates the potential for invalidated diagnostic testing elsewhere. l i
l The point at which the torque switch contacts open is dependent only on the setting of the torque sWch, rate of the torque spring, and spring pac;
- 1 i
3
' I ' i '
i '
i i '
i i '
i i I
o At te.t p h -
y - - c... e.o .ieio g .
~7 -
gi e -
33 .
- g I .i - US -
g .
. c C O -
, i , i , i , i , i , i , i , i , i , i e u e4 es e. is u u u u .e (Thousar de)
Syttom Pressute (psig)
Figure 9. Linear curve fit using industry equation' closely approximates Valve B running stem forces for opening without, flow.
! -i ..,,,,,,,,,,,,,,,,,,,,,
, a 4. .. -
.: -- - c.w. e.....ies:
?
3 ,
~_
g .
- og -
,3. j .. .
t g .s -
- l E -
O
.e .
..n . . . t , i i l- ,,
,,in . . i i , . . i .
- s. .4 .e u te s.e 34 (Thoveende)
System Proteute (peit)
Figure 10. Linear curve fit using industry equation closely approximates Valve B running stem forces for closing without ficw.
O
O O O
. i Valve B,1400 psig, 580 'F i i i g i i e i g . . .
i i i g 10 (
- - Actual ..
-- Calculated
~
5 -
- 0 a n
~
U C -
. e .. -----------___- -
o 68 -5 -
- u. o ,%.
O ~~~~, s
( gce _
's '
3- ~
" -10
's, -
's
's s .
.i %
I i
! -15 1
' i e i i i e i I e i i f i e i I i i e i
0 20 40 60 80 100 1
Percent Closed i
. sucoover i
Figure 11. Calculations using 0.3 disc factor and removing typical conservatises do not model final flow isolation forces.
i I
-.v, - ---%- .-,---,---v-- r_--.ri,.~ . . - - -,-,--w-.-- . , - - - - - - - , - - - - - , - - - -
. - .--- - _ - - . , . - - - - = - - - ._ -
preload and gap. No matter what causes the stem force to increase, whether flow loads, valve reaching full stroke, or even an obstacle in the disc path, the switch will always epen when the torque spring compresses to the
, O- predetermined point. The force at torque switch trip was used to trace the I function of the opert. tor from one test to another through various diagnostic equipment instt11ations and rer, ovals. Figure 12 shows the average stem compression at torcue switch trip for each of the 11 tests performed on Valve A, arranged in chronological order. The force measurements were made i using the INEL load cell installed as an integral part of the valve stem.
j During tests 1 through 7 the valve operator functioned consistently, '
'with a stem compression at torque switch trip of approximately 33,000 pounds. Tests 9 and 8, accomplished without operator diagnostic monitoring, showed consistent torque out forces, but at a significantly reduced level.
4 Here a drop of approximately 10,000 pounds appeared in the torque-out stem comwession, t j
Two different sets of valve operator diagnostic equipment were installed i
to monitor test 10, and the valve stem torque out compression returned to e
about the same leven as tests 1 through 7 (re-lubrication of the valve stem threads increased loads slightly). The diagnostic equipment was removed after test 10, and the results of test 11 show a similar reduction in force, even after the torque switch setting was increased from 2.0 to 2.5.
After the completion of testing for Valve A, the Limitorque motor operator was removed and partially disassembled by Limitorque i representatives with INEL personnel attending. The spring pack cover was ;
removed and the internal configuration inspected. The lock ring that t retains the torque spring and its locking set screw appeared to be properly t i
- /3 installed. The set screw was removed and a special tool was used to attempt U. to further tighten the lock ring. The ring was tightened almost one full
! turn before it reached its proper position. l l
Limitorque design records were used to correlate the loosening of the
' lock ring to torque switch setting and torque-out thrust. One full turn of the lock ring is equivalent to 19 degrees rotation of the torque switch. 1 One full torque switch setting is about 21 degrees. This loosening of the >
lock ring had the effect of backing off the torque switch from 2.0 (the actual setting) to 1.1 (the equivalent setting). From the torque spring l curve the loss of thrust was estimated at 10,600 pounds, very close to the l i
discrepancy in the measured data, j leproper positioning of the lock ring sometime after test 7 but before the next test would explain tne reduction in stem force after test 7. How It happened is not completely understood. None of the diagnostic devices
', installed prior to test 9 require the removal or adjustment of the torque spring lock ring; in fact, several of the devices are designed to diagnose spring pack gap, the result of improper lock ring installation. Review of !
the data taken by the various diagnostic devices shows no indication of I
! spring pack gap. Also, none of the devices are designed such that their
' installation would correct for this problem, with the exception of the Limitorque motor actuator characterizer (MAC) device, l f I
i O l l
l i
l NN % % \\ -:
N%NN%%Nb -
i ;
i NNN % Y --
]
j Aggx ..
j
. : x\wxwww -
2 1
s i N%%Nh%N-*f: i g j h%%%N* ~
- xNN % % % + L
e i.
\\\\ \\\ ~"
Nx%%\\W -
9'
_u
~
3 kkkkbb* (sputsnoyl) l (sql) uolsssJdwoo wels O !
k
.. Installation of the MAC device requires tne renoval of the torque spring !
lock ring to facilitate the installation of its spring pack load cell '
(] device. According to the Limitorque technician, the position of the lock v ring was marked prior te removal and the number of turns during removal was !
noted. The load cell device was installed and tightened to the proper position to provide the design spring prelocd. After testing, the load cell was removed and the lock ring instelled the appropriate number of turns to the oreviousiv marked oosition. We believe this explains the similar reduction in stem force before and after test 10.
- A correltted event was determlaed from a review of the Brunswick LER. j In this case, the HPCI steam line isolation valve (a GI-87 valve) had i successfully undergone several diagnostic tests using the MAC system.
Later, the valve motor failed on opening for an unrelated reason. During the subsequent motor operator check, greatly reduced torque out forces were observed. Investigating personnel discovered that a burr on the threads of the spring pack housing cover had prevented the lock ring from being fully ,
installed after diagnostic testing and had caused the lower-than expected torque readings.
Both instances of improper lock ring positioning could have been easily [
- diagnosed. A simple measurement of the lock ring position can be compared with both the position of torque spring transducer during testing and the -
r manufacturer design position in order to validate post-test valve 1
operation. Apparently this procedure was not completed for the l above described tests. !
3 Conclusions 1
O The desi9 n basis het water ble de a testia9 has shown that. 94vea eaoush thrust, typical gate valves will close against the high flow resulting from a line break. Proper operator sizing depends on correct identification of i
i the values for the sizing equation. Evidence exists that values used in the
- past may nct be conservative for all valve applications. The following ,
4 items need to be considered during sizing,of gate valve operators, t 1
- 1. Gate valve internal and friction-bearing surface design can have a i r
significant effect on the operatur force requirements for pipe break flow isolation.
f
- 2. The degree of subcooling at the valve inlet can greatly influence .
valve closure forces. Valve operator force requirements incre'.se [
as inlet fluid conditions approach saturation temperatures. ,
t
- 3. Industry trends toward using 100 percent system pressure for all l pre at-a terms in the sizing calculation are justified for
- high flow appitcations. - i j Improper operator lock ring installation following test n maintenance
- can invalidate in-situ test results and render the valve unable to perform !
its design function. This is important in light of the present trend by ;
O !
1 t
j I
l
r utilities to perform diagnostic testing of safety-related valve assemblies !
to answer regulutory concerns such as IE Bulletin 85 03. A final c;uality .
O check following diagnostic testing and maintenance must be made to ensure correct lock ring -installation.
Raferences !
i
- 1. ANSI B16.41, Functional Qualification Recuirements for Power Ooerated Active Valve Assemblies for Nuclear Power Plants. 1983. i
,2. . Brunswick Steam Electric Plant Unit '1, Inocerability of Hich Pressurt_. l Iniection (HPCI) S'ntem (E41) Due to Failure of HPCI Turbine Steam Inlet J
11glation Valve E41-f001, LER 87-023 01, June 27, 1988, f i
l i
f l
L O
J l'
1 i
i 4
l O
1 ____ _
SOLEN 0ID VALVES ON PR BLEM O
SOLEN 0ID VALVE FAILURES HAVE DEGRADED NUMEROUS SAFETY SYSTEMS FAILURES HAVE BEEN WIDESPREAD AND MANY APPEAR TO BE COMMON MODE FREQUEfiCY APPEARS TO BE INCREASit!G SIGNIFICAN E COMMON MODE FAILUP.ES HAVE POTENTIAL TC RESULT IN MULTIPLE TRAIN /SYSTEf1 FAILURES WHICH ARE BEYOND PLANT SAFETY ANALYSES CAUSES O . DESion DesiCieNCieS MAlliTENANCE DEFICIENCIES MISAPPLICATIONS ENVIRONMENT CONTAMINANTS m
E ,
i :
Q l AIR-0PERATED VALVE DIAGNOSTIC TESTING P
l (An example of work done recently at the ;
Davis-Besse Nuclear Power Station) ;
4 !
O :
t i
i l
t I
I Presentation to the ACRS Subcommittee on Mechanical Components October 27, 1988 f
l
,' l
- P. R. Wohld !
l :
j Consultant i O i
i
4 TESTING, DATA ANALYSIS At:D RESVLTS EVALUATION s
(j
( FOR REACTOR COOLANT PUMP SEAL INJECTION THROTTLING VALVE, MU-19
SUMMARY
MU-19 is a fall-open, pneumatic spring-and-diaphragm actuated, 1" globe valve with flow under the seat. Because the valve has been sticking in and mid travelthe because on packing going open, has with no system pressure present, recently been modified for live loading, there was a question on its operability.
To resolve the question, under zero system flow and pressure conditions, a test rig was set up to collect valve characteristic data from which an evaluation was done to determine if 1) the valve operating paramoters appeared acceptable, and 2) the valve could be expected to perform its function under normal system operating conditions.
The testing, data analysis, and evaluation indicate that the valve is operating properly and should be expected to stop short of fi11 stroke on opening without system pressure
")
y available. This is because the spring force available at the resulting from the s
end of valve opening is only 180 pounds, required bench set. Hence normal packing friction loads At 3000 greater than this amount aW111 stam prevent rejection full opening.
force of 1325 pounds pounds will system pressure, be exerted on the 3/4" stem, adding to the 180 pound This results in spring force ample macginavailable to openthe to overcome theactual, valve.863 pound dynamic !
friction load calculated from test data.
One concern raised by an analysis of the data is valve control stability. The static friction phenomenon, and its sudden relaxation at the beginning of valve stem motion, was noted to cause about a 0.15" jump in valve motion. When this amounts to a 33% chance throttling at 45% open 40.45"),If th:,s affect is not reduced during valve in valve stroke.
throttling and causes control instability, corrective mensuren will need to be considered (stronger actuator, '
control system adjustmenc, packing load reduction, etc.).
Table 1 shows parameters characterizing the valve and actuator that were determined through test data analysee.
The expected value is included on the table, when available.
e.
TABLE 1
(
v RCP SEAL INJECTION THROTTLING VALVE, MU-19, VALVE AND ACTUATOR CHARACTERISTICS DETERMINED FROM TESTING, 8-6-88 VALUE DETERMINED EXPECTED ITEM: FRe,t TEST DATA: VALUE:
- 1) Packing Friction:
Static - 1003 lb << 2492 lb Dynamic - 863 lb
- 2) Actuator Spring pate: 758 W/in 780 lb/in
- 3) Bench Set (at closuro): 15 psig 15 psig
- 4) Seat Load (under zero sys. pressure and flow): 2675 lb
- 5) Stem Travels approx. 0.4 in 1.0 in
- 6) Other: The valve stam was noted to jump 0.15 inches at the beginning of va've motion (on overcoming valve packing static friction).
O TEST SETUPt Test setup included two sensors, a 0-100 psig pressure transmitter (with a 0-5 VDC output) to monitor valve diaphragm pressure, and a spring loaded, lanyard potentiometer to sense valve stem motion. Motion was initiated by both manually adjusting the demand signal in the Main Control Room and by using a pressure regulator to directly change loading pressure on the valve diaphragm. An IBM compatible, portable computer based data acquisition store and analyze data from the system two was used sensors, Datatocollect collect [on was triggered by a small change in diaphragm pressure at the beginning of valve stroke.
4 DATA ANALYSIS ANC EVALUATION:
O Packing Friction Load (Dynamic Friction) i The difference in diaphragm pressure between an open and close stroke, at the same point of valve position, is indicative of twice the valve / actuator system dynamic 2
friction. It can be used to calculate friction forces, and, will be called packing friction herein since packing is the only friction source for MU-19. The dynamic friction load for MU-19 is calculated from the test data to be approximately 863 pounds.
e An(Static triction) evaluation of data from va.ve f.troking indicates that an additional force of 140 pounds over the dynamic friction load is needed to start valve notion.
4 Hence, then total static friction load due to packing is Lo03 poundra Both the static and dynamic friction forces are less than half the maximum number calculated by Engineering Design.
The force available from the operator is a pproximately 75 i times 60 (max. loading pressure times dia;aragm area), or 4500 pounds. This, together with an evaluttion of other valve loads indicates that the valve actuator is more than adequate to handle the friction loads.
1 i Spring Rate Spring rate is pounds force change per inch of spring travel ana is easily calculated from the linear portion of the diaphragm pressure vr. stem position data. This calculation resulted in a spring rate of 75s pounds per inch which I appears acceptably close to LEn 780 pounds per inch rate stated in specifications for the actuator.
Bench Set (at closure) ,
An interesting discovery in the development of the testing ;
, done was that the valve actuator This calibration Lbench set) can is partLcularly
' be ve*;1fied from th9 test data.
interesting bcenuse it allows verifying the bench set without :
' having to disconnect the actuator from the valve loads. ,
(While the technique is simple, the theory and technical ,
discussion gets somewhat lengthy and wil) not be covered ;
herein.) ,
Becauce MG-19 did not fully open during the test, the low !
bench sst pressure could not be observed in the loop. The !
upper bench set pressure was datormined to be 15_pshgt exsetly what it is called for on the valve calIErataon data :
sheet. Since the upper end la properly set, and because the :
spring rate appears reasonably c1cse to specifications, therc ,
is no concern for the lower and of the bench set.
()
i 1
- , ~ < - . . _ - -. . ,- - _ _ - _ _ _ _ _ _ _ . _ _ . . . - _ -m_-- _ ,- -,
I seat Lead The seat load is determined by noting the pressure at which ;
the valve is fully seated, determining the increase in ,
pressure beyond that re that increase by the di ired for seating, and multiplying hragm area. The set load determined during the test was 2678 pounds at a maximum pressure to the diaphragm during testing of 75 psig.
This seat load during testing represents margin in the actuator ca i operation. pability Half orto handle this additional is necessary loads during to overcome the system stun rejection force at maximum system pressure, leaving a significant margin to handle flow induced and differential !
pressure forces on the valve plug during throttling l conditions. (Leak tight closure and seat loads under system operating conditions are not considered important for this valve and were not evaluated.) .
L stem Travel '
Stem travel is measured directly by the lanyard potentiometer.
Actual stem travel varied somewhat with each stroke, not going full open, and was approximately 0.4 in.
4 The failure co go full open under static conditions was already knoiin. The testing described herein determined that ;
this is due to packin the actuate r spring. gThe friction stem exceeding the during rejection force capcbility systemof l
C* operating tionditions will be adequata to aid the spring in f ully openir q the valve.
, other I
As air pressure w3m increased to begin valve stroking in the close direction, a rapid stem position change of 0.15 inches l i
was noted. The reason for this is static friction which has .
i to be overcome prior to the start of motion. While this jump does not appear abnormal, it is an undesirsable feature for a !
control valve and could cause a stability problem during throttling on the flow control loop. i No unusual reatures were otherwise noted in the valve data.
i l \
L i
i I
.[
t t ,
3 I
l s
CONCLUSION The conclusions from the testing done on 8/6/88 are that
- 1) the actuator is set properly according to bench set specifications of 3 to 15 psig, 2) the valve and its actuator appse.r to be operating normally, 3) the valve can be expected to stroke fully open with the help of system pressure, and
- 4) there is significant margin in the actuator to handle dynamic loads from operating pressure and flow conditions.
What cannot be concluded is valve control stability. The fairly significant jump in stem position at the beginning of motion is cause for concern.
Short term action should be to check the system for control stability with operating system pressure and flow conditions.
Control system adjustments and/or a relaxation in packing gland load :an be considered if control is not stable.
Leno term action should include an evaluation of ongoing
- /alve experience and possible design improvements if problems continue. Also, it is apparent that techniques for valve calibration, troubleshooting, etc., need some re-evaluation to properly set the valves up, most efficiently assure valve reliability, and correct problems in a timely manner.
O 4
1 J
i j
- 1 1
O i _ _
i
O UnkelScape(c)MIT1984 Trace 1: 5.NI+N+- l l l l l Trace 2: 1 NE+NI y,, ,
SCANAT I A CLOCXTAULTS I .=
+ \
Trace 1: i i 6-1Mrsigliaphrase l\, \, yg,,,, 7;. 2 , f), ,,. g.,,,
rleN=e:u19d4 ' /- # "'#/v.,/6 Secon3OpenStroke I -
6-Aug-198813:29:47 i 1
+
Trace 2: I i >
OpenStrokePosit. / !
fileTr2 I
\' / ' p# # "'#'
FileMme:u19.d4 $ + 2 o /s%g.
O secondOpensirone ; i 6-Aus-198813:29:47 i Trace 1: .NE+MI Trace 2: .Mt+::: : -l :
- t (Trace 1: Tile ) .Nt+N Time (seconds) 5.NE+91 (Trace 2: Tile ) . MI+M Time (seconds) 5.NI+81 O.n f Shke I
I l
I
l l
O 5.88EfM: : : : :
Trace 1:
Trace 2: 1.Mt+MI Comeani (Move Cursors 1 I AssA<a
^
~
-- =
Up/ lown alters active cursorl Right/l.eftmovescursor. + !'
ISCReturnsToCommandMenu.I _
Cursor 1 + a i r SourceITeco I h #"" 8' "'6 V 5.87813E-81 Hot 1.37I+88+ .'
T 4.83344t+88 Pt 1: 881
& cmon I Cursor 2 i SourceITr21 +
V 6.56738E-81 Hor 1.31E+ MI i
T 6.42628E+ M Ft 1: 1171 l
Cursor 2-CursoriValues+
O V 1.48926t-81 Hor-5.86I-821 T 1.59284E+ M Pt 1: 291 Trace 1: .MI+ M Trace 2: .NE+M: : : : :
(Trace 1: File ) .Mt+M Tlee(seconds) 5.MI+81 :
(Trace 2: Tile ) .NE+M Time (seconds) 5.NE+81 C/use % k e-i
[
l l . I
- i
Q -
UnkelScope(c)litt1984 Trace 1: 5.NE6 . . . -. +-
Trace 2: 5.NI+MI SCM AT l CLOCXTAULTS !
, d Trace 1: I &
FINpsis Diaphrap i
A fileTri I TileMme:n19.d5 +
Q '
ThirdCloseStroke I of,, (r,s,,,) a ,.e 6-hs-198813:43:45 I i
I
+
Trace 2 i FINpsigIlaphrap I i III'IfI I Ji c/,s e ! I fileMme:n19.46 +
gg, ofc,,
O ThiraOpeeSirone i <- m l g
~
6-hs-198813:53:36 I w N l Trace 1: .NI+Ni Trace 2: .NE+N: : : :-- -~r "
(Trace 1: File ) .Nt+M C-0(Fr.vsPosit) 1.HI+M (Trace 2: File ) .Nt+N 0-0(fr.vsFosit) 1.NI+N Opw/C/os e. Sfake. Eys/ wests lwp l
1 e
O !
i e i
LT-1000 PORTABLE DATA ACQUISITION / DIAGNOSTIC SYSTEM "A PROBLEM WELL DEFINED IS HALF SOLVED" Minerva Renearch Corp.
l l
l i
l ,
1 ,.
I
, ~ . 6 o
l l j
> i i !
! O 1 l l l I l l I i I
j MAJOR APPLICATIONS )
l PROCESS DIAGNOSTICS - Solve problems of Quality, Productivity, Downtime by gathering data on key variables such as temperature I and pressure i PRODUCT TEST - Quality Assurance, Product Development, in vehicle Testing, Field Testing !
l l
KEY FEATURES t
- Portability
- Real time data display
- Use with any laptop
- No programming required [
- Complete software included
- 16 Channels ;
Y USE AS: STRIP CHART RECORDER - DATA LOGGER - DIGITAL OSCILLOSCOPE i
P.O. BOX 3448
- ALLIANCE OHlO 44601 + 216 821 5800
_ - _ ___ _ ._ ____ _ . _ _ _ _ . . . _ _ . _ L
4 O '
INDUSTRY INITIATIVES TO UPGRADE i VALVE PERFORMANCE I
I
, i a I
- I ED MOORE .
I Institute of Nuclear Power Operations :
i e
October 27, 1988 ,
1 I
! O i 1
i i I
r t
I i
f f
i l
1 i
l O l
i l
4 O OVERVIEW o Review of Previous Presentation to the ACRS l o Motor-operated Valves l o check Valves o Instrument Air / Air-operated Valve Failures O
O
l March 7, 1988 Presentation To The ACRS '
()
Discussed the following: ,
i o Industry events associated with motor-operated valve and check valve failures. l o Responses by the NRC to events:
--NRC IE Bulletin 85-03
--AE0D reviews of motor-operated
! valve performance
--NUREG review of check valves ;
t
! t I l l
l
! C) l
o INP0 actions for motor-operated valve and check valve events:
() --Experience dissemination (SERs, SOERs, 0&MRs)
--Evaluations / reviews
--Interaction / coordination with EPRI, NUMARC
--Update on M0V initiative
--Check valve activities to date
--Train maintenance personnel
--Maintenance Assistance and Review Team visits
--Promulgate results
() --Follow up by Operating Experience
()
Motor-operated Valves
() INP0 is continuing efforts to upgrade M0V performance industrywide. A review of NPRDS reveals the following:
o Limitorque models SMB-000 and 00 account for nearly 60 percent of all installed MOVs and experience an approximately 37 percent higher f a i l u r t. rate than do other operators, o Torque switches, limit switches, and motors account for approximately 70 percent of all MOV f ailures.
()
)
I i
. o DC motors do not track closely with the overall MOV failure rate trend.
() This indicates failure mechanisms other than torque switches and limit switches may be the cause of some DC '
motor failures.
o Several attributes tend to cause ,
increased failure rates: l
--Number of operator cycles
--High valve differential ;
pressure t
--Boric acid corrosion on packing 1 glands causing stem binding -
f
--Foreign material blocking valve ;
seats which affect torque settings i
I I
l l
()
I _ _ _ _ _ _ . . -_-.__-. __-__-_-_-- _.-- - - _ -
Increased MOV focus during plant evaluations /
assistance visits:
4
() o Reviews of the implementation of 4
recommendations in SOERs 83-9, 84-7, j 86-1, and 86-2.
I o Reviews of the station's overall MOV i
program, o Evaluations of on site maintenance
! activities.
2 o Reviews of CFAS and NPRDS data for M0V failure history.
1 l
l I
i
o P.eviews of preventive maintenance, MOV procedures, craft training, and post-maintenance testing.
{}
o Additional evaluator with M0V experience assigned to team during 30 of the 57 most recent plant evaluations and assistance visits to strengthen M0V program reviews.
--Resulted in recommendations for improving M0V programs at 24 of the 30 sites evaluated
--Most recommendations in areas of program control, physical condition of M0Vs, documentation and design control, procedures, training, testing, and preventive maintenance
()
C)
. o Several visits made to sites with low M0V failure rates to compile
() "good elements" of strong M0V programs.
--Key elements include M0V program control, design and equipment review, training and operating experience review, testing, preventive maintenance, predictive maintenance, and spare parts o Summary of industry M0V performance provided to all plant managers and maintenance superintendents during the most recent INP0 workshops, and to participants in the recent EPRI valve symposium. Copies also provided to plant personnel during evaluations.
()
CD
. Recommendations of SOERs 83-9, 84-7, 86-1, and 86-2 l
('l u-1 i
i o Train personnel on the theory of torque and limit switch operation and the design limitations of the motor operators (such as pressure, '
temperature and differential pressure). ,
o Develop maintenance procedures addressing assembly and aisassembly of actuators, troubleshooting, and !
preventive maintenance, o Maintain and periodically review history records that include torque and limit switch settings, stroke times, and current signature.
() o Develop post maintenance testing requirements that assure the adequacy of the maintenance activities performed. These tests l should be conducted under configurations expected during operational events.
l i
i I
i I
EPRI's Nuclear Maintenance Assistance Center (NMAC) is addressing MOV issues:
o Developing technical repair guidelines for limitorque valve motor operators.
o These guidelines will include detailed recommendations on topics ,
such as preventive maintenance, testing, equipment overhauls, and common industry problems.
o Initial guideline for the SMB-000 operator is expected to be available .
this fall.
o INP0 is working closely with EPRI in the development of these repair standards.
()
4 f
4 i
i i
o A review of NPRDS shows a decrease in the rate of MOV failures t,ince
() the beginning of 1987 (see graph).
o Industry initiatives have been effective in identifying areas where emphasis is needed to improve MOV performance and reliability, o Reviews have demonstrated that long-term reliability is dependent on a comprehensive MOV program.
o Future NMAC M0V technical repair guidelines should be of assistance in assessing and improving MOV programs.
()
b
O O O~
Limitorque AC/DC Motor-Operated Faikre Rates (December 1985 through May 1988)
E 0.0100-
.=
N0.020-E a 0.0060J -------------------- ----- ----- -- -
_ _ =
EW E 0.0040-g -
=
- = 0.0020 -
= .
s 20M i i i , , , , , 2 12/85 03/ 5 M/5 09/86 12/E 03/81 06/81 09/81 12/87 03/5 Tne Period *
-w wmk w -ap%
.x e w w.a % maes= g a
I Check Valves INP0 is continuing to evaluate and provide recommendations concerning check valve performance. A review of NPROS reveals the following: ;
o Larger check valves (>4") are ;
somewhat more subject to failure. 1 o More failures discovered from 0 to 3 :
years (due to design errors, i misapplications, manufacturing ,
flaws) and 10 to 13 years of service (due to worn out components).
o Service water, main steam, feed system check valves particularly l I
prone to failure.
() o Most frequent failure is stuck open f
(
valve exhibiting gross leakage. ,
l 1
I 1
l J
l l
l
- - . _ _ . . - _ , - ~ . . - - , . _ ~ . _ . - _ . . . _ _ . - . _ - - _ . - - _ _ - - . _ . - . . - _ , -
INP0 has evaluated the progress of the industry's check valve efforts with increased
() attention during 1988, o Focus has been on recommendations of SOER 86-3.
--Establishment of a preventive maintenance program
--Performance of design reviews l (size, type, orientation) o In December 1987, INP0 recommended I that all utilities use EPRI's '
"Application Guidelines for Check Valves in Nuclear Power Plants" to address recommendations of SOER 86-3. This position has not changed.
o Evaluated check valve performance on O all plant evaluations since April 1988. ,
o Personal contact was made with all EPOC's emphasizing the importance and urgency of the program. !
i f
?
l
)
o INP0's Technical Support Department is taking the lead on check valves
() during evaluations and assistance visits. The following areas are evaluated:
--Preventive maintenance requirements
--Evaluation of check valve applications
--Problems and corrective actions
--Performance monitoring techniques
--0verall reliability o Although still early, since mid-year l 1987, NPRDS shows a decrease in '
. check valve failures industrywide .
i
(} (see graph).
l o Utilities are incorporating EPRI .
check valve guidelines and SOER 86-3 ,
recommendations into their check
- valve programs.
{
l l l
. I l
CHECK VALVE FAILURE RATES (JANUARY 1984 THROUGH APRIL 1988) 0250; E A 7 0.00402 Y /
/ A iro m - ~
v y a
d o.W n
j 0.0010-e OM i i i , i i , , ,,iiiiiii
~ 01/M 04/M 07/M 10/M 01/504/507/510/501/504/507/510/501/8704/8707/8710/8701/504/E TuPied
+fa15eRdel-inddyhg, hisaciedabeiragd*sofr=piais.
Instrument Air / Air-operated Valves
()
o Many pneumatic valve failures are directly attributed to instrument air system problems.
o Instrument air system problems have been previously addressed in the following:
--SOER 80-1, "Loss of Redundant Emergency Diesel Generator Starting Air System"
--s0ER 81-9, "Dessicant Carryover to the Instrument Air System"
--SER 57-85, "Failure of ASCO Double Solenoid Valves"
() --SER 87-83, "Loss of Instrument Air Scrams" o SOER 88-1, "Instrument Air System Failures," issued on May 18, 1988, i
0 I
()
o Most common problem with instrument air systems is contamination by foreign particles, wear products,
(} water, or hydrocarbons.
--This has resulted in air-operated valves failing to open/close, operating erratically, or operating more slowly than required.
l
--Foreign mater 9als have caused flow restrictions, and degradiation of ,
valve seats, o Some air-operated valves have been installed with failure modes opposite to that assumed in safety analyses, o Supply and exhaust lines have been :
found reversed.
o Design deficiencies associated with O accumulators and check valves have l
l caused insufficient air reservoir ;
capacity for operating components, ,
such as air-operated valves, after a loss of air.
9 I
I r
() I t
- INP0 is tackling the instrument air problem by ensuring the following key recommendations
() of SOER 88-01 are met:
o coerations:
Procedures to assist in the identification, control, and recovery from partial or total loss of air events, o Trainino:
Providing classroom and simulator training on loss of instrument air events.
()
. o Maintenance:
() Periodically monitoring air quality at several points throughout the instrument air system.
Ensuring the preventive maintenance program provides for filter and desiccant replacement, and periodic air dryer maintenance, o Desian/ Analysis:
Verifying accumulators and associated check valves perform as intended on a loss of instrument air.
Additional training and SOER 88-01 evaluation guidelines have also been provided to evaluators.
() o HPRDS shows a decrease in air-operated valve failures industrywide since 1986 (see graph).
l
()
\
O V O
PNEUMATIC VALVE FAILURE RATES (JANUARY 1984 THROUGH APRIL 1988) 0.01 5 te 0.MEJ y v v b v m
v i V0.Re-
!s J :;
! g o.M-E l
E 0.0020-a 1 4 I I I i 1 I I I I I I I I I I 01f4 M/M 01lM 10lM 01l% 04l% 01l% 10l% 01l% '
Ml% 01l% 1050181 M]61013110l5101llE i TaePedad i
1 1
- -rdweue - u *yks l tysiseddhbeinngdalesdrmepedads.
l~
i l
+-
--~,- ,.m-me,-o r,-v m-m ,, w < w w,,-*~w- e,-w---- -e ,r- m--- -o ----,---w----wg,- ---o w-,~,w w---,m,w ---~--w- ,,-,,--e w-- -,---,,,-w . - - - - - -
O
SUMMARY
o Valve Performance Is Improving o overall Industry Response Has Been l Positive 1
o Evaluations Are Focusing on Valve Maintenance, Design, and Operability Issues l
O O
- -- --- - -- -