ML20248H123
| ML20248H123 | |
| Person / Time | |
|---|---|
| Issue date: | 09/20/1989 |
| From: | Office of Nuclear Reactor Regulation |
| To: | |
| Shared Package | |
| ML20248H118 | List: |
| References | |
| GL-89-10, NUDOCS 8910110181 | |
| Download: ML20248H123 (129) | |
Text
- - _
OR GRAL
~ '
UNITED STATES L' NUCLEAR REGULATORY COMMISSION i
l: awiss = = = = = == = = = = = = = = = = = = = = = = es = = = = = = = . - = = . m = = . . . . = . . . . - . . . . .
1 I
- NORKSHOP ON GENEFIC LETTER I
I. ' f Pages: 1 through 127 Place: Denver, Colorado Date: Sectember 20, 1989 W
===================r=======================
HEIIITAGE REPORTING CORPORATION onwatamatm 1228 L Sewt, N.W., Suke 600 Washington, D.C. 20005
)"?" F' i' n'
" ' ") F 2@6 M
'*'~"'
~
o
I I NUCLEAR REGULATORY COMMISSION
[
1 2- PUBLIC MEETING l
3 4
5
.. 6 7
43 8
9 Registry Hotel 10 Denver, Colorado 11 12 13 14 15 16 17 September 20, 1989 18 19
. 20 21 g 22 23
.. - . = - .
25 I
- - _- - _ __x
I LNEEE l 2 PANEL 3 Ted Sullivan Mechanical Engineering Branch - Nuclear Regulatory 4 Commission 5 Dick Kiessel
- 6 Division of Operating Events Assessment - Nuclear Regulatory Assessment e Jeff Jacobson 8 Division of Reactor Inspection - Nuclear Regulatory Commission 9
Tom Scarbrough 10 Mechanical Engineering Branch - Nuclear Regulatory Commission 11 12 13 14 15 16 17 18 s19 20 21 22 23
. _ = _ - _
24 25
- _ _ - - _ _ - _ _ _ _ _ _ - _ _ _ _ _ _ na
g 3 l
1 23pcygRINGS l i
2 (10:00 a.m.) l 3 MR. SULLIVAN: Okay. I'd like to get 4 started. My name is Ted Sullivan. I'm a Section Chief in 5 Mechanical Engineering Branch and we're the branch that's
+ '6 responsible for the subject. This meeting is the NRC meeting 7 on Generic Letter 69-10 which has the title of Safety-Related 8 MOV Testing and. Surveillance.
9 One thing I'd like to point out logistica11y 10 before we get started that we've put sign-up sheets in the 11 back, and we're interested in having you sign those sheets 12 because we're going to be producing a meeting summary which 13 I'll talk about a little bit later; and we'd'like to know who 14 all was in attendance so we can send it directly,to you.
15 We'll of course be sending it also to all licensees. But if 16 you leave your name back there, we can send it directly to ;
17 you and you won't have to wait for it to funnel.through your-18 organization down to you.
19 I'm going to make a short presentation, just 20 a few minutes. It's an overview type presentation on McV's.
21 A longer presentation will be made by Tom Scarbrough. Tom is 22 the one at the end of the table, this end of the table, and 23 Tom is going to talk more about the Generic Letter. In 24 addition to To ,
at the table is Jeff Jacobson. Jeff is also
- 25. an NRR person. He's from the Division of Reactor Inspection
_ _ _ _ _ _ _ _ _ _ _ . _ _ _ _ _ _ . - - _ _ _ _ _ _ _ .n
4 i
1 and Safeguards. And Dick Kiessel at the far end of the !
2 table. Dick is working in the Division of Operating Events 3 Assessment. Dick and Jeff have had a long involvement in 4 this Generic Letter and what led up to it and various MOV l 5 problems, and we decided we'd fill out the panel with a 6 variety of NRR people.
7 As I think was indicated in the Federal 8 Registry Notice, the intent of this meeting.primarily is to 9 respond to questions on the Generic Letter and its 10 implementation. We had similar meetings on a previous 11 Generic Letter that Mechanical Engineering Branch issued, 12 that was Generic Letter 89-04 on In-Service Testing. We did 13 similar meetings, similar format. We thought they were very 14 successful and we thought we should repeat for this 15 particular subject.
16 We've seen a number of variety of MOV 17 problems over the years and many of these problems have 18 related to torque, torque bypass, and limit switch settings.
19 other problems have included damage from operation, wear, and 20 thermal overload of the values and the motors. The causes of 21 these problems have been varied. They've included inadequate
~
22 engineering analysis on design, lack of coordination between 23 plant organizations, inadequate or weak training programs and 24 proced ure's , and Insufficient testing criteria.
25 operating experience has shown that MOV i
i E --- _ - - n1
I L
i q
5 .
t i failures have a potential for having severe consequences. !
6 L 2 MOV failures to open have caused a complete loss of feed ,
l 3 water at Davis-Besse, and in addition, failures of MOV's to I 4 close have occurred in safety related systems at Catawba and l.
5 Millstone.
, _ 6 The Generic Letter 89-10 will help provide 7 confidence in that the MOV's currently in use will operate 8 when needed. Significant; additional effort beyondEthis 9 Generic Letter is needed to resolve all concerns, and Tom's 10 going to talk about that a little bit more in his 11 ' presentation.
12 In addition to the NRR work and research work 13 that produced Generic Letter 89-10, NRC is engaged in a wide 14 spectrum of activities on motor operated valves. And my 15 purpose here is just to give you an indication that Generic li Letter 89-10 is not the only thing going on at NRC and of the-17 area. In NRR we, of course, had the Bulletin 85-03 work, 18 Generic Letter 89-10. Some related aspects are in Generic 19 Letter 89-04 which is in the In-Service Testing area which I
. 20 mentioned previously.
21 There's also a program going on in pressure 22 isolation valves. It's sometimes at NRR referred to as the 23 ISLOCA, Inter System LOCA problem. Right now it involves 24 pilot inspections that are mainly--actually, they're not 25 pilot inspections. They're audits. It's a pilot program
_ _ _ _ _ _ _ _ _ _ _ - - _ _ _ _ - _ - _ _ _ _ fl I
i 6
l t
i where a team is going out to about six plants and it's ! l t
2 basically an information gathering exercise. And out of l ?
3 that, the agency's going to evaluate whether to do follow-up 4 audits or inspections or to work on some generic 5 communications related to Inter System LOCA. So MOV's are 6 involved in that particular program. There are other EMEB, 7 other mechanical engineering branch activities that, as I 8 mentioned.a minute'ago, Tom is going to talk about.
9 Our Office of Research is also involved in 10 motor operated valves. There's research that's been ongoing Il and is still ongoing related to Generic Issue 87, failure of 12 HPCI steam line without isolation. This work is being done 13 by Idaho National Engineering Laboratory. It's a program 14 that I think is near complete called the Valve Performance 15 Program that involves MOV's as well as other valves, and our 16 Nuclear Plant Aging Research Program also has a component on 17 MOV's.
18 Then the Office of Analysis and Evaluation of 19 Operational Data, commonly referred to as AEOD, produced a 20 case study report back in 1986 that focused a lot of interest 21 on MOV's. They continue to be involved in the routine 22 analysis of the MOV events. And they also prepared a report 23 to the commission on motor operated valves.
24 ,
But the main thing I want to indicate is not 25 to try and give you an idea of what each and everyone of
_ _ _ _ _ _ _ _ _ _ _ - - n
l;
.7 ,
t
.c l' these activities is about, but to just give:you'an' idea that, I L g 2 the various program offices in NRC are interested and 3 concerned about this problem. We're approaching it in a
~
4 multi-phase' approach.. Our. efforts are not just concentrated
~
5 on feeling that 89-10'is going to-solve all of our problems.
6 On the industry-side, EPRI has comeTout with' 7 Ltechnical repair guidelines for Limitorque SMB-000-that came
~
8 out' earlier this year. I believe other technical repair
'9 . guidelines-are in preparation. There's an application 10 guideline.that_is being developed and that's going to'come
~11 out also being supported by EPRI.
~12 INPO is involved in MOV's from the point'of 13 view of plant visits. Th'ey also, in October of '88, issued a 14 document contained key elements for comprehensive MOV 15 program. It's a short document but it's a very comprehensive 16 good outline of key elements for a comprehensive program.
17 NUMARC is intending to produce guidelines for 18 implementation of the Generic Letter that we're talking about 19- today. Various vendors, testing vendors have come out with o
. 20 testing techniques. Diagnostics techniques I guess would be 21 more accurate. Various valve vendors are involved in varying degrees in_ cooperating in the assistance of MOV problems.
22.
23 IEEE came out with a maintenance practices 24 document. ASSE has recently had OM-8 on MOV's approved by 25 the Board of Nuclear Codes and Standards. OM-10 is a
__ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ __ __ _ __ _ __ ______________________________________________L
l 8 i
'l ' document that also has recently been approved by the Board of !
1
'2 Nuclear Codes and Standards;.and that is, essentially, the 3 in-service testing requirements for valves in general which 4 also covers MOV's and QME-1 is being worked on. This is a 5 document that is involved in the qualification of Mov's.
-- 6 All of these activities are very important l
7 but I would want to comment that it appears to us that no !
8 single one of these activities is sort of. leading the charge 9 or leading the way towards a solution of the problem. It 10 appears to us that industry efforts aren't being well
- 11 coordinated and integrated, and we think that's something 12 that needs to be thought about. l 13 I think I'd just like to conclude by talking )
14 a little bit about the format for this meeting, and I'll turn 15 it over to Tom next. Tom, as I indicated, is going to make a 16 - somewhat longer presentation on the Generic Letter itself.
17 Then what we intend to do is go over the questions that were 18 sent into us by a-number of utilities and organizations--I 19 think it's about six--and we will read the questions, go 20 through answers, and give you an opportunity to ask 21 additional questions to flush out the information. When we 22 do.that, we will have to insist that you go to a microphone 23 to do that. Otherwise, it will not be picked up by the 24 reporter. ,
25 When we're done going through all of those l
_a ------- - - - - - w.---,--- - - - - - - - - - - - - - - - , - - - --- - -- - - - - - - - - - - - -
i I
9 i
i questions that we have already received as a result of the !
t 2 Federal Register Notice, we're going to open it up to f 3 questions, just more general questions from the audience that 4 .hadn't been submitted in advance. We think it's appropriate 5 to go through the questions we already have received first.
, . 6 I believe there are cards on the back table that we would ask 7 you to fill out with your questions, give us a few minutes to 8 look at that,_ caucus if we need to, think about the answer.
9 So if you could do that and give it to us at lunch or one of 10 the breaks, that would be very helpful to do it that way.
Il We expect to break for lunch in about two 12 hours, break for roughly an hour, maybe a little bit more.
13 We expect to finish, based on this meeting and the way the 14 other meetings went, we'll probably finish around 5:00. We 15 have found that usually all the questions are answered by 16 then and also by then everybody's pretty tuckered out.
17 We do plan to put together a meeting summary 18 and I think based on our anticipation that that will meet NRC 19 management review, the meeting summary probably will take
. 20 about three months to get out. It's not going to be l 21 something that you should expect to see real, real soon. But 22 I do think that the meeting transcripts that are being taken 23 are a ve. Taitl;?r.1 representation of what our thinking is. j 24 And if you nee 5~to get information back on what happened at l l
25 the meeting, don't wait for the meeting summary. Get a copy i
o k s 10 L _
g I of.the transcripts. 'They should be available in about two !
2 weeks. Thank you.
3 MR. SCARBROUGH: Good morning. I'm going to 4 talk more about the details of the Generic Letter and how it 5 came about, some of the precisions in it, and where we're
. 6 going after the Generic Letter itself. First, a few words 7 about the background. !
~
8 The main event which raised concerns'for MOV 9 operability was' basically Davis-Besse back in 1985. To 10 refresh your memory of that event, it started with a loss of 11 of main feed water, a main feed water pump tripped, and the 12 operator went to initiate auxiliary feed water. It had 13 already actuated on low steam generator level. But when he 14 initiated the steam and feed water rupture control system, he 15 pushed a low pressure signal button and that automatically 16 closed the steam generator isolation valve. He realized his 17 mistake very quickly and he went to reset and open the 18 auxiliary feed water valves automatically. They would not .
i 19 open against that difference or pressure that had been
. 20- created across the two. After about fifteen minutes or so 21' they sent someone down and manually realigned auxiliary feed 22 water, and that resolved the situation. But it did indicate 23 that we were not aware of a problem that the auxiliary feed
- =: .
24 water valve ip MOV's could possibly not open under a high 25 differential pressure. The problem that was determined from ni
ll ,
I i
i that event was that the torque bypass switch was not set for j i
2 a long enough period to allow the valve to move off its seat.
l 3 That indicated that we were unaware that the torque was much 4 higher for a longer period of time to open valves under a 5 high DP. That led to Bulletin 85-03 in 1985.
6 The Bulletin focused on certain systems; 7 high pressure injection systems, emergency feed water 8 systems,<and the reactor core isolation cooling system, RCIC.
9 The testing focused there on sign basis testing to ensure 10 that the switch settings toward bypass were set properly to 11 accommodate those high differential pressures or high flows 12 that were seen during design basis events. The licensees 13 were to review the design basis for each MOV, establish a 14 program for selecting and setting those switches, perform the 15 differential design basis testing, differential pressure 16 testing where practical, and prepare procedures to ensure 17 that the switch settings were maintained. The program was a 18 two year program basically, and we had some very interesting 19 results that came in that led to Generic Letter 89-10.
20 A supplement to 85-03 was prepared in 1988 21 which clarified two aspects of Bulletin 85-03. It clarified 22 that all safe related MOV's and those selected systems had to 23 be included in the program. There was a problem; BWR plants
_ _ = .
24 weren't aware of the full scope of 85-03.
25 The second concern that was addressed in
_ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ . _ _ _ _ _ _ _n.
+# 1 1
12
. i
'l supplement 85-03 was that it was clarified that all MOV's in 1 1
'2- the selected systems must be capable of recovering from l
3 inadvertent mispositioning. They must be able to be 1 4 repositioned electrically in the event that they were 5 inadvertently moved to an improper position. ,
., . 6 I'll talk briefly about the support for 7 expansion of Bulletin 85-03 as a result of the responses that
~
8 were received when that bulletin was implemented. First, the 9 responses revealed that many more MOV's than had been 10 expected would not operate under design basis conditions--
11 roughly eight percent, which was much higher than had been 12 assumed in Probabilistic Risk Assessments at that point. In 13 addition, other MOV deficiencies were discovered that had not 14 been made fully aware of up to that point. As a result, NRC 15 contracted with Brookhaven to prepare an analysis to 16 determine cost effectiveness of extending the scope of 17 Bulletin 85-03 to all safety related MOV's, and the result of 18 that project was new Reg CR 5140. And that included the PRA, 19 Probabilistic Risk Assessment. And it found favorable dose
. 20 reductions as a result of extending the scope, and also 21 favorable in dollar savings. Brookhaven also raised some 22 qualitative arguments for the need for expansion of 85-03.
23 At the same time, tests were under way relating to Generic 24 Issue 87 which dealt with testing the valves from the reactor 25 water clean-up system, RWCU, at Wiley performed by Idaho
__ _ _ _ _ _ _ _ _ _ _ _ _ . _ _ _ . _ _ - . _ _ _ _ n>
13 l I l 1 National Engineering Laboratory. l 2 Several very significant problems were 3 revealed as a result of that testing. One of the major 1
4 concerns was that the valve thrust equation needed to be 5 reviewed. The results of the traces from the thrust that was
- 6 required were not following the standard industry equation 7 that was used to calculate the valve thrust. It also found !
8 that subcooling was a significant parameter, which, up to 9 that point, it had not been known that the the valve disk 10 factor was dependent on the subcooling. Valve design was 11 also found to be. The clearances between the seat and the 12 guides was found to be significant which raised concerns for 13 applying data from one valve to another when you assume that 14 ,
they were identical valves. It also indicated that the stem 15 factor changed with load. It was then assumed that that was 16 more or less a constant. And that was a part of the result 17 of this Generic Issue 87 testing.
18 Phase one of that effort which was discussed 19 in a public meeting in February of this year has been
. 20 released in the EG&G report. And for your information it's 21 SSR 8547. It was released in July 1989. Also, it's my 22 understanding that the report has been placed in the Public 23 Document Room which may be more readily available to you, or l youcancontackEG6Gdirectly.
~
24 25 There's some follow-up testin that is
_ _ _ _ _ _ . - - . - a>
i i
V i- !
EU 14' )
1 1 underway at this minute in Germanycwhich will also address !
2 the concern for additional thrust requirements. And from the 3 long distance telephone calls we've had and I've heard 4 second-hand, is that the preliminary results are confirming 5- the concerns that were raised in the earlier Idaho test. But
"- 6 there will be additional information released on those as we 7 know more about them.
8 Once the decision was made'to expand Bulletin 9 85-03, one of the main differences, and'I'm sure you're 10 aware, is the scope. Bulletin 85-03 addressed twenty-five to 11 thirty motor operated valves. The Generic Letter will 12 address a hundred to a hundred and fifty. -Also, not only do 13 you need to be aware of the safety related MOV's, but if you 14 have MOV's in safety related system that are not classified 15 - as safety related and they're position-changeable, then they 16 will have to be in the program also. And we'll talk more 17 about how you go about blocking inadvertent operation later.
18 But that's a significant point that you need to be aware of.
19 This is a very brief summary of the positions l
20 in Generic Letter 89-10, and you'll notice when you read the 21 Generic Letter is it tracks very closely to Bulletin 85-03 22 for a certain distance in what is among the recommended 23 actions. First review and document design basis for MOV 24 operation. , Same as 85-03. Establish and implement program 25 to review and revise methods for selecting and setting L_________________________ _ n
15 I
i -switches. Demonstrate MOV operability by testing a design j 2 ' basis differential pressure and/or flow, and document where L '3- design basis testing cannot be performed with the description 4 of the alternative testing. The next is, prepare procedures 5 to ensure that correct switch settings are determined and
'- 6 maintained. Eighty-nine oh ten emphasizes that the ASME 7 code, Section 11, stroke-time testing, is not oriented _to 8 verification of switch settings. So additionalcmeasures 9 beyond the stroke time testing will be needed to verify the 10 switch settings, diagnostic techniques. But the stroke time 11 testing is not going to be an acceptable means of verifying 12 the switch settings. The next one, analyze and document MOV 13 failures and corrective action. That was not in 85-03. And
'14 then periodically examine MOV data every two years for 15 trends. Again, that wasn't in 85-03. This program, 89-10, 16 is a continuing program and you need to be aware of that.
17 It's not a one shot deal.
18 Another important point which isn't on the 19 slide, but 89-10 states specifically that you need to address 20 degraded voltage. We'll talk more about that later and you 21 can bring it up in questions. That needs to be considered in 22 your design review when you're considering the operability of 23 the MOV agd the parameters that it will see when it tries to 24 operate. ,
25 The schedule was spelled out in the Generic C
o g.
l 1
1
'l '
- ,n 16 il K 1 Letter itself, but I thought it would be good just to show it. j- .
i 1 2 to you, indicate this is how you can summarize it without all I
.3 the extra wording that surrounds it in the. Generic Letter. ;
4 Program description and_ schedule,' operating license plants,
.5 within one year or one refueling outage, whichever's later.
6 Construction-permit holders with one year before the OL's 7 -issued, whichever's later.
i, 8 The initial program would be accomplished 9 within five years or three refueling outages for operating 10 licensed plants and for construction permit holders,'five 11' _ years or before the OL is issued.
12 Then the continuing portion. You verify the 13 switch settings every five years or three refueling outages, 14 whichever's later. Reporting within six months of the
- 15. Generic' Letter--the Generic Letter was issued on June 28,
- 16. which six months is December 28, advise NLC.that the Generic 17 Letter will be met or any modifications with technical 18 justification for that. And then, as a continuing 19 obligation, notify the NRC of any changes to commitments that
. 20 you had made originally. Then after initial five year 21 program is completed, within 30-days send a letter to NRC
^
22 indicating that that initial program has been complete and 23 now you're in the follow-on effort of maintaining switch
- ==-
24 settings properly. .
25 There's several activities that will be
_ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ . _ _ _ _ _ . - - _ - - - n
r- _--- _ _ - _ - . _ - - - -
17' I necessary in NRC to support.the implementation of the' Generic !
2 Letter. First, the work shops. And that's why we're here '
3 today. We had one in Chicago'in which we had a number of 4 - attendees and a significant number of questions. Technical-5 guidance for the staff. he're in the process right now of
-.. 6 working-on a temporary' instruction inspection manual chapter-7 that will be provided to.the regional inspectors. ~ we'll get
-8 their feedback, come up 'with :a consensus method: f or
- 9 monitoring implementation of the Generic Letter so that we're
]
10 all in a consistent mode there.
1 11 Audits of the program description. We're 12 going to start that next year; looking at the program 13 descriptions, how the programs are set up to see if there's 14 any problems in that area or if there needs to be additional 15 guidance provided on how to go about implementing the Generic 16 Letter. Following that, once we get into the actual 17 implementation of the programs, we'll be performing audits 18 and inspections, both from headquarters personnel and from 19 the regions on program implementation. And the last bullet
. 20 there is review of industry testing, and that's where the ,
1 21 industry gets together to perform prototype testing for I 22 generic use by licensees. We'll be involved in reviewing 23 that and making sure that that meets what we consider to be
- = - - .
24 acceptable testing. This is something that we're emphasizing 25 to industry and I don't think we can say it too few times.
l 4
l
__- - - - - - - -- -- - - _ nJ
18 o
1 We really are encouraging that.the licensees get together, 1 i i !
2 either as a group themselves, as utilities, or through EPRI, 3 and perform testing which'they canLapply to MOV's throughout 4 the plants. We've already sent a letter.to EPRI through 5 -NUMARC' indicating that this is an area that EPRI performed a
. ,- 6 . tremendous service to the utilities in assisting in the 7 implementation of. Generic Letter 89-10, and we'll keep 8 encouraging.this to" allow.a more prompt;and effective 9 implementation of Generic Letter 89-10..
10 The last. slide is just indication of where 11' we're going from here. Generic Letter.89-10 will resolve a.
12 concern for, at this moment, will those MOV's operate under 13 design basis condition. We've seen in the plants--have 14 mentioned a couple of them--cases where in actual situations.
15 the MOV's would not operate under a condition which was close 16 to its design basis. So there's' things to do beyond 89-10.
17 A rule change. We're working on a rule 18 change in 10 CFR 50.55A which will address in-service testing 19 and the need to ensure that testing demonstrates the
. 20 operability and reliability of valves or pumps as the case 21 may be.
22 Improvement of in-service testing criteria.
23 The stroke time testing in Section 11 of the Code doesn't 24 tell you much about the condition of the valve and I think 25 that's understood and accepted. We're going to be working
- _ _ - _ _ _ - - _ _ _ _ _ _ _ _ _ _ _ _ - _ - _ . n
],
19 j
- i
.j-
]j I' 'with ASME to iniprove that, come up with some better criteria j l .!
2 performing at a frequency which is appropriate--the ASME Code 3 has every.three months--come up with a frequency which is 4 appropriate for testing which really tells you something l
5 about those valves.
. .6 ' Regulatory guides on in-service testing. We !
7 plan to develop regulatory guides which will address the 8 acceptability of various in-service methods. ;We have some-
~9 good.research ongoing right now, as Ted mentioned, which is 10 coming.up with some information on test methods and the 11 performance of those test methods.
12 Design qualification of MOV's. This is 13 something which the Idaho test has raised a concern in this 14 area which we were not aware of before; particularly, 15 evaluation of the valve thrust equation. We've had meetings 16 with KWU where they indicate that we may be missing a term 17 out of that equation. But that is an area that needs to be 18 looked at further and that's one of the areas where we're 19 asking EPRI to do some research on.
.. 20 Thermo overload protection. That's an area 21 which is--there are positions on both sides on how quickly 22 the overload protection device will trip or will it allow the 23 motor operator to perform its function to its death. Ke're looking at an IiEE standard that's in the final stages right
~ ~
24 25 now. It's P741, which is--thermo overload, to see if this
.__- _- __ _ _ _ _ _ _ _ - J L
u !
- i. .
20
[
L
- l
< 1 will help in resolution of this issue. l 2 But those are some of the various activities 3 that we plan to perform beyond 89-10, so this is not--89-10 4 is'not the end-all to the problem. But we feel we have a 5 good plan of attack on MOV problems in general, and this is
- 6 our first wave. So with that, I'll conclude my initial 7 presentation. Now we're going to go over some of the 8 questions that we received and all additional questions.
9 I assembled these questions in order which 10 has no particular significance, just what was on my desk.
Il The first set comes from Tennessee Valley Authority. They 12 had indicated they were going to have individuals at all 13 three meetings, and they were at Chicago. Do we have any TVA 14 people here? Very good, very good. Okay. So we'll start 15 with your questions, and the first set was some we received 16 on Septemb e 8th. What I will do is, I'll read the question 17- and then give a response to those that I'm assigned to and 18 where my colleagues have offered to answer those, I'll direct 19 those to them.
- 20 The first question was, "Is NRC requiring the 21 licensee to address MOV operation outside design basis?" And 22 the answer is yes. In the sense that position-changeable and 23 that issu of being able to recover from an inadvertent 24 mispositioning may be outside design basis for your plant.
25 It's not considered to be something that needs to be inside n,
rn1 4
h i 21 1 e
i design basis. That has been addressed through the NRC. The l 2 'NRC has decided that mispositioning is something that needs 3 to be addressed, regardless of whether or not it's'inside or 4 outside the' design basis of that plant. So, yes, there are--
5 position-changeable is an item that could be'outside design 6 basis.
7 Now in terms of events, design basis events 8 that are in.your.FSAR, line. breaks, that is:only events.that 9 are already indicated to be in design basis. We're not 10 suggesting that you go and suppose new line breaks or come up 11 with new scenarios for design basis events. They will stay 12 the same. They should'be in your FSAR now. So-there's a 13 difference. The modes of' operation, yes, because of 14 inadvertent mispositioning. But in terms of events, no.
15 only the ones that are already in your design basis.
16 The second question, Dick Kiessel is going to 17 take that one for us.
18 MR. KIESSEL: Question reads, "When a valve
( 19 has its torque switch bypassed until the last two to three 20 percent from a design basis standpoint of travel in a closed 21 direction, can the valve be considered closed if it trips out
'22 on thrust?" The general answer is no. We would consider the 23 valve to be closed when it is in the seat and it is capable L 24 of meeting whatever your design basis leakage criteria are, j 25 and simply stopping at two or three percent open is not
__---___-_x
22 i H I sufficient in general. I i
2' MR. SULLIVAN: Okay. The third question was, l
l
'3 "Does Generic Letter 89-10 supersede Bulletin 85-03?" The
- 4. second part is, "Are the commitments for' Bulletin 85-03 still 5 monitored by the NRC7" And the Generic Letter states very e
co 6 clearly that 89-10 supersedes Bulletin 85-03. We are 7 expecting'you to live up~to commitments of 85-03 in the sense ,
- i 8 that those are the most important valves. That's why-85-03 i 9' was written including those valves as a first cut. If you 10 haven't finished.85-03, even though we hope everyone has, we 11 'would like to see those done as soon as possible. If you 12 .have a problem with that, discuss it with your project 13 manager and let him know what your schedule is. But that's 14 our position on 85-03 and how it relates to 89-10. There's 15 no formal effort right now to go out and monitor 85-03. We 16 expect to pick that up in 89-10. But we want to emphasize 17 that 85-03 valves are among the most important and we want 18 those done just as soon as possible.
19 The fourth question was concerning the KWU
. 20 Alliance testing in Germany. "How will the results of this 21 testing be shared with the utilities, industry groups, and 22 NRC, and will the NRC entertain feedback from this test?"
23 This is what I just mentioned earlier. There's a phase one
- =- _
24 report from the tests that were at Wiley conducted by EG&G 25 which is out, and you can get a copy of that by contacting Ri
- n. ,
2 j
, 23 I I 1 EG&G or the Public Document' Room. Phase two probably will !
1 2 have a report, the current ongoing testing in Germany. That 3 will have a report also. We don't know when,.probably early 4 next year sometime. The current testing that's ongoing is 5 very limited in scope and it's not going to be of tremendous
+ 6 help in resolving all.your problems in 89-10. So don't wait 7' for it. There will be testing done that needs to be'donesby 8 the. industry:itself-and you need to work-with EPRI or the j 9- other utilities.and get that underway. There's no' formal. )
10 feedback plan on the test results that are going on right now 11 in' Germany. There may be a meeting similar to the one that 12 -was held February 1st of this year. A transcript of that 13 meeting's also in the PDR which you can get a copy of. But 14- there's no plans to go out for formal comments on the test 15 .results in Germany.
16 l Then the fifth question Jeff Jacobson's going 17 to handle.
18 ~ MR. JACOBSON: "Can DP testing be waived on a 19 case by case basis for actuators with no torque switch in the
- 20. control circuit? This waiver.would be based upon requiring a 21 certain margin between the calculated thrust value required 22 for operation under maximum DP conditions and the actuator 23 thrust rating. The appropriate margin for each type valve
- - =-- .
24 could be established based upon plant specific and industry 25 experience." Generally the answer to that question is no.
_ _ _ _ _ _ _ _ _ . _ _ _ _ _ _ _ .- n
m <
24 L -
l
'l .Just because the torque switch is wired out of the circuit, I o
2 .that doesn't alleviate you from doing DP testing. And the 3 reason is, basically, that there's two' values you're trying 4 to come up with when you do the DP test; and one is,-what is 5 the' thrust required from the particular valve in question.
.. 6 And the other value is, what is the' thrust available from the 7 actuator that'1've got under design basis' conditions. iIf.you
~
8 don't do the DP: test, even though the torque switch.is wired 9 out, you haven't answered either one of those two questions.
10 lt may alleviate you worrying about where the torque switch 11 is set at, but you still have to verify the required-thrust 12 and the available thrust of the actuator. We'll talk a 13 little bit more about that later and how you can do that. I 14 .take it the question then is full DP testing. If it's 15 talking about just doing moderate DP. testing, the basic 16 answer to that is that there's going to be some testing 17 required on all valves, whether it's at full DP or low DP or 18 a static pressure, depending on what. prototype testing is 19 done along with it. So there's going to be some testing with
. 20 diagnostics required for all valves. And that's the answer 21 to that.
22 MR. SCARBROUGH: That's all of the first set 23 of TVA questions. Do you need any clarification or expansion 24 on any of thod answers?
25 MR. JACOBSON: If there's any questions on
)
__ _ ___ _ _----_-_ _ - - - n.
L 25 I any specific question as we answer them you can ask it at l 2 that time; we'll make it a little clearer.
3 MR. SCARBROUGH: We'll go on to the next set 4 of 7VA questions. These were submitted--the date was 5 September 12th. The first one was, "Does NBC have a
. 6 preferred diagnostic system?" The answer is no. That's up 7 to the licensee on what systems they consider to most fully 8 meet their needs and provide assurance that they monitor all 9 the various aspects of MOV to ensure its operability.
10 The second question, Jeff is going to take Il that one.
12 MR. JACOBSON: "The Generic Letter states 13 licensees should ensure that any test conducted using 14 diagnostic techniques, along with in-situ tests conducted at 15 conditions less severe than design basis conditions, will be 16 applied appropriately to ensure design basis operability of 17 safety related MOV's. What does apply appropriately mean?
18 Can we add margin above the required thrust to open or close 19 the valve in order to compensate for testing at less severe
. 20 conditions?" Basically, this is talking about what type of 21 prototype testing are we going to allow, and we've talked a
~
22 good deal about that amongst ourselves. The answer is that 23 we would like to see the prototype testing done at a differential p[ essure that would at least envelope the
~
24 25 highest pressure you're trying to qualify the valve in the s
_ _ . _ _ _ _ _ _ _ _ _ _ . . __ _ _ b
26 1 plant for. So it's possible that if the prototype is done at l 2 full differential pressure, the valve in:the plant could be 3 done at somewhat less pressure. Basically the way we see 4 this working is you'll have two identical valves, one in the 5- plant and one in a prototype test or in another train in the
- 6 plant. It will do the full differential pressure test on the 7 prototype and also do a test at some lower differential 8 pressure. And then on the'one in the: plant that you're'not 9 going to do the full differential pressure on, you'll do it 10 at some lower value to establish similarity between it and 11 the prototype. As long as the valve is performing similarly 12 at the lower differential pressure, you can use the thrust
.13 required valve from the prototype value for the one in the -
14 plant. That's a lot of words. Is that clear? I know some 15 of these things get a little long-winded.
16 MR. SCARBROUGH: Okay. Why don't we go on 17 to the next one. "What are some examples of when 18 differential pressure or flow testing cannot practicably be 19 performed?" One example is when it causes concern for plant
. 20 safety. If you have to put the plant in an unsafe condition 21 to perform this test, inject water to the reactor coolant 22 vessel which is unsuited for the reactor coolant system, we 23 have a concern of that nature. That's where you can look at 24 some other alternatives. The licensee should make this 25 determination. We don't have a list of what practicable
__ _ _ _ _ _ _ _ _ _ _ _ - - - - - - _ _ . _ _ - . _ _ _ _ _ . - - _ _ -_ . n.
't 3 27
~
l l
1- means. If you have a concern for plant safety in doing that l i
i L2- I test in plant, you can use a prototype. You still have to'do 3 the test on the prototype, full differential pressure as Jeff 4 said. You also have to go back to that individual valve 5 that's in question and perform low load tests on it, 6- diagnostic' type testing to make sure that the valve is set up 7 the same way. So you're not eliminating a need for some type U 8 of testing.of that valve in question or extrapolating.or 9 applying results of the prototype to that valve. But'where 10 you have a concern for safety and it's not feasible to do 11 tests in the plant, you're allowed to do it elsewhere in 12 another method.
.13. MR. JACOBSON: Let me expand on that a little 14 bit. A 1ct of what we're saying here today is differentithan 15 some of the things we accepted for Bulletin 85-03. I want to 16 make it clear that we realize that some of these positions 17 are more stringent than what we accepted for 85-03, and one 18 of the reasons is we've done a lot of research and so forth 19 since 85-03. One of the things specifically that's different 20 is that I think in 85-03 what was commonly done was, you had
. 21 similar valves or identical valves in prototype or in another 22 train and you really didn't have to show similarity by 23 differential pressure or testing at a lower value on the
- ===- .
24 valve in th,e plant. We had a question in Chicago 25 specifically--well, we had a lot of discussion about it--can c ,
L l
U 1 l
28
)
t I one in the plant be done at static or no differential
.{
2 pressure, and the answer is'you're going to have to determine !
l 3 at what pressure to test the one in the plant at so you can 1 4 justify that it's performing similarly to the one that was 5 tested in the prototype. And maybe that can be done at no
. 6 differential pressure or maybe you're going to have to put 7 some pressure on there to get the friction going in the valve 8 sso you can establish that the-valve factor and things like 9 that are indeed the same as the one that was tested in the l l
10 prototype.
11 MR. SCARBROUGH: Thanks, Jeff. Just to talk 12 about the Chicago meeting, we had a number of additional 13 questions raised at Chicago that we responded to. Many of 14 them different here, some of them the same. .But to get a 15 full picture of what the Generic Letter means and some of the 16 questions which may come up later, you may find the answers 17 in the transcripts of the other meetings. So just because we 18 don't say it here, there's a good chance we may have said it 19 at the larger meetings, the Rockville and the Chicago
. 20 meetings.
21 MR. JACOBSON: Did you have a question?
~
22 UNIDENTIFIED SPEAKER: The question I have, 23 it sounds like that we're going to have--if we pressure test 24 atlowervalue$,we'regoingtohavetodosometypeof 25 prototype testing to get that higher value. Is that right?
_ s
l , 29 i
l.' MR. JACOBSON: That's correct.- l-2l UNIDENTIFIED SPEAKER: There's no 3 ,extrapol'ation up to a higher value, conservative type i
4 er.trapolation?
5 MR. JACOBSON: 'We're going to look very hard
. 6 at any extrapolation. We're not going to come out and say 7 flat out you can't' extrapolate ten pounds on a thousand pound 8 valve or something. But until we get a.better handle on this 9 extrapolation, our: position is that it can't be done.
10 There's too much data out there that shows that extrapolation 11 doesn't work for some valves.
'12 UNIDENTIFIED SPEAKER: The reason I'm asking 13 is because one of our plants is an older plant and it's 14 pretty hard to find some of the old type of valves that were 15 back when the plant was designed.
16 MR. JACOBSON: Yeah. Well, do you have 17 valves in one train that are identical that you could test 18 maybe, and use those results for the other valves? Do you 19 have so many different types of valves?
.- 20 UNIDENTIFIED SPEAKER: We have quite a few 21 different types of valves. Our hardest ones are some of the 22 Velan type valves.
23 MR. SCARBROUGH: Well, that's where we're 24 encouraging 15[ustry to work together because you may have a .l 25 valve that someone else has in their plant as a spare or as a 2__________ _.__ n
H l
! l l 30
~
i 1 replacement, and they may have already run a test on'that and l-i L 2 use it in their plant and they can share that' data with you.-
l 3 So that's.really where we're. encouraging industry to get 4 together.- NUMARC may be able1to help in that area.to collect.
l 5 data Anr with individual utilities themselves.
6 MR. JACOBSON: This Generic Letter is going 7 to be very,'very difficult to implement without an industry -
8 wide response. It's going to be very costly for awsingle 9 utility to try to come up.with all this information.
10 You have another question?
11 UNIDENTIFIED SPEAKER: Thank you. Does the 12 prototype testing have to be done at a reduced voltage to 13 verify that the valve will run whatever the minimum' voltage?
14 MR. JACOBSON: The prototype testing does not-
.15 - have to be'done at the reduced voltage; however, if it's done 16 at full voltage and the one in your plant hee to operate at 17 reduced voltage, you're going to have to do the--let me tell 18 you who it would work. If you do it at full voltage on the 19 prototype, you're going to figure out what the required 20 thrust is for that valve at a given differential pressure.
21 And then on the valve in the plant you're going to have to 22 show that your actuator at reduced voltage can put out that 23 much thrust. So you can do it--you can handle the voltage i
- = - - .
24 calculational. You'll have to figure out what the valve 25 factor is on the prototype and verify by similarity that it's
-_ _ _ _ . _ _ - _ . n
< 1 31 j
i the same in the one in the plant.. Then, you do a calculation 2 to show what your actuator can put out at the reduced
! .[
3 voltage. Now if.you don't want to do'the calculations and' 4- - you can somehow simulate all the design basis conditions in 5 the prototype and.do it at reduced voltage, that may save you
. 6 come of the calculations.for the one in the plant. I think 7- that the best way is'probably just to do it at the full E 8 voltage and use the calculations to take care of the voltage.
9 UNIDENTIFIED SPEAKER: I just wanted to 10 clarify, if we have similar type of valves and similar type 11 _of operators, like we have six or seven groups and each group 12 has six, seven valves, do we have to DP test each valve or we 13 can select one or two valves and DP test those and we don't 14 have to test the others?
15 MR. KIESSEL: .First of all, what do you mean 16 by similar?
17 UNIDENTIFIED SPEAKER: Well, similar valve 18 vendor, same--valve and similar operator--the same horsepower 19 motor.
. 20 MR. KIESSEL: I like to use the word 21 identical for the valve.
~
22 UNIDENTIFIED SPEAKER: Identical, okay.
23 MR. KIESSEL: As long as we're talking 24 identical va165b your prototypical testing can be done on one 25 and then do a less severe testing on the prototype valve and G_
32 w
I
- l. also on the valv'es that you have in-plant so that you can l-2 ; verify that'two valves are acting in the same way and that-3 you have them set up in the same manner.
4 UNIDENTIFIED SPEAKER: So in other words, 5 you're saying we still have to go and test-each. valve.
6 MR. KIESSELs Oh, yes. You're going to do 7 some testing on the in-plant valves to demonstrate that that 8 valve has the same capabilities, shall we say, or is set the 9 same way, same characteristics.
10 UNIDENTIFIED SPEAKER: The reason I ask that 11 because we took an exception for 85-03 and.NRC accepted that i
12 position.
13 MR. JACOBSON: Well, that's why I said 14 there's a lot of positions we're putting out here that are 15- more stringent than what we accepted in 85-03. And I said 16 before, you're going to have to do some testing on every 1 17 valve that's covered by this Generic Letter. Now it may not 18 be a full differential pressure, but you're going to have to 19 do something.
20 UNIDENTIFIED SPEAKER: Are you guys going to 21 provide some guidance as to how far we can extrapolate or how 21 much we can extrapolate?
23 MR. KIESSEL: Yeah. The guidance right now
- ==- .
24 is zero. N,ow , if you can--and this is where we're going to 25 need the industry data. Right now we're looking at identical
_ ______ . _ _ _ _ _ . _ _ _ _ _ _ _ _ - - _ _ _ _ -n
c
~
33 i
1- . valves and possibly calculational differences within:the l-t
! 2' operator. We've already addressed the reduced voltage 3 concern. It can be handled calculationally. Different~
4 gearing arrangements probably can be handled.calculationally.
- 5. for-the operator. But the information that we have'today
~
. 6 leaves us very, very concerned about applying the data of.a 7 Anchor-Darling'six inch. valve to an Anchor-Darling eightlinch 8 valve of the same, pressure class. And gives us even~more 9 concern of applying that Anchor-Darling six inch valve data 10 to a velan valve. And unless we can get some good heavy data-11 out of the industry as to what are these ranges that we can 12 start talking about for similarity as opposed to identical,- i 13 we just don't have the basis. All we have is a lot of 14 concerns. ,
15 There's also a mike over on that wall and I-16 'think that one's working, too, so you don't have to fight 17 your way through the mob each time.
18 UNIDENTIFIED SPEAKER: On valves of the same 19 type, I'm in the same plant with the two gentlemen that just
. 20 came up. What we did for 85-03 testing, if we had ten 21 identical valves that performed the same function, say, ten 22 HPCI valves, injection valves, we DP tested one valve, 23 highest pressure we could get. We throttle back a little bit I 24 using our otheY valves there in the plant, research valve or 25 whatever, we got a second pressure. Say we're shooting for RI_____________________ n4
r -
e n-34 i,
^
t 1; 1920. pounds. We got 1900 pounds or 1890 and we extrapolated I.
' 2; to that design basis DP that was in our design ~ calculations.. ;
3 MR. SULLIVAN: We'd like to get : started again 4- and we stopped at the point of a failure of the system.
5 Could the gentlemen who was asking the question proceed with 6 his' question? Go ahead.
7 UNIDENTIFIED SPEAKER: My question was on the 8 DP testing ~ requirements of identical v'tves'within the plant.
9 And what I was discussing was the way that we performed our 10 DP testing for the initial 65-03 Bulletin. And what I mean.
11 by identical valve operator setup is that they are the same
'12 valve as like the manufacturer sells you ten valves to be 13 used in the system; the same operator, limitorque order 14 i number are the same, same gearing, same motor. Anyway, what 15 we did is we tested one of these valves under DP situation
~
16 under the required DP for this valve. Anyway, we
'17 extrapolated the DP information to the other valves at the 18 same time.
19 MR. JACOBSON: You didn't extrapolate it.
20 UNIDENTIFIED SPEAKER: Excuse me. You're 21 right. And now in some cases, without getting a line break, 22 we're'only capable of getting like, say, 1900 pounds out of a 23 valve; 1900 pounds DP across the valve without getting a line
- == : -
24 break. So we said that's as high as we could go and we
- - 25 extrapolated our last twenty or thirty pounds that was in our
_ _ _ _ _ _ _ _ _ - - _ _ - - - _ n
t 35 i
i design basis. And when we set the torque switch accordingly, !
i 2 we set the torque switch accordingly on the similar valve for 3 similar service. And we're asking if that's an allowable 4 situation for the 89-10 Generic Letter.
5 MR. KIESSEL: That seems a very reasonable
. 6 approach for the Generic Letter. The one point though that I 7 would like to do some fancy footwork around is the 8 extrapolation on the pressure. Twenty pounds out of two 9 thousand pounds seems like a reasonable extrapolation. How 10 much more than that? I ,:now we're not in a position that we 11 can say. I can tell you definitely that if you try 12 extrapolating a thousand pounds on an 1100. pound usage, that 13 you probably would not be met too favorably. Now where in 14 between a realistic arrangement might work I think is going 15 to be strictly on a case by case. But to answer your basic 16 question, yes, it sounds like what you did would be l 17 acceptable for the Generic Letter.
18 MR. SULLIVAN: With one addition which is 19 what Jeff and Dick are going to talk about.
l 1
. 20 MR. KIESSEL: I'm sorry. The different 21 degraded voltage which you didn't mention and really the 22 bulletin did not dwell heavily on it at all.
23 UNIDENTIFIED SPEAKER: Right. Degraded 24 voltage condiE[ons~as you cited earlier could be taken care 25 of by calculations prior to the test or after the test.
1 L
u 36
^
l
'l MR. KIESSEL: After the test, yes. l 1
2 MR. JACOBSON: One point to bring up in terms 3 of this extrapolation is the amount'that you extrapolated, it
.i 4 would be within the instrument and: accuracies that you're
]
l 5 using anyway. And I would think that as long as you're .{
1 6 still within the instrument and accuracies, that that might
'7 be a reasonable approach.
8 UNIDENTIFIED; SPEAKER: Okay.
9 'MR. JACOBSON: And that's another point I 10 want to bring up, is any time you're doing any--not j u 11 extrapolation--but trying to use data from one valve to 12 another, you have to take into account the instrument and 13 accuracies that you're using. Whatever diagnostic system c 14 that is,.you have to establish what the inaccuracies are and 15 add that to the thrust requirements'on the valve in the 16 plant. And we've seen some people that weren't doing that.
17 UNIDENTIFIED SPEAKER: Okay. So just in 18 summary, then, if we have-ten identical valves, we can DP 19 test one and use that information for the other valves.
20 MR. SCARBROUGH: You will have to DP test at 21 low load the valve in question in the plant by diagnostic-- i 22 UNIDENTIFIED SPEAKER: No. I'm not talking 23 about prototype testing. I'm talking about testing in the 24 plant. ,
l 25 MR. JACOBSON: What did you do on the other
__ _ _ _ _ _ _ _ _ _ _ __ _ _ n
37-f 1 nine in'the plant? I-l 2]- '
UNIDENTIFIED SPEAKER: We just took the DP 3 information that'we utilized on that first valve and we set 4 the torque switches'accordingly for the other nine valves.
5 MR. JACOBSON: You did static tests..
. 6 UNIDENTIFIED SPEAKER: Right. We did static 7 7 testing on all valves. We did static testing and we only DP B tested one valve.
9 MR.'JACOBSON: With diagnostics you did the 10 static testing?
11 UNIDENTIFIED SPEAKER: Right. With 12 diagnostics.
13 MR. JACOBSON: Well, that what was I was 14 talking about before, is that we really haven't assured 15 ourselves that static testing or testing at no differential 16 pressure is going to be acceptable. Static tests, you don't 17 really have any of the forces working that you do if tested 18 at higher differential pressure.- So I think we'd want you to 19 show us.why you feel the static test would establish a
. 20 similarity in the performance. We understand they're 21 identical, but there may be performance differences even
~
22 between identical valves. So I think we'd be asking for you 23 to show us why you feel they're similar.
24 UNIDENTIFIED SPEAKER: Okay. We justified 25 that in our initial letter, but the way you're taking off .
4 4
__n_._______ . _ _ _ _ _ _ b
ps 38 li right now, you.want to DP. test or: flow every valve in the 1 2 . plant?.
3 'MR..JACOBSON: We're saying we want.you-to, 4 on the valves that aren't tested at-full differential u
5 pressure--you can do one at full differential pressure like g - 6 you'did. The other ones we1want you to somehow establish 7 similarities. Now whether you can do thattat static. pressure 8 or no pressure.or whether,you have to have some: pressure'on 9 -there:toik) that, you have to show us how to do.that.
~
We're 10 not. going to tell you what you have to do in that regard.
II UNIDENTIFIED SPEAKER: Okay.
12 MR. JACOBSON: What we are going to tell you-13 is doing it at no differential pressure without any 1'
L 14 . justification is not acceptable.
15 UNIDENTIFIED SPEAKER: Right. And the 16 justification that we intend is to utilize the differential 17J pressure testing that we had on the identical valve under 18 identical service conditions, and that looks plausible to 19 you.
20 MR. JACOBSON: No.
21 UNIDENTIFIED SPEAKER: Okay.
22 MR. JACOBSON: You have to show on the valve !
H 23 that was tested at no differential pressure, why you feel
- == .
24 that valve was going to perform the same as the prototype at 25 design differential pressure. What you're telling me, you
{
s
_ _ _ _ _ _ _ _ . _ _ _ _ . _ . _ _ _ _ . _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ f__.__._ _ _ _ . . _ _ _ _ . _ _ . _ O
39 i
-1 think it's going to perform the same way because it's !
1 2 identical. But what we're saying is just because it's f
3 identical, doesn't necessarily mean that all the tolerances 4 or whatever else comes into play is not e factor. So if you 5 do a test that at least a lower differential pressure, you're
. 6 going to show that the forces at the lower differential 7 pressure, at least, are the same as the one in the prototype.
8 Now you can't be perfect unless you do both of them at full 9 differential pressure and we realize that, that there's some 10 point where you have to draw the line. But I think doing it 11 at zero differential pressure, none of the forces that are 12 involved in the thrust equation, other than the packing load, 13 comes into play. And that's not the force that we're worried 14 about.
15 UNIDENTIFIED SPEAKER: Well, the static 16 conditions are there and then what we looked at is you're 17 adding on the DP component. We did a static test on the 18 valve. You have that signature looking at it. And then you 19 do the DP test. You can see the difference in thrust
. 20 required.
21 MR. JACOBSON: Yeah, but we don't know that 22 the one in the plant, the difference is going to be the same 23 as the one that was prototype tested.
~
24 UNIDENTIFIED SPEAKER: Okay. That's all for 25 my questions. Thank you.
l
40 i
1 MR. SCARBROUGH: Thank you. To sunnarize one j 6
2 aspect--and at least we've hit around it may but maybe we l 3 haven't said it up front--if you take an MOV that you're 4 going to use to apply the results to another MOV or other 5 MOV's in the plant, that MOV--either you call it a prototype
. 6 or you call it in plant testing MOV, whatever you call it--
7 you do a full DP test on that MOV. Then you also perform a 8 low load type of DP test on that MOV also. So you have two 9 tests, two DP tests, full DP on low load type of DP. You use 10 diagnostics and know how a valve is set up and how it's 11 operating, the performance characteristics of that MOV. Then 12 you take identical--you have an identical MOV, identical 13 valve, particularly, actuator--you may be able to show that 14 they're closely similar that you can apply those results.
15 The valves in the plant that is in question, you perform a 16 low load or static, depending on your own justification for 17 ensuring the performance characteristic of that valve to make I 18 sure that it's set up properly to operate, set up the same as 19 your prototype in-plant testing, whichever one you're calling
. 20 that. Okay, so you have some diagnostic tests performed on 21 the MOV in question. So you're always looking at at least 22 three tests for two valves. You always have to test the MOV l
23 that's in question at some sort of low DP with diagnostics to make sure that It's set up to operate the same way as the one
~
24 25 that you're using as your prototype.
L__ _ n
L j
4 73 :41
~
l 1 MR. JACOBSON: That's our initial position l'
})
2 until we getimore information on this subject. Now a l 3 gentleman came up to me during the break and asked about, 4 well,lwhat if we test four identical-prototype valves at full I 5 differential-pressure. Now you're starting to get a little
. 6 more-data and if you' test four valves at full DP and at lcwer 7 DP and 'they were all performing the same:way, then maybe that 8 would be a . little Lbetter argument for :doing ithe .one rin "the 9 -plant:at static pressure. I'm not saying that's acceptable, 10 but certainly that's more acceptable than just doing one 11 prototype.. Now if the industry does a hundred valves at the.
12- same type at full differential pressure, then that's more 13 data we'have to go on. But what we're saying is if you're 14 -only going to test one at full differential' pressure, then we 15 really don't have enough there to establish similarities at 16 zero differential pressure in the plant.
17 MR. SULL1 VAN: None of this is to say that-18 we're not interested in full DP testing on a pair of valves.
19 If you could do it on one and you could do it on the other,
. 20 possibly I think we would probably want to say the best route 21 is to do full DP on both. But all of this discussion is 22 premised on an approach that is other than that approach.
23 UNIDENTIFIED SPEAKER: l'm hearing you say thatyou'rerA[u~iringdiagnostictestingonthevalvesandI 24 25 would think that if a full DP test is available, that it
_ _ _ _ ----_----_-----_---_ -- n
LO 42 R .
j
-1 would be acceptable to demonstrate that the valve can stroke i 2 under full DP without diagnostic.- ! i 3 MR.'JACOBSON: Only if you do it at degraded 4 voltage as well. If you don't do it at degraded voltage, 5' then you don't--without diagnostics--you don't know what the
<3.- 6 required thrust is or what the available thrust is and you 7 can't tell what you're actuator would put out at degraded 8 voltage. If you can do the in-plant test.at degraded voltage 9 at full:DP and meet all the other requirements--
10 UNIDENTIFIED SPEAKER: Well, I'm not saying 11 that I could do that. But what I am saying is,that I can do 12 a full DP test'maybe on-a valve and measure certain 13 electrical characteristics and then ratio those 14 characteristics back into finding the operator requirement.
15 MR. KIESSEL: --had an instrument at the 16 valve, how do you know that you've got enough margin left to 17 deliver the requirements?
18 MR. JACOBSON: How do you'know you've got 19 margin enough to cover the degraded voltage? How do you know
. 20 what your margin is?
21 UNIDENTIFIED SPEAKER: I'm not sure but the
~
22 correct assumption is that you are requiring diagnostic 23 testing.
24 MR. JACOBSON: We're not requiring it. We're 25 requiring you to validate certain things, and the only way I l
- _ _ - _ - - - - _ _ - _ - - - - - - - a
E q.'
143;
~
.I
- 1 know to do that right now is with diagnostic testing. You~
l-1 2 even mentioned-looking'at electrical characteristics. Well, 13 tluit's a: form of diagnostic. testing. There's.various types 4 ofLdiagnostic testing.
5' UNIDENTIFIED SPEAKER: Understood. But
-c 6 currently-in our environment diagnosticfis kind.of a.
- 7. euphemism for movats type setup or some other vendor, p 8 MR. JACOBSON: Well, there's people that:have 9 setups that are solely based on looking at. current. There's 10' all' kinds of th'ings out there and we haven't attempted to 11- -validate them all. But any system that can-be proven'to 12- validate the different things 1that we want to see is-13- acceptable.
14' UNIDENTIFIED SPEAKER: Have_you validated l 15 .any?
16 MR. KIESSEL: No. It's not our intention.
17 We're not going to produce a Q list of acceptable diagnostic 18 testing. That's not our job. That's the licensee's job.
19 MR. JACOBSON: Although when we find out y
. 20 about problems in a specific one we may act on them, during 21 the course of our inspections or whatever. However, it's not 22 our intention to validate the diagnostics.
23 MR. SULLIVANs I think we'd like to get on to
- =- -
- 24. the next set of questions from NUMARC, and I did have a 25 request from the audience that if the speaker could identify
_ _ _ _ _ _ _ _ _ _ _ _ _ - - - - _ - - _ _ - - -- -. -n!
I* .
)
i l
I L 44 L
- {,
1 .themselves:anditheir company--I think most people are doing l l-L > '2 :it. Just'to remind:you.
3l Okay. 1NUMARC submitted a letter to us dated k 1 i
Octob'er 14th,~and I.' won't read the whole. letter to you but-I.-
'4
]
5 .
will just read the high points here.. They're assumptions.
l 6 .They make a couple. assumptions that we'd like to address.
- 7 First'one wasLregarding position-changeable valves.. "It is o .,
'EL our interpretation that any position-changeable valve in a
'9. safety related system'should be included in the.MOV program 10 to ensure that"it will operate under the. service conditions 11 for which it was intended. 'Mispositioning of position-12- changeabic valves need not be considered unless 13- "mispositioning is part of the overall plant design basis."
14 And we addressed this earlier. Mispositioning may be outside 15- the. design basis plant, but it was a purposed intention to i.
16 include mispositioning as one of the aspects that must be
~
~
17 addressed in the 89-10 program. EIt's gone through all of the 18 NRC review process. It may be considered backfit,.but it is 19 one of the requirements or provisions of Generic Letter 89-
. 20- 10.
21 The second question or interpretation here !
22 was regarding degraded voltage. "It's our interpretation 23 that the effects of degraded voltage on MOV operations need
- =: . :
24 only be con,sidered if degraded voltage is part of the overall 25 plant design basis." And degraded voltage should have been
_ - _ _ _ _ _ _ - - _ _ _ n
'l s
- 45 i
= .]
i j l.
addressed in the design basis' consideration. Degraded }"
' 2- voltage 1 1n my definition is the-lowest' voltage'that the.
E . . j
- 3. plant would see without being: required to. shut down, then- j
,C 4 -subtracting off the cabling losses. So.the net result;is' E 5- what will the MOV see when it.tries to operate. And that is z.- 6 :a _ portion of 'J9-10 that may be beyond--it is beyond 85 j-7 possibly in your case. But it is'one of the' factors that has
, 8- been considered and'needs to be addressed;in 89-10..
'9 MR.-JACOBSON: Let me add a little bit on 10 this degraded voltage because it's a confusing issue for many L 11 people.. Basically, what we're looking for is.you're going to 12 -do your= design basis review and you're going to figure that 13 during certain accidents or certain scenarios a particular 14 ~ MOV has to perform. What you need to do is look at.what'is 15 the voltage that MOV-is going to see at the time it has to 16 perform. Now if the lowest voltage in the plant is 80% but 17 that's at a time that this MOV.isn't needed, then you may not 18' .have to worry about sizing :it for 80%. But you should know 19 what the voltages are going to be at the MOV at the time that
. 20 it's needed. And that takes into account cabling losses, 21 whatever.
22 MR. SCARBROUGH: The next one was regarding 23 design basis reviews. "It is our interpretation that a licenseemayh[EformadesignbasisreconstitutionofaMOV
~
24 25 application based on actual design and operational parameters O
1 I
l 46 i
l of the system in which the MOV is located rather than !
2 reconstructing the original procurement documents. In many 3 cases the valve vendor may no longer exist or procurement' 4 documents and specifications have not been retained by the AE 5 or other procurement entities, thus reconstructing the
. 6 original design and procurement information is not possible."
7 And this is a good point. We would prefer that you 8 reconstruct the condition that that valve will see. There's 9 no reason or need to go back to the original procurement 10 documents. They may not be based on current conditions in 11 the plant. You may not be able to track them down. And even 12 if you found them you still would have to validate them for 13 the conditions today. So we would prefer that you look at 14 the valve now and see what condition will be seen by that 15 valve and what conditions it must operate under.
16 The last one here in the NUMARC questions, 17 " Referring to alternatives to design basis testing in Item F, 18 we interpret that what is meant by ' appropriate design basis 19 test results on other MOV's' concludes test results could 20 come from tests performed by other utilities or 21 organizations; for example, EPRI, and not just from one's own 22 facility." And we've mentioned this, but we're encouraging 23 this effort, this joint effort by EPRI or the utilities
- - ==:
24 themselves in coming up with a data base, data they can apply 25 to their MOV's. You can use the test results from EPRI or
_ ---___ _____fL
'47
~
i
'l other' facilities as-long.as they're appropriate'and justified- l 2 -for. application to your. plant.. We're' encouraging that.
'3- We're not just' allowing that, we're encouraging it.
~
4' MR.'SULLlVAN: And, in fact, as Jeff. point 5 out earlier, we really don't see how this program can be
'. 6 accomplished without some level'ofLsharing, hopefully a broad-7' based level of sharing.among the; utilities.
.8- MR. JACOBSON: One point I'.dilikettolbring 9 out in-regards to the sharing of' data is that.if you're going
- 10. to use data from someone other than your plant, you have to ll .make sure that it's validated Appendix B data. Can't use 12 data from a-coal plant next door that a guy happened to be 13 'taking down:and'doing his' testing of1his valves unless it'was
'14- done in the proper manner.
'15 MR. SCARBROUGH: Okay. That'was it for 16 NUMARC.
17 Clive calloway, do you have any questions on 18 ;what we've said so far?
19 MR. CALLOWAY: Yes, this is Clive Calloway 20- with NUMARC. We are writing this guideline document and one i
21 area that we differ'is going to be in terms of scope, and it 22 looks like your feelings about the letter are not the same as 23 the authors.
.- ==.- ,
24 ,
MR. SULLIVAN: How so?
l L 25 MR. CALLOWAY: Well, in one thing, dealing
- _ _ _ _ _ _ - _ - - - --- n
pW
'~
'l
~}: i r.,
p . ,
48 p - i -. q '
1 with the position-changeable valves if it is outside of one's l 2 design-basis, and althoughlwe appreciate your concerns.for y 3 inadvertent operation'in this positioning and everything, the 4 scope of the lette'r is bounded and defined by;one licensing;
- 5. basis.
6- MR. KIESSEL: Clive,-we've been.down this 7' road.
8 'MR. CALLOWAY: Right. LI know.
9 MR.1 KIESSEL: I wrote the words..
10 MR. CALLOWAY: Right.
11 MR. KIESSEL : .' I wrote the words. I know what 12 I' meant by the words.
13 MR. CALLOWAY: Well, we know what you meant.
14 MR. KIESSEL: And I also know what CRGR- i 15 approved in those words and I also know what we told ACRS.
16 'And you're n'ot going to talk us out of position-changeable, 17- period. Now, next point.
18 MR. CALLOWAY: The next one I'd like to talk 19 you out of is dampers. Okay. Now, Dick, you and I have
. '20 been--we talked about this a lot during the meetings with 21 Mychleson and-Rothberg and Baer, and we concluded that 22 dampers were not in the scope of the Generic Letter.
23 MR. KIESSEL: Unfortunately that's another 24 wrong conclusion.
25 MR. CALLOWAY: I have the transcripts.
- - - ____________________________L
L i O 49 l 1 a
. I 1 MR. SULLIVAN: 'In some of the early l
' i 2 discussions, Clive, I imagine at the ACRS meetings, that is 1 1 L 3 what was talked about. But as this letter got closer to j 4 being issued and went through,.you know, the full scale NRC 5 management review, the dampers was specifically put back in.
.- 6 Maybe I shouldn't use the word back. But it was added to the 7 scope. It was a conscious decision. And early-discussions
~
I 8 that take place at ACRS do not have regulatory standing.
9 What has regulatory standing is what the letter says and what-10 constitutes the discussions and the meaning of the letter at 11 the CRGR meetings. In other words, you know, the final 12 management intent of the letter.
13 bul. JACOBSON: You brought up a point that we 14 hadn't addressed up until now and just for everybody's 15 benefit, that was from a previous question in Chicago and it 16 dealt with our air dampers that-have valves and actuators in 17 the scope of this Generic Letter and our answer was yes. As 18 long as there is a damper and it has a motor operator on 19 that, that that is also within the scope of this Generic
. 20 Letter.
21 MR. KIESSEL: Next question? -
22 MR. CALLOWAY: That's it for now, Dick.
23 MR. KIESSEL: All right. Sorry we were so rough on you. ~butlet'sfaceit; this is the second time
' ~
24 25 we've had a shot at you, so we know where you were going.
_ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ - - - - -. n -
. ; i .,
A 1 50' i-l' MR..ROSS: ~ Excuse me. I'm Bill-Ross from TU l
- 2
- Electric. Is there any guidance available from anybody in 3 the industry or the world on testing dampers?. Diagnostics.
4: can't be hooked up to them. They're a whole different 5 animal.
M 6 MR. KIESSEL: Well, let's put it this way.
7 When it was decided to put dampers back in, the concern was 8 .that we were using a piece of equipment--no--excuse:me. We 9 were relying on a piece of equipment to perform a. safety 10 function. And if the industry can't demonstrate that it's 11 suitable-for the proper use, why are they using it? 'Now that 12 was what was asked to us and why we were told to put those 13 back in. I quite honestly can't tell you that.Movats or 14 somebody else has a technique that currently can do what you 15 need it to do. But I have a feeling that the industry 16 working together can come up with such method in a timely 17 manner.
18 MR. ROSS: When you consider a damper close 19 enough in family to a valve as opposed to an SOV? I don't 1
20 mean to be short but it's quite shock. There's a deck of a 21 lot of work to be done.
22 MR. KIESSEL: Yes, I agree.
- 23 MR. ROSS: I can't say that five years is 24 going to get us to the point where we can test them.
25 MR. KIESSEL: But sc5aewhere along the line
_ __ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ . ___ n
r--" !
51 I somebody has got to have come up with some basis for putting i i l
6 2 it into their plant. l 3 MR. ROSS: But they do for other pieces of 4 equipment also.
5 MR. JACOBSON: I'm sorry?
- 6 MR. ROSS: But we do that with other pieces 7 of equipment also. The background that we're talking about 8 for the Generic Letter has got nothing to do with dampers. I 9 mean if you're talking about an air system that uses a 10 butterfly, that's to me an open for discussion situation.
11 But a damper?
12 MR. KIESSEL: The Generic Letter was written 13 for safety related MOV's. If it's in a safety related 14 system and it has a motor operator on it, it's covered.
15 MR. JACOBSON: I think it may be a' lot l 16 easier--I mean, we're all saying there's nothing that is 17 currently available to do this. But what we really need to 18 do is look at what are the factors involved and what are the 19 things we're concerned about with these particular types of
. 20 valves, dampers, whatever. And somehow do the same type of 21 thing that we're doing for the other valves in the plant.
22 They may not be the same thing that you have to do. It may 23 be a lot less. It may be you don't have to worry about a
^
24 valve factor. ~ Sou may not have to worry about a stem factor.
25 But you certainly have to assure yourself that the actuator
- e. >
RJ
c-m l
5 1
52 1 is set up to put out.'enough torque cr thrust that semelcody's I i I
2 calculated this air damper.is going to need. Right?
3 MR. ROSS: Yeah. A-damper to me is just as 4 equivalent to'a motor operated valve as an ADV is, It's just 5 not the same animal.
i 6 '
MR. JACOBSON: We realize it's not the pame.
7 And there's other things in the plant that may also need the 8 'same type of treatment, but we haven't got there.yet.
9 MR. ROSS: Why does the motor operated valve 10 generic letter throw in something that is not related?
Il MR. JACOBSON: We can't tell you why we chose 12 this'and not other things, but our management has directed us 13 that this should be included and that's the way it is. We 14 can't tell you why. The reason why is because it's a safety 15 related component and it may be subjected to some of the same 16 phenomena that these motor operated valves are. It's got an 17 actuator on there apparently that's the same, if there are 18 problems that we need to look at with the actuator that would 19 be apparent to this as well. We realize there's other 20 equipment out there that has problems, too, and we haven't 21 addressed that yet. But I guess what we recommend is that 22 the industry get together and at least look at this problem 23 and try tg quantify some of the types of things that are 24 involved. ,
25 MR. SCARBROUGH: That was it for the NUMARC
___-_-_____--________i
l 53 I
, 1 1 questions. The next set we have--well, we have eight-
. l !
1' 2 questions here from Consolidated Edison. We could break now l
3 for lunch and come back and start them or--we'll do a few 4- more.
l- 5- MR. SULLIVAN: We're probably not going to l . 6 finish this set but we kind of made some informal 7 arrangernents with the hotel. No guarantees, of course, but l -
8 we told them'that they ought to take a look at how1many 9 people are in here and make their own estimate of food and so 10 forth, but that we would plan to break about 12:00. We're 1
11 going to proceed for about fifteen more minutes.
12 MR. SCARBROUGH: Okay. Is Consolidated l
13 Edison here? They had indicated that they would have someone 14 here. That was an easy set to go over. Well, we'll do them 15 anyway. They said they would be here.
16 The first question was, "What method other 17 than testing at full design basis difference or pressure, if 18 any, are acceptable to approve operability of a given safety 19 related MOV7" I think we talked about that quite a bit.
. 20 About prototype and identical and making sure you had the 21 same valve and that different actuators you may be able to 22 justify if you have the data to back it up. So I won't dwell 23 on that one. But if you have any questions on it, bring it 24 upagainbefoE5we~quittodayandwe'lltalkaboutitsome 25 more.
I n
L .
i 54 I
l 1 Second question is, "Does NRC see any benefit l l I 2 to diagnostic MOV testing?" And I believe that's very I 3 obvious from our discussions here today and it's absolutely--
4 it's not only benefit, but it's almost essential to do some l
5 diagnostic testing on the MOV's. You use them to relate the 6 MOV that you're testing to the one that's in question if 7 you're not going to test the MOV's in question all the way up 1 -
8 to full differential pressure. Diagnostics is used in the 9 periodic verification of switch settings. And once you get 10 into that phase, where you're verifying switch settings, if 11 you don't want to do the full differential pressure test 12 every five years, diagnostics is a means of verifying those 13 switch settings in the continuing program. So there's 14 another area for diagnostics, it's almost essential.
15 The third question Dick Kiessel is going to 16 pick up.
17 MR. ROSS: If through the use cf type testing 18 or a specific identical valve we do a full DP test and a 19 reduced DP test, and we use the reduced DP test to acquire
- 20 that data to another valve, when we get to the periodic, our >
21 baseline has got to be reduced DP which means that every five 22 years I'm going to be doing DP tests on all of my related 23 valves and dampers. Is that correct?
24 MR. KIESSEL: You're going to have to come 25 back to the conditions under which you used--
I
_ _ _ _ _ . _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ n
R
(~ r I
L L 55 I MR. ROSS:- And at this time the commission is 2 not willing to accept a static test.
3 aug. JACOBSON: Well, wait a minute. You 4 could use--for the first test. you could do the test at low DP 1
f 55 ,and at no DP potentially, and then in the future tests it may
- 6 .be acceptable to do it in static pressure. And you'd have a.
7 baseline to compare it to.
r.-
8 'MR..ROSS: Okay. But if you're' concerned Y 9' about the full DP versus a low DP, if you're not duplicating 10 characteristics, those. characteristics won't show up, as you 11 said..except the vacuum load.
12 MR. JACOBSON: But we've already bought off on 13 them;on'the first verification e.nd I guess we can assume with 14 some certainty that.they he.ven't changed. Like I said 15 before, we can't be 100% on any of.these things unless we do 16 full DP tests on every valve. But we know that's not 1
17 possible, so we have to come to some median ground. And I 18 think where that ground is right now is that it's at a low DP 19 as opposed to'a no DP.
. 20 MR. ROSS: Periodically--a low DP and a 21 static DP, then static testing by technologies available 22 today should be able to monitor the parameters to ensure 23 control switch set points.
- == .
24 MR. KIESSEL: Say it again, Bill?
25 MR. JACOBSON: You're asking in future tests O
7_ ,
$ (l i
it' l '
55
. .b J. i
'jl
. if you could do them in no DP7 l 2- MR'. ROSS: Yeah.
3 MR. SCARBROUGH: As long as you have 4 confidence, if you're switch settings haven't been changed, 5 you can set up whatever program that you feel comfortable-6 with.
7 MR. JACOBSON: I think initially if youtdo it-i-
L 8 at low and zero and then in the periodic you do_them at zero, 9 I think that will be acceptable.
10 MR. KIESSEL: Bill, I think a classic example ~
-11 might be you've gone in and you've done some major packing --
12 if you have a static diagnostic test available as a baseline 13 and after your maintenance work run a static test again, you-14- got a pretty good indication what's happened to your packing-15 -and therefore you can make an engineering evaluation of 16- whether or not you have destroyed any margin you were 17 counting on.
18 MR. JACOBSON: Now if you go in there and~
19 rework the whole valve, you're probably going to have-to do E 20 at least a partial differential pressure test anyway. It's 21 not going to be acceptable to go and do a zero test after you 22 reworked the valve. If you just change the packing, the 23 packing's going to show up at zero so you can quantify that.
~
24 MR. ROSS: Okay. But the concerns that are 25 presented by the results of the preliminary INEL test and the
~
o
L
~. ,
l 57
. l
.l~ other activities and tests within the industry concerning l
~
2 valve factors and stem factors, that won't show up.in static
-3 test because there appears to be a phenomenon that says that 4' it's different under DP, so you don't know if it' changed -
5 -under DP. If you got to do a baseline of the DP, don't you 1 6~ have to revisit the baseline?
7 MR. JACOBSON: Well, I guess the assumption i 8 is that once it's verified, it shouldn't change unless 9 there's been some major. rework done on the valve.
10 MR. ROSS: Is there any basis for that?
11 MR. JACOBSON: No, other than that's our 12 opinion I guess. I don't have any basis for any other 13 position.
14 MR. SCARBROUGH: If you don't feel 15 comfortable with doing a baseline,-you're certainly welcome 16 to do a full DP test every five years. We won't have any 17 problem with that.
18 MR. ROSS: I don't but the Texas PUC does 19 because I'm going to spend a heck of a lot of money doing 20 that.
21 MR. SCARBROUGH: The current position right 22 now is that if you do initial full DP test and establish a 23 baseline at some' low DP, if you have it every five years do a
- === .
24 low DP type,of diagnostics that you feel comfortable with, 25 that's acceptable. Now as the results of the current
_ _ _ _ _ _ . _ _ _ _ . _ _ . _ _ . _ _ _ _ _ _ _ _ . _ _ . _ . _ _ _ _ . .. n .
l lj 58 t
1 testings going on, you know, it's always possible that we may l i
I 2 find out something or an accident may occur where that's l 3 shown to'be invalid. But right now we think that that is 4 enough to provide the confidence that they were offering.
5 Now that can change, but that's what we're allowing right
. 6 now. Now if an individual plant thinks that the continuing 7 baseline at low DP is not enough, go up to a higher DP. But 8 this is what we're doing right now.
9 MR. JACOBSON: I think when it comes time to 10 doing these periodic tests we'll be a lot smarter than we are Il today. We're talking five years into the future.
12 MR. SCARBROUGH: At least.
13 MR. JACOBSON: And we'll have a lot more data 14 to look at. And it's really kind of pointless to talk about 15 what we're going to accept seven years from now when we have 16 ! just a wealth more data to look at. It may be that zero will ;
17 be sufficient. It may be we'll have to change our position 18 and require full DP tests periodically. We don't know. But 19 right now that's our best guess.
20 MR. ROSS: So if I was comfortable as the 21 utility and doing static tests the first time, then the
~
22 static tests would get me along. I'd have to live with what 23 the audit finding was when it got to me.
' ~~
24 MR.- JACOBSON: That's right. We want to see 25 justification for whatever position you feel comfortable l
_______.___b
59
~
l I with, i i
2 MR. SCARBROUGH: The third question Dick j 3 Kiessel was going to take.
4 MR. KIESSEL: "What methods for calculating 5 required thrust for a given valve application are acceptable 6 to the NRC? There seems to be too many variations to 7 determine thrust industry-wide." The only method that we 8 will accept right now is one that is based on type testing of 9 that particular valve at the design basis conditions of 10 pressure temperature. The problem is that we have lost our 11 faith in the historic industry equation for determining 12 required thrust simply because we don't believe that what is 13 commonly referred to as the valve factor is appropriate for 14 many of the conditions under which these valves are used.
15 MR. SCARBROUGH: Okay. The next one Jeff is 16 going to take.
17 MR. JACOBSON: Let me add a little bit to the 18 last part; and it's not that there may be a lot of valves out 19 there where the old equation is conservative, but we don't
. 20 know which ones those are. And that's why we say we're not 21 comfortable with that equation.
22 The fourth question. "In cases where 23 differential pressure tests cannot be performed on a given 24 valve what }jink of documentation will be required? Will 25 50.59 safety evaluations be required?" I think we talked a
_- - __ n
60 I lot about what you have to do if you can't do differential L 2 pressure testing, and that's pretty clear. How'you're going 3 to_ document that is we would expect'it to be well documented 4- and any assumptions or whatever verified and so forth and'so
- 5. on. The 50.59 part of it, I.think that's going to. depend
. 6 .upon your individual plant procedures. As long as you can 7 assure yourself.that you haven't made a design change that 8 would effect the plant cause an unreviewed safety question.
9' I' don't believe;a'50.59 is required. Most plants do a 50.59 10- to verify that they really didn't need to do one, to verify 11 - that there wasn't an unresolved' safety question as a result 12 of a modification that was done. ~ And every plant is set up 13 differently on how they do 50.59's.
14 MR. SULLIVAN: If there are any questions 15 we'll take them and then go to lunch. If there aren't-we'll 16 go to lunch. Any questions on the floor on what we've done?
17 (No response.)
18 okay. Let's get back together at about ten 19 after one.
. 20 MR. JACOBSON: If you have any other 21 questions that you're thinking of, we'd appreciate it if you 22 could write them down so we could look at them and come to 23 some consensus position before we address them, rather than 24 take them cold right from the floor.
25 L
- l. .- _ _ _ _ _ - - _ _ _ - - - - n
61 I
I AFTERNOON SESSION l 2 (1:10 p.m.)
L 3 MR. SULLIVAN: I believe where we left off 4 was on the- question number five.
5 MR. SCARBROUGH: We have added a few more l t
.. 6 copies.of the handouts'and they're on the back table if 7 anyone missed them. And if we still don't have enough, let 8 us know; we'll get a few more of these things.
9 MR. KIESSEL: The fifth question from 10 Consolidated Edison is is, "Should other system components be 11 challenged in aligning systems-for a different pressure test?
12 What level of systems stress is acceptable to achieve 13 differential pressure testing?" Rather than trying to define 14 a level, let me tell you what we mean or what we don't want 15 you to do. We do not want you to place your plant in an 16 unsafe condition, period. We do not want you to violate 17 your--there are no magic numbers beyond that. Does that 18 answer the question? Are there any questions?
19- (No response.)
20 MR. SHARP: John Sharp, Florida Power and 21 Light Company. You say that you don't want to put our plants 22 in an unsafe condition. What about potential 3y damaging 23 other equipment?
'" #~
24 MR'. SULLIVAN: That'd be an unsafe condition, 25' wouldn't it?
I
t: i L;
62 e
i
'l MR. SHARP: Well, in most cases, yes, but not I i
2 always. I 3 MR. SULLIVAN: We're saying we. allow ,
4 alternatives.
5 MR. KIESSEL: Continto with your example. .
6 MR. SHARP: Well, during the testing for 7 185.03 for safety injection we dead-headed some of our safety 8 injection pumps.to achieve the DP that we tested under, and 9 that dia damage one of.those pumps when we dead-headed the 10 pump. Now I wouldn't say that that'put us in an unsafe 11 condition at the time. All we had to do was fix the pump.
12 We were in shut down and we weren't relying on that system.
13 It didn't put us into any kind of LCO. But we'did damage the 14 pump.
15 MR. SULLIVAN: What was the condition of the 16 pump when you started?
17 MR. SHARP: I'm not really sure.
18 MR. SULLIVAN: I think that you want to be 19 careful in saying that that was the cause of pump damage. It 20 may have just--the pump may have accumulated a certain amount 21 of wear from low flow testing or something and this may have I
22 just have pushed it over the edge.
23 MR. JACOBSON: The bottom line is we're
- ~==- .
24 allowing you to use prototype testing or testing of valves in 25 another train. That's the bottom line. There's no condition l-t
.. - - . _ _ _ _ - - _ _ - _ _ _ _ - . _ _ - n
_=_- _ _ _ - _ _ _ _ _ _ _ - - _ - _ _ _ _ - _
63 I
.I that we're saying you have to perform a full differential l L I L 2 pressure test-on that valve. Now when you look at all the
~
3 things involved, if you feel that it's impractical to do it, 4 then there's alternatives that we're allowing. So we can't 5 tell you when would be a good time to make that decision. If
.. 6 you feel you're going to damage your equipment, certainly you 7 should consider the alternative in that case.
~
8 MR. KIESSEL: Such as if your testing cycles 9 were to be along and;there was.a concern about dead-head for 10 a prolonged period-of time, I.think that would be a very-11 valid reason for trying to work out an alternative. And if 12 the 85-03 testing was in fact the total cause--I'm sorry you 13 didn't ask me sooner.
14 MR. SDLL1 VAN: Why don't we go on to the next 15 one. We hit this one before. Number six, "After initial 16 differential pressure testing or alternate testing, how often 17 should testing be repeated?" The Generic Letter does say to 18 verify your switch settings every five years, and that's in 19 Items D and J of the Generic Letter.
20 MR. KIESSEL: And you can use diagnostics as 2I one means of verifying the switch settings. Now the Generic 22 Letter itself--I'll read it to you- " Additional measures"--
23 and this is beyond the stroke time testing that we talked 24 about before ;ishoold be taken to adequately verify that the ;
25 switch settings ensure MOV operability." Now we talked about i
1_______ _ _ _ _ _ _ _ __ l
..g 64-I whether or.not you'could'do a static diagnostic test or low l
'2 pressure diagnostic tests. But you have to comply with the 3 words in-the Generic Letter to verify the switch settings u
'4 that you have confidence in whatever test that you're doing, S' ' static, low load, median DP, or if the conditions arise that 6 Lyou lose confidence in a low DP type of test in five years,
- 7. you may be forced up to a full DP. But the way the words are r
8- 'right now you have to verify those switch settings, and L
9 . diagnostics:is a means to-go about that. And whether or not 10 you can have that confidence and static condition test or low 11 DP-condition test, that will have to be addressed'at the five 12 year interval when you get to it.
13 MR. KIESSEL: Tom, don't forget; you got post
- 14. major maintenance.
15 MR. SCARBROUGH: Thank you, sir. The other 16' portion where you may need to perform DP testing is spelled 17 out in the Generic Letter in Item F. Design basis 18 conditions may not be repeated unless MOV's are replaced, 19 modified, or overhauled. So you would look at the
- 20 modifications or overhauling that was performed and make a 21 decision as to how extensive your testing needs to be to !
- 22. provide confidence that the MOV will still operate as it 23 would have before the modification. You may be able to do a 24 lowDPdIagnos~bicsonit. You may have to do a full DP 25 diagnostics. It depends on the type of maintenance, the type i
i 65 I of. overhauling or modification that you made to the valve to
! 2- determine the extent of testing you need t"o perform.
3 MR. SULLIVAN: Any questions on-that?
4 (No response.).
5' The next question Jeff is going to take up.
6 MR. JACOBSON: "Should~ design basis review 7 and documentation be performed concurrently or should all 8 design basis activities be completed prior to the initiation 9- of testing: activities?" ' Basically we really don't care in 10 what order you accomplish the requirements of the Generic
.11 Letter as long as they're all done within the five year time 12 frame. Doesn't have to be done in the exact order that it's 13 written out. However, it all has to be done within five 14 years. So if you're going to initially try to set your 15 switches, you're going to have to have some basis to do that 16 on and there may, before you need to do that, have to be some 17 -testing or prototype testing to figure out what-your valve
~18 factors and so forth really are so you can set the valve up 19 properly. So whatever order is most comfortable for you is >
20 okay with us.
21 MR. SCARBROUGH: Okay. The eighth question.
22 Ted's going to take that one.
23 MR. SULLIVAN: This question says, "Should 24 valves wh ch w e previously manually operated and have been 25 modified to be motor operated as a result of TMI lessons
- _ _ _ _ - _ _ _ _ _ _ - _ _ _ n
V 66 1 learned, be tested under full flow pressure conditions solely ,
2 due to the valve's function, i.e., safety injection pump 3 discharge valves?" The way I understand this question, I 4 would answer it this way. The valve's been modified and i 5 there's now an MOV on the valve. The valve, as I interpret
. 6 this question, has a safety related function, and.in that
-7 case, the fact that it was'previously a manually operated-8 valve doesn't change'the way the valve is right now. It 9 would need to be in the program.
10 MR. SCARBROUGH: Okay. That's it from the 11 Consolidated Edison questions that we received. We received 12 a note upon the questions themselves. They were going to.be 13 at the September 20th meeting. That's why we felt we should 14 go over these.
15 okay. The next set that we have came from 16 Nebraska Public Power District and I know their 17' representatives are here.
18 The first question is--there's a long 19 preliminary discussion that's before each question and I'm 20 not sure to what extent we'll read that. We'll just read the 21 question.
22 MR. SULL1 VAN: A lot of the discussion that's 23 precedes these questions is sort of setting the stage with 24 information tfiI anyone that's familiar with the Generic 25 Letter's already aware of. So I think reading the question l
t___---_______---- - - - - - - - _ - - - - - - - - - - - . - - - - - . -
- - - - - - - - fl
i 67 t
I will handle it for those of you from Nebraska. That's why i s k 2 we're doing it this way. l
)
i 3 MR. SCARBROUGH: Okay. The first question !
4 was, "In light of the statements, and they related to 5 position-changeable and single failure aspect, why is the 6 requirement for licensees to prove that an inadvertently 7 mispositioned valve can be repositioned outside the design 8 basis?" It may be outside the design basis. It was treated 9 as possibly being outside the design basis. And the NRC 10 reviewed that and determined that the concern for the 11 operability or liability of those MOV's demanded that 12 mispositioning be considered in the Generic Letter 89-10 13 program regardless of whether or not it was outside a 14 particular design basis for the plant. This was addressed in 15 a letter to the BWR owners group by Ed Jordan of the Office 16 of AEOD, and it said a major goal of the bulletin is to 17 ensure a higher liability of individua] safety systems 18 notwithstanding the fact that the safety systems are designed 19 to withstand a single failure. And that applies today just 20 as it did then, that mispositioning is inside the scope of 21 Generic Letter 89-10.
22 The next question is, "Since any of the MOV's 23 at Cooper can be mispositioned manually using the V clutch 24 and handle whedl mechanisms, can we assume that the
~
25 phraseology ' blocked from inadvertent operation at the valve l
l l
l l
L --- I
68
~
i l itself' refers to electrical operations only?" The simple !
i 2 answer is no. Blocking must be electrically and mechanically l 3 and this is where we talk about how you accomplish that. The 4 philosophy in Bulletin 85-03 and also in Generic Letter 89-10 5 is to prevent inadvertent operation. It's not to prevent any 6 deliberate act that's thought out by the operator or the 7 auxiliary operator. It's to prevent an inadvertent 8 operation. So you need to prevent this inadvertent operation 9 electrically or mechanically such that the operator in the 10 control room is able to take action with that valve and not !
11 have something happen in the plant--where it was 12 mispositioned by someone and now the operating control room 13 cannot move it back to its proper position. So the ways that 14 you can electrically prevent an inadvertent operation--these 15 are examples. There's other ways that you may think of.
16 First, you can rack out the circuit breakers.
17 That's a way. Second, you can have a key lock switch in the 18 control room as long as the key lock switch--the key was not 19 in the key lock switch in the control room. That satisfies
- 20 an inadvertent aspect. But it electrically has to be in the 21 control room and locally. You have to prevent an inadvertent 22 from both control room and local operation. Now mechanically 23 there's several ways to do that and we give you a couple of 24 examples here[ You may think of ways that are acceptable as 25 long as you justify them. You could chain the hand wheel.
l 1
-._____ ____ n
69 i
I You could take the hand wheel off. You could have a safety !,
2 seal. A little aluminum strip used in other situations where 3 you want to note that if you move this valve or if you move 4
this object, you will break this seal. It's not a chain. )
5 It's just a seal to ensure that the operator thinks about
, 6 what he's'doing and is sure of his action before he does it.
7 Another way would be a locked door to where these valves are.
8 That would take a specific action to go get the key to go in 9 and move these valves. Now those are examples of ways to 10 satisfy electrical and mechanical but you have to do both.
II Okay. That's an important point. Are there any questions on 12 that aspect before we go on to something else?
13 (No response.)
I4 It was just pointed out to me that the f 15 electrical--when you say control room and local, we're 16 including the remote shutdown room. That's electrical. Any 17 electrical operation, any place you can operate it 18 electrically needs to have controls to prevent that I9 inadvertent operation.
20 MR. SULL1 VAN: If no questions, let's go on 2I to the third one. Jeff will take that one.
22 MR. JACOBSON: Basically this refers to our I 23 use of the word safety related system in the Generic Letter.
24 I'll paraphras a little of this. It says in paragraph two 25 under recommended actions, on page 3, references made to any
[
_ _ - - - - - - - - _ _ - - _ _ - . - n
e.
I 70 I
l L MOV in a safety related system. The question is, if an MOV l 4
2 is a component of the safety related system and the MOV is 3 classified'and licensee's Q list is non-safety related, 4' should this MOV be included in the program? The answer to 5 that is if the MOV is in a safety related system, whether it 6 has a safety function or not, it either has to be blocked or 7- you have to prove that it's position has no consequences on 8 the safety related~ system. I think we had an example. I
- 9 think the pressurizer spray valve is in a safety related 10 system; however, you may not assume it's in any particular 11 position for'any accident scenario. And if that is the case, 12 you would not have to worry about everything with the Generic 13 Letter.if you don't care what position it's in. But just 14 because you may have it classified as non-safety related, 15 -then you would have to treat it as a position-changeable 16 valve unless it was blocked out or its position had no 17 consequence.
18 MR. KIESSEL: The fourth question actually is 19 a page and a half long of discussion and in order to get the 20 flavor--I'm not going to read the page'and a half of 21 discussion. I'm just going to try and highlight what I think 22 are significant points.
23 In the discussion they note that the initial
=: -
24 response is,due six months after the issue date of the 25 Generic Letter, and that we're having a series of public l
I l
- _ - - _ _ _ - . _ _ _ _ _ . - _ _ _ _ _ _ _ _ _ _ _ . _ .-_____________________A
71 i
i meetings and that the last of these won't be until September l 4
2 28th and that leaves only three months until the initial l
3 response is due. Then it goes on to point out that we allow 4 five years with three refueling outages and thus making a 5 target date of June 28th, 1994 going strictly on the five 6 year period. And they identify their refueling outages that 7 are scheduled during that time period. It goes on a little 8 further and notes that we, in the Generic Letter, allowed 9 that any refueling outages started.during the first six 10 months of the Generic Letter may not be counted in the three 11 refueling cutages that you have to complete the entire 12 program. And they go on to note that the licensee is going 13 to be a member of the BWR owners group and that the owners 14 group had their first meeting on August 29th to lay their 15 ground work, and that their meetings are going to be held in 16 the future and that, conceivably, this may delay their work.
17 And then they went on to note that INEL had concluded that 18 some of the diagnostic systems that measure both stem 19 thrusting and motor torque are best suited for predicting. I 20 think that hits the highlights of it. I want to get down to 21 the specific question.
22 "Does the NRC expect licensees to adjust 23 outage schedules and durations to accomplish Generic Letter 24 89-10 activiti
?" No. But it recognizes that at some point 25 in time this may become a critical item. But we're not l
l l
I l
l u__ _ _ _ _ _ _ _ _ . n i
72
-_ 'I l' -saying'that'you have to start now changing all your schedules l 2 so that you can-meet them. i 3 "If not, would the NRC consider making the 4 data the third public meeting as the start for the Generic 5 Letter to schedule rather than the issue date?" And there 6 the answer is no. The clock started running on June 28th.
7 They also implied that because their work 8 with the owners group and them-trying to get all the'various 9 parties together, that there might be a delay in.getting the l
10 work done. And.we' agree that in getting set up and getting 11 started there might be a delay. But we also believe that the 12 synergistic effect that is obtained by having a multitude of 13 licensees looking at these problems and delegating certain 14 licensees to look at one problem while another licensee is 15 looking at a different system and the farming out of work 16 effectively probably will expedite things. And so, the 17 simple fact that you are working with owners groups would not 18 be a valid excuse for why you need to extend the time.
19 Rather, we would tell you that it is an example of how 20 working together you probably can save time on doing many of 21 the aspects of the program.
L 22 To the people who wrote the question, did I l
23 fairly characterize your concerns and answer them?
24 -
=' SPEAKER: Yes.
25 MR. JACOBSON: Let me add a little bit there l
1 i
i Y _ _ - --___ _ _ _-- -_---.- l
d' 73 i
'l to make sure everyone's clear on this; and that is, just f
2 because.you cannot complete these valves within five years 3 during your normal course of your outages is not.a. valid
'4 excuse for not doing it. We don't require that you change 5' all your schedules around. If, however, the path you choose 6 requires that you have to extend some of your outages, then 7 we expect that. And that is one of~the reasons why the 8 industry is going to.have to get together. The more
.9 ' prototype testing that's done, the less impact it's going-to 10 have on people's outage schedules. If you have to.go and do 11 full differential pressure testing on every valve in the 12 plant, it's no doubtedly going'to extend the outages. Now if 13 you get together and do some good prototype testing, it may 14 be the outages won't have to be extended'at all.
15 MR. TAGGART: John Taggart,.Palo Verde.
16 Going back just a little bit on position-changeable and 17 blocking, did I understand you to say they had to be blocked 18 from both manual and electrical operation?
19 MR. KIESSEL: Yes.
- 20 MR. TAGGART: Well, the Generic Letter reads 21 block from inadvertent operation from either the control 22 room, the motor control center, or the valve itself.
23 MR. SCARBROUGH: Those words are interpreted 24 by the NRC to mean electrical and niechanical.
25 MR. JACOBSON: If you can block them j
_ _ _ - _ - - -_ _ _ n
.y 74 1- electrically and mechanically from just one of those places, !
i*
L 2 then that would be acceptable. But the goal is to have them 3 .b1ccked' electrically and mechanically. We understand that 4 the words may be.a.little misleading in this case.
5 MR.'TAGGART: Either/or and and are
'- 6 differently a:different interpretation.
7 MR. JACOBSON: That's why we're having this 8 meeting. We've talked about that position in great deal and 9 that is.the conclusion that we came up with. Doesn't make 10 sense just blocking them electrically if they can be operated 11 mechanically or'vice versa.
12 MR. TAGGART: Well, inadvertent operation in 13 the specific example you gave us, Davis-Besse, from the
'14 control room, not somebody going out and manually operating a
-15 motor operated valve. Usually have to take a little thought 16 to do something like that. That's why when we read this and 17 the n rds are fairly specific, either/or, we assumed that if 18 it had a key lock or some other--or you had the. breaker 19- racked out, that prevented inadvertent operation of the 20 valve.
21 MR. KIESSEL: Racking out the breakers can 22 electrically be energized from any control position, correct?
23 MR. TAGGART: It won't prevent you from
- ==
24 manually operating it. ,
25 MR. KIESSEL: I say it's going to
____z_____ -- n
pf ' ; ,} ,
a p To <
d4 C
75
.b- '
. I 1 1 electrically.. I i
2 MR. TAGGART: That's correct.'
- l. ;
- 3 MR. KIESSEL: It will prevent you.from r
-4 . electrically operating it. Now let's suppose that in.the 5 heat,of' passion somebody.goes racing down into'one of your 6- valve galleries, grabs the-wrong valve and closes,it, and
'7 that happens to be the valve that you electrically racked 8- out. And now we compounded the situation.
9 MR. TAGGART: Well, you're correct.
10l MR..KIESSEL: Yet, that's why we're.saying=
11- .that we want both mechanical and electrical. We don't want 12 -to'put the plant into the situation where conceivably it is-13 in a more. dan'gerous situation.
14 MR. JACOBSON: We don't see where this is'a 15 big item either, unless we're not seeing something.
o 16 MR. TAGGART: We have a great many valves 17 that are prevented from operating, like containment purge 18 valves, that wouldn't come under this because they are 19 blocked and not position-changeable. But we don't go out and 20 block those valves from manual operation because the only 21 real inadvertent or automatic actuation you get is electric.
22' MR. JACOBSON: The containment purge valves, 23 do they have a safety function?
24 ==MR. TAGGART: They are safety related.
~
25 MR. JACOBSON: If they have a safety function I
= _ _ _ _ - - - I
P 76
--. . . l.
1- you'can't block them out. You have to show that they meet l
c -
2L the design--they're under the Generic Letter. You can't 3- block'them out. ,
4 MR. TAGGART: .You're preventing inadvertent 1
.5 operation by racking out. breaks. You can't operate them l f
. 6- - remotely from the control room without racking the breakers 7 in and then taking a second action to operate thus.
8 MR. JACOBSON: I understand that but if they 9 perforza a safety, function, this rack out question.is not 10 applicable, because if they perform a safety; function, you i
11 .have to go through the Generic Letter and show that.they'll 12 'be able to de that, whether there's key lock switches on 13 there or not.- The blocking business only comes into' play if 14 .you have'a position-changeable valve which is a valve that 13 has no safety function.
16 MR. KIESSEL: And if disposition could render 17 a safety system inoperable.
18 MR. JACOBSON: So *.here's very few valves l9 that I know of that are even going to come into this 20 scenario.
21 MR. TAGGART: I think most of them might be 22 under our lock valve program already, but to read that into 23 this Generic Letter, I mean, it's a pretty gross step to go 24 from either/or*Io and.
25 MR. JACOBSON: Okay. We understand that.
l
_o
77 I
I MR. KIESSEL: And we apologize for the !
l 2 confusion. The intent was, though, for, as we said, to try 3 and convey the fact that we meant both mechanical and 4 electrical.
5 MR. SULL1 VAN: Got a question in back.
- 6 MR. CALLOWAY: I've got a question. If you 7 have a safety related valve, if your safety position is, say, 8 in the open position and that valve is mechanically or 9 electrically locked,4whatever, into the open position--now 10 it's a safety related valve.
t-Il MR. SULL1 VAN: What do you mean by safety 12 related' valve, Clive?
13 MR. CALLOWAY: A safety related valve in 14 terms of function. It has a safety actuation.
15 MR. JACOBSON: It's locked open you're 16 saying?
17 MR. CALLOWAY: Right.
18 MR. JACOBSON: It's locked open.
19 MR. CALLOWAY: For an example, the
. 20 accumulated dump valves.
21 MR. JACOBSON: Yeah. Those are normally open 22 and power's removed, right?
23 MR. CALLOWAY: Right. Now in a case of the 24 dump valves, ifthesearelockedopen,wouldyouthenhaveto 25 include these?
- - --___ _ __ ____ ______ _ _ C
. ,. ,; 9:
78 m
~
I l- MR. JACOBSON: You wouldn't have to include l 2 them'if they were manually blocked as well.
3 MR. CALLOWAY: Manually blocked as well. So 4 this.is a case where you're dealing with a safety related.
5 valve.
. 6 MR.-JACOBSON: Yes, p 7 MR. CALLOWAY:- That we can exclude from the 8 scope of the letter because it is locked'into a' safety 9 : position.
10 MR. JACOBSON: Yes. On this containment 11 purge valve I think he's talking about valves that are 12 normally closed but would be required to open in some 13 accident scenario.
14 MR. CALLOWAY: Right.
15 MR. JACOBSON: In those cases you would have 16 to put them in the program. Now a valve that never has to 17 change state'and it's open and locked open would not have to 18 be in the program if it's properly applied.
19 MR. CALLOWAY: Do.you agree with that, Dick?
. 20 MR. SCARBROUGH - Yes, that's the correct 21 answer.
22 Why don't we go ahead. We have question five 23 here. It has a very short introductory section so I'll just
~~~.
L 24 read it. ~'.
25 "It is our understanding, based on
. _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ D
l 79 )
I discussions at the ACRS Subcommittee meetings, that the i i I I
2 regulatory concern is an increase in the core melt 3 probability due to MOV failures and that the objectives of 4 the Generic Letter is to reduce this probability to an 5 acceptable level." The question is, "Would not this 6 objective be achieved by limiting the scope of the Generic 7 Letter to those MOV's in the direct injection flow paths ot l 8 ECCS? This would reduce the cost impact to licensees 9 considerably while still achieving the safety objective."
10 Bulletin 85-03 addressed high pressure lines 11 injecting, fighting cooling. The results of 85-03 raised a 12 generic concern for MOV in all safety related functions.
13 That's how we came to 89-10, to address all of those 14 concerns. Now the question was relating to direct injection 15 flow pads of ECCS is a good reason for doing those MOV's 16 first in your 89-10 program. But just doing those doesn't 17 resolve this generic concern that's risen as a result of 85-18 03. So the question provides a good reason to go ahead and 19 doing those direct injected flow pads first since those, by 20 the question itself, indicates that those are high core melt 21 risk valves. But the program is still covering all safety 22 related system, those MOV's and those safety related system.
23 MR. KIESSEL: And I think this might be a 24 good point to interject that when we talk about looking at 25 your design basis scenarios, we're not just talking about the I
n
80 'i l
I I initial firing off and getting water flowing into the reactor !
2 to replace'back which is lost from a loca or~any other. i 3 initial portion'of the scenario. We are talking about the
~
4 entire scenario that.is.necessary to be followed to bring the
! 5 . reactor to a cold shut'down and to keep it there.* We're
- 6 talking about being able to keep the plant in a safe
(-
7 condition for long' post accident. And in looking.at your-8 safety scenarios we, in other. sessions,-had a number of 9 questions.along'the lines of, gee, the purpose of'this valve 10 is to top off an immediate leak and let me flow water into ll' -the reactor vessel and then I don't have to worry about it, 12- right? .W ell, somewhere in the scenario there probability is 13 a phase where that valve gets closed because we're going to 14 do some realigning of systems, or that valve gets closed 15 -because we're going to use the system to regulate a level if 16 we have just a small loca. .Or at some future date the 17 emergency operating procedures may call for massive system 18 realignments and'it's following.that entire train of thought-19 that is necessary to bring the plant from that accident
. 20 initiation to a cold stable condition and must be considered.
21 SPEAKER: Based on what you just said, the 22 case with the accumulator isolation valve that was locked 23 open during operations, some plants take credit for that 24 valve closing sY tfiey can initiate RHR during an accident so 25 they can get the plant cooled down in a safe shut down
_ _ _ _ _ _ _ . _ _ _ - - . - = n
,}
81 l
1 condition. In that case,-then, that valve would have to be !
2 included in the program.
3 MR. KIESSEL: That's correct. In that 4 _ scenario it would have to'be and that's why it gets very 5 difficult when we sit here and say does this particular
- 6 valve--do I have to consider it opening, do'I have to 7 consider it closing, do I have to consider it both opening 8 and closing, because we're not the experts on what your 9 scenarios are.
10 MR. SCARBROUGH: But for a valve like that we
- 11. can only operate down the line in the emergency operating 12 procedures. You have to qualify for conditions at that 13 point. You wouldn't be talking about something where it was 14 actually discharging into the coolant system and try to shut 15 it at that point.
16 MR. KIESSEL: And was I correct in hearing 17 earlier that the valve was racked out until the point in time 18 where it's needed? Did you say it was electrically racked 19 out? And it would be at that point in time you put the
. 20 breakers back in that scenario. That's when you start ,
21 looking at what are the pressures and temperatures and 22 conditions that you would anticipate. I think that came from 23 the other Chicago meeting. They identified a valve that was 24 on the RNR sys[dm. Some of the valves are intentionally 25 racked out to prevent their inadvertent operation during
_ - - _ _ - . -. - n
(- ( .
d 82 o,*
L l either plant operations or accident scenarios, but then later i
2 on operator in action was required to re-initialize those 3 valves so that they could then'be used for the long term 4 cooling mode. And, again,.it's only during that window of 5 where the valve operation is necessary that we want you to' 6 consider in looking at your design basis. But make sure that 7 you consider the entire scenario.
8 MR. SCARBROUGH: That was it for Cooper.
9 Do the Cooper: people have any questions on what.we've 10 answered so far that we can maybe clarify for them before we 11 go on to the next set?
12 MR. KIESSEL: Or have we dug you deep into 13 the well.
14 SPEAKER: Yes.
15 MR. SCARBROUGH: We'll move on then to Gulf 16 States, River Bend Station. Are those gentlemen here? Gulf 17 States? River Bend? There's only three. We'll go ahead and 18 answer them. The first one concerned single failure question 19 on mispositioning and we already addressed that. The second 20 one, Jeff was going to take that one.
21 MR. JACOBSON: "Has the NRC done an analysis 22 with respect to specific vendor's valves' ability to meet 23 original design basis requirements?" The answer is no.
~
24 However, when b find out about specific problems with 25 specific valves we may do the appropriate thing, such as l
l l
l i
_______.m__ ____ _ _ _ _ _ _ _ - - _ _ _ . _ _ _ _ _ _ _ _ - _ _ _ _ -
k l 83
'I write in.the information notice which one was just written on I
'2 some Borg-Warner _ valves. We don't intend on coming up with a _l 3 hit list of problem valves, at least not at this time we 4- don't.
5 MR. KIESSEL: The third question, "Will the 6 .NRC provide results derived from industry-wide Bulletin 85-03 7 responses?" It's unfortunate that the questioner is not here 8 because I'm not sure what he means by the results. .I'll tell 9 you.what-we're planning.on doing. We're planning on having 10 our contractor prepare a status report. The status report is 11 going to identify all the responses that have been made, all 12 of the. inspections that have-been done and documented in 13 inspection' reports, and that's basically going to be the end 14 of it. We may get into a little bit about the final summary
~
15 report, but he has not been tasked at this point in time to 16 do that. This document will a new-- and if it.is, as are all 17 of the documents that my branch issues--on the close out or 18 status of bulletins--and if it is as I initially envision, 19 simply addressing the responses and the inspection reports, 20 it probably.will be out in about six months. If it gets 21 expanded to include summaries of the final reports that were 22 submitted, it May stretch out for a year or so.
23 MR. REYNOLDS: Bob Reynolds, Arkansas Power 24 and Light. ,on the same line with regard to our responses to 25 85-03 that we've already made, if we make any changes or have 1
- _-_-__________-x
84
~
I madeLany changes in preparation for the' Generic Letter >89-10, l r
2 do we need to.come back.to'you as far as changes in our 3 iriitial response' to' 85-03?
4 MR. KIESSEL: No. The Generic Letter 5 specifically turned'off any further reporting requirements 6 under the bulletin. Rather what'it said was, if you were 7 going to make a bulletin response, keep it on file. That 8 information then becomes part of the package--defining your p
9 . program.or on defining the specific information. But, no.
10- We did not want to continue to burden licensees with making 11 responses to the bulletin while at the same time trying to 12 get their act together for the Generic Letter.
13 MR. SCARBROUGH That's it on that set. The 14 'last resubmitted set of questions we have came from Texas 15' Utilities.- The first one addressed position-changeable and 16 what we meant by blocking schemes. And I think we've
- 17 hammered that one and I won't go further than that unless 18 someone has a particular question.
19 The second question, Ted is going to take 20 that one.
21 MR. SULLIVAN: This question says, "If a 22 utility takes exception to the schedule or recommendations 23 contained in Generic Letter 89-10 in the initial response, 24 will the NRC pFovide a response or is the utility permitted 25 to implement these exceptions as .'.ndicated for subsequent
_________...___m__ __ _ l b . . _ . _ . _ . _ _ . . _
85
, I!
I revisions.to the schedule or commitment?" This question is l i
2 bringing us back to referring to paragraph'L on page 6 of'the ,
3
]
Generic Letter. It's under the Section that's titled l 4 Schedule. And I want to spend a minute a talk 'about the 5 recommendations area first and then talk about the schedule 6 aspect.
7 I think the answer that I'm going to give, 8 and it applies to both, but I did want to comment on the 9 recommendations--a lot of the purpose of this meeting was to 10 get' feedback for you. Was to give you feedback on the 11 exceptions or difficulties that you envision with respect to 12 meeting the technical aspects of the Generic. Letter as far as 13 1 -alternatives go. And that's largely what we've been talking 14' about. Any alternatives or other technical areas that you 15 envision under this question you really ought to bring to our 16 attention at these meetings. That's largely what they're 17 .for.
18 With respect to the schedule, I want to 19 basically start out by reiterating some of the things that 20 Jeff and Dick were saying. That the schedule was carefully, 21 thought out. We spent a lot of time at the NRC working on 22 schedule aspects and we're not going to be particularly 23 receptive to extensions without strong justifications.
==
24 Now to try and get to the heart of the 1
25 question, this paragraph L that was I referring to was really C________ _ _ _ _ _ __ n
5 1
86 1 not envisioned to be a prior review and approval process. It 1 i
2 just isn't what we had in mind. And I think if you look at 3 various aspects of the Generic Letter, you get that flavor.
4 That this is something that we want you to go do. It has 5 someuhat different character from 85-03 in terms of 6 reporting. This is basically not a prior review and approval 7 process. In general, we don't p1'an to provide responses to 8 the information you submit under L. Part of the reason why l
l 9 we. wanted that information was so that we would be aware of ,
l 10 what's going on in general. We wanted to know if there are 11 any trends that we ought to be aware of. In saying that 12 paragraph L does not imply a review and approval process, 13 we're saying that the licensees should impicment the programs i 14 without waiting for NRC feedback. There are however 15 attendant risks that are involved in this philosophy.
16 Licensees who av put in or take exceptions need to be 17 confident that they can support those reasons. We are going 18 to be have the licensing project managers look at these 19 responses. He're going to be giving them some level of
. 20 review, and these reviews could result in some challenges. I l 21 don't want to say that this paper is just going to come .in 22 and is not going to be looked at. I'm merely trying to 23 dispel the notion that it's a prior review and approval 24 process.
25 I want to point out one last thing. If a l
_ _ _ _ _ _ _ _ _ - 0
(.-
l L
87 l 4 1 s l' utility believes that it's taking such significant exception !
2 to the Generic Letter, you ought to bring it to the attention 3 of the PM and get some dialogue going.
4 Does that help or thoroughly confuse the 5 matter? Any questions on these points?
6 MR. ROSS: Bill Ross with Texas Utilities.
7 Part of that question in my mind relates to the long range 8 program which you all have talked about is needed by the 9 industry. I personally do not believe and 1 don't think I 10 can recommend to my management that we can get through a two Il phase approach, as commented in the Generic Letter, in five 12 years. I strongly believe that implementation of programs 13 for MOV general improvement, including the design basis 14 review as outlined, training, spare parts, and a test program 15 to assure general condition is there based on the knowledge 16 that we have the schedule is probably reasonable. But to I7 solve the problems that the Generic Letter raises and 18 requirements of DP versus non-DP, a rate of loading, stem 19 factors, coefficient of frictions, those sorts of thingc, 1 20 that we're probably talking about an industry test program I
21 before you get to the plant specific situation of a five year 22 direction. ,
23 MR. SCARBROUGH: I think the concern there is J 24 85-03 raised t issue of the operability of MOVs and design 1 25 basis conditions. And we've seen that at several of the i
____ ___ n
l 88 i l 11: plants where they weren't able to operate those MOV's in !
' 2 design basis, that is the real deriving force here. It's o
'3 great that!we're going to have improved maintenance and 4 ensure the training, and that's all part of the whole 5 program. But the concern today is will those'MOV's operate 6- under a design basis condition when'they're called upon,'and 17 right now-we don't have a lot of confidence that they're 8- going to do that.
9l MR. ROSS: By the commission's own 10 admittance, though, you also don't have any confidence today 11- .that there's any way to address these questions, and it's 12 going to take time, just as your letter to NUMARC encouraged 13 the industry to get together, to get EPR1 to deal with this.
14- That's one phase of a test and you probably got to get'in the 15- parameter test and then you got to get into the grouping of a 16 valve specific. application and get type testing done. And to 17 accomplish those three things needs to be done before you go 18 into your plant to be able to say this is the design basis 19 test and it's a reasonabic test. Because any test less than
. 20 that is at risk.
21 MR. KIESSEL: The Generic Letter specifically 22 recognized that there may be certain valves for which true 23 type testing information wou1J not be available, and it
- ===: .
24 offered as an alternative for that a two step phase in which i.
25 you would make a conservatively best guesstimate at.it and c__ _ _ _ _ _ _ _ _ _ . _ _ _ - - ._ _n
89 1 set your plant valves up for that. And then if the bad news 2 came, you would then have to reset. So thero is an 3 alternative and it is recognized. But it's also hoped that 4 we're talking about a small population or a small percentage 5 -MOV valves. There are a lot of them that are used in cold 6 water service that are still very, very vital to your plant.
7 But I think you ought to be able to have easy access to data 8 to verify that those valves will function under the design 9 basis conditions. When we start getting into the bad actors 10 so far that we know of is a reactor water clean-up system--
11 isolation valve, it may not work. You're going to have to l 12 make your best guesstimate now and hope that you can come up.
13 with a definitive answer within that five year period. If 14 not, we're going to expect to hear from you. !
15 MR. ROSS: To get answer that we haven't got 16 an answer.
17- MR. KIESSEL: I'm sorry?
18 MR. ROSS: And to get an answer that we j 19 don't have the answer yet. As I understand the Davis-Besse
. 20 event valve is still not solved. So you have to pull the 21 valve out to do some more testing.
22 MR. KIESSEL: Again? I know they replaced it 23 once.
. :-- l 24 MR. ROSS: They took it out. They took out 25 the bad actor and went to go and test to try and figure out
_ _ _ _ _ _ _ _ _ _ _ _ _ __ n
l 90 I what the problem was. There's more than bypass problem. !
! i 2 MR. KIESSEL: Yes, and they couldn't figure l 3 it out. So what they did was they replaced it, period. And 4 MR. JACOBSON: We understand that there may 5 be a few valves out of a population of a hundred or a hundred 6 and fifty that due to some peculiarity you may not be able to 7 completely complete.the program within five years. However, 8 I've seen some plants that have done essentially this'whole 9 Generic Letter in quicker than a five year period.
10 I think this is a good time to talk about 11 something else; and that is, if during the course of your 4 l
12 design basis review or in the course of doing this Generic 13 Letter you find out that you have a generic problem at your 14 plant with undersized valves or undersized actuators, you 15 don't have five years to wait. You have to take immediate 16 action. And we've shut down a couple of plants on MOV's and 17 we're going to continue to do inspections, maybe not 18 specifically to the Generic Letter. But as part of our 19 inspecting at headquarters we are now looking at MOV's on
. 20 just about all our SSFI inspections and inspections like 21 that. And if we find out that you have an MOV problem at 22 your plant, you're going to have to address it at that time.
23 This letter was written with the assumption that most of the i
~
24 valves in your p1 ant are okay. That's why we're giving five
~
l l
25 years to complete it. We realize there may be some plants l l
l
_ _ _ . _ . _ _ d i
O'
- 91 I
l- out there that'that's not the case, j I
~2- MR. KIESSEL: Bill, did that answer you or 3 did we completely turn you off?
4 MR. REYNOLDS: I suspect that there--
5 extension requests.
6 MR. KIESSEL: I would greatly surprised if we 7 did not receive at least one. In fact, I might even bet on 8 two or three.
9 MR. SCARBROUGH: Okay. The last question 10 here in our set of resubmitted ones from Comanche Peak, 11 - '"Will the NRB publish answers to utility questions submitted 12 for these workshops? If so, when?" And we touched upon this 13 but since this is the last question, I'll reiterate. The 14 transcripts of all three meetings should be in the FDR in a.
l'5 ' couple weeks. We're also going to prepare minutes by conbing 16 the three meetings, so we have one set of minutes for all 17 -three meetings. Those minutes will focus on the most ,
18 important questions. It won't hit every single question but 19 focus on the most important ones that what we consider need 20 to be addressed. But the review of those meeting minutes up 21 through the NRC chain will take a while. Could take three 22 months. So don't wait for the meeting ndnutes before you 23 start on the program. The transcripts will be available in
- = = = - .
24 the PDR. Y,ou can get those. I would encourage, if you're 25 going to read--if you came to this meeting you'll probably w_ = _ _ _ _ m
"4
~
92-F l' .not:want to read this meeting, but read the other'two as -1
'2' .well. . The'other.two. meetings are larger, they were.more 3 . questions subeftted in' advance, there's a lot of good l stuffs 4'
~
in those transcripts'which could be applicable to you. So-5 that basically is the line of the meeting minutes'.
6 .Was there any questions on Comanche Peak'in 7 general that.we.could respond.to?-
8 (No response.)
9 MR. SULLIVAN: What we'd~like to do at this
- 10. point, we have' half a dozen oriso cards with questions. We
,+
' ll- would like to caucus to go through and-decide who's going to 12' answer which question. Make sure'we've read them at'least 13 once before we start' talking. We would strongly encourage 14: you at this point in time, if you have more cards that you'd-r 15 like to write questions on, please 'do so; otherwise, we're
-16 basically going to be concluding the meeting after the cards
' 17 and floor questions. So why don't we get back together in 18 about fifteen minutes.
L 19- (Off the record.)
. 20 MR. SULLIVAN: We took a much longer break i 21 than we expected but we'd like to get back to the questions ,
22 we have on cards at this point.
23 Are there any other cards?
24 (No response.)
( ,
25 A lot of the questions on the cards are new l
l l
l
____1___- - _ _ _ _ _ . _ _ _ __ n i
l 93 i
i i questions that have , come up before. Some of them are I l j i
l 2 believe just writing down for our purposes what we've already ll 3 talked _about. But most of them are basically new questions.
4 MR. KIESSEL: The first question that came in -
1 5 I want to read to you the exact question. Then I'm going to 6 rephrase the question and answer it t'e way I think it was .
7 meant. "What level of detail is required for the program 8 description and schedule that must be submitted by June 9 19907" Okay. On the surface, the way the question is 10 written, there is no program description and schedule that 11 must be submitted by June 1990.
12 Now, let me answer the question that I think 13 was meant. What level of detail is required for the program 14 description that must be available on site by June 19907 15 That is enough detail for us to be sure that you are going to 16 and are. capable of meeting the requirements of the Generic 17 Letter. That means that you have also--there is scheduling 18 information that is required. That means that we see a 19 reasonable schedule for which you are going to be handling
. 20 these valves. And it's not that you're going to do all 900 21 safety related valves during your last refueling outage, but 22 that you've got them spread out. And that it looks like it 23 is an attainable schedule and that you haven't bunched everything to 5he end.
~
l 24 The program description is going to l
25 describe the conditions under which you were going to
)
_ _ - - _ _ n
')
')
-)
- , ga j
- i 1 establish'the similarity, and demonstrate for the valves that t 4 j 1
- 2' have not been full design basis. tested and how you're going l. j 3 Jto demonstrate that they are going to be available. I'm not 4 saying.that you're going to have'to talk'about each valve.
15 But.you're going to have to be more specific than broad M. 6' generalities. You're going to have to be more specific than
.7 .the broad general descriptions that you submitted for ,!
8 Bulletin 85-03. You're_ going to have to start getting almost 9 into your procedures that-people are going to start to use..
10 But you should be one step ahead of that. Should be roughly 11 the same procedures that you were using for a lot of your 12 bulletin testing.
13 And I think that answers the question, or 14 does it? That cover it? Did this raise any questions in 15 your mind?
16 (No response.)
17 Next question. "On Item H, what is the NRC's 18 intent for the scope of the two year Mov data review?" At 19 bare minimum we'd like to see you reviewing your valve
. 20 records to look for the valves that are acting up. A look-21 see at the industry data for valves that you have in your 22 plant would be useful. But we're not going to be asking for j
23 great detailed examination and, you know, five hundred page 24 reports ind th'ihgs'like this. But you have a program that
~
25 you are aware of what has been going on in the industry and
--- - - - - - - - - - - - - - - -- - - - - - - - _ - - - - - - - - - - - - - - - - - - _ - - _ l_
L !
)
95
't i;- l' in your. plant, and you are starting to flag the--that'may j i
.2 need early attention.
3 "On Item H, what is the NRC's view of an 4 appropriately modified NPRDS?" We have no view. First of-5 all', we have no control over NPFDS. ' Industry wouldn't'want 6 us to have control over NPRDS. Secondly, you are the people 7 who are best suited toward determining what is the type of.
4 8 information.that you yourselves.are going to need. -And for 9 the long" term care, remember, we're talking about a program 10 in which you'are recycling through all of your valves every 11 five_ years or three. refueling outages unless you can come up
.12 with a good data base and good data that says you don't have
- 13. to do it that frequently. I doubt sincerely that any 14 -individual. utility can acquire'that kind of data in the 40 15 year-life ~ span of their plant. But I think all one hundred 16 and some odd operating plants pooling their information 17 probably can come up with some good data. 'And so what do we 18 envision NPRDS doing for you? As much as you can get it to 19 do? As much as you want it to do. We're not going to i 20 dictate what it must to do for you. We cannot dictate what 21 it must to do for you.
22 Any other questions along this line?
23 (No response.) ,.
- - . = -
24 ,
MR. SCARBROUGH: I'll go ahead and take a 25 couple here. This question earlier today was stated that
- _ _ _ - _ _ _ _ _ _ _ _- _ _ __ .n_ \
n; s ,
H I
E p 96 o .g. j il : mispositioning of valves was in theiscope of,. Generic Letter !-
.j l
-2 i 89-10 even if it'was outside the plant-design basis.. "Has a' ) 1 34 backfit analysis been performed? If'so, is it available to l
4 the industry?" Yes, a backfit analysis has been performed.
5 It's b'een addressed by the connittee .to' review generic 6 requirements CRGR and NRC that' analysis can be found in the .r
-l 7 Public Document Room within the Session Number.8809150045'..
8 -And that is there. And there is alsonanother letter.which-9- may.also be in:the'PDR--I don't have a' Session' Number for it-10 .-September 29th memo, 1988, from Eric Beckjord, Director of 11 Research. That-may be in there also. So there'has been 12 analysis. performed. It's been reviewed thoroughly by the NRC 13 through the process for backfit, and it was decided that
'l4 mispositioning'was an important enough issue that will be 15 -included within the' scope of Generic Letter 89-10.
16 I'll go through the Washington Public Power 17 L Supply System questions that we. received. They're pretty IS- short here. A number of them have already been answered.
19 One here is, "Are any balance of plant valves 20 'ncluded in the five year requirement?"
i No. The Generic 21 Letter 89-10 only addresses safety related or MOV's in safety
- 22. related systems or associated boundary type of safety related 23 valves. But balance of plant is not required in Generic
- =: .
24 Letter 89-1,0. It is encouraged that licensees look at those 25 balance of plant MOV's that are considered most important I
___________._m _ . _ _ _ _ . _ n'
97 I from a plant availability point of view and determine which !
i 2 ones it would like to see ensured of their operability. But 3 the NRC has no plans right now to require any balance of 4 plant MOV's to be included in the program, i l
5 Another one here is, " Provide an example of j i
- 6 when a design basis test must be repeated." And I think we j 7 addressed that already, about the five year switch 8 verification. That has its own associated provisions with 9 it. When a design basis test must be repeated. They must i
10 verify every five years. And how they do that depends on Il their technique. They can use diagnostics or if they can't 12- verify it, then they would have to go beyond that. But I 13 think we've discussed that and I don't want to get into it i
14 too much.
15 The other aspect of design basis testing was 16 the maintenance, okay? But I just wanted to finish that 17 question.
18 MR. MASSEY: I'm Jim Massey, Washington I9 Public Power. The question is at what point do we have to do
. 20 differential pressure testing for overhaul? You mentioned 21 overhauls in the Generic Letter. Does that mean every time 22 that I have to tear apart an operator and replace the grease i 23 that I have to do a differential pressure test on it?
' ~
24 ~ MR. SCARBROUGH: It depends on the type of 25 maintenance, the modification that was performed. If your I
98 t
i competent that what you've done has not affected the ability .!
2 .of that valve to perform under design basis conditions, no.
3 l You could use a low DP type of diagnostic tests to make sure 4 i that everything's working the way it was before you did that 5 overhaul.
6 MR. JACOBSON: If you didn't tear the valve 7 apart, I don't see any need why you'd have to do a full DP 8 test. If you just did some work on the actuator you could 9 probably do a static test to verify that at torque switch 10 setting it's putting out the thrust that it's supposed to.
11 MR. MASSEY: Okay. That's what I needed to l 12 know.
13 MR. KIESSEL: Specifically, the Generic 14 Letter left that to your discretion as to when you feel that 15 it is necessary. You know, when you have done the work that 16 the licensee feels will have affected the valve's 17 performance, then you must do the testing commensurate to 18 that level to re-establish your confidence. You go and you 19 change the packing, fine. Just make sure the packing loads
. 20 have not exceeded some threshold valve that you have 21 determined. If you tear the valve apart and completely 22 rebuild it, then I think we're looking at some kind of design 23 basis.
^ ~
24 ~~MR. MASSEY: Okay. But tearing the operator j 25 apart shouldn't--
t
r-I'd ;
99 I
-l MR. KIESSEL: If you are sure1you're going.to. 'l-
.I 2 get outlof the operator what you were-getting out.before, j
'3 :then you don't need to do a DP test on that because your
'4 diagnostic 1 equipment.will tell you where you're tripping.
5: MR. JACOBSON: There's many things that this E 6 Generic' Letter's meant to verify and you have to. figure:out' 7 what the maintenance did, which things it could effect,~and
-* 8 then re-verify'those-particular items. 'You don't have to 9 verify things ~that couldn't possibly have been effected.
10 MR..BARTA: Brad Barta, Stone and Webster.
11 An extension of Jim's question. In the cause of an audit 12 three orffive years down the road, would the commission 13 expect to see justification for each one of these' cases?
14 Say, they went in'and saw a battery.of maintenance orders
~
15 against a certain valve. Should we have a program that's in 16 . place saying if.you've done-this then do a-justification. If
> 17 you've'done this--but you. don't need a justification. Pretty
- '18 - good break down of typical maintenance activities. Is that 19 the kind of thing we should see in program development or 20 should we have a justification on a case by case basis as to f
21 why?
22 MR. KIESSEL: I would prefer not to see it on 23 a case by case basis but rather something that is fed back !
- == - . ;
24 into your program description. You know,'I can see where 25 when the program was originally written you thought of ten 1
. . _ _ _ _ _ . _ _ _ __ _ n _)
100
~
l 1 different types of maintenance that-you might do on the valve l '
l 2 and you put the commensurate level Df post maintenance 3 testing on it, and five years in the future you suddenly come 4 up with an eleventh'one, I can see that getting fed back into 5 .your program. And as long as it is documented in the program, 6 the only justification you need in the valve package is that 7 QA that is necessary to show that you were following the 8 program.
9 MR. JACOBSON: When we do an audit, we're 10 going.to look at the maintenance that was done.and the retest 11 that was done. If we feel there's a problem in that the
- 12. retest doesn't adequately cover the work, we're going to.ask 13 justification. Now, where that justification is, whether 14 it's in the program or in some other documents somewhere will 15 depend upon how your particular utility.is set up.
16 MP. K1ESSEL: I would like to think that they
- 17. put problematic type things all into the program, rather than 18 having repeated a hundred different times for each valve. I 19 guess each licensee has the freedom to choose.
20 MR. JACOBSON: Number six, "Is motor current.
21 analysis after adjustment adequate for operability?" That's 22 a very general question and I can't give a specific answer to 23 it. It depends on what maintenance was done. In some cases motorcu$ rent 5Iyheacceptable.
~
24 I'm not that familiar with 25 the different motor current techniques that are out there
- -_ .__-______A
u.
N 101'
- 1) rightfnow. 'If you can show that the thruct requirement'
~
f 2 hasn't changed by looking at the motor' current, then that
-3 . would be a good way of'showingLthat your thrust hasn't i
L- -
- 4. changed. Whetherfthere's instruments out there now that can 5 do that~or'not, I don't know. ;I know there's instruments 6 that try to do that. Whether they're successful.or not I'm 7! not sure of 9 .#-
8 MR. KIESSEL: . You'reLtalking about instrument-
"9- sensitivity--
.10 MR. JACOBSON: Yes. There's a lot of 11 different' instruments out there now'that do a spectral 12' . analysis and motor current, or just looking at motor current 13 and power factor, motor current and voltage. There's a lot:
14' of different combinations out there. 'What each one can do I 15 don't know.- You'll have to look at that yourself.
16 MR. SCARBROUGH:- Okay. The last one from 17 . Washington Public Power Supply was, "Since the NRC' agrees 18 that industry--groups, for example, EPRI, are'necessary to 19 complete the scope of the Generic Letter with an allotted 20 time frame, might it be considered to extend the time frame 21 for testing completion since these' programs have only i
22 recently begun?" No. I think we've talked about this. The 23 five year schedule was set up with the-type of aspects in 24 mind. We a,re aware that there is going to need to be 25 cooperation between the industry groups. There's a lot of l
W
- 1 F >
I i n 102 1 I
i ,
1 valves out'there in the plants right.now that.will not need l
.I g 2 to have.this type of cooperation. You.can go test them now
- 3. at their design base conditions. You don't have-to go and 4 run and do a prototype, things of that nature. And if that's-i 5 the most direct way.of implementing 89-10, go ahead and do 6 it. Do qet those valves out of the way. As time moves on 7 and you'get EPRI involved and the industry groups together,
-8 you can start working on the more difficult ones. At this 9 . point theJfive year programrwas set.up with this problem in 10 mind where the industries have to work together..
11 MR. JACOBSON: Let me add something in terms 12 of that. This motor operated valve problem is not a new 13 problem. It didn't occur after we wrote the Generic Letter.
14 The industry was aware of this problem for many, many years
.15 and really.hasn't jumped on it until just recently in the 16 last couple of years. And still, we do not see a very 17 -consorted effort in the industry in addressing this problem.
18 And you're asking us now whether we can extend it because the 19 industry is just now getting together. Well, we haven't
. 20 really seen the industry get together on this and start 21 addressing a lot of these problems to the extent that I think 22 it needs to be done yet. Maybe in the future, if they really 23 do put forth the effort and there are still problems that 24 aren't bsing re solved, we're going to take a look at that.
25 But we think there's a lot of work that needs to be done in
- _ _ _ _ _ - _ _ _ _ - _ - n.
103 i
I that area. j 2 MR. SCARBROUGH: Okay. That's it for the 3 Washington Power Supply System. Was there any clarification 4 or expansion of any of those answers?
5 (No response.)
6 Dick, you want to go ahead and do those, 7 please?
8 MR. KIESSEL: Okay. I've got two questions 9 that came from--please excuse me for butchering your name--
10 Javed Haider from APS.
II First question, "Do we have to consider MOV's 12 that are key locked, disabled at the MCC or change locked for 13 position-changeable scenarios?" I think we've been through 14 this several times. We're looking for mechanical and 15 electrical blocking of operation.
16 Question number two. I'm going to read the 17 question and then I'm going to ask you to expand on it a 18 little bit if you would, please. "Do we have to consider the 19 effect of one mispositioned MOV on the other MOV in the same
. 20 system?" Could you expand as to exactly what you're meaning 21 so I can answer your question?
22 MR. HAIDER: What I'm trying to get at there 23 is we have some MOV's in the same system and if I disposition 24 one, do i hav5 to account for the effect of that 25 mispositioned MOV into the thrust equation for the other i
_ _ _ _ _ _ _ _ _ _ . _ _ _ _ _ _ _ . _ b
104 i
1 MOV's in the system that are downstream or whatever? I i
2 MR. KIESSEL: And the answer there is no 3 because we're assuming--
4 MR. HAIDER: I was hoping you would say that.
5 MR. KIESSEL: --put this mispositioned valve 6 into your program that you can recover if that then restores 7 your system to where you think you should be for the 8 operation of the other MOV's that are required to operate.
9 MR. HAIDER: Okay.
10 MR. TAGGART: John Taggart, APS. Do you have Il to consider the worst case time that valve could be 12 mispositioned. We're talking about during the blow down or 13 the maximum DP it could ever see during inadvertent 14 operation.
15 MR. KIESSEL: Yes. In other words you have 16 to be able to recover from the mispositioning and at the 17 worst condition that that valve would see because of its 18 disposition.
19 MR. JACOBSON: Are you talking about a valve
- 20 that's called mispositioning? It's in the disposition 21 category or a safety related valve that could be 22 mispositioned?
23 MR. TAGGART: That's inadvertently operated.
' ~
24 You have to c sider the worst case time that they could 25 inadvertently operate?
_ - _ _ n
/ i
?^ ,
105 j
"I MR. KIESSEL:. That's'right. And it has to be- j-I
'2 Jable toibe restored back.to its proper position'that'any time
- 3. during.the scenario ~where--
4- MR. TAGGART: That's' going to greatly.
- j. 5 ~ redefine the design b' asis.
.. 6 MR..JACOBSON: Why. don't we. hold on, Dick.
7 We need to' caucus.on this one because we were talking about' 8 this before. Why don't we get back, John.
9 MR. K1ESSEL: To make sure I understand, 10 you're saying that you have a valve that you assume p'ositions 11 during the'first five seconds of the scenario and the max DP 12 that it would receive during that, for sake c*i argument, is'a 13 -hundred pounds. -But that later on in the scenario, if you-
'l4 have not recovered it at that point in' time, it might see 15 five hundred pounds.
16 MR. JACOBSON: If someone dispositions it-17 -later on you're talking about, right? Say the DP now is a 18 'thousand pounds. You're asking do you have to size it for 19 the thousand or only size it for the hundred where you 20 thought it was going to work at.
21 MR. TAGGART: That's correct.
22 MR. REYNOLDS: Bob Reynolds. What I'm 23 hearing I'm questioning, too, because the way we've 24 considerdd it-Es'if you disposition one, that's considered s 25 a single failure and we try to design for that single
= - _ _ - - _ _ .:____- _ _ - . _ _ _ l
.l1 106
~li _ failure.. ]
i
'2 MR. KIESSEL - We've thrown single failure out-3 of this.. We will not' accept hiding a mispositioned valve-4 within your single failure scenario.
5 MR. JACOBSON: We have to get back and talk, 6 amongst ourselves concerning the valves that are normally in 7- the program,:the safety related valves, at what conditions 8 you'have to size them.for. Whether it's at any-time or if. I 1
9 it's just-at the time they're called upon in your design
- 10. basis scenarios. I think that's the question and we're, going II' to talk'about that. Okay? We're not giving you a definitive 12' answer. Not now.
13 MR. KIESSEL: Did I. cover your point about 14 the mispositioning and single failure? .
- 15 .MR. REYNOLDS: No, you didn't because if you-l.
16 -you know, even though we can recover a valve, if we consider l'
I 17 that as a single failure, that could give us a worst case l
18 condition on another valve that valve might have to operate.
19 at. But what you're telling me now is that if I don't have 20 to consider a disposition as a single failure, that means the 2l valve that I'm going to size in another area in our system 22 doesn't have to operate at the higher pressure that we 23 originally sized it for because you're not considering that 24 disposition,as a single failure.
25 MR. KIESSEL: Because we're saying you now L_______________ _ _ _ _ _ _ . .A
y .. .
j L t, n j 107 I
E 1 can be assured that you will recover that valve before the l
. i b 2 other valve is required to operate.
l 3 MR. SCARBROUGH: You do not have to look at t
4- the effect caused by the mispositioning on all the other L 5 ~ systems and piping and everything else. Mispositioning is f
a 6 only-intended to be able to move that valve back where it p 7 should be. You don't have to do an analysis of piping stress 8 -analysis downstream as a result of that valve going to an 9 improper position. But on single-failure, single failureris 10 not a reason to exclude these valves or exclude 11 mispositioning.- That's already been addressed at the NRC and 12 it's been to notify the industry, notify single failures.
13 It's not a way to include these valves.
14 UNIDENTIFIED SPEAKER: Then we. agreed Monday 15 that a mispositioning scenario only needed to be carried 16 forward in time as long as the safety system's required to 17 fulfill its function?
18 MR. F'.ESSEL: That's right.
19 MR. JACOBSON: That's for sure. If it's not
.- 20 even required, then you don't have to worry about it at that 2l time. There is one point that's still open and I think we l
22 need to talk about it.
23 MR. KIESSEL: Next question came from John 24 Sharp, Florids bower and Light. "Can we have some 25 clarification as to what constitutes a ' motor operator' with I'
1 I
n p.
108
~
i l regard to dampers? We have two types of dampers; those !
I h 2- actuated by limitorque which we call MOV's, and those. f
.3 actuated by a motor with a gear and two snap-lock switches; 4f no spring pack, no torque switch, no bypass. Will this 5 latter group be included?" Unfortunately, the Generic Letter 6 is not a limitorque MOV generic letter. It is an MOV Generic 7 7 Letter. It applies to any valve that is operated by a motor.
8 SPEAKER: Is that meant to include solenoid 9 valves?
10 MR. KIESSEL: I don't consider a solenoid a 11 motor.
12 Second question from John. "Even though we 13 are saying that no extrapolation will be acceptable, the fact 14 that extrapolating twenty pounds is acceptable seems to 15 indicate that this will be on a case by case basis. Is this 16 the case? If this is so, will the percent of design basis be 17 the only parameter or will the amount--the available exceeds 18 the required also be considered?" Oh, golly. I knew I 19 shouldn't have tried to pick two extremes and to define the 20 very broad range where I wasn't going to give you an answer.
21 Without retracting the fact that I personally don't think 22 twenty pounds out of two thousand pounds is a significant 23 difference, yes, it will be on a case by cane basis and, yes, a- ==-- _
24 not only wi,11 percent of design basis be a parameter, but 25 also what is the margin would be an excellent argument in Ma a,h-,-mm-.mu,- . - - - - . - - - - - , - - , - - - - - - - .- -- -- --
L !
109 j 1
j {
l there also. But again, please don't try and pin us down as l
, 4 2 to an exact number. It's very difficult and I just proved l 3 - it.
- 4 MR. SCARBROUGH: The first is type instead of 5 prototype and the type versus prototype is concerning--in 6 prototype, what we mean there is an MOV outside the plant 7 which is used to ensure the operability of an MOV in the 8 plant. Now in my opinion a prototype is an identical valve 9 with an identical actuator which could possibly justify a 10 different actuator under certain conditions. And then that Il valve is used to perform tests on it. We talked about the 12 low DP and the high DP and established the performance 13 characteristics of that prototype and apply it to the MOV in 14 question in the plant and perform a low DP or static DP test 15 on that MOV to transfer that information from that prototype 16 to the one in the plant. Now that's my interpretation.
17 Bill, you want to expand on what you meant by that?
18 MR. ROSS: What I'm addressing there is 19 basically all Class lA equipment a nuclear power plant has to 20 go through qualification testing. That is typically done by 21 type testing. It's a semantics thing as opposed to prototype 22 and when I think prototype, I think of a first time ever 23 built design that will go into production. And in fact, 24 qualification of nuclear power plants and MOV's specifically, 25 I think it's IEEE 382, are required to be type tested to meet i
_ _ _ _ _ - - - - - - - _ - - _ _ m
f 110 1
(
'l - their qualificational' requirements. i i'
2- MR. SCARBROUGB: Kell, wt're using the word 3 prototype because it was in the Generic Letter. They have 4 used that word to mean an MOV that is tested which is-5 identical or near as identical as possible, then transfer
- 6 that information to the MOV in question in the plant. There 7 wasn't any intent--a prototype in the sense of a prototype B1 8 bomber. We're talking about a valve-that you do the test on, 9 then you transfer that .information to a valve in question.
10 MR. ROSS: In the semantics of the nuclear world, type 11 testing by statement means something in qualification. In 12 . fact, what we're doing here is monitoring parameters or 13 establishing from a type test that in fact something in a 14 plant meets those same qualification requirements.
15 MR. SCARBROUGH: My understanding of 16 environmental qualification testing is they take one actuary 17 maybe made by limitorque and they do some tests on that. And 18 then they apply that to every limitorque actuator as an EQ 19 qualification. We're not doing that here. We're not getting i
20 into the similarity because it's the same manufacturer. That 21 is not acceptable to establish prototype transfer. It has to 22 be identical. We're talking about the exact same valve and
. 23 then the type of actuator which you can be competent that it
~
24 will work the-EEme way as the one in question. That's what 25 we're talking about here. Now they differ somewhat from the
- - - . _ - . . _ _ ____ _ _ _ _ _ _ _ _ . _ _ . _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ L
i e
i 111 1
I past use of type and environmental qualification. testing; but !
2 what we've said here is how we interpret the Generic Letter. '
3 MR. ROSS: Under qualification testing it has 4 to be--matched the installed condition.
5 MR. .JACOBSON: What qualification test?
- 6 MR. ROSS: Excuse me?
7 MR. .JACOBSON: Which qualification testing
~
8 are you referring to?
9 MR. ROSS: For motor operator valves IEEE 10 -382. The valve manufacturer, likewise, on safety related 11 valves typically is required to demonstrate the flows against 12 the design basis and we are discovering in the past five 13 . years his equations may not be right, his stroke testing may 14 not have been against pressure, and this is the reason we're 15 here today. And I don't want to belabor type versus 16 prototype. To me it's a semantics thing, that the nuclear 17 industry thinks type testing for qualification and not 18 ' starting out a new prototype operation.
19 MR. JACOBSON: We weren't trying to reinvent 20 the wheel. Barring the semantics concern, are you clear with 21 the definition that we were using or attempting to use?
22 Whether we call it a prototype or a type?
23 MR. ROSS: Yes. The prototype type testing 24 that you're ts Eing about thougn to me, like some other 25- areas, challenges the way that weve been doing business for i
^ - -" - -,---_._----___,_a__a_ _ , _ _ _,_,_,___ _ _ _ _ _
l 112 i l I a long time. It challenges the qualification across the f l
El board and I think that the Generic Letter references the 3 requireme.nt to make sure that it's qualified. Is that 4 invalidating all the qualification documents? The installed 1
5 configuration has to match the design basis test that it was 6' qualified to or justify by exception?
7 MR. JACOBSON: We're essentially saying that 8 the initial qualification testing that was done for these 9 things was inadequate. That's why we've got the problem that 10 we've got today, because it was inadequate and we recognize Il that fact.
12 MR. SULLIVAN: It's my understanding that a 13 ' -
lot of the type testing as used in the nuclear industry is 14 similarity based with analysis or some sort of qualitative 15 arguments that justify the similarity. And I think what 16 we're saying is that that isn't going to be acceptable in 17 this arena and that's one of the reasons why we avoided the 18 word type testing.
19 MR, SCARBROUGH: Let me go through your 20 second question here. "In order to transfer type test valves 21 to in-plant testing, is determination of valve factor and 22 stem factor conditions for the in-plant valve assumed?" And 1
23 this is up in the area of technical differences of opinion on 24 , whether or not you actually have to come up with a specific 25l valve factor from your prototype testing or can you do it j
l l
- - _ _ _ _ . _ _ _ _ _ _ _ _ _ . . . n
i i
113- 1
, -l 1
I without coming up a number--says, here's my valve factor. I ;
2 But the point is, you want to be able to have enough
- 3 information on that prototype MOV to be able to transfer the q 4
knowledge and results of that testing on that prototype type 5 MOV to the MOV in question. Whether you do that with valve 6 factors or you do that with some sort of margin in your 7 traces and your profiles, that's up to you. But you need to n .
8 be able to make that transfer. And that is the purpose of 9 -the Generic Letter. Whether:or :not you need to come up with 10 .a particular factor that's up to your results. i Il MR. JACOBSON: You need to_ justify the way 12 you do it.
13 MR. SULLIVAN: Could you, just for my 14 benefit if no.one else's, expand a little bit on how you mean 15 the word conditions in your question? You're talking about 16 transfer and you're using the word conditions, and I'm not 17 exactly sure what you mean. You want me to read the question I8 again?
l l9 MR. ROSS: Yes.
. 20 MR. SULLIVAN: It says, "In order to transfer 21 type test valves to in-plant testing, is determination of 22 valve factor and stem factor conditions for the in-plant 23 valves assumed?" I'm really not clear on what you mean.
' ' ~
24 MR. ROSS: If, in fact, in order to meet 25 design operability requirements, if the type test identifies j
i i
I
_ - _ _ _ _ - _ - - _ _ _ __ n i
l 1
114' !
i I question marks that are there in the industry today for valve j ,
i 1 2 factor and stem factor, and rate of loading questions, DP 3 versus static, if you have the type test not in-situ, you got 4 to have the same condition identifiable in the plant test.
5 The conditions have to be duplicated somehow.
6 MR. SCARBROUGH: When you say conditions, are 7 you talking about temperature and pressure, flow?
8 MR. ROSS: No.
9 MR. JACOBSON: He's talking about the 10 physical condition.
Il MR. ROSS: Determine the physical conditions 12 in regards to the valve factor and the stem factor, DP versus I
13 static.
14 MR. JACOBSON: The prototype testing, those 15 types of conditions need to be modeled on the prototype to do 16 the test, but you'll be able to transfer that information to 17 the MOV. You need to have the right water temperatures and 18 the right DP across it.
19 MR. JACOBSON: He means the physical 20 condition of the stem. Do you have to ensure that it's the 21 same on the prototype as the one in the plant.
22 MR. ROSS: If in the type test you determine 23 through the type test that you have a valve factor on this 24 type valve ,to be X, have you got to determine what the valve 25 factor is on the in-situ test?
_ _ - _ _ _ _ _ _ _ _ _ . _ _ _ _ _ n
e.
115
, w.. Ik- .
I l MR. JACOBSON: You have to show similarity I 2 between the installed valve and the prototype valve. -l 3- MR. SULLIVAN: I think we talked a lot about 4 this this morning and I think, Bill, you were coming at this 5 this morning from a different angle. I think what we were E 6 saying was what we'were willing to accept at this point. And 7 we acknowledge that there's a certain amount of judgment in 8 what we were saying we would accept. And I'm not sure that 9 we can expand on that anymore than to say that, y'es, it would 10 be better to do full DP testing on all valves, then you 11 wouldn't have to worry about your question. I'm not sure 12 we're going to be able to shed anymore light on this subject.
13 We'll think about it some more between now and the meeting 14 next week. If we come up with anything different, we'll 15 reflect it in the meeting and in the summary.
16 MR. ROSS: In that case,I would just close 17 with a plea for understanding of our judgment.
I8 MR. SULLIVAN: Understanding of?
19 MR. ROSS: Of the judgment used by the
- 20 utilities when they're doing that because we ire still 21 discovering things.
22 MR. SULLIVAN: We understand that.
23 MR. KIESSEL: " Attachment A of Generic Letter i
24 89-10, Summary of comment, MOV deficiencies, Item 3, 25 Unbalanced Torque Switch. There are several conflicting k .- _ _ _ _ _ _ _ __ _ _ . . - . _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ . _ . _ _ _ _ - _ _ _ _ _ _ _ _ _ . . ___ . _.________._____________m. - . _ _ _ _ _ _ _ _ _ _ _ _ _ _ . _ _ _ _ _ _ _ _ . _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _
116 I
I- definitions as to what unbalanced torque switch means; e.g. l 1 i
2 torque thrust, tolerance, balancs. What is the NRC l J 1
3 definition of unbalanced torque switch in the Generic 4 Letter?" Do we have anyone here from Movats?
5 Great. What do you guys mean by unbalance 6 torque switch, because we pulled it right out of your list of 7 discoveries.
8 MR. CIESIELSKI: My name is Dave Ciesielski 9 and I work for Movats. Our definition of an unbalanced 10 torque switch related directly to the actual movement of the 11 spring pack in order to trip the torque switch at equal 12 settings between open and closed. And the boundary we put on 13 that was 100 millivolts for ten thousandths of an inch. And 14 it was strictly based on spring pack movement at equal torque ,
15 switch setting.
16 MR. JACOBSON: In other words, at a given 17 thrust in one direction, you would have a torque switch 18 setting, three in one direction--a balanced torque switch 19 would be one where a setting of three would give you so much
. 20 thrust in one direction and then an identical setting of the 21 three would give you the same amount of thrust in the other 22 direction. An unbalanced torque switch, the same setting 23 would give you a different thrust in one direction than the 24 other. s thE[ right?
25 MR. CIESIELSKI: That's basically correct.
C______________________ _ _ _ _ . _ _ . _ _ _ _ _ _ _ _ fl _
ii I
117 i
D j l Again, getting back to really--we're taking thrust out of~
. '1 2 this because of-that' ten thousandths of an inch. It could 3 still be considered balanced and not be'giving you identical 4 thrust in each direction. And I.think those that have been 5 looking at.this question of unbalanced torque switches 6 realize there may.have been considerable more.importance 7 placed on a balanced torque switch previous than we placed on 8 an' unbalanced or' balanced torque switch now. What 1 mean by 9- that is as.long as you're getting the thrust required-to 10 overcome the' differential pressure, you really don't care if Il the torque switch is set at one and a half open and two 12 closed or one and three quarters opened and one and three.-
13 quarters closed.
14 MR. JACOBSON: Previously it was a problem 15 because people didn't'do the testing and they didn't know 16 what they were getting. They assumed it was the samq in both 17 . direction.
18 Fa. CIESIELSKI: They were setting the torque
'19 switches to a numerical valve, two and two, so that's when it
. 20 became important. ;
21 MR. KIESSEL: Did that answer the question?
22 Great. Thank you very much.
23 Final question I've got is on page two of 24 Generic Letter 559-10. There is a statement, "INEL test has 25 concluded that a device measuring thrust in motor torque
---__---c__.___-_--m__ . _ _ _ - _ _ _ _ _ _ . _ _ _ _
I 118
. a 1 -would be best suited for..." and then it goes on..."is there l 2 a' typographical error in that statement, i,e., motor torque 3 should be. actuator. torque?" Yes, there is a difference 4 between motor torque and actuator torque. But what they were 5 aiming;at is that they felt that it's necessary to know the 6 thrust in the stem that is required and was being delivered, 7 and also the torque in the operator. Now whether you're 8 calling it motor torque or torque at some other point in the 9 gear train, prior to getting to the worm gear relationship, 10 would be a mute point. The idea being that you had some way 11 of measuring both torque and thrust would give you an ideal 12 system for measuring or for the diagnostic work. And this 13 was as opposed to seam of the diagnostic systems that only 14 measured thrust and some of that really only measured an 15 indirect thrust by measuring a torque or a position.
16 Did I answer that question for you?
17 (No response.)
18 Okay. That's the final one that I've got.
19 MR. SCARBROUGH: If I may make a statement 20 here. In Chicago we had a question come in from Duane 21 Arnold, and we hit all around it but maybe it would be good 22 to let you know what the answer was in that situation. The 23 question was, "Is it necessary to calculate opening and 24 closing dif,ferential pressures for valves whose safety 25 function is only one direction--in other words, opened or
- e p
119.
!: j 1 : closed--or.for. passive valves which do not change position-- [
, 2. in,other words, already in their proper. safety position?"
b 3- 3: And the answer to that is each MOV in a safety related-system 4 must be able to open, close or both depending on its safety 5 function or required position. All MOV's may not be tested.
,y 6 in.both. directions; however, if it's relied on in the design 7 . basis or in the emergency operating procedures,-then it must 8 be able to perform its function in that direction. 'So the
- 19 conservative approach would.be to go ahead and test in both.
10 directions. You don't have to worry about inadvertent 11- mispositioning and it's all taken care of. Eut that is how 12 you can'look at the. question'of MOV's and safety related 13 systems'and whether or no they need to be tested in one.
14 direction or the other.
15- Are there any questions on that?
16' MR. TAGGART: Repeat-your answer?
I7 MR. SCARBROUGH: Repeat it? Okay, MOV's in 18 safety related systems--don't worry about whether right now 19 whether or not it's classified as safety related valve or
.- 20 classified as a non-safety related maintenance valve that-21 happens to be in a safety related system. But any MOV that's 22 in that safety related system must be able to open or close 23 or both depending on the safety function. That would be a 24 safety r61ated=Dalve. Or it's--safety position. That's a 25 valve that just happens to be in the safety related system
120 !
! I I that may not have a safety function. So you need not test l I i 2 all MOV's in both directions. But if you rely on them in 3 your design basis for some function or in the emergency i l
4 operating procedures, then it must be able to perform its
- 1 D function in that direction. That may help you over the !
6 - mispositioning question.
7 MR. TAGGART: If the valve is only required O
8 to function in one direction, if you disposition it, then it 9 has to go the opposite direction, right? So you have to test 10 it in both directions because you have the potential of Il mispositioning it because it's position-changeable.
12 MR. SCARBROUGH: That's right.
13 MR. TAGGART: Which is just opposite of what 14 you just said I think.
15 MR. SCARBROUGH: No. If it has a safety 16 function in a direction, it must be able to perform in that 17 direction. If it has a required safety position, it must be 18 able to be in that position. To-get to that position. And in 19 this case if it's a valve that's just sitting there and not 20 performing any function and it's inadvertently moved to an 21 improper position, it has to be able to get back out to its 22 safety position, the position that allows that safety related 23 system to function properly. Say it's a valve in a dead-end 24 loop or a line of the system, it has no relationship on the 25 function of that safety related system, then we don't care
_ _ _ _ _ _ _ . _ _ _ . _ n
i 121 e ;
j l about it. It doesn't have a safety related position. You I e 1 2 don't care whether it's open or closed. An example we gave 3 in Chicago was the pressurizer spray valve. If it has no 4 function in the analysis of whether it's open or closed, you 5 don't care about whether it's open or closed, and so, you 6 don't need to perform all the review on things of that 7 nature. It can be opened or closed. That's a valve where 4
8 the operability, you don't care. So what we're just trying 9 to do is get away a little bit from the safety related 10 position-changeable dichotomy that is in the Generic Letter 11 because it can cause confusion as to what you're supposed to 12 do. You really don't need to look at whether it's safety 13 related or position-changeable. You can'just look at it as a 14 valve in a safety related system. If it has a function, if 15 it's normally shut and has to get open, you have to be able 16 to open it under the worst case condition. If it's a valve 17 that's always open and has no function, you never need to be 18 able to close it, you don't want it to close, it.
19 inadvertently closes, then you have to be able to get it back
, 20 out. That's all. We'll keep saying this and we'll say it 21 again in Rockville and we said it in Chicago, and that in my 22 opinion a more direct way of looking at the issue of which 23 MOV's are in the program and how you need to address them.
- ~ ~
24 MR. REYNOLDS: Bob Reynolds. In regard to 25 that, if you've got a valve that's supposed to be open, let's
- _ - - - - - _ - - - - _ ---- ----- - - - 1
I L
122 i
I say, and your torque switch--let's say your valve is not l j i ;
2 necessarily sized to go closed and somebody inadvertently j 3 closes it, you could trip out possibly on your overload 4 heaters, right?
5 MR. SCARBROUGH: Repeat the last part?
6 MR. REYNOLDS: If you're not sizing it down 7 to go closed and its position is to open, and somebody 0
8 inadvertently operates that valve, you could trip that valve 9 out on the overload heaters.
10 MR. SCARBROUGH: That's correct. We realize 11 that.
12 MR. REYNOLDS: So that basically, then, you 13 can allow some time frame to go in and reset that valve up 14 instead of from a control room standpoint in reopening that 15 valve.
16 MR. SCARBROUGH: Yes. The Generic Letter 17 assumes that if a valve is mispositioned inadvertently, it 18 gets there--okay, we're not worried about getting there. We 19 assume it's there. It's in that disposition position. That
. 20 disposition. We don't know how it got there. It just got 21 there, and now you have to get it back to its safety 22 position. So, yes, it could have tripped out. You may have 23 to go down there and flip it back to reset it, to be able to 24 move it back 65I. ~All the Generic Letter requires is that 25 it's down there in that disposition, you have to get it back fl j
a.
123-i-
'l: out to'a-safety position. -And that's where you.have to be [
2- able to' ensure the operability, to move it back to its proper 3 _ position.
4' MR. REYNOLDS: =And it does not have to be 5 Efrom a control room standpoint.
6- MR. KIESSEL: I think you're starting to push 7 into an area that we, ourselves, have not really fully 8 explored and just as we did on an early question,.I think the ..
9- best: answer to-your question is that we're going to have to 10 think about it and get back to you as soon as we can.
II - Hopefully by the next meeting.
12 MR. JACOBSON: We had'a lot of discussions on 13 'that particular scenario.-
l4 MR. KIESSEL:- You're starting to get into 15 grounds that were not really considered. You raise a very.
16 valid question.
l 17 MR. JACOBSON: You may burn the motor out if l
-18 you don't have a torque switch or thermo. overload.
19 MR. KIESSEL: And the next question is, how
.- 20 long are we going to give you to change out the motor?
21 MR. JACOBSON: We have to revisit that. l 22 MR. REYNOLDS: Okay. Thank you.
l 23 MR. SULLIVAN: This is something that we're 24 going to spend some time this afternoon, definitely on ,
25 Monday, and to the extent that we can define this more I
i w____-_____- .- _ n'l
p 1
124 l
I clearly, we will on Tuesday. Nevertheless, we will flush it !
i 2 out as clearly as we can in the meeting summary. We realize l 3 this is a question that the Generic Letter isn't clear on and 4 we need to think about it some more, 5 Clive, you had a question.
6 MR. CALLOWAY: The justifiable ways to block 7 a valve from inadvertent operation, as far as the electrical 8 methods, last Monday I think you said that a status board 9 alarm was an acceptable way of electrically blocking the 10 valve.
Il MR. K1ESSEL: No. It was in lieu of a ;
12 mechanical lock on the valve. Would we accept the status 13- board alarm that was always available as an indication that 14 somebody had mechanically tampered with the valve.
15 MR. CALLOWAY: I see.
I j 16 MR. KIESSEL: And if I remember the question, )
i 17 it also had a separate electrical block so that electrical 18 operation was also prohibited.
19 MR. CALLOWAY: Right. Okay.
. 20 MR. KIESSEL: You know, somehow that you know !
21 there's something going on and somebody made the mistake and l 22 you immediately get it back, correct that, I think was the 1 23 intent.
1 .
' ~
24 ~~~MR. CALLOWAY: Thank you.
25 MR. SULLIVAN: Are there any other questions?
R_
f H 125 t
L 1 (No response.) l
> ,- i 2 ; Okay. I think we've sort of_ identified one [
3 area that we're going to be working on, in' addition to t
~4 preparing for the questions for the_ Crown Plaza meeting. The
$ 5 question that we've had some dialogue with Bill'Ross on.
6 MR. KIESSEL: We're going to stay away from 7 the semantic aspect but you raised a very valid question.
8 What are we looking for in the comparison between the valve 9 that was tested atemax conditions and the valve that's in the 10 plant.
ll MR. JACOBSON: These are the three items I 12 think we were unclear on. One of them concerns, what are the 13 acceptable test methods for the valve that is prototype 14 tested in the plant? Do you have to do it a some DP or can 15- static DP be used? And also on the repetitive tests after 16 the first one, can those be done at static DP? I think we're 17 going to talk a little bit more on that.
18 The second question is the thing we just 19 talked about which is, do you have to ensure that a valve can 20 be mispositioned without being damaged in order to be 21 repositioned to the proper position? I think that's a good i 22 summary of it.
23 The third question is--this is for a safety 24
- related valve--when looking at your design basis review, do 25 you only have to look at the conditions in effect at the time L
126
~
I I that it's assumed that valve will be stroked in your design !
i 2 basis, or do you have to account for stroking at any. time due l 3 to the fact that it could be mispositioned at any time? I 4 think that's the third area of clarification.
5 MR. KIESSEL: Does that agree with what you 6 think we're taking home?
7 MR. JOHNSON: One other item I'd like to 8 revisit is when you do your testing inside, your DP testing 9 inside_your unit--prototype testing you might say--what 10 degree of DP testing do you feel is required for the other 11 valves of identical service inside that plant?
12 MR. JACOBSON: That falls under the first 13 one.
14 MR. JOHNSON: Because it really doesn't fall 15 into prototype testing because you're using same valve line 16 up.
17 MR. JACOBSON: When I say prototype, I've 18 been using that to mean either outside or an identical one 19 inside.
, 20 MR. KIESSEL: Perhaps a better way of 21 phrasing it is what type of testing do you need to do to 22 apply a full design basis test on one valve to another valve.
23 MR. JOHNSON: Right.
^ ~
24 MR.~ JACOBSON: Whether it's in the plant or 25 out of the plant, it's going to be the same relationship.
n
127 I MR. JOHHSON: Well, I think that there's a 2 greater degree of quality inside your plant. You're working 3 underneath your quality program.
4 MR. JACOBSON: You better be working within a 5 quality program if you're doing it outside of the plant, too, 6 or else the data's not going to be that valid.
7 MR. JOHNSON: But also you're using your r
8 pump, your piping system, and your valve right there in your 9 plant. And so, if you have identical valves that are lined 10 up and parallel, you're bound to get better results I 11 believe. Anyway, if you could look at that area.
12 MR. JACOBSON: We'll look at it.
13 MR. JOHNSON: My name is Henry Johnson.
14 MR. SULL3 VAN: They're the three areas that 15 we'll try and flush out and clarify in the next transcript.
16 You can look for them there. They will also be addressed in 17 the meeting summary. And if there aren't any further 18 questions, it would appear appropriate to adjourn the l9 meeting. Thank you very much. ;
20 (Whereupon, the above-entitled matter was 21 concluded.)
22 23
. ..=.
24 -
< 25 L
i f
_ _ _ _ n I
l l CERTIFICATE 2
This is to certify that the attached proceedings before:
NUCLEAR REGULATORY COMMISSION 3 In the matter of: GENERIC LETTER 5
s ,
At: DENVER, COLORADO Date: SEPTEMBER 20, 1989 0
were held as herein appears, and that this is the 8 transcript thereof for the file of the department.
8 DJ I 9 M4Oflic'ial &/
Reporter 4
11 12 13 14 15 16 17 18 19 e 20 21 22 23
. =. .
24 25 i
_ _ _ _ _ _ ____-_-__________________-___-_---_-_m