ML20149D511
| ML20149D511 | |
| Person / Time | |
|---|---|
| Issue date: | 01/05/1988 |
| From: | Advisory Committee on Reactor Safeguards |
| To: | |
| References | |
| ACRS-T-1632, NUDOCS 8801130016 | |
| Download: ML20149D511 (131) | |
Text
{{#Wiki_filter:Ad/24'7~ /432 ~ UMIED STATES " q, g ~ NUCLEAR REGULATORY COMMISSION O-IN THE MATTER OF: DOCKET NO: REACTOR OPERATIONS MEETING O LOCATION: WASHINGTON DC PAGES: 1 - 129 DATE: JANUARY 5, 1988 jpp;Q p":r gg p q qtr ' % ;r , : k,.f 3 g & s 3010IliGIT10V8 [iom EnR; S 0a1c0 ir m O 4 Heritage Reporting Corporation N\\ Of0cial Reportres \\ 1220 L Street. N.W. ) wanungton. D.C. 20005 (202) 628 4444 ( 8801130016 880105 PDR ACRS T-1632 DCD
-1 PUBLIC NOTICE BY THE 2 UNITED STATES NUCLEAR REGULATORY COMMISSION'S 3 ADVISORY COMMITTEE ON REACTOR' SAFEGUARDS O 4 5 6 7 The contents of this stenographic transcript of the 8 proceedings of the United States Nuclear Regulatory 9 Commission's Advisory Committee on Reactor Safeguards (ACRS), 10 as reported herein, is an uncorrected record of the discussions 11 recorded at the meeting held on the above date. 12 No member of the ACRS Staff and no participant at 13 this meeting accepts any responsibility for errors or 14 inaccuracies of statement or data contained in this transcript. O 15 16 17 18 19 20 21 22 23 24 0 2' Heritage Reporting Corporation (202) 628-4888
1 UNITED STATES NUCLEAR REGULATORY COMMISSION 2 ADVISORY SUBCOMMITTEE ON REACTOR SAFEGUARDS 3 s ) 4 In the Matter of: ) ) 5 REACTOR OPERATIONS ) G
- Tuesday, 7
January 5, 1988 8 Room 1046 1717 H Street, N.W. g Washington, D. C. 20555 10 The above entitled matter came on for hearing, 11 pursuant to notice, at 1:00 p.m. 12 ACRS MEMBERS PRESENT: k_,,/ I3 MR. 7 ESSE C. EBERSOLE Subcommittee Chairman 14 Retired Head Nuclear Engineer Division of Engineering Design 15 Tennessee Valley Authority Knoxville, Tennessee 16 MR. CARLYLE MICHELSON 17 Retired Principal Nuclear Engineer Tennessee Valley Authority 18 Knoxville, Tennessee, and, Retired Director, Office for Analynic and Evaluatior 19 of Operational Data U.S. Nuclear Regulatory Commission 20 Washington, D. C. 21 MR. CHARLES J. hYLIE Retired Chief Engineer rw 22 Electrical Division U Duke Power Company 23 Charlotte, North Carolina 21 25 Acme Reporting Compony .na, u.....
2 1 ACRS COGNIZANT STAFF MEMBER: 2 Herman Alderman 3 NRC STAFF PRESENTERS: 4 Tad Marsh Ted Sullivan 5 John Huang 6 7 8 9 10 11 12 O 13 U 14 15 16 i 17 18 j 19 j 20 21 o o. 23 21 25 Acme Reporting Company ,,o,,.>.....
3 1 MR. EBERSOLE: The meeting will now come to order. 2 This is a meeting of the ACRS Subcommittee on 3 Reactor Operations. ,3(-) 4 I am J. Ebersole, Chairman of the Subcommittee. 5 The ACRS members in attendance are C. Michelson and C. Wylie. 6 The purpose for this meeting is to be briefed and 7 discuss in-service testing at nuclear reactors. I believe 8 it pertains especially to valves and pumps. The topic NRC 9 action underway scheduled for 2:45 p.m. will be closed to the 10 public because the staff is anwilling to discuss this in an 11 open meeting. 12 lierman Alderman is the cognizant ACRS staf f member o 13 for today'c meeting. j j 14 The rules for participation in today's meeting i I 15 have been anncanced as part of the notice of the meeting 16 which was published in the Federal Register on December 22nd 17 1987. 18 This meeting is being conducted in accordance with 19 the proviuions of the Federal Advisory Committee Act and the 20 Government in the Sunshine Act. 21 We have received no written statements or requests 22 to make oral statements from members of the public regarding (m ) w/ 23 today's session. 2/, it is requested that each apeaker first identify 25 himself or herself and speak with clarity and volume co that Acme Reporting Company
4. .R..' I he or she can be readily heard.- 2 -I don't have any comments to open the meeting. I 3 would like to ask the-other. Subcommittee members-- '4 MR. MICHELSON: Is:there some simple' explanation as 5 to why we have to close part'of the meeting? 6 MR. MARSH: Yes, there.is. There is some pre-7 decisional information we want to discuss. 8 MR.'MICHELSON: Predecisional? 9 MR. MARSH: That's right. 10 MR. MICHELSON: 'We. discuss predecisiona1 in' forma- ~ f 11 tion all tho' time ~it seems like to me. This is wh'at we might 12 be doing. This is~what we think we are going to do. 13 MR. MAR 3H: Well, there are'some actions underway .i 14 that I don't think are appropriate to talk to the public 15 aboute 16 MR. MICHELSON: These are not' personnel ~ kinds.of 17 actions. I mean that I would understand.- But predecisional 18 in the sense'of what we might want to do here is somehow-- ~ 19 MR. MARSH: No, there are some actions underway l i 20 within the staff that have not yet received management con-s 21 currence that are not appropriate in my judgement to discuss 22 - with members of the public here. O MR. MICHELSON: Just a matter of curiosity. I don't 23 i 24 have strong feelings about it. l 25 MR. EBERSOLE: I.think, Carl, that--I gather that i l 9 Acme Reporting Company [ ~ ,,o,,.,.u.. -... - -. s ~. - - - -.
5 1 ve are getting in at an early stage of things which I endorse 2 before things get more congealed. 3 MR. MICHELSON: That part is fine, but I wac just a f i \\_/ 4 little curious as to what the sensitive area was that we 5 were trying to protect. 6 MR. MARSH: Well, maybe you will-- 7 MR. MICHELSON: Maybe I'll understand later wher. I 8 hear it. 9 MR. MARSH: Maybe you'll understand further as we 10 go on. 11 MR. EBERSOLE: Just to start with, I think I have 12 the current view that valve testing is already inadequate, )) 13 and I gather that the topic may well be that whatever we do 14 already is difficult, and it's with some reluctance that I 15 have the feeling that we-~I may act in my standard role as a 16 complicator, Carl. 17 MR. MICHELSON: Well, thin Section XI testing has 18 very little to do with our real concern and that is whether 19 valves can perform their function under accident conditions. 20 It has very little to do with-- 21 MR. MARSH: Maybe that will be clarified for you 23 too, but you say it's unrelated. (3! s _- 23 MR. MICHELSON: Well, I know what Section XI does. 24 It's been established for a long time. Now, what it might do 25 in the future might be a different matter. Acme Reporting Company .sv,........
6 l' MR. MARSH:- What it's intended to do is another-2' matter. 3 MR. EBERSOLE:- Well, anyway, let's proceed with the 4 . presentations and-- 5 Ted Marsh, you've'got.the' floor. 6 MR. MARSH: I'm Ted' Marsh. ' And on my right is Ted 7 - Sullivan and on my lef t is John IIuang.- i 8 We'd like to discuss with you today the state of 9 IST, where things are within the~NRC, our perceptions.of 10 problems and our plans for resolving these problems. 11 Let me just go ahead and start. 12 By way of introduction, the objective of in-service () 13 testing of pumps'and valves is that together..with the techni-14 cal specification requireme'nts, they ensure the operational-l 15 readiness.of cafety-related pumps and valves. 'And when I say i ~ 10 together with the technical specifications they go hand in 17 hand. There is an expl'icit reliance in the technical speci-l 18 fications to do'in-servicestesting in accordance with Section ) 10 11 requirements. 20 MR. MICHELSON: What do we mean by operational 21 readineso? ~ 22 MR. MARSil: They are ready'to perform their. safety-23 related function. 21 MR. MICHELSON: Now, by that you mean if the pur-25 pose of the valve is to isolate a pipe break, then this-test Acme Reporting. Company uo,,.,..... -m-----i- ..-._.,,m ..,...,,,...,_. - _..._,.y.-
r 1 periodically' confirms that that valve-could isolate a pipe 2 break if a break were to, occur? 3 MR.' MARSH: The purpose of the test is to ensure A+ 4 that valves are supposed to do their safety-related function. A 5 There are inadequacies in the program, many of'them, and you 6 have hit on one of them. i 7 MR. MICI!ELSON: In fact, isn't it true that the only 8 purpose which really is obtained is the valve.is:just' as; good D' as it was when you started up uith it?- It certainly is no -l 10 better? 11 MR. MARSH: You-are going to learn aflot more-12 about valve testing and how.it'sidone. () 13 MR. MICHELSON:.Okay. Well, maybe~we'll' hear it' 14 all and I'll just reserve all my comments'for the end. 15 MR. MARSH: Don't do that. 16 MR. MICHELSON: I have a ' lot of tlusn. 17 MR. PARSH: But I do want to leave you with the 18 impression, the strong impression, thatlthe objective is-to s 19 ensure that these valves.arui pumps are able to do their 20 safety-related function. And that means to pass flow if 21 they are supposed to. Isolate if they are supposed to. Pass 22 the right flow from a pump into a system. They go hand in . O 23 hand with the test specs and they have as their bases the 24 safety analyses. 25 A little more background. In-service testing is Acme : Reporting. Company i m,........ a. -~..,a
8 1 required by the regulation under 10 CFR 50.55A. It requires 2 that pump and valve testing programs be constructed in ac-3 cordance with the ASME Code, Section XI. The regulations x> 4 also require that IST programs be updated every ten years to 5 the latest code and addendum. And also the regulations ex-6 plicitly allow for relief from those requirements which are 7 impractical and the regulation is not clear on what exactly 8 impractical means. 9 Let me, before I get too much further, I want to 10 go to the next slide, and I'm going to come back to the status 11 but I want to go to the next slide to put into context testing 12 of safety-related pumps and valves. ,,(J 13 We tried to characterize testing that is done on 14 pumps and valves and we put them into basically three bins 15 baced on the time frame in which they are done. 16 There is the initial and pre-service testing phase 17 of pumps and valves. And that may take place in the shop, 18 in the pre-operational mode of the plant, in the functional 19 testing of a plant, that is, the cold or the hot, during the 20 startup or the power ascension phase. That portion of valve 21 and pump tenting in important in that it verifies that what l l ( 22 you have bought and inctalled is what you want it to be. J ,3 LJ 23 And that it is performing its function properly. We see that 21 by octablishing initial reference values. 25 MR. F:BERSOLE : Let f.io ask you a question about that Acme Reporting Company .,0,,
9 l' phase.of testing. Is at that time, and maybe only.at;that -2 time, you can confirm that the_ valve,_if it's a valve,-or.may-3- be a pump, will do what11t is supposed to do'under emergency . f') Ev 4 duress conditions ~that you will never really: experience in 'T 5 the life of the plant? 6 MR. MARSH: It may be. Very special tests daring. ~ 7 the initial phase of the plant that you may never duplicate. .R. EBERSOLE: And then you index those tests. You i 8 M 9 have to get some parameters out of them~that you can verify 10 continuity of in the in-service program, don't you? 11 MR. MARSII: That's the initial reference values. 12 MR. EBERSOLE: Yes. () 13 MR. MARSH: But I don't want you to'think that 'i 14 these three tests, these three testing periods--that's the 15 initial, the throughout life, and the as required--together 16 all verify that those pumps and valves are able to do their E.. safety-related function. -There are weaknesses here_and we 17 18 are going to talk more about them. 19 MR. EBERSOLE: Okay. 20 MR. MARSil: So you do some special tests'during 21 the initial phases of the plant but it is not complete. 22 You do other things during these tests. You also 23 look for systems interactions, any fibra'tions, or any system 24 peculiarities that could have happened during construction. 25 I am going to skip to the last one. The "as Acme Reporting Company ,,os,.,..... .r ---.r,- ,,Z. ...-,rv ....r,,4.-,-,,,,,-
~ 1 i 10 ~1 required" tests are any tests that may be done because of 2 maintenance on pumps and valves, any special NRC initiatives, 3 -any. vendor or NSSS requirements, go check to make sure that 4 this pump.is._ installed properly, things thati are initiated ~ 'S not from.the licensee. 6 And then what.I think is the major' block in' terms 7 of testing and in terms of importance is the "throughout-8 life" test. That's the conter part. That isLthe routine 9 in-service test and technical specification test that verify 10 that throughout the life of.this plant that the system and 11 the pump is able to-do its intended job. That is, to miti-12 gate events or safely shut down'the reactor. 13 So I wanted to put into context testing, in-service 14
- testing, where it is, how it's done, in terms of time i
15 throughout the plant. ;You have first the initial pre-service, i la then the throughout life. That is the life of the plant. 17 ' And then some routine as required testing. Not routine. i 18 As required. 19 Go back to the first slido. Now, the status of 20 the in-service testing programs at plants. 21 The bullet says that there are only approximately. 22 thirty plants that have received safety evaluations for their 23 in-service testing program. And many of those SERs are no 21 longer current. And what that means is that first, everyone 25 has an in-service testing program. Okay. I don't want this i -l Acme Reporting Company
+ 11' 1 -.to-imply to you that there are. plants that don't have pro-- ~ 2. grams.- But I do want you to understand that ~there are r.:c.y 3 plants, the bulk of plants have not received explicit review .O v 4 and safety evai.uations. 5 .MR. MICHELuGN:- Are there explicitly stated re-6 quirements that these programs are: supposed to meet? 7~ MR. MARSH: There is IWV and IVP. 8 MR. MICHELSON: No, I mean the regulatory require-9 ments. 10-MR. MARSH: Well,'that is the regulatory require-11 ment. 12 MR MICHELSON: There is the Standard Review Plan .( ) 13 which I gather in the regulatory requirement. Is that right? 14 MR. MARSil. 50.55A endorses IWV and IWP, so it is 15 a regulatory requirement. Also the Standard Review Plan. 16 The Standard-Review Flan though--I don't know how much you've 17 read it. It's not a very explicit document. 18 MR MICHELSON: Well, I've read it. It doesn't 19 take too long'to read it. 20 MR. MARSH: No. And if you read IWV and IWP it 21 won't take you long to read that either. 22 MR. MICHELSON: It doesn't take long to read these 23 either and they are not much further help when it comes to l ~ 24 accident or adverse loading conditions and so forth. 25 MR. MARSH: They are not. Acme Reporting Company aa,,.>....
12 1 MR. MICHELSON: They just are not addressed in 2 here. It's hard to'tc11 if they go beyond what I'd call ~ 3 normal loading in the in-service-inspection program. 4 MR. MARSH: You can't tell-by that. 5 MR. MICHELSON: They don't address at all such 6 problems as the maintenance procedures that are used. In 7 other words, the adjustment of torque switches and'this sort 8 of thing. 9 MR. MARSH: Absolutely.not.- 10 MR. MICHELSON: Which have a vast'effect upon the. 1 11 ability of the valve to respond to a particular loading 1 12 requirement. 1 () 13 .MR. MARSH: Absolutely,right.- -14 MR. MICHELSON: They are justDsilent in'all of 15 these areas. 16 MR. MARSH: There are many parts of that requirement- ~ 17 XR. MICHELSON: Yet the Standard Review Plan says l 18-as long as you meet the code, that's all you need to do. j 19 MR. MARSH: Yes, sir. 20 MR. MICHELSON: That's what it says'very expli-j 21 citly here. 22 MR. MARSH: I will not disagree with you at all. ] 23 MR. MICHELSON: In fact, it says that if.you meet a 21 the code you also satisfy criterion 37, 40, 43, 46, 54 Like 25 this is all you need to do. And it simply is not adequate. Acme Reporting Company l . m........
13 1 MR. MARSH: That's right. That's right. 2 MR. EBERSOLE: We're going to hear how that's going 3 to be fixed, Carl. (-(_) 4 MR. MICHELSON: Yeah? I'm not being very patient, 5 I know. 6 MR. MARSH: That's all right. Okay. 7 MR. MICHELSON: But I thought 6.he purpose of the 8 meeting was to have a dialogue of-- 9 MR. MARSH: The purpose is to tell you problems, 10 okay. The purpose is to tell you problems, and you can-- and 11 we'll raise problems to you and you can--I do mean this. We 12 do need feedback on these problems and the perceptions that () la we have and the course that we're taking. And that's really ns 14 what we are after. 15 MR. MICHELSON: You are going to tell us later I IG guess in this closed session the course you are taking? 17 MR. MARSH: We're going to tell you the course 18 that we are on right now. That's right. l 19 I do want to leave you though with the terms of i 20 the number status. That is, the number of plants that have 21 received explicit review is few. And that's only thirty of cs 22 the approximately one hundred. And I say many of them are I b 23 no longer valid or current. What that means--the current l i 21 state of affairs is this. A plant works on an IST program. 25 They develop a program. They submit it for review. It has ] Acme Reporting Company a u,.,.
4 14 1 taken us many years to refenicate and in many cases we have -2 never completed the review of.that program. There are some-3 plants that are in their second intervais without'ever having ~ 4 received this SER in their first interval. 5 MR. MICHELSON: Well, having completed al review 6 doesn't seem to give ne much comfort though because all I 7-guess you can do is review with the Standard Review Plan and 8 all it says is you are supposed to meet the code. And so you 9 read the code and as long as the vendor--I mean as all as the 10 licensee meets the code, that's all you can check against. 11 MR. MARSH: I do take comfort in the review. There 12 is something beyond what you see there. (])- 13 MR. MICHELSON: I know that there are serious-in-14 adequacies in the code. 15 MR. MARSH: There are inadequacies there and many 16 of them are--when you do the review, you can pick up these 17 inadequacies in the code. 18 MR. MICl!ELSON: But you can't do anything about 19 them. 20 MR. MARSH: Well, it depends on how you read that 21 code. 22 MR. MICIIELSON: Lecause you meet the Standard Re-23 view Plan by meeting the code. 21 MR. MARSil: There are some things that you can take 25 care of. I don't want to be in the posture of defending Acme Reporting Company y .,o,,.>.....
15 1 excessively where we are, I do war.t to say though that the 2 program review has merit. ' mean we do review the programs 3 and many of these things we do pick up. There are inade-n (' / 4 quacies there. There are ir. adequacies there and we are on a i 5 curve trying to get healthy. And there are things that need 6 to be improved there and the Standard Review Plan needs to be 7 improved and the process needs to be improved. There's a 8 whole spectrum of areas that have to be improved. 9 And I don't want to dwell on any one as the cource 10 of the problem because it's a many-area problem. 11 MR. MICHELSON: Are you also going to tell us a 12 little bit about the escarch programs that perhaps got to 13 be done before one can really put together this kind of a (a) 14 program? 15 MR. MARSH: No. I'm not going to tell you about 16 research programs. 17 MR. MICHELSON: Pecause I'm not sure you have the 18 data base with which to draw-- 19 MR. MARSH: I'm going to tell you about requests we 20 have made of other organizations to help us put in the right 21
- programs, f3 22 MR. MICHELSON:
That would be helpful. U 23 MR. MARSH: And the complaints that we've made to 21 the code and what they are working on, but I can't tell you 25 explicitly what some of those other programs are. Acme Reporting Company ~u.~.
^ l 16 1 MR. ESERSOLE:- In the thirty-plants.that have.zo-- 2 ceived SERs, I_just deduced that--you think'that's certainly. 3 better than the plante that have not received SERs.: 4~ MR. MARSH: That's~right. 5 MR. EBERSOLE: Is there in fact ~meanintiful.dif-6 forences between those thirty plant. programs thhn'the'ones 7 that are not yet-- 8 MR. MARSH: Let me ask Ted to answer that, because 9 he's-closer to those exact ~ programs. 10 MR. SULLIVAN: Well', you can' look at thatlin a i 'l number of different ways. One is that alth'ough' thirty planti l 12 have received SERs, and most of the rest or maybe fifty of () 13 the other seventy are' 'in some : state of review. Where we are 14 interacting through our contractor with the-utility and try-15 -.1g to get them to understan'd, albeit that the code is 10 weak, what is meant by that code. That it is a component 17 test, for examp3c, and.that.a system test doesn't necessarily 18 test different components. i l 19 So I don't think there is as much diffence between 20 the thirty and the seventy as it might appear because we are. 21 working on many of the' remaining-- 'l 22 MR. EBERSOLE: In fact, I was thinking it might 23 be sort of an inversion principle, that the SERs that have i 24 been issued were issued under cuch compromised conditions 1 25. without actual real requirements, you know, to make the tests Acme Reporting Company
1 17 I do what they should do. That it is implying a blessing on 2 an' inadequate-set of programs'in the thirty that you've.is-3 sued. ()- 4 MR. MARSH:- I think you are right. ~ I don't'want 6 t.o--thirty is better than zero. 6 MR. EBERSOLE: It may be wo'rse. '7 'MR. MARSH: What you are seeing here is a new 8 management philosophy that's here too.- This is a new empha-9 sis on IST that hasn't-really been there before. And I think 10 'IST programs have improved a lot within the last~six or 11 eight months, the ones that we have reviewed. That's not to 12 say that programs in the past were inadequate. But I'think-() 13 that there have'been improvements in those'progr'ams.that 14 we have reviewed. 15 MR. EBERSOLE: In this connection, I attended this l 16 Orlando meeting of the maintenance executive group not long 17 ago and one of the outputs of a good program like this is to 18 develop the information for--not preventative maintenance-but l 19 predicted maintenance. 20 MR. MARSH: Absolutely, absolutely. 1 l 21 MR..EBERSOLE: Which is certainly going to be the 22 road we are going to follow. 23 MR. MARSH: Absolutely. And I was in-St. Pe'ers-24 burg, Florida, talking to the O&M Committee about maintenance 25 programs and the increased NRC and industry attention towards l Acme Reporting Company . m..a.....
l' - I 18 , plant life extension'and maintenance,,and;it behooves them 2 to come up.with good IWV and IWP requirements so that.they 3 can legitimately look'for degradation in their equipment. () 4' MR. EBERSOLE: It not only is going to be better. 5 It's going to be more economical to do predicted maintenance. 6 'MR. MARSH: Absolutely. 7 The'SERs that we say.are out of date is because-of 8 another chink.in the system. And that is that plants can 9 chcnge their in-service testing program without the NRC ex-10 plicit safety evaluation itself. Although the technical' 11 specifications say don't implement any= relief requests, any 12 relief frcm the code requirements until.you've gotten.NRC O 'a "ve' "" - 14 And then another problem in this process is.that' i 15 there are, of the remaining seventy, there are'approximately-10 only thirty-two that have been given explicit interim relief. 17 That means that they have been given a short letter that says 18 that it's okay to implement these requirements--implement 19 these relief requests although we haven't finished the review - I 20 yet. 21 Interin approval is a pill. It's a way of getting 22 you past a hurdle and over a more basic problem and a more k y baF*c problem iS procesD roquirements, resources and trying gg te get programs reviewed. 25 Okay. What I do want to leave you with from my Acme Reporting. Company i,o,,...o..
P 1 19 1 part of the presentation at,the beginning though-is.the im-2 portance of IST.. That lt'has as its basis safety-related 4 3 pumps and valves. I want to leave you that it is required ~ by regulations and that there arefat least!-administrative 4
- 5 5'
deficiencies in the. number.of= plants that have received re - 6 views an'd there-are those that'are implementing relief re- ~! 7 quests without explicit NRC approval although they'should have 8 it by technical specification requirement. 9 John Huang is the next speaker. He is going-to 10 talk in roore detail about programs and how the review is done s 11 and some of _the explicit requirements on 'pmnps and valves. { ) 12 MR. HUANG: I am John Huang. I'm currently one () 13 of the many IST program reviewers in"the Mechanical Engineer-14 ing Branch. And today I am going to go through briefly on f t 15 how the reviewer is doing his reviews. l 16 And you can turn to the next slide which summarizes 17 the IST review process. 18 As you probably know, IST review consists of'two 19 parts. The first part of the review is to make sure that I 20 all safety-related components, in this case, pumps and valves, i i 21 are covered in the program. i I 22 The second part is to make sure that all. test--all 23 .of the required tests are performed' correctly in accordance 24 with ASME Section XI requirements. We may allow comments on 25 the adequacy of that requirement, but today as a reviewer, t Acme Reporting Company .ao,,. l
~ 20 1 we'll try to follow what is given to us'. 2-And in'the second part of the review,'if the'licen-3 see cannot meet the' testing requirements as specified.in the
- p'
' s-4 code, they will'subm'it requests for relief and sometimesLthey ~ 5 .may propose alternative-tests-so the reviewer will review the 6 adequacy and= acceptability;of those relief requests-and/or '7 alternative test-method.- 8 Now, if you look at the1 slide--the first about the ) 0 review is to identify what system we wanted to be considered 10. in the program..And.to do a-good review on.that, the're-11 viewer, in my judgement, must have very good understanding 12 of the system, the functioning of the system, and associate () 13 safety analyses. 3 14 MR. EBERSOLE: Are we just talking now about mecha-15 nical fluid systems, air systems? 16 MR. HUANG: Most of the systems we are reviewing 4 17 today is the fluid system. 18 MR. EBERSOLE: That doesn't mean air systems. 10 .MR. HUANG: Yes. The air system currently by 20 terminology is not covered because they are associated with j 21 compressors. 22 MR. EBERSOLE: This is mechanical enginee' ring,. .() 23 valves and pumps, handling liquids. That's what we are-talk-24-ing about today. 25 MR. HUANG: Correct. j 1 l i l Acme Reporting Company , m...... m. .._,_7..,.... -m.,,- .y w.~ ...~,v---,
21 1 MR. MICHELSON: I never sensed that in reading the 2 Standard Review Plan. Did I miss it somewhere? I thought 3 they--the air system, if it's safety-related air system, its i ? 4 valves get in-service inspection just like anything else. 5 MR. HUANG: As far as the valve is concerned, as 6 a reviewer now, I'm concerned, will include this valve. 7 MR. MICHELSON: Well, I thought compressed air 8 systems were always included if they were deemed to be safety 9 related. 10 MR. MARSH: What John is saying is that those 11 portions of the air systems that are safety related are in-12 cluded in the program. That rarely involves the compressor. n() 13 MR. MICHELSON: And have always been. 14 MR. MARSH: I can't say has always been. I can't 15 say that. 16 MR. MICHELSON: Well, if it's an essential com-17 pressor, it's in the program too. 18 MR. MARSH: If it is a safety-related compressor? 19 MR. MICHELSON: Yes. 20 MR. MARSH: No. 21 MR. SULLIVAN: The compressors are not in the 22 program. ,q v 23 MR. MICHELSON: I sure don't get that out of the 21 Standard Review Plan. Nor do I get it out of the code, for 25 that matter. Does the code explicitly rule out air? Acme Reporting Company . u..... 1 i
~ 22 .1 -MR. MARSH: No. + 2 MR. MICHELSON: I didn't think it did. ~ 3 MR. MARSH: No. Os.. 4 -MR.-MICHELSON: Then this'is something that's 5-neither in the Standard' Review Plan or in the code, that's. 6 done by the reviewers. t 7 MR. HUANG: We need a clarification for the re-8 viewers there because what I' inherited from the previous .i .+ 9'- reviewer, we do not' consider compressors.. i i 10 MR. MICHELSON: I'm not surprised,-but on the other. 11 hand I don't'see~how it implies what we think as the program, f i 12 the Standard Review Plan. O ta xa no^"o= roaer e* 1eeet we'11 erv to cover 1 14 the valves in-the air system, to clarify that. And if we 15 feel there is a safety-related;funct, ion in the air system 10 for that particularly valve, we include it'in the program. 17 MR. MARSH: You've got to help me. I: don't think 18 there are very many safety-related air compressors. I think-10 there is instrument air and service air andJthere are cer-20 tainly portions of there systems that are safety related and 21 those are included in the program. 22 MR. MICHELSON: Some of them have safety-related S.h 23 air compressors. 24 MR. MARSH: Compressors, okay. 25 MR. MICHELSON: Oh, yes, oh, yes. .The seismic, the Acme Reporting Company i,o,,.......
m 23 i i 1 whole' bit. ' Depending on'the plants. l 2 MR. MARSH:- Okay. 3 MR..MICHELSON:.And how they happen to arrange 4 .them. 5 MR. EBERSOLE: That wouldn't include refrigeration t 6 systems. 7 MR. MICHELSON: It might. .I didn't want to compli-8 cate the issues by-getting into chill-water systems and 9 freon. systems,-but, yes, I'think--they have pumps too. ' Ab- { 10 solutely. 11 MR. EBERSOLE: Well, everything is interrelated. 12 We have a big flap going.on on the matter of the influence f l () 13 of control systems on safety systems. 14 MR. MARSil: Right.- 1 15 MR. EBERSOLE: And here you'run right straight into 10 it. -!n 17 MR. MARSH: Right.. i 18 MR. EBERSOLE: And one of the first things you run j 19 into, and I notice it's not covered in here, even the test'ing j j 20 program is never mentioned. Whether in examining a pump or i 21 a valve or both in combination with each.other that you j 1 22 verify the extremes of operation that are possible with a () 23 pump such as--run out operation, i i l i 24 MR. MARS!!: You do not, i l I 25 MR. EBERSOLE: Or stall flow zero conditions'. I Acme Reporting Company .>,,sa.....
24 1 MR. MARSH: You shut off that. You do not. 2 MR. EBERSOLEt And yet it's fundamental.in examin-ing the interrelationship with control of safety systems, 3 4
- you determine the ultimate potential of pumping pressures, 9
5 or wl.atever. 6 MR.. MARSH: - The code assumes that--you. test those 7 systems initially throughout its range and you shut off ' 8 within that flow initially. And it assumes that-if you can \\ ~ ~ it's either far out.on the 9 establish a reference valued, 10 curve or near up on the curve with some flow. That if there 11 is degradation, you'll detect it there. But you don't have 12 to test the pump throughout its entire flow range. () 13 MR. EBERSOLE: But you do have'to know what it's 14' capable to do. 15 MR. MARSH: You do that and you should do that as 16 initial-- l 17 MR. EBERSOLE: So all pumps are given a test at 18 zero flow and a rile. 19 MR. MARSH: Initially. 20 MR. EBERSOEL: That's what I'm asking, i 21 MR. MARSH: Through the life of the plant. 22 MR. EBERSOLE: And then you have to infer'whether i () e 23 that is retained as a characteristic. 21 MR. MARSH: You have to assume, you know. In many I 25 cases that's not a good assumption. Acme Reporting Company .au.s...... m. ,,,,__,,_,.,,.,.,y-.....,,,_ ,m_
~ 25' ~ 1-MR. MICHELSON: The code 4 is pretty clear I'think 2 when it says that for the reference values =you are looking ~ 3 at the equivalent of when it's known-to be operating accept-O 4 ably. 5. MR. FARSH: Yes. 6 MR. MICHELSON:.But the code'does not talk about. 7 looking at the range of operations that the equipmentLwill 8 be exposed to, like the valves shut.off on the discharge.of. 9 a pump until it can. deliver. water'at lower pressure into the 10
- system, 11 MR. MARSH:
That's right. '12 M R.' MICHELSON: These reference values don't include 13 the full range that the pump will see during an accident. 14 MR. MARSH: No. 15 MR MICHELSON: It just includes one' point in 16 fact. 17 MR. MARSH: That's true. - 18 ' MR. MICHELSON: Which is the normal operating point. 19 MR. '4ARSH: That's right. The' recirculation flow, 20 whatever. 21 MR. MICHELSON: And,it's up to the user to specify 22 what that is and.just make sure that later on the pump's 23 characteristics still look good at the normal operating point. 24 MR. SULLIVAN: Actually the code doesn't infer that 25 it has to be a normal operating point. Acme Reporting Company ,ns,s,....
} 26 1 MR. MICHELSON: Well,'it not only says when it's-2 known to be operating acceptably-- 3 MR. SULLIVAN: Right. 7"N (_) 4 MR. MICHELSON: You take it. 5 jMR..SULLIVAN:.The test is often'done using a 6 . mini flow line, which is another technical problem. 4 l 7' MR. MICHELSON: Sure. You can pick any point you 8 want-- [ i -9 MR. SULLIVAN: That's not a very,-good point. 10 MR. MICHELSON: .I know.- You are right. But.it 11 doesn't pick up these end extremes, Jesse, like the valve i 12 wide open. Or the valve fully closed. () 13 MR. EBERSOLE:; But he did give you some-comfort in 1 14 saying that they do determine what the pump will.do at run i ~ 15 out. { j 16 MR. MICIIELSON:~ No. i 17 MR. MARSH: No,-not~through the in-service test. 18 I'm saying initially. l 39 MR. MICilELSON: Well, I don't know if they do it j i t 20 then. j i 21 MR. EBERSOLE: I always thought sure that they did. l j on MR. MICHELSON: In some cases they do it only at i (:) ~~ 23 the vendor's plant. And they may not even do it there except l e for-- { 94 25 - MR. EBERSOLE: Well, I'm just saying. I don't know Acme Reporting Company l ,,o,
1 .~ 2 7 - 1 ~that you ever do it. Because I' don't know thatLthe different 2-specs even say that you have to. j 3 MR. SULLIVAN: I. don't think that there.is an-NRC 4' requirement'that it be done. 5-MR. EBERSOLE:- Well, if there's no requirement, 6 you can be assured that it's not going to be-done 7 MR. SULLIVAN:, Not universally. 8 MR. EBERSOLE: That's true. ~ 1 9 MR. MICHELSON: The standard' specifications that 10 You buy pumps to call for a certain commercial code to be i 11 used and in there it calls for determining this pump run out 12 condition and so forth. The NPSH curves. It's all pre-j () 13 scribed in that. 14 MR. EBERSOLE: I think though, Carl, in-those you 7 15 get when you buy the pump, it's important-to verify at least j 16 once what the ultimate capacity of the pump 71s at run out and l 17 at close discharge. Because that represents _its operational 18 damage potential or loss of function potential. And whether 10 or not you can reproduce that later on-in the< service life is i 20 another issue. 21 MR. MARSH: We'll check that. My impression is go-i 22 ing to be that like all plant status the early ones may not { O 23 have tested it that way and the later ones which were pur-21 chased to other more up-to-date specs may have tested it. { 25 We'll check that for you'. Acme Reporting Company ,,o,..n....- i
28 1 MR. EBERSOLE: Ted, we '.ve recently seen a few re-2 ports from the field that indicate when some control function 3 is lost. ,b 4 MR. MARSil: Yes. 5 MR. EBERSOLE: Due to the excessive pressure and G capacity of certain pumps. Hitters are overstressed because 7 the relief valves are not set to handle the higher flow with 8 the uncontrolled condition. You know, when a diaphragm failn l l 0 in a pressure regulator and you go to full purap flow at full 10 pressure. You can overstress a manifold beyond the capacity 11 of the relief valve because there is no coordination between 12 the relief valve setting and malfunction of the control de-l [G 13 vice. 14 MR. MARSII: What valve is it that we're talking 15 about? 16 MR. EBERSOEL: Well, this is a valve that was 17 feeding I think a feedwater 1.tne, a feedwater heater line, 18 where some control malfunctions occurred. 19 MR. MARSH: Well, that's safety related it sounds 20 like. 21 MR. EBERSOLE: Well, it was feedwater. I think 22 it was--I could pick it out for you. But it happens--you L-)J 7 23 see it lots of times. Some regulator fails and then there 21 is revealed to you a design error of the relief function to 25 take care of the failure--it didn't take into account the Acme Reporting Company .,e,,.,.
29 1 full capability of the provider. 2 MR. MARSII: Right. 3 MR. EBERSOLE: Of the fluid. ( t ~/ 4 MR. MARSH: And it's probably true when you look 5 at the IWV and IWP. It's not explicit there in testing it G throughout its entire range too. 7 MR. EBERSOLE: See, I could apply this to voltage 8 in electrical systems, air, whatever. 9 MR. MICHELSON: The point is that there is no re-10 quirement in the code to test it over a range. 11 MR. MARSH: That's true. 12 MR. MICHELSON: Only one point. Whatever you think ('s 13 is what they call here, "known to be operationally accept- 'us) 14 able." You pick that point. You establish your reference 15 values and you reconfirm them on a periodic basis. 16 MR. EBERSOLE: That's so antique that it sounds 17 like iron horses. 18 MR. MICIIELSON : Well, it's not antique. It's very 19 purposefully wricten. 20 MR. EBERSOLE: I know. 21 MR. MICHELSON: To make it simple to do. 22 MR. EBERSOLE: That's mechanically engineering of (') L-) 23 the 1800s. 21 MR. EBERSOLE: Yes, but it was done to make it j 25 simple and palatable and probably a code committee agonized Acme Reporting Company ,,va..>.....
30 1 for days or weeks over those. 2 MR. MARSH: My impression is that it was written-- 3 we've spoken to the folks. It was done as a starting point. e's ) 4 Okay. It was not done as the best way to go. It was done as s-5 a starting point and the problem is that it never progressed. 6 MR. EBERSOLE: The start was the end. 7 MR. MICHELSON: Nobody ever challenged it, so why 8 do more? 9 MR. MARSH: There are challenges either. There 10 aren't inquiries to these things either. 11 MR. EBERSOLE: Ted, coming out of this, I think 12 that the guidelines or the bases for your tests are really [") 13 out late. They are obsolete. They are antiquated. And x-14 you've got to go to the roots. 15 MR. MARSH: Yes. 16 MR. MICHELSON: The unfortunate thing is though 17 when you read the Standard Review Plan, it cites the proper 18 design criteria that should be met. But then it says that 19 the code takes care of these design criteria, but you go 20 back and you read the criteria and the code doesn't begin 21 to take care of what the criteria asks for. For instance, 22 its operational range is clearly prescribed in the GEC. But 7s i \\ V 23 it's not even mentioned in here. 31 MR. MARSH: Without trying to defend that, I think 23 that too is meant to be something that was going to be based Acme Reporting Company ,,e>..<.....
31 1 on a living document. The code was meant to be something 2 that was going to grow. It was going to llaprove. It was 3 going to be better with time. t') ? ks 4 MR. MICHELSON: Yes. This was done back in 1981, 5 the last revision. These words may have appeared several G years before that. 7 MR. MARSH: Probably so. 8 MR. HUANG: -- Actually took out the existing, 9 the current PIDs, and followed flow paths, flow paths after 10 flow paths, and tried to identify all the pumps and valves 11 that would perform safety-related functions. 12 MR. EBERSOLE: Are you going to refer to this dia-() 13 gram here now? u 14 MR. HUANG: Yes. I would take you through an 15 actual review of safety injection system later if you are 16 interested in how we actually ask questions. Everytime that 17 we see a pump and a valve. 18 MR. EBERSOLE: Well, before you do that, I'd like 19 to call your attention--you say here "emergency core cooling 20 system." 21 MR. HUANG: Yes. (~x 22 MR. EBERSOLE: And if the connotation is correct, q_) 23 that's for cooling in an emergency. 21 MR. HUANG: Yes. 25 MR. EBERSOLE: I'm really in fact more interested Acme Reporting Company
32 1 in the this operational challenge all the time. 2 MR. IlUANG : Okai'. I picked the easy one. I made 3 a mistake here. r~(_) / 4 MR. EBERSOLE: And this one is--I would just gather, 5 it's on standby of some sort ready to do a certain function 6 on call. 7 MR. HUANG: With exception to RHR system. The RHR 8 system would do something-- 0 MR. EBERSOLE: Would this, for instance, have ser-10 vice water on it somewhere? 11 MR. HUANG: Somewhere, yes. 12 MR. EBERSOLE: And does it have the removal of (~3 13 residual heat when you are normally on line, which is of K/ 14 course is a part of the standard removal process additive to 15 the feeding heat process removal? 1 16 MR. MARSli: If you are going to ask if this program 17 has the heat removal capacity-- 18 MR. EBERSOLE: Yes. 19 MR. MARSH: The answer is no, it does not. It is 20 only--what the requirements are to test the operational readi-- 21 ness and that does not include heat removal capacity. j 22 MR. EBERSOLE:
- Okay, 23 MR. MARSH:
And component cooling water is attached 21 to RHR and service water is attached to complement cooling 25 water. Acme Reporting Company .ma,,.
33 1 MR. MICHELSON: Does it do anything to check opera-2 tional sequences either? 3 MR. MARSH: Sequences. Do you mean cycling of g (l valves upon loading or something like that? 5 MR. MICHELSON: No. System sequences. This valve 6 opens and then that valve opens-- 7 MR. MARSH: No. 8 MR. MICHELSON: And that sort of thing. 9 MR. MARSH: No, it does not do that. 10 MR. MICHELSON: It's part of what the GDC talks 1 11 about. Y '.u do two things. It tests components and you also 12 test systems including their operational sequence. ['J N 13 MR. MARSH: I don't think--IST is meant to go with w y the technical specifications. I don't think IST is meant 15 to test sequencing of valves. 16 MR. MICHELSON: And I don't think so either, but-- 17 MR. MARSH: I think it's meant to make sure that I 18 that valve is doing its job. Then the tech specs in some 19 integrated--there is an O&M standard that is being worked on 20 at this point to do an integrated test routinely on ECCS l 21 systems. l i l 22 MR. MICHELSON: That's where you'd pick up the ,D sv 1 23 sequencing. l 21 MR. MARSH: That's where you should pick up so-l 25 quencing. And there are technical specifications Acme Reporting Company . ~ ~.....
34 1 requiring load sequencing tests, where you should be testing 2 valve cycling, diesel generator load shed and pickups. 3 MR. EUANG: And the programs only tested the compo-o (_) 4 nents of it. 5 MR. MICHELSON: Yes, and then only partially G tested the components. 7 MR. MARSil: If it were a healthy program, if it 8 did what you wanted it to do then, you could build on the 9 component test and say the system is able to do its job by 10 building on a component level test. 11 MR. MICl!ELSON: With a littl sequential testing. 12 MR. MARSH: With some sequential testing as a () 13 type of specification test going along. 14 MR. IlUANG : After we identify all these valve com-15 ponent P& ids, they are compiled into a list that is submitted 16 to us from the licensee. That is the part we call a continu-17 ous review. If we see any valve missing, then we will 18 generate a question and ask the licensee later why the valve 19 was not included in the program and that completes the first 20 part of our review. 21 Next, we'll move into the relief requests, because 22 even in the submittal itself, the licensee recognizes in many 7,t iLJ 23 areas in these instances they could not meet the full testing gi requirements as indicated in the code and that they were com-25 ing with relief requests and in some cases alternative Acme Reporting Company e a,.
35 1
testing methods, to gather the information.
So the staff 2
will review that, the basis for them, and of course in many 3
cases we come up with questions on the adequacy of those
(%,
U 4
bases.
5 MR. EDERSOLE:
On your selection of safety-related G
valves, let me ask you.
There are many specialized kind of 7
valves that don't always even appear on normal--what I call 8
normal full diagrams.
I wanted to give you a "for instance."
9 The long shaft--you know, emergency cooling water pumps, 10 they have to have air relief valves at the discharge of the 11 pump itself.
In essence, to clear the air out and also to 12 break back in when they are done.
And if those valves don't
()
13 function properly that pump's function is lost.
Generally 14 the bypass is so large that you don't get enough water in it.
15 Now, how do you pick up--do you pick up those kinds 16 and do they get those under the ISI program--
17 MR. MARSil:
T.
IST.
18 MR. EBERSOLE:
What?
i 19 MR. MARSH:
IST.
l 20 MR. EBERSOLE:
T rather.
Do you pick those up and 21 get them on your list or do you just not look for those kinds 22 of valves?
>....
36 a
3; in this area and they do get involved a lot in details like 2
that.
3 MR.-EBERSOLE:
Okay.
1 4
MR.~ MARSH:
And I want.to give you-another' example 5
of an area.
6 MR. MICHELSON:
These valves, by the way, have 7
stuck by QA programs--
8 MR. MARSH:
-I don't want to imply that everything 9
is perfect, that you make sure you get them all, because you v
10 don't.
We got an inquiry from Region 5 on a plant on the 11 west coast and they said we were off walking around this 12 plant and we noticed that there was a valve on the top of the
()
13 containment cooler and it's an' air relief valve.
And it 14 makes sure that once you start the' pump, you bleed the air 15 off to make sure you can flow things through it.
Is it in 16 the program?
And we looked and EG&G looked and we came to
~
17 the conclusion that, no, it was not in the program and it 18 should be in the program.
l 19 MR. MICHELSON:
But did it even show on the P&ID?
20 MR. HUANG:
This one isn't shown.
21 MR. MICl!ELSON:
It didn't show there?
But it showed us some kind of little thing with a little--
oo
()
23 MR. HUANG:
With a little small component--
24 MR. MICHELSON:
And it wasn't obvious.
It wasn't 25 in a flow stream as such.
It was off to the side.
Acme Reporting Company
,,o,,.......
37:
1 MR.-MARSH:
But that's where you_can rely on some-2 body's experience, if they've seen.this kind of problem before t
3 and they know about the system,-then lots of times they can 4
pick it up.
5 But I don't want to imply'that, yes,' sir, we'get G
every one of those little valves, because th'ats probably not 7
true.
8 MR. MICHELSON:
They aren't little.
In the case 9
of those raw water _ pumps, they are about eight' inches,.and 10 these are very large~ valves.
But they are little in~ terms of l
i 11 the P&ID.
12 MR. HUANG:
And that can only come under our current l
13 review process.
We could have missed a few..Decause'if the 14 system is far away from the-main' flow path and we are not 15 following through, then we may miss them.
This one happened 16 to be something in the main flow.
It's on the air cooler--
17 MR. MICHELSON:
It's exactly the same principal 18 valve except there, they are fairly small-because the problem 19 is much less there, j
20 MR. HUANG:
Okay.
After we did all the review, we l
21 usually come out with a set of requests for additional infor-1 22 mation and we send it to the licensee for response answers, 23 and usually we give them sixty or ninety days, and after that 1
]
28 we meet with the licensees'to resolve all the open issues in 25 that request for additional information and we plan to resolvo r
Acme Reporting Company e,, u.....
1
38 1
all of those issues in our meeting.
And following the meet-2 ing, usually, not always though, the licensee will come up 3
with a satisfactory revised IST program.
(\\
(_)
1 MR. MICHELSON:
Now, on the thirty plants that you 5
have SERs for, I got the impression that they were rather 6
old SERs.
7 MR. SULLIVAN:
No.
I would say that most of them g
have been written in the past four years.
9 MR. MICHELSON:
Do you consider all of those then 10 already reviewed with this more current kind of thinking?
11 You are not going back to look at those thirty?
12 MR. MARSH:
I would say we have about ten SERS f ')
13 since the reorganization, about ten are the most current.
v y
They reflect the latest thinking.
15 MR. MICHELSON:
Now, do those ten pretty well re-16 flect the kind of thinking you are doing today yet?
17 MR. SULLIVAN:
I'm not sure what your question is.
18 MR. MICHELSON:
Well, you are beginning to under-19 stand better what ought to be on a list--
20 MR. MARSH:
That's right.
21 MR. MT.CHELSON:
--for safety-related valves, for 22 instance.
( \\
( /
~~
23 MR. KARSH:
That's right.
3 MR. MICHELSON:
And as you think more, do you go 25 back and see if these were picked up on your last ten SERA?
Acme Reporting Company
.va,.
~_.
39 1
MR. MARSH:
No.
2 MR. MICHELSON:
Or are they.already out of' date :in
~
3 i terms of perhaps being inadequate scope?
?
4' MR. SULLIVAN:
You know, I-think thau we have such
'5 a' tremendous backlog and the whole activity in this. thing in
~
6
'very active.
On any given: day, the group leader of EG&G 7
and~myself may'be on_the phone talking about two or three 8
different plants.
It's all'we can do to'try and_ reflect--
~
9 MR. 1*ICHELSON:
Are you going back to look at some 10 of the SERs or none of.them?
Tf you issued an SER, is that 11 already--that's not a focus of your attention now?'
+
12 MR. SULLIVAN:
We're trying to forward'look.
13 MR. MICHELSON:
By forward look, you_mean you'are 14 trying.to look at the other seventy plants that you haven'.t
~
15 got SERs on~.'
16 MR.: MARSH:
That's right.
17 MR. SULLIVAN:
That's right.
18 MR. MARSH:
I don't think at this stage.it's not 10 productive to go back and look--
20 MR. MICHELSON:
Now, some of your SERs might be 1
21 pretty old already.
Five, six or coven years..
22 MR. MARSH:
That's right.
r O.
~
23 MR. MICHELSON:
And you are still not going to look 24 at them?
25 MR. MARSH:
No.
-Acme Reporting. Company.
,,e,,o,.....
7
'40 s
,i; 1
MR.'SULLIVAN:
There are some extra problems when-
.2
.you.say you are going ~~to go backLand.look at them.- Ano let's
~
'3 just skip ahead a little bit to.my part of the presentation.
O. R 4
What.it has to'do with is that-utilities very fre-l r
~
5 quently send in revised programs.
As soon as:they do that, i
6 they outdate the SER.
Because they change. things.
7 MR. MICHELSON: Even-beyond what they revised?
I
~ 8 mean do you go back and look at the whole situation again or 9
-just what they-propoce to-change?
l 10
~MR..SULLIVAN:
What we've'been.doing since we've
-l 11 gotten new SERs out since April is we'.ve been trying to keep'_
l 12 all of them current.
A utility sends.in~a new~ program, 5
r
()
13 additional relie f request or a. plant modification, we go ' back
'i 14 and review it for that change and try.to keep up--
15 MR. MICHELSON:
Just for that' change.
I 16 MR. SULLIVAN:
Yes.
.17 MR..MICHELSON:
Okay. 'You don't'go back and review 18 all of what was covered by the SER?
l 19 MR. SULLIVAN:
No.
I l
i 20 MR. MICHELSON:
Just the change?
i l
21 MR. SULLIVAN:
We try and have them identify where 22 the changes were and we review for those changus to try and j
'(:).
1 23 keep those SERs current.
24 MR. MICHELSON:
Now, when you do that, then you j
25 review those changes against your current thinking?
o Acme: Reporting Company
[
a o,..,. 4...
-41' a
1 1!
MR. MARSH:. That's right.
2 MR. MICHELSON:
But you do not reviewi_thefrest of..
]
^
3 the plans at all.
4 MR.. MARSH:
That's right.
5 MR. MICHELSON:
'So.although?it's a little better, I
~
G it's still not saying that because it's got an up-to-date 7'
SER that it's necessarily.your up-to-date thinking?
i 8
MR. MARSH:
Tru'e.
'I 9
MR. EBERSOLE:
Well,-then, in that. connection, let; 10 me ask you.
I was just getting around to this.
The industry 11 I think, everybody hopes it will lift itself up by its own I
12 bootstraps to pick up the needs such as we're talking about.
()
13 Do you advise when you issue an SER and you approve it witn 1
14
'the present ground rules, do~you advise the industry at-large i
15 that there are deficiencies and it's up to them to pick them
'G up or have them forced on them' sooner or later?
l l
17 MR. SULLIVAN:
We haven't been doing that.on a l.
18 utility-by-utility basis., We've been trying to do that by j
19 working through the~O&M committee.
i 20 MR. EBERSOLE:
It's really just to assemble the 21 information and then watch the industry's response to the i
22 need?
.O.
23 MR. MARSH:
That's a reactive war.
That's right.
24 And we have taken some steps which we hope are going to be 25 nore uniform or generic, more broad--
Acme. Reporting Company j
. o0,..,......'
l 42 1-MR. EBERSOLE:
Politically they love to say--
2 MR..MARGH:
They love it.
3 MR. EBERSOLE:
Oh, they love-it', but'it real life
,nq
- t 1
x/-
4 that's not the-case at all.-
5 MR. MARSH:
Well, the' industry's reaction is kind 6
of interesting.
They love--I don't want to say-- the current 7
situation, that is, the backlog is of no great problem to the 8
industry.
It really isn't.
9-MR. EBERSOLE:
No, because; you are.doing :nothing.
10 about it.
11 MR. MARSH:
That's right.
We are not-doing enough 12 fast enough.
O t3 "n
888aso's=
vee-14 MR. MARSH: If'they can revise th'eir programs ~.and 15 implement them without our concurrence, who cares?
Who 10 cares?
But if you say "no more,"
and now.the programs are 17 going to square away and we're going to.be on concurren'ce 18 and you can't change it, that's another matter.
19 MR. EBERSOLE:
Then you'll get their attention.
20 MR. MARSH:
Then you are going to get attention.
1 21 That's right.
22 MR. HUANG:
To answer your-question in part'I.think g-V) 23 the Rule 50.55A, we had a ten-year' update, automatic up the 24 requirement in there.- So every ten years we get a chance to-25 re-review the program.
And at that time we review with our Acme Reporting Company
,,o,,.,.....-
43 i'
k 1
new thinking.
Th'at's onet.of the reaso's we-don't'want to go-n 2.
back to the old programs..
-3
.Okay.
After we receive the-licensee'sJresubmittal,a
'f most likely we will concur and issue the SER.
f'
\\J/
4-5 MR. EBERS'LE:
Now, it's a't that issue point that O
6 I think you might put in.a little addendum and.say.we issue f
7 this in the face of known deficiencies.
8 MR. HUANG:.As a reviewernwe'can't say'that.
' i 0
-MR.
EBERSOLE:
We= fully expect.the industries would l'
10 pick up these matters and take care'of them;on an appropriate I
11 schedule.
12 MR. MARSH:
I think we pick it up and make it as li
()
13 good as we can,but we don't say, "known deficiencies."
i 14 MR. EBERSOLE: -But see, if 'you don't do'the, you i
15 give a blessing of sorts to something that ought not to be l
l 16 blessed.
17 MR.. MARSH:
I agree.
18 MR.-EBERSOLE:
'And that gives everybody great 10 comfort in saying, "Oh, I abide by the regulations, therefore r
1 i
20 I am perfectly safe."
21 MR. SULLIVAN:
That feeds into the whole philosophy l
l 22 of back fit.
(
l 23 MR.-EBERSOLE:
Yes.
l l
24 -
MR. SULLIVAN:
When we try'to make these things:
h
(
25 better, we're going to back to call them back fit.
Acme ~ Reporting Company-
, ?
,,o,..,.....
1 44 1
MR..EBERSOLE:
If they do,'_theyEdon't have to cal),
2 it.
3 Mk.7 MARSH:
Or if'.we work:through the code, get.the
((
~
4 code to change, and'ge.t them to update themselves to:the:later 5
code version, then that's not a'back fit.
6 MR. EBERSOLE:. Yeah.
7 MR. MARSH:
But the practical ~ matter-is the. code 8
hasn't improved.
The code isn't any better.
And when you 9
update'to a later code version, you haven't improved.
10 MR. EBERSOLE:
Well, what:about the folks that-11 write the codest 12 MR. MARGH:
We're working.
([ }
13 MR. EBERSOLE:
Are they involved in realization of 14
-needs?
15.
MR. MARSH:. We're getting there.
16 MR. EBERSOLE:
Okay.
17 MR. MARSH:
We are getting there.
18 MR. HUANG:
If you are interested, I will take 19 you through a review of one of the systems.as indicated on 20 the second slide.
21 I selected safety injection system from this P&ID 22 because that's the simplest way to go through and if you see 0
23 some circle in that flow' path, that's the flow pathfI'm going 24 to take you through.
We can start off under refueling-water 25 storage tank coming down.
And you'll see the valve there.
I.
Acme Reporting, Company
,m,.>.a...
45 1
circled.
That's a motor-operated valve normally open..So 2
since this.is a safety-related system, Class II,;and auto-3 matically'we know that valve to be in the program.
By.the way,
.4 we ask a.
few questions before we know what kind of test, 5
you know, we wanted to do on that valve, because.this valve.
G is normally open and so we. wanted to exercise this valve.
7 And so.if you'll. move'to the next valve, that's a check valve and -we know the ~ i:he'ck valve should perform'its 8
9 function in an open position and normally this valve probabiy 10 closes just by gravity.
11 MR. EBERSOLE:
I'd like to make a comment.
We are 12 looking at slides.
I doubt that the audience has--or'do they
[]
13 all have--
14 MR.. HUANG:.They should have a copy.
If they 15 don't--
16 MR.;EBERSOLE:
If we have any complaints, we'can 17 do it on the board.
Okay.
All right.
l 18-MR. HUANG:
Always wecheve a frequency problem 19 associated with all the valve testing, and in this particular 20 case, generally speaking, we have a frequency problem be-21 cause this is a aafety-injection system.
We cannot inject.
22 the flow into the reactor vessel during normal operations
~
23 so most of the frequency will be postponed to shutdown'and 24 refueling outage.
With a couple of exceptions I will discuss 25 later.
l Acme Reporting Company aos,.....
i-
e 46 1
And if'ycu follow the flow path--
2 MP. MICHELSON:
Well,1that still--that docsn't ex-3
'cl'ide a particular valve, such.as.-90ll from being on;your:
~
~
4 list.
It.just1means that'that particular valve can only be 5
tested under certain' circumstances.
C LMR. HUANG:
Right'.
7 7
MR. MICHELSON:
I notice you'didn't circle that-8 valve.
Does'that mean it wasn't on your list?
9 MR. HUANG: ' Key lock' valves.
10 MR. MARSH: He's only going'one path at a time.
11 MR. HUANG:
I'm just going:down one path.
I 12 MR.-MARSH:
This is not meant to be all inclusiv'e.
I
()
13 MR. MICHELSON: Okay.
The'n we'll take the other
{
i 14 path.
You still end up with more check valves and so forth I
15 getting into the legs;that you didn't circle.
j I
16 MR. MARSH:
That'at least'would have helped.
I 17 MR. HUANG:
It onlyfgoes~through a wind flow path-,
- t 18 MR.'MICHELSON: Yeah, but:you didn't circle the rest i
19 of the valves then.
20 MR. HUANG:
I did.
Because there's--let me see.
21 MR. MARSH:
Which one do you have in mind,. Carl?
22 MR. MIChELSO'.h Okay.
I see now what you are doing,
' i 23 Yeah, it would be helpful-- you picked it up.
I see now.
' j 24 MR. HUANG.
I think I picked up--
)
25 MR. MICHELSON:
The only thing you didn't do is' l
Acme P.eporting Company
,m,u....
~-,o,,c
,-,,x----n,
.,an
. ~.
1 circle the-key lock valve,-and that's because'.'you didn't---
2 That is on your list though,cis th'at right?
- i 3
MR. HUANG:.Right, it'sion-the list. 4 MR. MICHELSON: 'Okay. 5 MR. HUANG: But-I take you-through the common ~ 6 line here. I did not take'that flow path with key lock valve, 7 MR. MICHELSON: Yes, I got-it now. 8 MR. HUNAG: All right. 9 MR. MICHELSON: I'm just slow today. 10 'MR. HUANG: So'if you follsw it step by step, you 11 will know now which flow path I'm taking you through. 12 TheLnext valve, go up a little bit, is the flow-() ~ 13 8923. That's a valve, again a motor-operated valve, and: 14 normally open. So that automatically will be in the program 15 and it will be exercised with the progress of time. 16 And then next is the pump. We don't even ask~any 17 questions about a pump. We automatically. listed,'that pump ~ 18 into our program. 19 And then you go next to another check: valve there 20 with a P-1, a square.with P-1 next to it. That's the flow ~ 21 path I'll take you through. That check valve again, the-22 function there is to open and deliver water. And so that 23 valve will be exercised open. And normally again wouild re-24 main closed just by gravity. 25 And then I take you through the common dischargeL Acme Reporting Company ,m,.......
'48 1 fline that goes-through the valve.9010. That; is ; another motor-2 operated valve and is automatically in the program. And then you'll follow through the 8802. That's another valve in the 3' r~ Ek v l 4 flow path. And again that's another motor-operatedLvalve'.- 5 I don't know-whether it shows--okay. I.put a CIV there. This isolation 6 valve also perforns_-according to GDC 56, 7 valves function. That's why I put a note there becauce we 8 will decide on.what kind of leak retesting would be required 9 for them. 10 If you will follow.that flow path and then turn up ? 11 a little bit you will.see another valve there and I put a i 12 "no" there because that is one. of thevalves:that is.not in i (). 13 the program. And that's the valve used to prevent a. pump ~ 14 run"off. And when they run the pump and they adjust the i 15. flow to the position that you are comfortable with, and then i 16 they take the pinwheel off-that valve. 17 MR.-MICHELSON: It's a controlled orifice. i 18 MR. HUANG: It's almost like an-orifice'after that. ) 19 MR. EBERSOLE: Is it sealed or marked in_anyicalli-j i i 20 brated way that it's partially open-and-- l .21 MR. HUANG: Partially open. It's in one position. 22 MR. EBERSOLE: And does it stay in position just fmb ' Z3 in friction or is it locked or what? 24 MR. HUAMd: They took the wheel off. ~I.think it's. 25 a different means of lock. Acme Reporting Company a w >..... -l
~ '49 'l - MR. EBERSOLE: So that's the anti-run out system? 2 MR. HUANG: Yes. 3 MR. EBERSOLE: All.right. . : (~) M 4 lMR. HUANG: .So next to that is another, check valve. 5 I put a PIV and CIV both there. 6 MR. MICHELSON: Now, if you don't have'that valve. 7 on your list, do.you ever check it? 8 MR. HUANG: Today? 9 MR. MICHELSON: Yes. How;do you know that it's-10 going to prevent run out~five years from now? 11 MR. HUANG: That's a good question. I don't think 12 it's in our program and I hope that there will-be a test-- () 13 MR. MICHELSON:' And there is no--those generally 14 are hard to mark' depen' ding on whether it's a rising stem-or 15 whatever,_but it's generally hard to mark the valve in some 16 unique way so you know if it's been: readjusted for whatever 17 reason. 18 -101. HUANG: Today we can figure'out a reason to 19 include th'at, but it's not in our program yet. I will get 20 into other manual valves _later with you. Even if it's a 21 manual' valve, we want it to be exercised. 22 The next one is the check valve and also an isola-23 tion valve and also performs in this: case pressure isolation. 21 That's why I put a PIV there. Because once I put a PIV there 25 that valve will be required to be-leak rotested for the Acme Reporting Company aos,......
L l 50. 1-reactor. cooling system leakage. 2 Iflyou follow through the; path-there', there's 3' another check. valve. Also performs PIV. And-after that is 4 a discharge'into the core -- ~ 5 .So that comp 1etes our entire flow path review for 6 this portion. 7 MR. EBERSOLE: The' logic of that is if I put two 8 valves in series I claim redundancy for pressurized isolation 9 and containment isolation, right?.With two check valves. 10 Two twin checks. l 11 MR. HUANG: We could trade--let me see. We re-12 quire two PIV in series. So these two check valves- ~ () 13 MR. EBERSOLE: .Yes.- I'm saying that they are in 14 series. So they are redundant. 1 15 MR. HUANG: They are in. series. i 16 MR. EBERSOLE: Okay. Now,.these'are"two check 17 valves in series. 18 MR. HUANG: Yes. 1 19 MR. EBERSOLE: Over-the years they do what? How 20 do you know they are working? I 21 MR. MARSH: That's why you've got to check them. 22 MR. EBERSOLE: And how are they checked? 23 MR. HUANG: There is a test now in between these 24 two check valves. 25 MR. MARSH: There should beianyway.- Acme Reporting Company ,,,,,.a.e...
t.. 51 1 MR) HUANG: It's.now showing in this-- 2 MR. EBERSOLE: So;you confirm that the first one-3 is not7 leaking? G Is / 4-MR. HUANG: We. work separately, individually.- Both. i 5 of those two valves-will not leak. I 6 MR. MARSHi Let me;say this. We make'sure that ~ 7 both valves are included in the program. Both of them. j 1 8 MR. EBERSOLE: Whatever the problem is. 9 'MR. MARSH: Whatever the--the program lists.both 10 valves, and that means to us that both valves should be 11' tested independently, individually, not in series. But.we 12 firid cases where plants don' t in fact go and check.indiv-13 dual' valves in series. l .() _ 14 MR. EBERSOLE: All.right. Let me say that11 have 15 a pump operating.and I blow the. casing on it. s f 16-MR. MARSH: You'mean like a safety injection pump? ] t ~ 17 MR. EBERSOLE: Yes, or whatever. I lose, you know, 18 I have massive. failure. ~ 2 19 MR. MARSH: Okay. ] 20 .MR. EBERSOLE: What's the performance you premise 21 on those PIVs and swing check mode in attempting to close i 22 against a higher reverse flow? 23 MR. HUANG: Today if the position.has been changed 24 by injecting fluid through the check valve, requiring ~it to' test'to verify the valves are reclosed and-. 25 do an Acme Reporting - Company n - -..,..J
~ ?, ( 52 i s l' MR.'EBERSOLE: Well, howLdo~you know'.thatLin clos-- l 2 ing.though with-the violent reversible flow:that:they would: L3 experience.that they won't even stay.in place at.all? ~ 4 MR. MARSH: You've got-to be careful. You sort of' violated'one of'the rules here. 5 '1;e I? j 6 MR. EBERSOLE: U 7' MR. MARSH: One of the rule's is you don't postulate s 5 8 a design' basis accident which is a LOCA which' started the l 9 safetyLinjection pump in the first place. i 10 MR. EBERSOLE: True. 1 4 11 MR.. MARSH: And also'a passive failure in the'sys-12 tem which is another-passive fail'ure--case of blowing up. l f") 13 MR. EBERSOLE: It's valid right.- .w/ 14 MR. MARSH: You can postulate. that tlie system-- l 15 MR. EBERSOLE: Yeah, I'll find other' valves to do 16 this. [ i 17 MR. MARSH: There is th'e same thing somewhere. else. 18 MR. EBERSOLE: Yes. This is a little hard'to do, j i 19 Yeah, okay. Carry on. { i 20 MR. HUANG: Okay. This' wind flow path that carries i 21 us through the refueling water storage tank to the core l 22 elect dischargc. However, since we are also interested in .( c, 23 the testing of the pump,-I will take you through the mini i 24 flow line there. 25 So following the pump.before--you see another line ] 'I Acmef Reporting -Company .na, si..... .1
l. -53; / 1' coming down. That's the. mini flow line for the-pump. And' ~ t 2 we will also review now each-and every' valve.on that mini; 3 flow line. So iflyou'll-go down a little bit you'll see a 4y + \\/ '4 check valve.- So we inclu'e that check valve there in the d 5 . program because we need that' mini flow-to protect the~ pump. 6 And also enhance the pump performance.. 7 However, once-- 8 MR. MICHELSON: When you put it in your program, 9 do you. identify what'the function is? LIn this case', those 10 check valves are really--that first check valve on the mini 11 flow line is there in case that you have~one, pump only run-12 ning, you don't want to back flow into the'other pumps.. f () 13 And so really theLfunction of that valve is to 14 prevent' reverse flow and not necessarily to open,.although i ~15 it also has the function I guess of having to open when you 16 want mini flow on that particular pump, and it's'there to l d 17 prevent reverse-flow if the opposite pump is running. 18 So do you put both of those functions.down on your 19 list as-- 20 MR. HUANG: Okay. Now, to answer your question-- 21 MR. MICHELSON: It has to work both ways. This has 22 got to be tested both ways.
- O 23.
MR. HUANG: Yes. We will verify the closure of 21 that valve. Okay. We will also verify any opening-- 25 MR. MICHELSON: Any opening, okay. And that's on Acme Reporting Company-e a n.>....
54 1 your list that you have to verify? 2- 'MR. HUANG: Yes. By the way, we will.not do a. 3 leak testing on this. valve. ?
- A /
4 'MR.-MICHELSON: No. Leakage would-not be a con-5 sideration. 6 .MR. HUANG: We will veri #v in either>way in this 'l s 7 case. j 8 MR. MICHELSON: I-see. Okay. ~9 MR. SULLIVAN: The way..they do that is they'cate-i 10 gorize that valve as a Category C valve. It's just a code ~ l.I 11 nomenclature, and they should. indicate in their program that 3 12 it's tested in both open and closed position. l () 13 MR. MICHELSON: I think it's a Category B,'isn't 14 it? Or is it a.C? 15 MR. HUANG: No, this is a C... Check valves are 16 always C. t 17 MR.- MICHELSON: Check valves'are always1C. . Yeah, j 18 that's right. .But the code doesn't differentiate opening e 19 and closing or does it? l -i 20 MR. SULLIVAN: Yes, it does. I 21 MR. MICHELSON: It must. Yes, it would have to. 22 Okay. - () 23 MR. SULLIVAN: So what we do is we look at the l 21 program and see what they have specified and if we think that l l 25 they haven't specified both positions-then it.becomes a l l Acme Reporting Company l ,,o,,.>.......
55 1 question for the meeting. 2-MR. MICHELSON: Okay. But sometimes it'olllyihas 3 to open, other; valves only have to close and some have to.go ' 4; both ways. 5 MR. MARSH: ~Both directions. 6 MR. MICHELSON: And'you will have that on this 7 master list?- 8 MR. MARSH: This PIV is a bi-model valve too. 9 MR. MICHELSON: You don't have a sample of a master 10 list here, do you?- 11 MR. MARSH: No'. 12 MR. MICHELSON: That would'have cleared up a lot of 13 what you are-- O 14 MR. MARSH: We can' send you an old program. We've i 15 got lots of them. 16 MR. MICHELSON: Oh,'I'm sure you've got myriads 17 of computer printouts.. 18 You've answered the question. That's good enough... 19 MR. HUANG: This is the utilities submit at:the 20 table for the valve. What type of valve, what kind of test 21 it is that you perform. And what frequency. 22 MR. MICHELSON: Let me ask you a question about 23 safety injection pumps in an opposing. context. What do you i 24 look at in the combined testing program ~here that.looks at J I 25-both the pumps and their related valves to ensure that you j Acme Reporting Company ,,w.>.... --m
56 .1 don't lose your safety' injection. pumps eitherfthrough' loss 2 of suction, which i's common, a common possibilityrhere,, or 3 . deadhead? I.have~bo'th these possibilities in this diagram 4 because they.both focus on a common discharge and they have 5 a common suction. So'I've got a' double barrel' potential for 6-having lost' suction or deadheading. And how do you now han-7 die the. problem-- 8 MR. HUANG: Okay. My understanding of testing of 9 the -- First of all,Jwe have.to. measure in inlet pres-10 sure. The inlet pressure measurement _would give you a sound 11 indication-- 12 MR. EBERSOLE: That's the normal condition. You've ) ~( ) i 13 got the head of the storage tank on it. 14 MR. HUANG: Right. Make sure now that there is 15 available hat in the' pump. 16 MR. EBERSOLE:.But then-- 17 MR. HUANG: The. mini flow is the one. prevented the 18 shallow head. The damage to the pump. This is why.we. verify 19 the mini flow path here. 20 MR. EBERSOLE: Well, what about the single failure 21 criteria in here for the loss of both pumps, which is by ) 1 -l 22' closure of that 8806? Don't you need to know hou'long you .J O a 23 can--that's the supply line from the RWST. I 24 MR. HUANG: Okay. That's a system design question. ] 25 This is why this valve always remains open normally, and j .I Acme Reporting Company ~ d so,,.,. o.. J
57 1 there's a single--okay. I probably should explain a little 2 bit. We have an S here. 3 MR. EBERSOLE: I know. Y 4 MR. HUANG: They will receive a signal to open. 5 So if there-- 6 MR. MARSH: But it is open anyway. 7 MR. HUANG: It is normally open and they will re-8 ceive a signal to open and from a system design point of 9 view-- 10 MR. EBERSOLE: It is open. 11 MR. MARSH: This does not get around system de-12 ficiencies like that. [) 13 MR. HUANG: No. v 14 MR. MARSH: It can't do that. The same thing with 15 the 8814 and 8813. 16 MR. EBERSOLE: Yeah. I'7 MR. MARSH: It doesn't get around design deficien-18 cies. 19 MR. EBERSOLE: Right. You could get around it by 20 providing a trip sheet on it. 21 MR. MARSH: Sure. But it is not intended to be a 22 redesign review. It's intended to make sure that the system em, L) 23 is doing as it was designed to do. 24 MR. EBERSOLE: Okay. 25 MR. HUANG: And a lot of the times when we review Acme Reporting Company , m,.,.....
58 l' the relief request is because of design deficiency.in-thel 2 system. But still it's a minimum requirement of the' safety analysis. And so.there are cases when we have to compromise. 3
- O N/
4 MR. MARSH: Help me, John. 8806'I'think is-a valve 5 that's.normally and doesn't even have' power on it.. I think 6 it's normally a depowered valve, is that right? 7. MR. HUANG: I'm not quite sure. Because I only 8 checked it to the extent they.have a signal to.that valve. 9 MR. MICHELSON: You've got the same problem on the 10 return though. If you have"too high a pressure in the sys-11 tem, depending on.whatithe head of these sub-pumps are and 12 one of those two return valves fails to open-- () 13 MR. MARSH: That's why I think their-- 14 MR. MICHELSON: There _. no isolation when you 15 start recirculating hot soup.. You don't want to go back to 16 the refueling water storage tank. 17 MR. EBERSOLE: Here's when they come up on the'need 18 to have known what is the ultimate pressure potential of these 19 pumps with zero flow because that defines the loading on the 20 pipes and valves that they are going to have to be loaded 21 against. I don't even know whether the valves are designed 22 for those heads or not. You know, whether we have a couple 23 of designs. 21 MR. MARSH: I'd hope so. 1 25 MR. EBERSOLE: I should think so. 4 Acme Reporting Company .,ou,...... m
k 59 i l .r 1 MR. SULLIVAN: I'think they should be because I 1 ~ 2 specified.the design pressure.of the. piping hopefully based 3 on shut off-head. l if]~ 4 MR. EBERSOLE: I should hope so.. But I don't know. ~! i 5 of a rule'that works on that. But it must be understood, i ~ 6 MR. MICHELSON: The tricky one.to watch for,is when~ j 7 you get around to turbine-driven pumps ' and you had better -8 watch what the mechanical trip speed is or the; electronic. t -9 Usually you use the final mechanical trip because that pres- -l 10 sure rating goes way the heck up before the trip finally l 11 occurs. You get quite high pressure because as the speed i 12 goes up the pressure goes up and it's unloaded. But here i {) 13 with constant speed motors-- '14 MR. EBERSOLE: Yes, it will just leap'back. 15 MR. MICHELSON: Nobody would miss'thatione. I I 16 MR. EBERSOLE: Is there a developing rationale for' t ~ I 17 protective trips on both suction _or'high discharge. heads. I i l 18 notice where pumps are expensive like main feeds you generally 19 have these things. But when you come down to these little a 20 bitty but important you don't. r l 21 MR. MARSH: We've talked about this before in 22 another-context. RHR-system design and AR45 design and 23 things of-that nort, and the only place I know where some- ] 24 plants have initiated trips or on some RHR systems, some RHR 25 pumpo have automatic automatic low-section pressure trips. i Acme Reporting - ~ Company aa,,....... m
>N h - 60 I 'Most-don't. Most don't. 2 MR. EBERSOLE:.There's been a rash of daily reports 3 .about the potential:for common mode failure of both pumps g. '(d 4-due to' common: discharges being. closed'recently. 5-MR. MARSH: Yes.. 6 MR. EBERSOLE: What's your interpretation of what' ~ 7 ought to.be done about that? 8 MR. MARSH: -I guess I would need to know.more about 9 the problem'before I could comment on that. -10 MR. EBERSOLE: Well, I think in.some cases it's 11 just the accidental-- 12-MR. MARSH: Manual valves or MOVs? (f 13 MR. EBERSOLE: I don't know.- I can't remember. - 14 How do I know? .l 15 MR. MICHELSON: Generally ~they are MOVs because 16 they have to. isolate--in' fe.ct, there are often times like in 1 17 this diagram there are two MOVs. 18 MR. EBERSOLE: You've_got a double chance to do'it. ~ 19 ~MR. MARSH: Mini flow 7research is-aLproblem. 20 MR. EBERSOLE: Are there any specifications for 21 duration of operational ability of pumps on deadhead or on 22 deficient suction? 23 MR. SULLIVAN:. Sometimes manufacturersihave-- 21 MR. EBERSOLE: You'know, it shouldn't fail in the 25 first second or two. Acme -Reporting: Company ,an, u.....
61. 1 MR. MARSH: It depends. You know,.some pumps'you 2 can go,and it will. handle this in.a minute.1 Some pumps you: 3 can go in^ thirty seconds and RHR pumps-- 4-MR. EBERSOLE: Well,'I.-think those differences 5 .ought to be related/to the design protective"features. You 6 know, you'can't wait-for the. operator to respond in' ten 7 seconds to keep the. pump from blowing.up. 8 MR. MARSH: No.' I don't want tx) cross ever into 9 other areas. 10-MR. EBERSOLE: I know. 11 MR. MARSH: It's hard to do'that.. IST is supposed 12 tx) make sure that the "as designed" system is doing what it ( )- 13 is supposed to do. 14' MR. EBERSOLE: Piece by piece.. 15 MR. MARSH:. Not to add.new designs,inew. protective 16 devices. There's enough problems with the'IST program as it' 17 is. l 18 MR. EBERSOLE: Yes, I knew that was,the limitation 19 on it. 20 MR. SULLIVAN: You' guys are trying to integrate 21 and we try to compile. Somewhere we meet in the middle. 22 MR. EBERSOLE: True. 23 MR. HUANG: Anyway,'in order to complete these 24 flow paths wit! v. 25 MR. MICHELSON: Let me ask before you go on with + Acme Reporting Company. ,,o,,.,.....
r-4 .62 c 1 that though, this question.of check valves and what the code 2 requires. I was a little confused whenfIl read the code be-3 cause it talked about normally. opened checks and normally )' -4 closed checks. 5 MR. MARSH:. Right. 6 MR. MICHELSON: But it didn't talk about.the case 7: where it had to be both normally opened and normally. closed 8 ' depending on the particular-function it was performing'at 9 the time. 10 MR. MARSH: Right. 4 11 MR. MICHELSON: Now,.how does the code address 12 these valves which on occasion have to be cure to be open /~T and other occasions have to be sure to be closed? Do they 13 V 14 combine these two tests? Because these two tests don't seem 15 to fit the fill. 16 MR. MARSH: The way we approach it is--I don't.think 17 we directly relate to those words "normally open" and "nor-18 mally closed." We look at what the safety function of the 19 valve is. If'the valve ~is required..to stroke open to-fulfill 20 its safety function, then we want it to be tested to verify 21 that it fully opens. 22 MR. MICHELSON: But you say you are going to use 23 the code for your test. Because that's what everything so 21 far I've read says. The code is what I read to find out what 25 the test is, so-- and I have both of them to perform. Do I Acme Reporting Company < m,....... 1
63 1 go to the normally open part and then go to the normally 2 closed part? 3 MR. SULLIVAN: Yes. 4 MR. MARSH: Well, you car. say that the safety 5 function of this valve is bimodal. And you therefore should 6 test this valve in both modes. And the code tells you how 7 to test the valve given its safety mode without the code 8 being explicit if it has two modes of operation. 9 MR. SULLIVAN: The code misleads people I think 10 because sometimes people don't identify both functions. 11 MR. MICHELSON: Yes. The code doesn't even recog-12 nize that there might be a valve-- ( ) 13 MR. SULLIVAN: You might interpret that to mean 14 the code thinks there's only one function. 15 MR. MICHELSON: But everybody that uses it under-16 stands this? 17 MR. SULLIVAN: If they don't, we explain it to them. 18 MR. MICHELSON: Okay. 10 MR. EBERSOLE: You have a method of checking that 20 both positions were--it's come to light through a violent 21 path of a revelation that we need to know how it's opened. ,w 22 That doesn't mean it's flapping either. (_) 23 MR. SULLIVAN: Right. 21 MR. EBERSOLE: Are you doing anything to confirm ) 25 that when it's open it's really stable and open and not Acme Reporting Company a,+,....
64 1 standing there hammering itself to death? 2 MR. SULLIVAN: I would say we're not. 3 MR. EBERSOLE: No, the code in fact says all you x_ 4 have to do is open 50 percent, which is probably not a good 5 Position to operate check valves at normally. 6 MR. MARSH: Right. 7 MR. EBERSOLE: But the test only says you see that 8 you've got at least 50 percent sea port. 9 MR. MICHELSON: But now we know it had better be 10 open and streamlined and steady. 11 MR. EBERSOLE: It may be that the 50 percent would be 12 very unstable. ^ MR. MARSH: Is it a minimum of 507 (v') 13 14 MR. MICHELSON: Well, it says at least 50, the 15 same thing. 16 MR. MARSH: Well, that passes the desired flow at 17 a minimum of 50. 18 MR. EBERSOLE: No, it didn't say passes the desired i 19 flow. It just showed that the disk move is sufficient to 20 provide a flow area of at least 50 percent of the area of 21 the sea port. In other words, opens-- 22 MR. EBERSOLE: It takes no account of the dynamic /._s, N.) 23 failure. 21 MR. MICHELSON: No, it doesn't address whether that 25 io good enough or anything else. The assumption is if it will Acme Reporting Company ao,,.
2 65 4 / - 1 open that.much it-must be-' good.enough. 2 .MR. MARSH: I'm sure you know about the^ problems 3 of check. valve testing in general. 'And this is one place to- ' () f4 criticize what's going.on:in'other; places, industry. efforts, i f 5 and then other major check valve testing deficiencies. In 6 my opinion it's another examplelof IST program problems. f 7-Check valve testing in general. Whether you come back to the s 8 code or whether you come back to the SRP_or whatever-- 9 MR. EBERSOLE: Well, just how active is the code .i 10 .in picking up these deficiencies? -Are they-sleepy, moving l 11 along, you know, like a frozen' river or what? I 12 MR. MICHELSON: Well, they are used'to working in l 13 terms of years to change someth'ing. 34 MR. MARSH: A code is careful and slow, let's put i 15 it that way. 16 MR. MICHELSON: Glac'ial'is angood word. 17 MR. MARSH: We:have. talked to'a lot of utilities 18 and we asked them why don't you bring y~our; concerns to the 19 code. 20 MR. MICHELSON: Well, take some outfit like North-21 east. 92 MR. MARSH: We talked to Northeast. 23 MR. MICHELSON: .They would be the kind of utility 24 that would probably advance.this cause better than others. 25 MR. MARSH: Well, we've talked to other utilities Acme Reporting / Company ,,o,..,.....
66 1 who have been--they've said, "Wnat shall we do that for? If 2 we do that, it's just going to be glacial. It's going to 3 take us years and years and years. Why do it that way when \\_) 4 we can do the testing that we want to do without having it 5 take ten years to get into the code requirements." 6 MR. MICHELSON: Yes. Just a good regulatory guide, 7 for instance, would do it. There isn't a regulatory guide 8 on this at all. 9 MR. MARSH: Not yet. 10 MR. MICHELSON: And there never was. 11 MR. SULLIVAN: There was a draft. 12 MR. MARSH: But there was a draft regulatory guide s [~3) 13 that went out for comment and a lot of comments came back x 14 and said, "We don't need this. We don't need more guidance. 15 Leave us alone." 16 MR. SULLIVAN: I'd like, if I can, to go back to 17 your point about the end of paragraph 35.27. 18 MR. MICHELSON:
- Yeah, 19 MR. SULLIVAN:
We really don't know of a way that 20 you can show that the disk area is 50 percent of the air at 21 the cea port. 22 MR. MICHEISON: I wasn't going into that, but I ,-3 E.] 23 wondered-- 21 MR. SULLIVAN: What we basically do and we end up 25 arguing a lot with utilities on this point is we say that Acme Reporting Company ,,o,,.,....
67 1 you've got two ways that you can mi';e sure the valve is 2 fully opened. One of the paths--the minimum required design 3 flow through that valve. The other is to take the bonnet n ) 4 off and make sure that it is-- 5 MR. MICllELSON: Is free to swing. 6 MR. SULLIVAN: That it is free to swing, which is 7 addressed in here. That is one of the exercising methods. 8 MR. MICIIELSON : Yeoh, I wondered how they saw the 9 50 percent, and on a lot of systems you can't get full flow. 10 A lot of systems the mini flows are no way near full flow, 11 MR. MARSH: And that's a problem. That's another 12 Problem. There have been pump failures because of that. (~) 13 MR. SULLIVAN: You can take a system like contain-x2 14 ment spray and there are some check valves in that system 15 that exercise of flow, you wet down the containment. 16 MR. MICllLLSON : Yeah. 17 MR. SULLIVAN: So the only alternative really is 18 disassembly. 19 MR. EBERSOLE: Let me take the case of a check 20 valve which has always been interestir. to me. It's on a 21 boiler. It's the main feed check valves. And as yo" know, 22 these discharge through a protected region so you don't--if p) \\_ 23 yc a happen to have a main steam line failure, they have to 24 close. Tnef are the first thing that closes. The other 25 valves will try to close, but they'll never make it in time. Acme R e p ort;.7 9 Company e,,,....
n.. ...~ ~. a f 68 1 The reverse flow checks will close first'. Therefore they'll 2 see the full brunt of'the rapid flow reverse and the shocking 3 load imposed on their. disks,:while they are trying.to get !(k 4 ' shut, and you can argue rather easily that the pieces of.tha ~ 5 check valve will go into the control: valves down line and 6 the lines will' simply stay open. And if the steam doesn't 7 continue to discharge into working equipment areas, you'll 8 still make.up water and it'll pour'down-eventually, but 9 you will have lost the entire inventory of primary fluid 10 right out in the world. 11 I-don't know what is done by the industry at large 12 here to validate the performance of these check valves in i . {) 13 performing this violent closing function.- I see a: variety to of operations or designs in which some designs have in-15 corporated extended shafts with hydraulics numbers to impede 16 the rapid closure. Others don't do anything. I've seen no 3 '17 ' need described in regulations to do a-dynamic analysis for i j 18 the violence of closure. 19 And, Carl, you well remember this from Sequoia 20 days. There is a need to require, there's'a regulatory re-21 quirement, some sort of proof of. performance,if it's no more 22 than analytical, that you won't tear the valve up. Because j ~ 23 you expect it to close. That's its major safety function. 21 .MR. MICHELSON: That's a procurement problem l 25 though I think in light of what you are looking at. i' Acme Reporting Company a n...... . I
69. 1 MR. MARSH:. Well,-that's-- 2 MR. MICHELSON: You don'.t-intend to~do dynamic. 3 testing of check valves as near as I~can tell-from reading--
- O MR. EHERSOtE
We11, vou eueht te either ene1vze-5 it or. test it in-the plant. . ell, but I mean that's beyond W 6 MR..MICHELSON: 7' the scope of ISI. i 8 ' M R. :-1A R S H : You've got to do better than you are 9 doing. 10' MR. MARSH: IST. 11 MR. MICHELSON: IST, excuse me. 12 It's got to go.beyond the scope of the' code. 13 MR. EBERSOLE: The code.was made for-sewage lines t-14 - or-something like that. 15 MR MICHELSON: No. It just addressed dynamic-J 16 testing of check valves, as.near as I~can tell. It addressed 17 the functionality in terms 1of making sure;they can swing open 18 if they need to or swing closed if.they need to. 19 MR. MARSH: The safety function'of that valve is 20 to perform an isolation function and that ought to be em - 21 bodied at least some way in that code. Now, that doesn't 22 mean you should do dynamic testing though. O 23 MR. MICHELSON: To meet the GDC I think ysu've have 21 to, i f 25 MR. MARSH: Right. Acme Reporting Company ,u.a,,.>.....
70 j MR. MICHELSON: But not to meet the code. 2 MR. MARSH: To meet the regulatory requirements. 3 MR. MICHELSON: Well, the regulatory requirements, ,() 4 as I read them, it just says meet the code. That's what the 5 Standard Review Plan clearly says. 6 MR. MARSH: Go back to the reg. To the what the 7 reg says. To do the safety-related function you've got to 8 do the code. Don't do the SRP. That's not the regulatory 9 requirement. 10 MR. MICHELSON: The GDC are the regulations, of 33 course. 12 MR. MARSH: And the 50.55A. (J~] 13 MR. MICHELSON: And they clearly spell out you've ~ g4 got to do this. 15 MR. MARSH: Right. 16 MR. MICHELSON: But the Standard Review Plan says 17 what the reviewer is supposed to do and decide if he meets 18 the regulations. You meet the code and you've n.et the regu-19 lationc. That's what it very clearly says here. Even cites 20 the criteria. 21 MR. EBERSOLE: I don't know how that--the basis for f-22 that. I don't know where that ever came from. But I'd be \\ ) 23 curious to find out. gg MR. MICHELSON: Well, it's a very clear statement. 25 It's very clear. Compliance with the code will assure Acme Reporting Company o ,,e,.....
71 1 conformance with 10 CFR Part 50, Appendix A-- 2 MR. EBERSOLE: %ho is the author of that I wonder? 3 MR. MICHELSON: It also includes 50.5 SAG. ("'; ) 4 MR. MARSH: I didn't write it, but like most -m. 5 Standard Review Plans it assumes that the code that's behind 6 it is either good or is going to be good. 7 MR. MICHELSON: Or good enough. 8 MR. MARSH: Or good enough. 9 MR. MICIIELSON: It might have been good for burning 10 coal. 11 MR. HUANG: Let's finish this flow path quickly. 12 And there's another valve in the mini flow lines (v") 13 which is again is only there fer permitting the pump run off. 14 That's why it's not in the prograia. 15 MR. MICHELSON: Okay. 16 MR. HUANG: And then there's two more operative 17 valves. They are automatically included in the program for 18 exercise and the stroke timing. 19 And that completes our wind flow path review and 20 we'll make a list of the pump and valves in that review and 21 then after we've reviewed the entire system, we make a whole 22 list and then we compile it into two, the licensees list to [s\\ v 23 say, hoy, you know, your review is complete or it's not. 21 Then we'll start making a list of questions. 25 MR. MICHELSON: The problem I had with these valves Acme Reporting Company , m, u....
A 72' =>
- 1 that you said were'there for preventing. pump run.off and 2
that is that'they'are also there to facilitate maintenance. ~ Say, when I've got to take another valve out'or take a pump. -3 i O ~ 4 out, I take that: orificing valve and I screw it closed. 5 MR. HUANG:. Oh, I thought you were going to use 6 81-- 7 -MR. MICHELSON: No',- sir..You check carefully. A 8 lot of-times you have to use these,vslves for: isolation ~. 9 MR. MARSH: Well, the ' tech specs say,~at~~1 east for 10 the--I'm not-sure it's this~ plant--the generic-Westinghouse 11 tech specs say that if you ever do main.tenance on this system .} 12 such.that the flow characteristics of the system can-be i 13' ' changed, you will do a flow test of the entire. system. I 14 MR. MICHELSON: -That was what I was.trying to lead 15 to as I said. So once you've done your work and you've i 16 opened this valve up, now what controls how you reset that i t 17 valve. .i [ 18 MR. MARSH: It's a flow balance test. f i 19 MR. MICHELSON: I mean what-- 20 MR. MARSH: How do you do that? l 21 MR. MICHELSON: What requirement sets up this flow f i 22 control test that you were talking about? This is not the 23 IST. 1 24 MR. MARSH: No, this is an FSAR kind of a valve. l 25 This valve here is designed to make sure that l'f you have a Acme Reporting Company m,..,.....
73 1 break in any one of these lines and you have a single failure 2 in one -- 3 MR. MICHEL3ON: Well, I was pointing out to you r's (.) it's used for other things. And when it is, then I'm asking 5 how do you assure iz gets back in its proper orificing posi-6 tion? 7 MR. MARSH: I'm trying to get you there. Okay. 8 The tech specs say that if you modify any cotc.ponent here in 9 a maintenance way, yceu. overify the flow characteristics of 10 that system; i.e., maka i' an 50.36 in the FSAR I think it's 33 supposed to be_ 12 MR. MICHELSON: You mean like a pre-op test - l () 13 MR. MARSH: Go do a pre >p test on it &gair.; (_/ p MR. MICl!ELSON : To nase sure. that you've got the 15 valve back-- i 16 MR. MARSH: That you've balanced
- properly, i
17 MR. MICHELSON: In other words, you don't do the 18 code test. You go back and do some other test. And this is clear to the utilities. It's not the code test you now use ig 20 but-- 21 MR. MARSH:
- Right, 22 MR. HUANG:
Se'. tion XI-- es ( )( MR. MICllELSON: Well, these code tests wouldn't 23 21 get you back there. 25 MR. HUANG: Yes, because you cannot do it, any code Acme Reporting Company . m. u.....
, v. 10 1 requirement on this. 2 'MR. MICHELSON: Because theseLare not system re-3 quirements. These are individual valve. requirements and'you '4 can'tiset the valve.. Youve got to set the syst'em.- 5 MR.. HUANG: So there are individual' component' re-6 quirements. 7 MR. MICHELSON: But let me ask you though..They 8 do operate as a system. They are' parallel, for-instance. 9 MR. MARSH: Sure. 10 MR. MICHELSON:' Do you validate in the testing' l 11 program or in the plant,'in the manufacturing plant, how 12 well these pumps oparate as parallel pumps;. Differences in () 13 characteristics can easily cause-parallel pumps--one pamp to. 14 really act as a buck on the other onc and you think you are 1 3 15 getting parallel flow when one pump is at stall flow condi-t 16 tions. And yca are ruining it while you appear to be running. ] 17 MR. MARSH: There.has been recent concern on that. 1 l 18 That two pumps An par.lel, you may end.up with one pump 19 deadheaded. Because there'is a we~aker pump. I 20 MR. MICHELSON: It's an ancient problem. 21 MR. MARSH: There's been a Part 21 notification on 22 it from Westinghouse recently that says it's specifically.on Z3 the hdR pumps. And there is an internal NRC action trying to 24 address that concern. 25 MR. MICHELSON: How do you handle that at the Acme Reporting Company t . m,.,.....
.~ /- f ' 75: i i component level?- 2 MR. MARSH:.You have:to make sure..that the flow,- 3 - the recirculation flow, and the pump.difforence allowed by {C UE 4 the IST testing is not so great'thbt;that=along with the re - S . circulation.fitw won't deadhead the pump. It's a system i 6 characteristic. 7 - MR. MICHELSON: -It's a system characteristic re-t 8 vealed'by a component test. 9 MR. MARSH:.That's right. 10 MR. MICHELSON: Well, are you doing that? i 11 MR MARSH: Well, there's internal things going.on 12 being considered about how to address that. Whether it's 13 an inforination notice or a bulletin or something,. there are 14 things being considered on that. And it's working it's way 15 through the NRC process. { 16 MR. MICHELSON: I take it it's a universal problem? 17 MR. MAP.Sil: It is. It was identified on a group of 3 18 plants on the RHR system, but it was brought because it can 19 occur for any parallel pump system. 20 MR, MICHEL3ON: Yt.s, any parallel pump. 21 MR. HUA.:G : Let's move on to tell you a little bit 4 22 about Section XI ar.d the tests are, you know.. First is the i 23 pump and then the v51ves. And then you will see how limited 24 these tests can bc. 25 And then next slide is the ASME Section XI' Code l Acme Reporting Company i .n,,sa..... ,--e. r- ,~..,,-nne,-....nn.n.,, .- n c,rv.-,-,,.,c, nn,,
76 i Requirements for Pumps. 2 we run quarterly test on all the safety-related 3 pumps to measure the following parameters. (q _) .i MR. EBERSOLE: Could I interrupt you just one mo-5 ment before we get into this. 6 We just left check valves. I've heard over the 7 years a peculiar thing, a difference between the Westinghouse 8 and GE plants. Is that GE always complained they had to put 9 in exercisers on their safety-related check valves and West-10 inghouse and CE and AW went along and never had to bother. 11 They said it was a political problem. What was the problem 12 and why is it so? ) 13 MR. MARSH: I don't know-- 14 MR. EBERSOLE: What's the basis for putting exer-15 cises in and why if you have to put them in over here you 16 don't have to put them in over there? 17 XR. MARSH: I can't answer. Can Joel or Frank, do [g you know? 19 MR. EBERSOLE: If anything, I should think the BWRs 20 would need it worse than the clean water boilers. 21 MR. CHERNEY: Frank Cherney from Research. I don't-- 22 we know that that is the case, but I don't think we know why ,U it is the case. 23 21 MR. MARSH: Jesse, are you cure that it was us re-25 quiring the:a to do that? Because in general BWR systems are Acme Reporting Company l
77 1 better able to be flow tested. 2 MR. EBERSOLE: Well, my experience was only the 3 fact that I saw them first on boilers and then I started es ( ) 1' 4 looking for them on BWRs and didn't find them and I found 5 this interesting. 6 MR. MICHELSON: You can't post test them through 7 these checks. 8 MR. EBERSOLE: So I found this peculiarity and I 9 think I--I certainly suggest you investigate why you either 10 have to have them or you-- 11 MR. MICHELSON: Just a nice neat answer on how you 12 knew how well the check was doing. And he thought it was , ~) 13 important enough to spend the money. ( 14 MR. EBERSOLE: He didn't think of it. The regula-15 tors thought of it, according to them. 16 MR. MICHELSON: No, no. I don't think so. 17 MR. KARSH: Well, we'll see what we can find out. 18 I'm not sure we're going to be able to find an answer, but 19 we'll try. 20 MR. EBERSOLE: It brings its measure of hazard. 21 You might exercise them and they stay propped open. 22 MR. MARSH: It's better than disassembling. f3 t 23 MR. EBERSOLE: True. Okay, go ahead. 21 MR. HUANG: In the pump test we measure t'oe inlet 25 pressure. We measure the discharge pressure. Then we Acme Reporting Company ,,0,,
78 1 calculate differential pressure. across the pumps, and we 2 measure the flow rate. We also measure the fibration of the 3 Pump today at the code and vibration amplitude. It's a better I') Es 4 technology used by many licensees. 5 We also observe and verify lubricant level and the 6 pressure during the time and we also measure bearing tempera-7 ture. But bearing temperature, we only measure it once a 8 year. 9 MR.MICHELSON: But you also should point out 10 clearly that this might be under a very nominal flow condi-11 tion that you do all of this. There's no relationship to 12 what it might have to do when you have the accident. I) 13 MR. HUANG: Yeah. v 14 MR. MICHELSON: And the bearing temperatures might 15 indeed be very low for this nominal flow and it yet again 16 under the loadings of a full load it might go up. And on 17 the other hand it could go down. You just don't know. 18 MR. EBERSOLE: Well, the bearing temperatures will 19 follow the temperature of the fluid being pumped anyway. 20 MR. MICHELSON: Well co some extent that's true-- 21 MR. HUANG: Unless there's a-- 22 MR. EBERSOLE: I notice also you say you checked U, y 23 bearing temperature. Not by checking bearing temperature but 24 by detecting the flow of the bearing lubricant someplace in 25 here. Acn;e Reporting Company m, n,,...
79 1 MR. HUANG: Yes. 2 MR. EBERSOLE: And I thought, well, at many pumps 3 they don't have circulating lubricant. What do you do then? (_) 4 MR. HUANG: We rely on the better of vibration l 5 technology. 6 MR. EBERSOLE: You don't do the bearing temperature 7 check? 8 MR. MARSH: The latest code, should it be approved 9 and implemented, it's OM-6, and it takes out bearing tempera-10 ture as a parameter that's being measured and is much better 11 in vibration amplitude by measuring velocities. 12 MR. EBERSOLE: Well, on vibration amplitude, I ( ; 13 couldn't figure out from what point to what other point you 14 measure vibration amplitudes. It seems to me there would be 15 a broad selection or why you couldn't alter vibration ampli-16 tude by simply screwing the foundation down tighter. 17 MR. MARSH: It's not good there. That's an IWP 18 you are looking at. 19 MR. EBERSOLE: Yes, that's the only thing I had to 20 refer to. 21 MR. MARSH: It doesn't tell you very much about how 22 to do it there. g-.) 't j 23 MR. EBERSOLE: Does the amplitude in fact reflect 21 the shaft movement within the journals? 25 MR. MARSH: hot necesarily. Not necessarily. But Acme Reporting Company < m..>.....
80 1 you do it on the bearing housing. 2 MR. EBERSOLE: It's the whole thing? 3 MR. MARSII: It's the whole machine. r~'s (,! 4 MR. EBERSOLE: Well, I can fix that by tightening 5 it up a little bit. And so I don't know what the interrela-G tionship is. 7 MR. MARSH: State-of-the-art vibration monitoring 8 is to use biaxial on the shaft itself, proximity detectors 9 on the shaft, the velocity profiles, spectrum analysis. 10 MR. EBERSOLE: Are we going to go to state-of-the 11 art? 12 MR. MARSH: It's working that way. I'd 13 MR. EBERSOLE: By the way, this is a heavy part of \\_f/ 14 this--what do you call it? Not preventative but predicted 15 maintenance. 16 MR. MARSH: Is velocity? Is vibration-- 17 MR. EBERSOLE: Is vibration measurememt. It's a 18 very large part of the predicted maintenance. 19 MR. MICHELSON: Vibration gets much more severe as 20 you work down towards the minimum flow end of the pump. It 21 gets quite violent with some pumps just before your mini flow 22 line opens. There you have to do a little judgement as to w> Lj' 23 when to--how to do the timing. But I don't have to test any-24 where near that point. If I were smart, from the utility's 25 viewpoint, and never wanted to flunk one of these tests, I'd Acme Reporting Company ac.u....
81 1 play my machine through its full range of operation and find 2 where it-- the code says that's acceptable. Find where it 3 runs, because most of those pumps will have interesting points (-) 4 of flow where it vibrates severely and have other points 5 where they run smooth. 6 MR. EBERSOLE: Just like the main turbine, Carl. 7 They all vibrate at some stage. 8 MR. MICHELSON: Yes. So if you're smart, you'll 9 find the point to do your testing at which you know you will 10 always pass. 11 MR. EBERSOLE: And that may be way off from the 12 operating point. (a) 13 MR. MICHELSON: It may be way off from what it has 14 to do when it goes into an accident. 15 MR. MARSH: I think a bigger problem is the mini 16 flow. Rather than a vibration is to mini flow-- 17 MR. MICHELSON: Well, you get the high vibration 18 at the mini flow end of the operation. 19 MR. MARSH: Just before you open up recir. But my 20 point is that the code allows and we allow and plants are 21 designed so that they test in the mini flow mode. And mini 22 flow we know is a problem. Mini flow has caused degradation. ,3 r(; 23 Mini flow has resulted in problems. 21 MR. MICHELSON: Some pump vendors didn't specify 25 enough minimum flow requirement. 4 l Acme Reporting Company ,,0,, A
82 1 MR. MARSH: Right. Or some pump vendors did and 2 it didn't get implemented at the plant. And they had S per-3 cent flow rather than 20 or 30. /m i i \\/ 4 MR. WYLIE: Did you say the code was being trained ~ 5 on the temperature? G MR. MARSH: Yes, sir, right. OM 6 is the latest 7 cede version which is supposed to replace IWP and it takes 8 out bearing temperature as a parameter to be measured. It 9 loosens the windows on differential pressure and on flow and 10 the vibration amplitude which is in velocity terms is much 11 better. There is a much heavier reliance on vibration. 12 MR. EBERSOLE: But haven't they lost the antici-im vs) 13 patory aspect of temperature rise and now they are waiting 14 for actual physical damage? 15 MR. WYLIE: But that's a different cubject though. 16 I mean that's for monitoring performance throughout the-- 17 MR. MARSH: There are people that say that bearing 18 temperature is not a good parameter to monitor. Because if 19 you really have a problem with that thing, you are not even 20 going to be able to see it in bearing temperature. Because j l 21 the thing is going to go by the time you see it. As a matter r~s 22 of fact, if you have got a journal, the bearing temperature k-] \\ 23 will go up, up, up, up and all of a sudden as the bearing 21 wipes, the temperature is perfect. 25 MR. EBERSOLE: But isn't there only a pre-damage Acme Reporting Company ,,ca,.<..... \\
83 1 condition that you can detect is rising temperature. 2 MR. MARS!!: But the argument that bearing tempera-3 ture is a useful parameter when you test the pump every p' 's ' 4 quarter is not good. 5 MR. SULLIVAN: That's not a good-- 6 MR. MARSH: That's not a good parameter for pump 7 performance. Vibration is a better indication of pump health, 8 MR. EBERSOLE: Well, vibration is actually obtained 9 damage though, isn't it? You've already-- 10 MR. MARSH: It's a state. 11 MR. EBERSOLE: You've already-- 12 MR. MARSH: It's a state. It's on its way. It's () 13 between health and death. 14 MR. SULLIVAN: But you are measuring every quarter. 15 MR. EDERSOLE: Right. But you measured it at a 16 level where you can continue to run a long time with it 17 existing. 18 MR. MARSH: Maybe. 19 MR. SULLIVAN: That's the hump. 20 MR. EBERSOLE: Yes, that's the hump. 21 MR. SULLIVAN: In terms of bearing temperature, I 22 think that the code group had a choice of either deleting it ~.k) 23 or making it continuous. I think the annual requirement which 21 is what's in the code now is generally regarded as not being 25 of any benefit. Acme Reporting Company ,,u,,.,.....
84 1 MR. MICHELSON: Well, most of these devices have 2 some kind of measurement on a continuous basis. You know, 3 either embedded thermocouples in the journal or in the oil if ,a (-) 4 it's an oil cooling system. Or sometimes in the water. But 5 they are not very accurate because they aren't--the embedded 6 one is probably the most accurate one. When that journal startr ; 7 getting hot you've got a good indication that something has 8 gone wrong. But that's an expensive process to put those 9 journals thermocouples in there-- 10 MR. EBERSOLE: I suspect, Carl, the reason is that 11 there's no system that detects minor deviation from standards, 12 No matter what the temperature is, you need a real sensitive (v) 13 detector. 14 MR. MARSH: State-of-the-art, as I say, in rotating 15 machine health is basically vibration monitoring on the 16 shaft itself. Proximity detectors with spectrum analysis, 17 looking at the first and the second harmonic. And trending 18 that. That's really where you'd like to be. 19 MR. MICHELSON: Now, we we did that under accident 20 flow conditions that would be pretty good. 21 MR. MARSH: Pretty good. That's right. 22 MR. MICHELSO:i: But we don't do it under those %) 23 conditions. 24 MR. MARSH: No, we do not do that, no. We do not. 25 MR. MICHELSON: The next best--we could probably Acme Reporting Company +,n....
85 i squeeze it down to the mini flow line and see if it's behav-2 ing there. 3 MR. MARSH: One of the reasons why the code I ,(,) 4 sense is not willing to go that far is the perception that 5 it's black magic. You know, it's an air form that's out 6 there. It's like ISI. It's so complex that it takes special 7 education to do that and there aren't very many people that g are around. The fact that there isn't a good inspection 9 certifier process in the IST. ISI, you know, they have levels 10 of certification. You have an inspector that gets trained. 11 He gets a patch. It's a big deal. It's a mark of excellence 12 in the inckhstry. (~) 13 MR. MICHELSON: The first limitation is simply v 14 that they can't compute full core accident flow conditions 15 and they can't go much beyond minimum flow in some PWRs. 16 BWRs are in fair shape. But the PWRs are in very bad shape 17 on getting high flow for testing. 18 MR. MARSH: That's true. Let me go back 6nd 19 finish my point about inspectors. One of the problems in 20 IST is that there is not a certified inspector. There is a 21 certified ISI inspector and he has knowledge and proficiency. 22 There is no similar thing in IST which is more complex, ,b 23 There are more different aspects of this procesa and there are different levela of degradation. The code hasn't thought gg it important enough or we haven't as an agency pushed them 25 Acme Reporting Company e,. u.....
L 86 1 far enough to have there be certified inspectors. And it's 2 another symptom of the problem. 3 Now let's go down to the valve requirements. (m. \\-) 4 MR. HUANG: Another point I'd like to make about 5 the range--after we do a test, we always compare to the 6 reference value with a -- given for the pump parameters. 7 And as you see from that code it ranges from.9 up to 1.04. one pump can run at a.09 8 So if you have a two pound 9 and the other pump could run at 1.04, there's a further 10 10 percent difference. This could be one of the reasons that 11 you have one pump running against a deadhead. 12 Okay. Let's move to the checking on the valves () 13 that are intended to perform here on the next page. And as 14 you can see, I had specified in here--I made a table from 15 that IWV. There are several tests that we would like to per-' 16 form on the valves. One is exercise. Exercise including 17 exercise at open and exercise at cloce. This answeru the 18 question on the check valve. It depends on what function 19 that check valve is performing. 20 Now, also you indicated a manual valve here. And 21 in our recent review we found that there are some fifty 22 related system separated by a manual valve, and according /y d 23 to the code, you don't have to do anything about that manual 2.; valve, but we feel this is important enough to exercise that 2i manual valve because at the time when you need it, you want Acme Reporting Company ,m.......
87 1 to be able to open it. So if you never open and lubricate 2 it, you may run into problems. 3 MR. EBERSOLE: That brings up the question about (_) 4 certain manual valves that almost never open. And someday 5 you have to open it and this thing is a solid bucket of rust. 6 MR. MARSH: That's a problem. We've seen that in 7 a couple of operational events where the fact that these 8 valves had never been tested because they are quote unquote 9 not safety-related valves and not relied on explicitly in to safety analysis, but when the operator has to go stroke them 11 for some reason, they are rusted closed and cannot go and 12 that's a problem. And I agree that that's a scope issue at (^') 13 thic point, j 14 MR. HUANG: And there are certain valves that re-15 quire to be exercise and stroke timed. That's always applied 16 to power-opereted valves, motor or air operated. 17 And then we have another category of-- 18 MR. EDERSOLE: We already know that stroke time 19 under this code test is not necessarily-- 20 MR. HUANG: It's a static type of test and we have 21 no problem with the test itself. 22 We have another type of-- t 23 MR. 1BERSOLE: Wait a minute. Before you leave 2 the stroke timit.g, is there some relationship between when 25 you tighten up on the packing glands and when you do the Acme Reporting Company .n,,.,.....
88 1 stroke timing teut? 2 MR. HUANG: Yes. 3 MR. MARSH: There is a requirement of the code that n 4 '/ 4 says that if you alter packing or if you use something that 5 can perform, you can alter the stroke timing after you check G it. 7 MR. MICHELSON: I thought the code excluded packing 8 tightening as a requirement for retesting. 9 MR. MARSH: It does--if you stay within the manu-10 facturer's-- 11 MR. ? The code doesn't talk about that. 12 MR. MICHELSON: I thought somewhere-- (,) 13 MR. MARSH: We do. 14 MR. SULLIVAN: It's on page 181 of the version I 15 have. 16 MR. MICHELSON: 1817 17 MR. SULLIVAN: It's Article 3000. 18 MR. MICHELSON: I don't even have 181 in my set. 19 MR. MARSH: All right. Let me read it. I'm sure 20 it's in yours. It's in the first page of IWV 3000. It says, 21 "Adjustment of stem packings, removal of the bonnet, stem 22 assembly, or actuator and disconnection of hydraulic or 7S U 23 electrical lines are examples of maintenance that could affect 21 valve performance parameters." 25 So it doesn't exclude-- Acme Reporting Company . ~,.,.....
89 1 MR. MICHELSON: No, it wouldn't exclude-- 2 MR. MARSH: --adjustment of stem packing, but this 3 footnote does cause an operational problem. 4 MR. EBERSOLE: What about the case where I've heard 5 some plants to stop simmering or just slightly they run along 6 and retighten the valves after they've been closed on torque 7 switch to the point where you may not open it? g MR. MARSH-You mean you engage the operator and 9 turn the hand wheel? 10 MR. EBERSOLE: They manually tighten the valves 11 after the torque switch has done whatever it did, but not 12 enough. (~)T 13 MR. MARSH: Shame on them. L 14 MR. EBERSOLE: I don't know of any exclusion of do-15 ing that except common sense. It's got to come out again on 16 the impulse law. But I don't know of any prohibit on opera-17 tors going around and saying-- 18 MR. MARSH: "Let's torque up the - 19 MR. EBERSOLE: "Let's get it shut tight." 20 MR. MARSH: Bad practice. You defeat the torque-- 21 the closure switch. 22 MR. EBERSOLE: Right. And the calibration on-- (V 1 23 the opening stroke. But I don't know of any exclusion-- 21 MR. MARSH: No, I don't know of any exclusion. It 25 is tough to regulate for practices like that. I mean Acme Reporting Company a n sa....
90 1 operators can always defeat tests and safety systems if they 2 so choose. 3 MR. EBERSOLE: Well, you can shoot them or some-o 4 thing if they did or fire them or whatever and fine the com-5 pany. That might be an appropriate thing to do. If we find 6 the operators doing that you have some punitive action. 7 MR.SULLIVAN:It's pretty hard for an inspector to 8 catch. 9 MR. MARSH: It sure is. If on subsequent test the 10 valve doesn't stroke properly, like the first or second 11 hammer blow doesn't it lift it off its closed seat-- 12 MR. EBERSOLE: You don't get a second one. (a) 13 MR. MARSH: You'll get it when it starts to go and 14 then it'll come around and hit it with a second blow. 15 MR. EBERSOLE: Will it hit it a second time. I 16 thought the first blow took the starting switch out of the 17 circuit. And you didn't get a second shot at it. Am I 18 wrong? 19 MR. MARSH: Help me. 20 MR. EBERSOLE: I think the opening stroke executes 21 the hammer blow and from then on you've got no-- r3 22 MR. MARSH: You've got no torque. Until the switch U 23 times out. The old cwitch times out. 21 MR. EBERSOLE: Right, right, right. 25 MR. MARSH: And then your torque is-- Acme Reporting Company s,,.,.....
91 1 MR. EBERSOLE: I certainly don't know of any 3 successive hammer blown. But maybe there are. Do you know 3 of, Carl? (g _) ,g MR. MARSil: First and second hammer blows on opening ? 5 Is there just one? g MR. PAGE: One hammer blow. Then the actuator 7 comes around and engages. It's approximately 160 degrees 8 in turn. 9 MR. MARSH: Just one shot. 6 10 MR. EBERSOLE: b'ait a minute. And then what? 11 MR. PAGE: You were talking about the opening-- 12 MR. EBERSOLE: The statement was made that there pd 33 were two hammer blows. 34 MR. MARSH: Is that wrong? 15 MR. EBERSOLE: I don't know, but there's a lot of 16 things I don't know. Someone suggested that maybe we had an 17 impact wrench. There is only one hammer blow to get the 18 thing open. You have multiple hammer blows, but I don't gg think that's the case. 20 MR. PAGE: I think you just have one hammer blow. 21 MR. EBERSOLE: One hammer blow. I don't know why an impact wrench design wouldn't work very well. It would on (% ~~ repeat until it got clear. New design valve. 23 gg MR. PARSil: Go ahead. 25 MR. HUANG: Okay. It's also required to leak test Acme Reporting Company ,2v,,.<.<...
92 1 certain valves if they perform certain safety-related func-2 tions in the area. 3 And also we heard a lot of fail-cafe type of a (') '/ 4 valve and we also test in quality to verify that function. 5 For safety-relief valve we try to verify the set-6 point pressure on a staggering basis within five years. 7 That sums up all the testing required by the code. 8 Thank you. 9 MR. EBERSOLE: We have a fifteen-minute--no, wait 10 a minute. No, carry on. 11 MR. MARSH: If you want to, it's a good place to 12 break. / !.ss) 13 MR. EBERSOLE: Is it a good place to break?
- Well, 14 you don't want to break until you are ready to close.
Let's 15 carry on. 16 MR. SULLIVAN: The next slide talks about problem 17 areas and I would say we've done a pretty good job packing up 18 the first problem area which is technical problems. 19 I've given them a slightly different cut on this 20 slide so instead of getting into detail problem areas maybe 21 we could just talk about these concepts here. g3 22 The first concept, we've really hit on pretty L.) 23 heavily. There's we believe inefficient or deficient test-21 ing requirements in the code. Given the code and the way it 25 is, we also have another shortcoming which we're going to talk i Acme Reporting Company .,n
93 1 about later and that has to do with the fact that you really 2 need some sort of an implementation guidance available for 3 how we believe the code should be implemented. An example is (_) 4 the one that we talked about a few minutes ago with IWV 35.22. 5 What does a "normally open" and a "normally closed" valve 6 mean in the regulatory atmosphere. What do we mean by that? 7 he land up being in the situation where we explain 8 to utilities at these meetings we have what we mean by it. 9 If they haven't obviously caught on. But that's a tedioun 10 way to go about doing your review. 11 There ere many, many other areas where the code 12 need some sort of interpretive guidance. (~) 13 MR. MICHELSON: Well, isn't there a more fundamen-x- 14 tal issue here on testing though and that is the NRC has not 15 Prescribed what kind of tests should have been done initially. 16 MR. MARSH: Right. 17 MR. MICHELSON: Including pre-operational tests. 18 It talks about procedures and everything but it doesn't talk 19 about what kind of test you do on a valve as such. So the 20 IST was intended to be a repetitive thing to reverify so to 21 speak that the initial test done way back at the factory or 22 wherever was good, was still being validated. >( g3 MR. MARSil: Right. 21 MR. MICHELSON: But we never described what kind of 1 25 tests should be done to begin with. l Acme Reporting Company { .a n.>.....
94 1 MR. MARSH: That's true. 2 MR. MICHELSON: And that isn't necessarily a part 3 of your program I guess but it ought to be a part of some- ,O \\-# 4 body's program. 5 MR. MARSH: Well, our program is based on that pro-6 gram. Our program is based on you asserting bealth. You are 7 proving health initially. And then you establish a reference 8 value. 9 MR. MICHELSON: Are you trying to then determine 10 what kind of tests should have been done initially somewhere 11 or are you just trying to determine what should be done with 12 uhat is in the plants today? /^\\ ( ) 13 MR. l'ARSil : What we are trying to do is end up-- 14 MR. MICHELSON: It's two different kinds of pro-15 blems. 16 MR. MARSH: --with an IST set of programs which 17 means requirements, regulatory positions, technical specifi-18 cations, code activity, and a whole umbrella to make sure 19 that through the life of the plant the pumpa and valves are 20 able to do what they are intended to do. 21 MR. MICHELSON: ISI to me infers in-service in-gs 22 spection as opposed to pre-service inspection or testing. \\. ] 23 It's really testing, not inspection. 24 MR. E B'.RSO LE : Implicit in IST is that there was 25 a pre-service testing program at one-- Acme Reporting Company
- 95 1-MR. MARSH
- - That's right.
~ 2 MR. EBERSOLE: 'And you really have no fou1dation. 3 MR. MICHELSON: And then you tailor the *:STLto. fit ..[ v. 4 the pre-service-- l 5 MR. EBERSOLE: It's not unusual, you k ow. Where 6 ia the' party that does the PST7' 7 MR. MICHELSON: Yeah. 8 MR. MARSH: The PSI? The pre-service inspection 9 kind of thing. The pre-service testing? (0 MR. EBERSOLE: Yes. Can we reasonably extrapolate 11 your function to include that? 12-l MR. MICHELSON: It's one way to'do it. I think.it~may () 13 be a sufficient way. ~ 14 MR. MARSH: It's not in the scope at this point. 15 And I'm just left not having a goo'd answer at this.poitit 16 whether it should be expanded to include that. 17 MR. EBERSOLE: I would just endorse it. 18 MB. MICHELSON: That's getting back to the earlier le question I asked about. I think there's'some research that_- 20 needs to be done to establish what is a proper way in which' 21 to do valve testing to begin with and industry doesn't 22 necessarily do it properly. ' Industry does very little as a 23 ratter or fact in terms of other than simple job test like 24 putting a full differential on a valve and seeing if you can. 25 crank the gate open. That's the kind of testing that they do 1 Acme Reporting. Company. aw..... . ~.,.,.. =
.96 4 - 1 ats the. factory.. And thei , do tliat on prototypes, but: real 2 real tecting has'not been thought through. Now, the best ~ 3 that we have is the B-31-- you know, the1 valve code.- Ive. i. (:) 4 lI.. forgotten.- They-keep changing the."2mbers on me. 1641. ' ~ 5 That's as close as we--I used to work.on that committee. 6 And that's as close I think as we.ever came to trying to 7 prescribe testing--now, that wasn't meantito be in-service-. 8 That was meant to be factory. validation-type testing. But 9 there is no tie in with that as I could sence in what we are 10 looking at here-today. 11 But that is an important element that you pick up 12. on and carry through in-service, i(]) 13 MR. MARSH: Right. 14 MR. MICHELSON: But without that element being 15 established by the NRC, it's hard to see aow you get a good 16 in-service testing plan. 17 MR. MARSH: If you err.ed up With a test program. 18 that established health of an MOV today, and-that means.if 10 you could devise a program to make sure that it could do-20 its safety function today-- 21 MR. MICHELSON: By that you mean-- ~ 22 MR. MARSH: At an accident condition. 23 MR. NICHELSON: As it.came out of the factory. 21 MR. MARSH: I mean just today. You go out to a 25 plant and-you look at an MO torque and you nald, "I've got a ' Acme.Reportins.L Oampany <as,,.r u... ,s v-, as-.-,, 4.,...,,-. n,'-,-, _ _ r--,.n ,,.,,e..~,~,-,,,,-, ---..cr..,,
197 1 testing program. I'm going to connect. leads to his valve. 2- .And I'm going to do an analysis on it. And'I'm going to make 3 sure:that once having tested it, it's.able to do its; function. 4 throughout the life - " 5 MR. MICHELSON: . Practically'you can't do that. G-In some_ cases you can and in.come cases.you can't. 7 MR. MARSH:. Some you can and some you can' t. You 8 can't do it all. You can certainly do betterLthan stroke 9 time testing, I know that. 10 MR.'MICHELSON: Now, that-is.a more elaborate test 11 that you do once to get a good set of reference. values and 12 from that you would decide how~to go and pick a particular ' O 'a "" der e=a reveet te-14 MR. MARSH: You would pick current signature on 15 current and you would trend current to make sure that tlua l 16 valve is performing as you initially did it. I 17 MR. MICHELSON: But you have'to pick a condition 18. under which you are going to take your signature and see, 19 this is the kind of thing that isn't developed and I wasn't f 20 sure if you were going to develop that or whether somebody 21 else was. 22-MR. EBERSOLE: I'll give you a wild dream that I 23 once had that a prony brake ought to be put on:every. valve 24 shaft. 25 MR.-MARSH: A what? Acme Reporting Company . m,.a...... y
L .98 L 1 MR. EBERSOLE: .A prony' brake. At d.iscrete inter-2 vals you tested'under:the re'alistic' emergency' loads that'it' i 1! 3 would have to follow. That would be a costly addition to a [ ( ) l 4-valve. 5
- MR. MICHELSON
But even.it'would-not be--- I G MR. EBERSOLE: It would only arbitrarily: define 7' what that load was. .] 8 MR. MICHELSON: It wouldn't-be complete. It.would 9 be better than.what we're doing though. ~ .t i 10 MR. EBERSOLE: It would actually load the-valve' I 11 'under its test condition, which you don't-.do now. They just-j 12 flap back and forth and the most horrible part of that is O ta v e seaerete ra^ seetietice waeu tuer ao#'t neve a=v 1oea o= 14 them. I 15 MR.' MARSH: Right. I 16 MR. MICHELSON: ;Which has nothing to do with how t 17 they operate under, loads. 6 1 18 MR.-EBERSOLE: Which makes people who use PRAs, t 19 visionaries that think they;are dealing with.the real world, i 20 MR. MARSH: Let us agree that the~ program that we 1 i 21 have low and the framework that we have now isn't where we r want it to be and it needs to go in a certain direction. 22 23-And I'm not sure whether we should or whether another group 24 - 'should look'at reestablishing the baceline as part of this 25 thing. Acme Reporting Company-j .n,,.,....: ,-,.m-_ w e.--....u-.
99 1 MR. EBERSOLE: Wouldn't you think it would'be more .2 logical that you~ integrate that with your own effortu'so you 3-know your own baseline? Rather than have some other' bunch. )- 4 doing-that? 'i .s 5 MR. MARSH: Yes, but you are going.to hear about I 6 my. plans. 'I want-to keep my. plans at-least' realizable.- 1 .i 7 MR. EBEPSOLE: - Okay. ) i 8 MR. MARSL:. Within some kindLof time. frame. And 9 if I start making it'too big, I'm not going to'.get done what< 10 we need to get done.. l 11 MR. EBERSOLE: Okay. Well,Jbefore you start keep-12 ing the health of the child-- (~}- 13 MR. MICHELSON: Even incthe 1641, partlof this is-14 done by analysis, because even there everything can be done-- l 15 MR. MARSH: Can be' tested. ' i 16 MR. EBERSOLE: The reference is made,-Carl, to the r 17 hilt, you know, and you are really keeping the' health-of a- ? 18 deformity in the first place. And I don't think.that that's 19 what we wore after. l 20 MR. MARSH: Right. 21 MR. SULLIVAN: Well, what I'd like to do is mcVe 22 on to the legal problems. I think you are very well awarc of ( 23 the technical problems. 21 The legal problems center around what these various 2a regulatory documents say. We've got technical specifications ~ Acme Reporting : Company ..,o,,.,..... ^ v -,a,- -.-4.- -,,,,,e+,-,,w-rem-a s ,m .,-,..m,,,,,-m- ,,r ~~r,, ,,,,,pc.-n.s..--w--e +,,,y
~ ~. - 5-{ .10 0 ~ l' and Standard ~ Review Plan and a. rule. 10 CFR 50.55A. 2 If you'look at.the technical specifications. I '3 think I would ~ like to startL with -that fecause that kind of' E 4 drives the' review in a certain way. It:says that you should- ~ L 5 follow ASME code requirements excapt ithere written relief 6 has been gianted, okay. 7 And what that really means is that a utilities: 0 8 program should be ' reviewed before they put it into effect. 9 Because every-utility needs to have these relief requests 10 granted. At least some portion of -them. Some. fairly l'arge fraction of them do turn out to be reasonable to some ex-11 12 tent. () 13 When you look at 50.55A it doesn't say;anything li like that. It basically'says and people even argue over. 15 what this says, but the preponderance of: people seem to agree ~ 16 that what it says is.that within. twelve months of putting a ~ 17 program into effect, you need to.have your relief-requests 18 approved oy the staff. Okay. So that gives a-different~ 19 window. 20 MR. MICIIELSON: When you talk about a relief re-21 quest. The code talks about where practical you do sonnethir.g. 22 I can be in compliance with he ccde by not do_ c ac~ 2 og . O' i 23 because it wasn't practical, but do I.have to'ask ror at 21 relief request anyway? 25 MR. MARSH: Yes.. We have'to buy that. J -Acme Reporting ' Company an...... <,.,.,2,. ..J,,,... -.s,.,-,, L- .--..,,1,, 4 s,n. .,,-r
101 1 MR. MICHELSON: Okay. In other words,l,every case 2 where tne code-says where you can do'it, they have to.do it- '3 unless they get relief? 4 MR. MARSH: Yes,-~that's right,.They have'to'be 5-given-- t ~ G .MR. MICH'ELSON: ,You go'beyond the code'in that,re-spect. And everything is a requirement un.less reliefEis 7 8 granted. '9 MR. MARSH: The tech specs say you do what the '10 code says unless you get specific written relief. 11 MR. MICHELSON: .tes, but :he code doesn t say you 12 have to do it. The code says if'you canIdo,it. 13 MR. MARSH: By prac'ticality. The code also says ~ 14 where we have bought off into practicality..The code.has l 15 wording and language in it about-this to the NRR, the D'irec- [ 16 tor of NRR approving.where practical. 17 ~ MR.:MICH3L6CN Not the code doesn't. i 18 MR. MARSH: The regulation. I'm'sorry. The regu-: 19 lation. 20 NR. MICHELSON: But,the code--seo, it'doesn't s r.y - 21 p u nave to do certain things. -*t just Les if you can.do 22
- them,
.O j 23 MR. MARSH: Right. WeL1, the regulations-- 21 Met. MlCHELSON: And I was justitrying to-mJ:ke aure 25 that if they can't for any.eason, that's what.you are Acme Seportin31 Company,< .j ,,a,,.......: W .. e a. -, -- .+.h---,- .J4,*. s. ..a- .-.-..._.,.4
___= '102' r .1 talking.about-- 2 MR. MARSH: Impracticality. That's been the source 3 of problems because the regulation wonr' ' 7. ;^u 4 judge pract'icality. That'would not ba fne place where'you- .5 . would say that. But.it's very interpretive ~in there and we { 6 don't have'a policy internally on what is impract - 1.- So 7 it's basically been ad. hoc. As time.goes on, the criteria 8 changes. Utilities don't kno'w what'our-criteria is.becau~se' 9 it has changed. 10 MR. MICHELSON: It's not in'the Standard Review 11 Plan. 12 MR. MARSH: It's not in the SRP. It's not-in'the i () 13 code. It's not in the regulations. 14 MR. MICHELSON: It's just in people's heads then. P 15 hR. MARSH: That changes with time. That's bad. l 16 MR. MICHELSON: Yes. , j 17 MR. SULLIVAN: Okay. This tech spec we believe j 18 is appropriate. We-think-we should be reviewing relief ra- 'f 19 quests particularly because we don't have & guidance docu-20 ment out that talks about how we think the code should be 21 implemented. i l 22 However, because we have such a backlog of reviews, ( 6 23 the current situation is that the utilities and tht regions have basically had to treat that tech spec as if it's not 24 25 o f fective. And'that's a problem. .i Acme: Reporting Jompany L o u. w.... j
. ~ _. - 103 s 1 MR. MICHELSON: It's cn.the books but you'are not 2 enforcing it, so t'o speak. 3 MR. SULLI/AN: We really can't. Elght. Not yet () ~ 4 anyway. That's one of-our-- + .5 M R '. MARSH: It's a, problem.'.That is a problem, 6 because here is an inspector and he goes to Plant X and he i 7 sees that they are doing something.tha'. ir, h'is judgement i 8 they shouldn't be doing. And he looks through their books l 9 and finds that they submitted a' relief request and it hasn't f t 10 yet been acted on. Where is he in enforcement land? Is he 3 1 IJ supposed to enforce? Is he. supposed ix) allow them ix) con-12 tinue with a practice that's not in his judge.ac:. goou vr () 13 not? 14 Lately they get on thc phone.and they'say, "Hey, l ~i 15 -Ted, what do you think.about thin?- Should.we b6 doing that r 16 or not?" We get into a dialogue and' frequently we'll act on 17 it. We'll say.yes 'r no formally to what-they are danig. o 18 MR. MICHELSON: Yes, but you are ad hocking'a 19 situation you haven't looked at in a comprehanaive sense yet.
- t 20 MR. MARSH
- - That's right.
i 21 MR. SULLIVAN: We had a situation recently on a I i 22 plant where one particular relief requist~gc ? be a flap. (A. _/ i 23 It turned out that we did settle that one part_cular flap j i 24 but the rest of the program still has not been totally acted j 25 on, j i Acme Reporting Company l .na,.n..... b .,--...,_,r
104 1 MR. EBERSOLE: Well, is 'the' looseness-in-the code 2 iSself give you room to do this or do you ",ast do.it-. anyway? 3 MR. MARSH: Oh, the code?- You mean the regula- -4 tion. The code doesn't really get into.enforement type areas. 5 The regulation isn't. specific. The tech spec is what's the! 6 specific part that in at issue. J 1 7 MR. BBERSOLE: Well, part is the loosoness in the 8 code itself, where practical, et cetera. t '9 MR. MAR $H: Butithen the. regulation says you've 10 cot.to--we have to buy impracticality reasons. 11 MR. EBERSOLE: You are going to have to identify 12 impracticalihr. O 22 i
- ^"$":
"is"'- ^"d '"e" the'*e " "ve 8 14 the regs say we have.to buy it, the.. tech spec seys dcn't you 15 inplement it until we. buy it. p3 - MR. EBERSOLE: Yeah, okay. l 17 MR. MARSH: That's the enforcem'nt. process. e 18 MR. EBERSOLE: So you go beyond the code. 19 FJt. MARSH: Right.. Into:the regulations and into 20 the tec. specs. But there is an inco'aistency 'in theway- ~ 21 things are structured and you have the ASME code and then n you have the regulations and then you have the tech specs. '23 And they are not all on the same wave length. 24 MR. SULLIVAN: Okay. Another problem that we have-1 25 ja just the way this rule is written.. We believe that'it's= Acme Reporting. Company ao,,..... +
~ . 105 1 been written largely around ISI.- A lot of people refer' to 2 ISI when they mean ISI and IST. { 3' MR. MICHELSON: Yes,'I sense that.: l 4 MR. SULLIVAN: . Basically'IST appears to'be an 5 afterthought'in the writing of this regulation.-- It.seems a '6= to be a tag,on.. That particula'r regulation causes.us.some h 7 difficulties-in terms of what we believe to be the proper 8 scope of these programs..The~regulatior. in some places j i 9 strongly infers that the.only thing that needs.to.be! included i 10 in IST is ASME Code Class I, II and III pumps and. valves. 4 i 11 Yet in ather places, it says things like in-service l 12 test to verify operational readiness of pumps and valves i () 13 whose functions are required for safety. i 14 El..EBERSOLE: At this point let me ask you this.- 15 You are a mechanical engineer talking about pumps.-.h.J valhes. f f 16,ll Many of the things you say here are applicable to many other f 17 tnings like electrical apparatuses.- Who is taking-care:of a 18 that part of the barn? f 19 MR, SULLIVAN: Good question. There are not com- 'h i 20 parable tests for electrical equipment like there is-- [ 21 MR. EBERSOLE: Some of them are quite similar. I i 22 Over pressure, over voltage, overflow, over current.- l 23 MR. SULLIVAN: But I mean there isn't a regulation l 24 that says go and test your electrical equipment commensure-25 .with an IEEE standard. There is~not that.- Acme Reporting. Company c ao s,.u. .--y-.._,
1061 y 1 MR. EBERSOLE: Is that a void'in-the ISI business? 2 MR. MARSH: ISI? q 3' MR. EBERSOLE: I.mean IST. 4 MR. MARSH: I don't-know.; Right now-IST is sort 5' of--it's sort of a narrow--notLnarrow, but-it's focusedJin 6~ on-pumps and valves. Hydraulic performance of pumpstand.- 7 . valves. Making sure that it's able.,to do its intended mecha-d nicaliperformance function. And-that part~needs1to be. fixed: 9 up. 10 We're also' talking to'. electric people abSutLtheir 11 areas too. But beyond that, that's as'far as we're going. 12 MR. EBERSOLE: Well, I noticed that some of the. () 13 language in the test, it says isolate the driverifrom the" ~ 14 valve, and that's another -- out there. It sure is the 15 thing that runs either the valve,or the pump. .It has its 16 own set of problems. 17 MR. MARSH: You do some, of course, limited testing 18 of the electrical equipment by.doing the. hydraulic testing 19 too. If you test a pump, you test it's motor. 20 MR. EBERSOLE: For instance, if you are coming up 21 on some kind of a load on a valve or pump, you don't know i 22 what the margins are on the valve. 23 MR.' MARSH: That's true, that's true. ] 24 MR. EBERSOLE: You"know that came out in the Salem l 25 trip problem. We don't know margins or functions. Acme Reporting : Company .,a r, s a. i
w, -107 1 MR. MARSH: You know maintenance' policy is a big 2 issue these days. And IST.is'just one; area of maintenance ~ 3 and how its~ things are deficient. There is electrical-, ) 4 MR. EBERSCLE: It goes back into design though. 5-MR. MARSH: Being ' designed for it you mean?' 6 MR. EBERSOLE: Yeah,'I-could say the same things ~ 7 about, you know, about the mechanical-area'. The fundamentals 8 that start in' the factory..And the capabilities. The ' root-9 capabilities are not a part of the in-service testing.. Not 'I 10 now.- 11 MR. MARSH: That's true. I can't disagree. 12 MR. MICHELSON: -Well,. things like, say, under ( 13 frequency trip. There are cer'tainly tests that you can per-14 form on the components that do the tripping,'but there's no 15 requirement.that you periodically go in and do such tests. 16 It's set by utility policy as to how often you check those. .17 Isn't that right, Charlie? \\ 18 MR. WYLIE: I was'trying to wrack my brain. 19 MR. MICHELSON: There's no regulatory document I 20 that says you've got to go in every three months and check 21 your under-frequency trips. 22 MR. MARSH: I don't think so. O-MR. MICHELSON: But utilities do it on some periodic 23 - 21 basis. 25 MR. WYLIE: They do it periodically. 4 i Acme Reporting Company j .l <>as....... )
4 %.m 4 k3.x& e 4 ad-au+ r4.-J &~..a-4 p->+ e A h A 4-- J 4 w* 4 108 a I MR. MICHELSON:. I don't know if it's~ yearly _or 2 . what, because by experience they felt.that'theyLshould do it .3 about that often tofprotect their investment. l' 4 MR. EBERSOLE: On.the other side'of the coin,-I:
- j 5
' don't think there's even the existence of the over. frequency i 61 trips. 7 MR. MICHELSON: They don't even put them.in not 8 alone test them. 9 MR. EBERSOLE: Right. 10 MR. MICHELSON: It eliminates the' testing ~ problem. 11 They are not there at all. .i i 12 MR. MARSH: Some, things are tested byl technical () 13 specifications. Some parameters of that. sort on some elec-j 14 trical equipment. And some tech specs will say,'for example, 15 go and test to' complete trainEfrom pressure transmitter.all q 16 the way through to. actuation. And that includes. logic. That 17 includes everything.. 18 MR. MICHELSON: But it doesn't include under fre-19 quency, for instance. 20 MR. MARSH: No, no. I don't mean'to say that. 21 MR. MICHELSON: To do that, you've-got to go to 22 the component and do it. g-)) 23 MR. WYLIE: Oh, sure. There's no tech spec on 24 something that doesn't exist in the.first place. 25 MR. MICHE', SON: Under frequency does. Acme Reporting' Company ,,a,,.n..... , ;~..
.~.
- 109-1 MR.'WYLIE: 'Under frequency is a case.
2 MR. MICHELSON: That's what I was referring to. ~ 3 MR. WYLIE: Probably'not excess. ~ ~4' MR. MARSH: Over' frequency you mean? .5 MR.-MICHELSON:. ?h). Over' frequency isn't-normally 6 protected against-- 7, MR. EBERSOLE: .Over-frequency is~a: good field to 8 examination. It needs to be pursued. 9 ' MR. MICHELSON:' And.under voltage trips and so-10 forth. You don't testLthem' routinely. You do ~it by special-- 11 MR. EBERSOLE: It's supposed to be good stuff, Charlie. 12 It lasts forever. ~() 13 MR. MICHELSON: No, you te'st them periodically. 14 MR. WYLIE: What's that? 15 MR. MICHELSON: Your under voltage trips. 10 MR. WYLIE: Oh, sure. 17 MR. MICHELSON: 'But you doithat again:byfutility 18 standards. Whatever'the utility decides is the. frequency they ought to test them on. Nothing to do with' safety. 19 20 It's done by experience more on the importance of it. 21 MR. EBERSOLE: Well, I just wanted to reflect the 22 fact we're dealing in a part of the world here. O. 23 MR. MICHELSON: Turbine-tripping is another example, I 21. .I don't think that turbine tripping ever got in the tech l t 25 spec. f Acme Rep.orting. Company-aos,.>.... .z ..1
110 i MR. MARSH: It did. l 2 MR. MICHELSON: Did it get in there? Yeah, maybe 3 it got in there. 4 MR. MARSH: It depends. 5 MR. MICHELSON: By tech spec requirements or by 6 utility practice? 7 MR. MARSH: It depends on the vintage of the tech 8 specs. It depends upon the extent to which they rely on 9 the turbine trip for over pressure protection of the cooling 10 system. If it was taking credit for it on the safety analy-ji sis, it should have gotten in there. 12 MR. MICHELSON: Okay. Under voltage gives you a () 13 go d example where they don't put that in the tech specs. 'q/ 14 MR. SULLIVAN: Well, one of the points I was making 15 was that we end up arguing with utilities over what is the 16 scope of this program. Is it just ASME Code Class I, II and 17 III, pumps and valves? Or is it all pumps and valves that 18 are required for safety? 19 We have a host of other varied intimately related 20 problems. Some of them we've talked about already. Under 21 administrative problems we have SERs that we have put out f-22 that get put out of date as soon as the utility sends in O 23 another program. 24 We have from time to time issued interim reliefs 25 based on a quick review of the progritm. We may have, say, Acme Reporting Company .na,.,.n..
111 1 not reviewed the scope, but we may have reviewed the relief 2 request to make sure they all appear to be reasonable. And 3 issued an interim relief. Well, that's a problem in and of N 4 itself because it's not a complete review and they get simi-5 larly outdated by revisions being sent in. 6 The job of doing a review takes something on the 7 order of about eight staff weeks of EG&G time, in addition 8 to our own. It's a very tedious reviz, very involved. 9 The SERs we turn out are something on the order of fifty to 10 seventy pages long. It's quite a bit of work to review all 11 these relief requests. 12 We may get anywhere in the neighborhood of fifty () 13 or so relief requests for each plant. 14 Another administrative problem we alluded to pre-15 viously that I have on this slide has to do with the fact 16 that this tech spec is largely ineffective. And we are going 17 to be talking about that more later. 18 MR. MICHELSON: Are you using the standard tech 19 spec now in all this? 20 MR. MARSH: Most plants have that standard tech i 21 spec that gives that wording that I just gave you. That is, I 22 don't you-- 1 em 1 I \\_/ ) 23 MR. MICIfELSON: But a lot of plants still use 21 plant specific tech specs. 25 MR. MARSH: Yes, right. l Acme Reporting Company <ao,, si.....
=- 112 1 MR. MICLELSON: And how-do you handle'those? Do 2 you have to go-- 3' MR. MARSH: We're not trying to change the t'ech 4 spec now, and my' judgement is and I haven't. verified this, y 5 but about 60 or 70 percent of the' plants have that tech' spec G-in intent, either that exact wording or words to;that'effect. 7 And so it's operative at most of the' plants. And so it's 8 a problem at most of'the plants'. EIt may.not be-the exact 0 wording but it is still there as a requirement. 10 MR._ MICHELSON: Thelrequirement is-there on every 11 plant. It's just a matter of what tech-spec has'been written 12 in to them. () 13 MR. MARSH: That's right. 14 MR. SULLIVAN: The goal is the same. If we turn 15 out an.GER on a plant that doesn't have that tech spec, our 16 goal will'still be to keep that,SER current. 1
- 17 The basically have--again, these things are very l
16 intimately related. We Just tried to separate these things-10 out into different bins. 20 But we have an enforcement problem because,~you 21 know, Ted was talking about this earlier. Without an in-g 22 spector that will go to a plant and there's no SER for him 23 to inspect against to find out what we thought about a par-24 ticular. relief request. So he's kind of in the situation of 25 having to use his.own judgement. Acme Reporting Company aea,sa.....
l- -113 1 MR..MICHELSON: In the case of motor-operated 2 valves, you are thoroughly familiar with the problem of'the 3 torques and the limit switches and-so'forth by now.- Yet a 4 velve can pass _this periodic test-- 5 MR. MARSH:.Without any problems at all. 6 MR. MICHELSON: Under.Section XI without any pro - 7 blems at all. So how'are you picking up_that aspect or are - 8 you even picking up that aspect of the. motor-operated.va'lves? 9 MR. MARSH: You know, there wasLthe bulletin that 10 said we want you to pick three' systems, RHR, LPSI-and safety-11-injection system and it went to all plants-and it said, "How 12 do you test the stroke time? How do you test that valve? ~ () 13 And how do you ensure that theLtorque' rating in proper for 14 the switch?" And it says how do you basically ensure that 15 things are all ri'ght. 16 And it basically went'beyond--it did what the code 17 should have done. It said to plants, you should be-testing ~ 18 these parameters for this valve. You are not. Eki how ' are 19 you makingf sure that it's all-right. 20 A lot of responses came back to show that there 21 wasn't adequate testing going on, and so there is other 22 action being planned inside. Subsequent bulletins or,some-O 23 thing to make it more--go and test these valves. Make cure i 21 you do this correctly. 25 MR. MICHELSON: Well,how is that testinig interacting Acme Reporting Company aaos,......
114 1 with your code, if at all? 2 MR. MARSH: First, we know it's going on and we 3 are making sure that we are aware of it and IST land. We \\/ 4 are carrying the message to the code committee. It's 5 recognized as an IST area as well as--there are many areas 6 that are legitimately IST that are not being handled by the 7 IST people. 8 MR. MICHELSON: So you are saying sooner or later 9 it ought to show up-- 10 MR. MARSH: In the IST program, that's right, 11 MR. MICHELSON: But not now. 12 MR. MARSH: But there are hurdles to go through to ,\\(,) 13 get the code to improve, to get the regulatory bodies al-14 together and getting them approved. 15 MR. MICHELSON: Are you going to tell un later how 16 you think that the staff is going to ensure that this is 17 getting integrated together? 18 MR. MARSH: As best we can at this point. I'm not 19 going to hold out great specificity at this point. 20 MR. MICHELSON: But it's certainly part of our 21 scope? 22 MR. MARSH: It's a part of what we want to end up g3V 23 with. 24 I just want to add one more thought to Ted's dis-25 cussion. And that was that IST was an add on to ISI. That Acme Reporting Company t20/# A/4 4h59
115 1 nessage-- 2 MR. EBERSOLE: What's the chronology of this sort 3 of thing? i i V' 4 MR. MARSil: IST was an add on, so-- 5 MR. EBERSOLE: When? 6 MR. MARSH: '74, 75 time frame. 7 MR. EBERSOLE: So it's an add on some years back? 8 MR. MARSH: Right. 9 MR. EBERSOLL: Way after the IST. 10 MR. MARSH: ISI. 11 MR. EBERSOLE: In the years then it has gone on 12 since then, this has been a growing thing, right? () 13 MR. MARSH: Well ISI has grown. IST has not 14 grown. ISI, in the requirements, in the code, and in the 15 programs at plants and quality at plants and in inspector 16 programs-- 17 MR. LBERSOLE: I'm just trying to get a feel for 18 the rising difficulties. How long have they been with us? 19 Are they peaking now? 20 MR. MARSH: I think we are much more aware of the 21 problems. And an awareness in the importance of maintenance 22 rx in plant life extension a1d reorganizations and a lot whole %-] 23 lot of things coming together to say we need improvement 24 here. i 25 MR. EBERSOLE: How did they cone into being? Under Acme Reporting Company 9024 628 4888
1116 1 auspices?' Did ASME do it? 2 MR. MARSH: Largely,a. change of'NRR..There was.a 3: change in management philosophy..There was a more regional 4~ people who were.much more familiar with these problems,.came1 =5 to management areas and said we need-to getlbetter. ~ And l ~ 6 that propogated down. 7 MR. EBERSOLE: Now, at the front end oflthis is I 8 going to be some maintenance programs?' 0 MR. MARSH: Right, right. 10 MR'. EBERSOLE: Okay. 11 MR. MARSH: But I do want to say;that this add on 12 philosophy is evident at the plant level too. It's not just () 13 here. It's not just in the regulations. It's_not just in 14 the code. It's also down.at the plant. When you go to a 15 plant and you say, "We want to talk to your IST coordinator. 10 We want to get all of the IST people together. .Let's talk i 17 about IST, problems, perceptions, how you do it." They sort l 18 of look at each other and they say, "You mean ISI?" You -l I 19 say, no. No,'we mean IST. And they say, well, okay, let's 1 20 see if we can get some maintenance guys here together. So l Ehe maintenance guys get together and in'one plant, one b'ig 21 22 utility's case, it was the first' time that the IST people \\./ 23 from the individual plants had ever sat down and talked to 24 each other before. And they looked'and they said, wo have ~ 25 never done this. Acme Reporting Company .n,..,...~..
117 1 MR. MICHELSON: Now, IST is not'a term referred to 2-in the regulations. 3 MR.. MARSH: Yes, it is.. LIn-service testing. In-f- / 4 service.preservices, pressure testing, in-service insepction - 5 .MR. MICHELSON:. Yeah, but I'm looking.at 50.55A, i 6 -Section G. And I never find it. I find plenty of;ISI,'but 7 I never find the words IST as such. But ma'ybe'I didn't have 8 time to.really look. Is it in-there somewhere? 9 MR. MARSH: Yes, it's Gr(i) and (ii). 10 MR. MICHELSON: Okay. t l 11 MR. MARSH: It's.at the bottom of page 473. 12 MR. MICHELSON: 473?- (). 13 MR. MARSH: You've got a different. version. 1 14 MR. MICHELSON: Well, you've got a different ver- ~ 15-sion than I'm-looking at. 16 MR. MARSH: It's the same wording. 17 MR. MICHELSON: Yes, the wording is the same. 18 Oh, down here, okay. But.that--the word-"in-service testing" 19 here is used in a different context. ' Performance of in - 20 service testing of pumps and. valves-is not in-service test-21-ing programs as such. 22 MR. MARSH: That's right. 23 MR. MICHELSON: 'It only refers to in-service in-24 spection programs, is all that's referred to in this section. 25 There is no in-service testing program. Acme Reporting ~ Company .,n......
i-118 1 MR. MARSH: True. 2 MR. MICHELSON: There's no differentiation made. 3 It's all ISI if you read the regulations. There's no IST. 4 IST is.something you--and I understand. 5 14R. MARSH: It does talk about in-service / testing. 6 MR. MICHELSON: NotEas a program requirement. 'If 7 you look at'the--if you look under G. In-service inspection 8 requirements. That's all under G. 9 MR. MARSH: Okay. Which is in-service inspection. 10 MR. MICHELSON: Which is also in-service-testing. 11 Or the testing of components in service. Aidifference. But 12 I don't find it as a regulatory-- .() 13 MR. MARSH: There's'no separate section for in-14 service testing. That's true. That's a problem. But you 15 get that into the wording under G and you find that there's 16 lots of words about in-service tests. There's also words 17 about in-service' inspection. There's also words about pres- -18 sure. 10 MR. MICHELSON: But there's no words about'in-20 service testing requirements. 21 MR. MARSH: Right, true. That's a problem. That's 22 a problem. 23 MR. MICHELSON: It's a small point, but--- 24 MR. MARSH: Well, that message is carried in the 25 industry. Hey, there's not a section in the regulation on ~ Acme Reporting Company . m,.,.....
119 1 in-service-- 2 MR. EBERSOLE: Th'at's also carried in the code 3 because there's only two sections that deal with pumps and 4 valves and the rest of Section XI, it all has to do with 5 all the other stuff. 6 MR. MARSH: Section XI is titled "In-service In-7 spection." 8 MR. EBERSOLE: Yeah. 9 MR. MICHELSON: Does in-service inspection have 10 anything to do with dynamic components in the plant? 11 MR. MARSH: No, no. 12 MR. MICHELSON: Well, I can argue I must look at [l 13 a machine that's running and determine it's not shaking it-v 14 self to pieces. 15 MR. EBERSOLE: That's not a code that's written. 16 MR. MARSH: Look at it from this perspective too. 17 The ISI is aimed at ensuring that yot r pressure bounda.ry is 18 okay. 19 MR. EBERSOLE: I know. It's an antique. 20 MR. MARSH: Bat it's also passive. It's making 21 sure that the passive part of the plant is okay. 22 MR. EBERSOLE: It's somehow related to the large -w i 'u J 23 LOCA era. 21 MR. n,'.R S H : Yes, it's the old idea. And the amount 25 of emphasis in the rega and ia the codes for the active Acme Reporting Company ,20,, si....
120 1 equipment as opposed to passive is less. 2 MR. EBERSOLE: Oh, we still suffer from the absence 3 of getting the heat out. [ \\- 4 MR. MARSH: Let me go through the policy procedure 5 change. I'm going to go through this slide quickly. And 6 then ue'll end up after this. And these are statements 7 that we made to the ASME code and things of that sort, and 8 I'm going to follow up this after the break with the closed 9 portion where we are going to talk about what we're actually 10 doing. 11 But thece are questions that the NRC is asking our-12 selves in tenns of policy and procedure. (v) 13 Because of the problem with the ASME code, should 14 the NRC continue to rely on Section XI t - the O&M to develop 15 pump and valve IST standards? 16 Should we require prior written approval for relief 17 requests? That is now what the technical specification says. 18 To what extent should updates to a later code 19 version be required? The whole precept was that you would 20 update yourself every ten years because the document was liv-21 ing and it was going to get better and you would grow. And 22 that's not been reality. Reality has been very different. -,yU 23 To what extent should use interim relief? And 21 I think I've expressed our problem with using that pill 25 philosophy. Acme Reporting Company ,,n..r....
121 1 Should the "exigency" policy be used in_ handling 2 emergency relief requests? The Agency has a policy for when 3 we will' entertain relief from our regulations and yout have'to ~ And that basically means 4 satisfy the exigency requirements. -5 that it has to.really.be'an1 emergency. Something that you 6 just found out about.- Something'that is going'to precipitate ~7 a plant shutdown. Should we use.that policy. 8 Should the NRC IST efforts be more inspection ~ 9 oriented anc less program and relief requect review 'riented? o 10 In other words, less office time looking at paper,,More en-l I 11 suring that the programs that are-there are doing what they 12 are supposed to be doing. (J 13 MR. MICHELSON: Well, how can you do that when you 14 haven't provided any guidance to the inspectors? l 15 MR. MARSH: You need to do that first. You need to 16 make sure you have sufficient guidance so that the programs-i 17 that are out there are good. 1 18 MR. MICHELSON: And then you can rely on'ihspectors, 19 MR. MARSH: Then you do the inspection. 20 -Should IST programs be reviewed by NRR in detail? 21 That's the same question as the previous one. 22 Should licensee be allowed to modify and implement 23 IST program revisions without NRC're*Lew? 24 Should the NRC hold regional and industry IST q ^ 25 symposiums? Structure?- This is an item _that the code was Acme Reporting Company _ .ao,,.>..... e
e v 122' 1 very interested-in..It;got me good feedback on that. And y 2 it's something that we are thinking very seriously about. We 3 have. good success'when we.do these' things..And it does'-tend 4~ to. normalize philosophies and approaches and pass.along good 5' 'information.tothesmaller'utilitiesandwethink~thatcouldk 6: be a very good idea. 7 And'how should we proceed with.these changes? 8 Should there be rule changes? Should we use generic letters? 9 Should there be a reg guide? New reg and schedules? 10 These are the questions that we consider when-11 taking the next step forward. 12 MR. MICHELSON: We've-talked a lot. And I want-() 13 to make sure I still' understand. What'is basically the 14 problem today? Is that'we just don't foresee'ever getting. 15 out of all this backlog? 16 MR. MARSH: It's that. 17 - MR. MICHELSON: You don't now how to cope with that 18 MR. MARSH: We're not getting~anywhere. 19 _ MR. MICllELSON: Or are we dissatisfied with :all 20 the paper that's been established, the test that establish 21 because they don't prove anything in practice? Or what'is 22 the driving problem here or is it a combination of a lot of 23 problems? 24 MR. MARSH: 'I don't want to say the driving. .But 25 one of the main concerns is from the resource perspective. Acme Reporting Company ,n,,.c.....
123 1 It's a wacte. It's using up a lot of contract dollars. A 2 lot of Agency time. 3 MR. MICl!ELSON: It's just not accomplishing much. ') 4 MR. MARSH: It's not doing whc' it's intended to 5 do. 6 MR. MICHELSON: And that's what's driving you to 7 want to-- 8 MR. MARSH: There are more reasons than that. 9 That's real basic reason that I have. It's a mess P.co. It's 10 a mess. Anytime that we can say that the technical specifi-11 cations are having to be ignored for discretionary enforce-l 9 12 ment, there is something basically wrong. 1. s ( ) 13 MB. MICHELSON: Shaky. 14 MR. MARSil: It's something really shaky here. 15 And when we rely on the code to structure itse]f and it's not 16 doing that, that 's a problem too. 17 MR. RICHARDSON: I think it comes down to the 18 basic question that we don't hc.ve confidence that safety-19 ielated pumps and valves will perform their uafety function 20 when they are supposed to. We don't have that ceafidence 21 MR. MlCliELSON : But only part of what we talked P' .i today will help me because you are only doing some c z.. i.c-:. tion tes ting-- MR. MARSH: That's right. We are not doing design 25 varlfication. Acme Reporting Company <sw.,.... I
124 1 MR. MICHELSON: You are not going back and really 2 determining that the component can really perform its 3 function to begin with. f i 11 4 MR. RICHARDSON: That's qualification testing. 5 MR. MICHELSON: But that's part of it. 6 MR. RICHARDFON: We're not doing that. 7 MR. MICFELSON: I know you are not. But having 8 qualified the equipment and then it's a verification problem 9 though to determine that it's staying in that ctate through-10 out the life-- 11 MR. RICHARDSON: We're not going to be able to 12 complete close the -- for just one program. 3 ( ) 13 MR. MICHELSON: I don't know how you can help but 14 go back. If you don't know if it works to begin with, how 15 can you kqep verifying thit it works. 16 MR. MARSH: We need to verify that regardless of 17 this reference value that you have picked, is the pump and 18 valve able to do its safety-related function? And if that 19 single point, checking all the parameters around that single 20 point don't give you that assurance, then you've got to re-21 think where you are. 22 MR. MICHELSON: And that's somebody else's scope. ,-) (/ ) ~? MR. MARSH: I'll be honest with you. I want to be 24 honest with you. We hadn't thought about that part of the 25 problem until you brought it up. That's the truth of it. l Acmt, Reporting Company n
125 1 MR. MICHELSON: Well, the two got to go togather. 2 MR. EBERSOLE: Let me ask about the reality of the 3 tech specs as you look at them and look at them carefully. k._J 4 Tech specs are funny things. You see, for instance, sharply t 5 defined times at which you have to shut down. It's arbi-6 trary. 7 MR. MARSH: Arbitrary. 8 MR. EBERSOLE: And ought one to examine the ar-9 bitrarinuss of the tech specs and examine each point of 10 arbitrariness? And argue v' y does it have to be that way? 11 What's the reality of that? 12 MR. MARSH: There's more--I think that's more f) 13 judgement than testing frequencies. v 14 MR. EBERSOLE: That's a far bigger scope than 15 we are talking about today though. 16 MR. MARSH: I think there's more data available 17 to help you define adequate testing frequencies and techni-18 ques than there is on shut down. You see such things ae 19 "you've got to do something in 96 hours." 20 MR. MARSH: That's another source of the problem. 21 MR. EBERSOLE: Who says 96 hours? 22 MR. MARSH: Let me. ell you where the 96 hours ceme -,s V> 23 from. We bring this up in everi mne of the code meetings. 21 96 hours was based on a long weekend concept. It was based 25 on their having sufficient time so that if there is a long Acme Reporting Company w. u..... I
126 1 weekend there would be enough window available to evaluate 2 on Monday morning. 3 MR. EBERSOLE: Oh, come on. fx i t 1_/ 4 MR. MARSH: True. 5 MR. EBERSOLE: I said it was arbitrary. G /.R, MARSH: But there is a little bases and it 7 isn't good but that's it. It's a long weekend concept. 8 MR. EBERSOLE: A sentence in here s a;'s, "The 9 resistance to the system shall be varied until either the 10 measured differential pressure or the measured ficw rate 11 equals the corresponding referene values." And of co'rse 12 they mean "and" and not "or." m(j') 13 MR. MARSH: No, they mean "or." s 14 MR. EBERSOLE: Ch, come on. 15 MR. MARSH: They mean "or." 16 MR. EBERSOLE: That's a paired parameter. 17 MR. SULLIVAN:No. You set one and then you measure 18 the other. i 19 MR. MARSH: You march alon., j l 20 MR. EBERSOLE: They are a mutched set. 21 MR. MARSH:
- Yes, i
s., 22 MR. bOLLIVAN: Sure, t hey are. V 23 MR. EBE iOLE: They indicated thov were separate. 21 MR. SULLIVAN: That's how you get referenced 25 values. Acme Reporting Coinpany .~...v..
127 1 MR. EBERSOLE: What they eliminated was the word 2 that you preset one of them. 3 MR. MARSH: Right. ,,\\ ,' s' 4 MR. EBERSOLE: Well,anyway, there's other evidence 5 of arbitrariness all through there and I was just sort of 6 thinking, well, maybe the tech specs, you could look at them 7 and loosen up things by treating them a little bit more 8 realistically. 9 MR. MICHELSON: I'm confused, Jesse. What does 10 that have to do w!th whether the valve could perform its 11 function or not? 12 MR. EBERSOLE: Well, it doesn't. Because that's-- }) 13 MR. MICHELSON: It just tells you that having de-14 cided it can't perform its function, how many hours or days 15 do you have to get it back into compliance. 16 MR. EBERSOLE: Yes. Tech specs only define the 17 health of whatever was as is. And they don't get back to 18 what it was in the first pince. 19 MR. MARSH: Lots of times the shut down times, 20 you know, are what is a reasonabic time to shut down the 21 plant with some analysis time'up front. It's not the commen-22 o surate risk in taking that long to do it-It's driven by b, 23 how long it takes to safely-- 21 MR. EBERSOLE: Valves with remote position indica-i 25 tors shall be observed two years, every two years. j A:me Reporting Company au,.<....
h J 128/1; 9l l -1 MR. MARSii. What's.the basis-for two years?' j '1 2 But I'll tell you that we would like to know. I'llftell you.- I 3 that we have asked for some rethinking of some of these positions so we can come up'with a more' technically d5fensi-4 -5 ble. testing program. 6 MR. EBERSOLE: Well,.I'll be glad when they get 7 -rid of just the passage of time to define when we do, things.. 8 I'd rather have a' seriesJ of events integrated. 9 Shall we have a fif'.een minute break? Are you 10 at the end of your fit-at part? 11 MR. MARSPt l ' '. at the end of my first part. The 12 plan was that a u ared--to have a break and then,have the e. ~T 13 closed session after that. (J l' 14 MR. EBERSOLE: All right. The meeting will be 15 closed after this. 16 MR. MARSH: Then they can come back.in about-- 9 17 MR. EBERSOLE: Come back in fifteen minutes. And 18 are we going to have any more open session? 19 MR. MARSH: Yes. 20. MR. EBERSOLE: Okay. Come back in fifteen minute 3. l 21 (Whereupon, there was a fifteen minute break.) () 23 j 24 25 -l Acme Reporting Company- ) .a.a.u.... a
~.. - 1 CERTIFICATE 2 3 This is to certily that the attached proc'sedings before_the 4 United States Nncleak kegulatory Commission in the matter of: (]) 5 Names Reactor Operations 6 7 Docket Number 8 Place: Washington, D. C. 9 Date Tuesday, J9nuary 5, 1988 10 were held as herein appears, and that this !s the original 11 transcript thereof for the file of the United States Nucleat* 12 Regulatory Commission taken stenographically by me and, 13 thereafter reduced to typewriting by me or under the direction l 14 of the court reporting company, and that the trans'cript is a '15 true and accurate reco{d of the foregoing proceedingc'. 16 /S/. akrr b /24bf ~~'~ ~~~~~ 17 (Signature typed)e IRUIN COFFLkE 18 Official Reporter 19 Heritage Reporting Corporation 20 21 22 23 24 25 () Heritage R9 porting Corporation (202) 678-4888 , -.}}