ML20247C359
| ML20247C359 | |
| Person / Time | |
|---|---|
| Issue date: | 06/13/1989 |
| From: | NRC |
| To: | |
| References | |
| GL-89-04, GL-89-4, NUDOCS 8909130336 | |
| Download: ML20247C359 (185) | |
Text
v - , 1 r
L September 8, 1989 NOTE TO: Jim McKnight Document Control Branch, IRM From: Tom Scarbrough Tm S Mechanical Er.gineering Branch, NRR
SUBJECT:
PLACEMENT 07 MEElIN3 TRANSCRIPTS IN NUDOCS At t a: r+d ar e the transcripts for two public meetings on Generic Letter EC-N. 'En.iaante o n Levelop ino Accept able Inser vice Tes ting Progr emt . ' As I ctist ussed wi th 32m Elorton eartier today. I would li6e these trar,stripts to be enteiec into NUDOCE for wide avallebility. From my discussion with Teresa Neville. I have alr eady sent copies of the transcripts to the Pub:ic Document Ecom 50 as tte have them publiziy available as scori as possitale.
If vo.i have any questions concerning this request, I can be reached at XE0916.
i o
l l
C909130336 t4ISC 090 PDR I 9130336 l
r _
^Tm ri
( w 3 J .\l P: _
UNITED STATES NUCLEAR REGULATORY COMMISSION ,
l i
INSERVICE TESTING NRC-342-710 GENERIC LETTER 89-04 i
Pages: 1 through 181 l
Place: Chicago, Illinois l I
Date: June 13, 1969 {
t
( HERITAGE REPORTING CORPORATION j OficialReportm 1220 L Street, N.W., Suite 600
<f>
Washington, D.C. 20005 (202) 628-4888
7_
. 1~
' NUCLEAR REGULATORY' COMMISSION'
. INSERVICE ~ TESTING 4
NRC-342-710 GENERIC LETTER 89-04 Tuesday, June 13, 1989 Bost'on Room O' Hare Marriott 2535 W. Higgins Road Chicago, Illinois'
' APPEARANCES:
TAD MARSH TED SULLIVAN HORACE SHAW HERB ROCKHOLD TOM SCARBROUGH CARL PAPERIELLO i
HERITAGE REPORTING CORPORATION (202)628-4888 c - __ _ _ _ = - - - _ . _ _ = - - - - _ - - _ - - - - - - - - - - - - - - - - - - . ----------------------A
i..
2 l.
1 PROCEEDINGS t
2 10:00 a.m.
3 MR. MARSH: Good morning. My-name is Tad Marsh.
4 We're going to have a series of speakers this morning and 5 we're going to discuss a lot of inservice testing issues.
6 I'd like to introduce Carl Paperiello who is the 7 Deputy Regional Administrator for Region III. He's going to 8 be giving some opening comments.
l 9 MR. PAPERIELLO: Good morning.
10 I'd like to welcome all of you to this meeting to 11 discuss Generic Letter 89-04. We have representatives from 12 both the region and our headquarters offices to discuss any 13 questions yu have on the generic letter, to give you a i 14 brief presentation on the generic letter.
l l 15 This is actually the second time I've been 16 involved with this. As some of you may know, I nappened to 17 be the regional representative, or I was until last Friday.
18 The regional representative on the NRC committee to review 19 generic requirements, so I had a very long presentation by 20 NRR on this generic letter in order to get our committee's 21 approval to issue it.
22 Inservice testing is a very important area. I 23 want to make something very clear. You don't do inservice 24 testing to keep the NRC happy, or if you are, you're doing 25 the wrong thing. You do inservice testing to make sure that i
HERITAGE REPORTING CORPORATION (202)628-4888
3 1 the pumps and valves that make up your plant, the critical 2 components of your plant work when you want them to work.
3 It's in both of our interests that these devices work when 4 they have to work.
5 We look at it from the viewpoint of safety. I'm 6 not saying you don't, but our primary interest is the safety 7 aspect of it. I, by the way, am not an engineer. I happen 8 to be a physicist. In fact, I'm a health physicist. I 9 spent quite a few weeks at Three Mile Island in 1979 and I 10 never want to go through that again.
11 So the important thing is we have devices out 12 there that we want to iork when we want them to work. There 13 are a couple of ways you do this. One is you need a good 14 maintenance program. We'%e not here to discuss that today, 15 but that's part of it. And part of the NRC's initiatives on 16 maintenance. This goes for everything, not just pumps and 17 valves, but the maintenance side of it. Then you have the 18 testing side of it. But the goal, ultimately, is to make 19 sure things work when you want them to work. Some of these 20 pumps and valves that are part of emergency systems, you I i
l 21 don't normally use, and the first time you know they don't '
22 work is when you call upon them to perform.
23 You have concerns which I guess we don't have, and 24 I want to make it clear. We understand them, except when we 25 get into it it's not something we have to deal with every HERITAGE REPORTING CORPORATION (202)628-4888
4 1 day. That is, when do you test and how do you test and how 2 do you optimize maintenance, because you have costs, and we 3 recognize that. I can test it, I've got to shut my plant 4 down, and that's an unreasonable thing to do. So there's a 5 lot to this.
6 In the past the NRC has done detailed reviews of 7 inservice testing programs. Frankly, we don't have the 8 resources to continue to do that. Anybody who reads the 9 newspaper for the last ten years, or at least since 1980, 10 knows that the theme in the federal government is to down-11 size. We just don't have the resources to do the detailed 12 reviews of inservice testing programs that we would like to 13 do. So basically, we're going to give it back to you. The 14 purpose of this generic letter is to explain to you what our 15 expectations are.
16 Another thing I want to point out, if we issue 17 something that isn't right, whether it's a regulation or a 18 generic letter, you should tell us. I can't promise you 19 we're going to change it because how it looks depends on a 20 lot of considerations. But you have an obligation as a !
21 regulated industry to tell us when we do something that's 22 stupid because we don't want to do that.
23 Keep in mind what our goal is. I think we share a j 4
l 24 common gcal. We want the devices to work when they're J 25 called upon to work.
HERITAGE REPORTING CORPORATION (202)628-4888 i
- _ _ _ - _ _ - - a
p o
5 1.
1 With that I'm going to turn it over to Tad Marsh, 2 and hope you have a good meeting.
3 Thank you.
4 MR. MARSH: Thank you, Carl.
5 Let-me first go through some logistics with'you. g 6 We have attendance sheets in the back of the room. Please 7 . sign. If-you're an NRC licensee, you needn't put your 8 address. We know what that is. If you're not an NRC 9 licensee, please give us your address. The reason for that 10 is going to be: clear in a minute.
11 We also have name tags in the back of the room for 12 your convenience if you'd like to have those.
13 There are cards in the back of the room for you to 14 indicate questions. We're going to ask that as you develop 15 questions during the meeting you write them down, pass them 16 up to the front of the room, and we'll get.them here at the 17 front of the table and we'll answer them.
18 I ask also on the cards with your questions, 19 please put down your name, your company, and the question-20 itself.
21 We will be taking minutes, as you can tell. This 22 is the third meeting we've had of this sort, third in a l
23 series of four. We've not done this before. We've not had 1
24 a court reporter before. We were in Philadelphia last week 25 and Atlanta last week, and we were busy trying to take HERITAGE REPORTING CORPORATION l (202)628-4888 1
6 1- minutes. It got to be real tough. We felt like the
.2 questions were so important and the answers we wanted to 3 make as clear as'we could, we asked for us to have 4 transcripts taken from the meeting.
5 The purpose of the transcripts and the purpose of 6 the court reporter is to help us in distilling everything 7 that's happened in coming up with meeting' minutes. The 8 transcripts will be available to you for clarification if 9 you.want through the public document room, but the idea is 10 we will distill from the transcripts answers to questions 11 and answers and write meeting summaries. We feel that's 12 important because we're going to write meeting summaries 13 -from all four meetings into one ummary, and that's going to 14 be sent to all participants. That*s important because you 15 may not have had a chance to ask a question that you think 16 of later on or may answer later on.
17 We will answer all the questions that we get, 18 either here today as best we can, or if not today, in a 19 subsequent meeting. Or if not, through the meeting minutes.
20 So you ask the question and we will do our best to answer it 21 today, but if not, it will be answered. The idea is we will 22 send everybody meeting minutes from all four meetings.
23 What I'd liks to do is follow a format that seems 24 to have worked pretty well in the past meetings, and that is 25 to first go through those questions that have been sent to HERITAGE REPORTING CORPORATION (202)628-4888 L__________--___--____ - - - - - - - - - -
7 1 us in advance. . We've.gone through them, we've got the 2 clearest answers, and we think that's probably on the-3 foremost of your minds anyway. After we're through with 4 those we'll then go through the cards that you have written 5 today with questions on them. Don't hesitate to write 6 questions on cards, really. We get to virtually all of L 7 them. The last couple of meetings we've had 40 or 50 8 questions from the. audience during the meeting, and we work 9 hard and we gat most of them done.
10 If the card routino is cumbersome for you, please l
11 use the microphone. Or we're a small audience, you can 12 stand up or use a loud voice and give us your questions from 13 the audience.
l 14 Let me spend just a minute talking about our 1
l
- l. 15 agenda. We've got opening remarks, 15 minutes or so. I'm 16 going to give some background on the generic letter and 17 background on inservice testing, about 15 minutes. Ted 18 Sullivan is going to give the majority of the presentation l 19 that's on the generic letter itself, how it's structured, 20 what it's meant to do, some detaile af it.
21 We're going to then have two sessions where we go 1
22 into questions and answers. Atlanta, we needed the entire 23 lunch period to go through questions, so we didn't take a 24 lunch break. We just ate as we were talking and working.
25 We'll see how this one goes. If we need a lunch break we'll HERITAGE REPORTING CORPORATION (202)628-4888 l
8 1 take it. If we need to keep working we'll do that too. 1 2 We didn't need to do a staff caucus at Atlu. r.a .
3 We've got this down as an agenda. If there's something we 4 really need to caucus on, to think carefully about, give you 5 an answer today, there is time allotted to do that. But we 6 didn't need that at Atlanta, so we're going to play that by 7 ear as well.
8 The idea generally though, is question and answer 9 sessions, two of them, with breaks in between. That's the 10 major thrust.
11 Let me give what we believe to be the basic 12 purpose of incervice testing. Its objective is to assess 13 the operational readiness of safety-related pumps and 14 valves.
15 I drat your attention to the word safety-related. I draw 16 your attention to that because that is what we believe the 17 basic purpose to be, and it differs in some respects from 18 what's in the regulation, per se wording, and also what is 19 in the ASME code.
20 There is a disparity between what regulation now 21 says, class one, two, and three components, and also what is 22 in the ASME code about the scope in IWB and IWP in that it 23 relates only to class one, two, and three components. For 24 those of you that saw advance copies of the generic letter, l 25 did anybody see advance copies of it? Good. Remember 1 HERITAGE REPORTING CORPORATION (202)628-4888 L - -
m- _ __
1 9
1 position one. It had in it the scope of inservice testing 2 was safety-related pumps and. valves. A lot of feedback from 3 the industry saying that's not right, go back and read more 4 carefully what the regulation says. As a result of that-5 we've modified position one and it's now position 11 in the 6 gene ic letter.
7 We believe basically, though, that that is the B intent. We understand that the regulation and we understand 9 that the code now limits us to class one, two, and three 10 equipment. That's why the generic letter is so structured.
11 But we do think that regulation changes, other 12 clarifications may be in order in the future to correct what 13 we believe the basic objective to be.
14 10 CFR 50, 55A requires pump and valve IST 15 programs in accordance with the Code in Section 11. That's 16 IWB and IWP. It requires that you update the IST programs 17 to the current code edition and addenda every ten years.
18 For those of you that have been familiar with the code, I'm 19 sure you're aware that it hasn't changed much in the last 20 ten years. It's been a pretty stagnant document.
21 For those of you that are again, familiar with it, 22 here it is. This is the ASME code as it pertains to IST.
23 It's about 15, 20 pages worth of technical guidance, 24 clarifications, requirements, etcetera. It's not much.
25 It's not much in the way of content. It's not much in the HERITAGE REPORTING CORPORATION (202)628-4888
10 1 way of clarity. It's not much in the way of guidance. It 2 doesn't really tell you what to do.
3 If you compare this to inservice inspection, 4 there's at least a ten-fold change between this document and 5 what the inservice inspection document is. Much more 6 guidance is provided there. Many appendices, some of which 7 are mandatory, non-mandatory. The code itself is deficient 8 in terms of inservice testing.
9 The regulation does allow granting of relief 10 requests for code requirements that are judged to be 11 impractical without giving any guidance as to what is 12 impractical. We find that a lot of licensees believe many 13 of these requirements are impractical, although there's not 14 much guidance in here for what that basic requirement is.
15 Let me give a moment's status. As a result of 16 many, many programs being developed, many revisions, many 17 added relief requests, many within the ten year process, 18 changes to the IST programs, there's a lot of volume to 19 inservice testing programs. There's a tremendous volume of 20 them. And in the past this has been a tremendous backlog.
21 I look for the cause of the backlog as being a climate.
22 It's a climate within the NRC. It's a climate within the 23 industry. It's a climate within the code. It's a climate 24 across the board. I think we all share some responsibility 25 in the situation of inservice testing as it is now.
HERITAGE REPORTING CORPORATION (202)628-4888
l 11 1 Many requirements that are in the code itself 2 haven't been clear, what is it that's actually supposed to 3 be done. Hence, a relief request gets generated. NRC hears 4 this plus 100 other relief requests from others, a little 5 bit different twist, and it's untenable.
l 6- Licensees, because in many cases you have to 7 implement relief requests. You have a plant to operate.
8 We're aware of that. You've made an equipment change or you 9 have to do some kind of testing on this equipment, so you 10 ' implement the relief request without its being specifically 11 approved in advance. That's contrary to the thrust of the 12 regulation and the tech spec. We know that and it's a 13 situation that we're trying to remedy. We think the generic 14 letter is going to fix much of that.
15 Some of the problems that we've seen in inservice 16 testing. We have seen a variety of what we believe to be 17 inadequate testing requirements in the code itself. For 18 example, the MOV's are allowed to be atroke time tested and 19 that's basically the extent of their performance tests in 20 the code itself. You know that AC induction motor powered 21 MOV's can't be adequately tested only by stroke time 22 testing. That's not sufficient.
23 Pump vibration testing in IWP is inadequate. OM6 24 is better in that regard. It's a much more explicit set of 25 requirements. The performance objectives are clear.
HERITAGE REPORTING CORPORATION (202)628-4888
12 1 Nonetheless, the current' code requirements-are weak.
2 Safety valves in class two and three systems dol 3 not have to be tested by either the current code or by the 4 latest code. Even the latest code doesn't require 5 explicitly class two and three system safety valves to.be.
6' tested.
7 Check valves. The code implies, leads you almost 8 directly down making a definitive decision. Does a check 9 valve have a safety function as either forward flow or does 10 a check valve have a safety function in the back flow? Does 11 it allow for bimodal, dual function safety performances? We
'12 know.that many safety valves have safety functions in both 13 directions.
14 The trending requirements of the code are weak.
15 There are other requirements. I think the point has been 16 made. Given that the code is weak and lacking, the NRC 17 could have and perhaps should have, given some explicit 18 requirements. What do we want? In the face of this 19 relatively weak baseline requirement, let's tell you what we 20 want. We didn't do that. That was a mistake. We should 21 have given ;fou core explicit guidance. We should'have said 22 this is what we want you no do.
re* wide
.23 We tried at one point. There was a r$d'quy'that 24 was written. It was just a draft. It was good. It had 25 some good stuff in it, but it never got formalized and was HERITAGE REPORTING CORPORATION (202)628-4888 l
I
13 1 never issued.
2 I've already talked about the huge volume of 3 programs, revisions, and relief requests and the backlogs 4 itself.
5 Many licensees don't submit changes to their 6 programs because the NRC has never said in what case do we 7 want to see a program, what types of revisions are okay 8 without submittal, what types are not, so the guidance 9 hasn't been there for you. And there's been the volume 10 itself.
11 I've already discussed the relief request having 12 to be implente,ed without prior explicit written approval, 13 contrary to the regulation in 10 and also the tech spec.
14 As you can see, you can understand clearly, if you 35 don't have a safety evaluation report, you don't have 16 explicit guidance from the NRC, in the face of weak, code 17 guidance, inspection efforts are very difficult. When 18 inspectors come to your plant and they try to see whether 19 you've got a good program or not, whether you're doing 20 things correctly or not, it's very difficult to say 21 definitively yes or no if there hasn't been clear guidance 22 to you from us, or from the code.
23 IST program implementation varies. Sometimes 24 poor. We've seen a lot of this. In fact when the latest 25 organization at the NRC happened about two years ago, the HERITAGE REPORTING CORPORATION (202)628-4888
14 1 first thing we did was come around to all the regions. .We 2 gatnered Commonwealth Edison. folks from this region, Duke 3 from Region II, large utilities from each one of the 4 regions. We sat down and discussed what's it like? What 5 kind.of IST problems do you have? Where can we improve 6 things?. How are things going?
7 In some of these meetings we saw cases where there 8 was not good corporate coordination, not good plant 9 coordination, the implementation varied from plant to plant 10 within the same ccmpany, so it was a mixed bag all over the 11 case. In some cases we saw no central focus of the plant.
12 In general, inservice testing is lumped in with inservice 13 inspection. You get a group that has both expertises there, 14 and they're very different expertise. Very different. A 15 guy that does an ND inspector is a very different guy than 16 that does a pump and a valve test, but yet they may be in 17 the same group of people that write the programs.
18 That's because, in my opinion, everybody views it 19 as a Section 11 requirement. Section 11 is a-code 20 requirement. And the code has it in Section 11, IWB and IWP 21 is just one of those sections that are in Section 11. So 22 therefore, let's have a group that corresponds with the 23 Section 11 group.
24 The NRC never said do it differently. Have a 25 separate group for inservice testing. It's a different l
HERITAGE REPORTING CORPORATION (202)628-4888 l
_--_=-__-_____ _ A
15 1 function.
2 We find many cases where junior people are given 3 the inservice testing program development as a task for them 4 and it becomes a training ground for them. And when they
'5 get expert in pump and valve testing they move on to maybe a 6 . systems expert, maybe operations, something different. So 7 you don't in many cases retain expertise within IST. You 8 use it as a developmental area and then move on to something 9 else.
10 Sometimes, maybe more frequently than sometimes, 11 we see poor organizations at the plant itself. We see that 12 the tests are performed by one group. We see the programs 13 developed by another group. We see the lines of 14 responsibility between those two groups are very fuzzy. You 15 should be able to go to a plant and say who has the 16 responsibility for inservice testing? Who is he? Who is 17 the man that's making sure the tests are developed right, 18 run right, and if there's problems they're factored into 19 plant operations? It should be closely coupled within all 20 the facets of it.
21 And I want to reemphasize something Carl said.
22 Many people, many plants view inservice testing as nothing 23 more than one of those things you've got to get done to 24 satisfy those guys back in Washington. Well, it's a lot i 25 more than that. It can be a true benefit to you. It's an HERITAGE TsEPORTING CORPORATION i (202)628-4888
)
16 1 excellent program as a feed into maintenance.
2 When you do your routine tests, when you do your 3 training, when you can see equipment that's heading in a 4 certain direction, it ought to send a flag to you that you 5 can do something before you get to the required action 6 range, before you have to declare something inoperable.
7 It's been given short shrift for a number of years.
8 Given those types of problems with the code being 9 inadequate, with the backlog of stuff, and the lack of the 10 NRC having said anything, we set out to give you better
~
13 guidance, and the generic letter is the first step in a get 12 well program.
13 What do we think it does? We think it does three 14 things. We think it provides generic guidance on 11 15 significant IST problems areas that we have seen before.
16 It's divided up into three specific areas. It's basically 17 alternatives to the code requirements, technical 18 specification / interpretation, and code interpretation.
19 There are three categories of those 2...
20 Having given you the guidance on what we want in 21 some inservice testing programs, it then says how do you 22 develop the program, once we've given you the guidance that 23 says what to do? Having given you the guidance and having 24 told you what to do, it then says whether you're approved or 25 not. So in our mind it clarifies the status of all these HERITAGE REPORTING CORPORATION (202)628-4888
17 1 back programs. It says we've taken 11 of the most 2 significant issues that we've seen out there. We've given 3 you technical guidance, policy guidance, administrative 4 guidance on what to do about those 11 issue. Now go fix 5 your programs to addrese these 11 issues. If you do that 6 and certify to that effect, the program is approved.
7 Then it says what to do in the future. If you 8 change beyond that program approval, what do to. It 9 basically resolves the 405 question for you.
10 As I said, Ted Sullivan is going to give you much 11 more detail on the generic letter, the workings of it, and 12 start into the question and answer frame, but let me go on 13 to talk about what things are going in the future.
14 I hope you're aware that there are new standards 15 coming out. OM6 and OM10. They are approved through the 16 BNCS including the NRC, we're a member of the BNCS, and they 17 are about to be endorsed in the regulation this summer, 18 maybe winter. About ready to be approved in terms of 19 Section 11 endorsing.
20 I hope you' re familiar with those new standards.
21 If not, there are some opportunities for you to become 22 familiar with it, but there are technical changes there.
23 There are improvements.
24 We have under consideration modification of the 25 regulation, 10 CFR 50. 55A (G) . There are two things that are HERITAGE REPORTING CORPORATION (202)628-4888
w w
)
18 1 being thought about at this point. One is a simple 2 regulation change. If you read the current regulation you 1
3 can tell that IST is an add-on. It's sort of a paperclip 4 that got stuck in there. The regulation wasn't developed 5 with IST in mind. The regulation was develooed with ISI.in 6 mind, Section 11 in mind. When IST got added the regulation 7 was just, there were a couple of phrases added to it, and 8 it's very difficult to read. You can't tell what this thing 9 is all about because it's just sort of add-on sentences 10 there, so it's poor and needs to be clarified just from.an 11 administrative sense making two regulations, one that 12 pertains to ISI and the other pertains to IST. Para:llel 13 wording, no technical changes, but clarification in terms of 14 what this regulation is all about. That's the first thing 15 that's being thought about.
16 The other is the longer range plan. The longer 17 range plan in terms of what do we want IST to be in the 18 future? What types of requirements do we want? Are we 19 happy with the way this relief request approval process is 20 working? Are we locked into generic letters from now on, 21 articulating what we need, what we want, putting into place 22 some type of generic approval. The regulation changes that 23 we're thinking about here are taking a longer view at that.
24 There may be future generic letters. This is the 25 first generic letter. It had 11 technical issues that we HERITAGE REPORTING CORPORATION (202)628-4888
L I l
l 1
19 1 felt were the most important and we could resolve and give i
l 2 you guidance on the most expeditiously. There are other l I
3 issues we want to address. We couldn't do it in this kind 4 of timeframe.
5 I want to advertise just for a minute the 6 symposium happening in Washington this summer, August 1, 2, 7 and 3. It's a jointly sponsored ASME/NRC symposium on 8 inservice testing. Invited papers on testing techniques, 9 testing problems, program development. We're going to have 10 the head of the ASME and the head of.the NRC, Commissioner 11 Rogers will be speaking at that time, along with the EDO.
12 It should be very informative, very good discussion 13 sessions.
14 There will be an invited paper session and another 15 session where other papers will be coming in and presented 16 on a variety of issues. The O&M people will be there to 17 talk about OM6 and OM10, discussing what these programs with 18 this new guidance is like. And to give you a white paper 19 that's been developed by them. The white paper takes OM6 20 and compares it to IWP line forline and tells you about the 21 differences. Why the code went to this way, what the 22 technical reasons are. The same thing for IWP and OM10.
23 That's the end of my pitch. Ted Sullivan is now 24 going to discuss with you the generic letter itself, how 25 it's divided, what it's meant to do in more detail.
HERITAGE REPORTING CORPORATION (202)628-4888
20 1 Questions, longer range stuff?
2 Okay, thank you.
3 MR. SULLIVAN: As Tad Marsh said, I'm going.to 4 start going through and discussing the generic letter and 5 some of our thinking in terms of how we put it together, why 6 we put it together. But I'm also aware that from the two 7 previous meetings we had there will be a lot of questions 8 about the generic letter. We already have questions from 9 three or four.different utilities, some of which are aimed 10 at how the generic letter works, what our thinking is.
11 Others are strictly technical questions. So I think between 12
~
what I'm going to say and the questions, we'll hopefully be 13 able to give a full discussion of what's behind the generic 14 letter.
15 As you realized when you read the generic letter, 16 there are basically three groupings of plants. There are 17 plants that are listed on each of twc tables, and then the 18 plants that aren't listed on any table. I thought I'd start 19 out by discussing what our thinking was when we put together 20 those tables, because we've already gotten quite a few 21 questions even before these meetings about why certain '
22 plants appeared on certain tables, and so forth.
23 (Q) Do you have any additional handouts?
24- MR. SULLIVAN: No, we don't.
25 MR. MARSH: We'll get some more copies for you.
HERITAGE REPORTING CORPORATION (202)628-4888
s < ~
21 1 MR. SULLIVAN: *If there are some of'you, multiple 2- attendees from a;given utility, you might be willing to 3 share them. We.are going to be. putting these overheads in f
4 the meeting summary.
1 5 MR.. MARSH: How many do you need? I 6 (Pause)
.7 MR. SULLIVAN: As the generic letter indicated, 8 the Table 1 plants are the plants with an SER that was 9 nearly completed. What we meant by that is the following. l 10 As you are aware, when we do an IST review, or when we've 11 done them in the past, the basic scheme has been to review i 12 the program, then send a list of questions to the utility 13 through the project manager. Then there's an IST review
-14 meeting set up. Customarily these have been at the plant 15 site and two day's duration. At the conclusion of the 16 meeting we generally ask that the utility revise the IST 17 program to respond to the issues raised in the meeting.
18 When we get that response, we sometime later issue an SER.
19 I know that that's a process that has frequently 20 encountered a lot of difficulties. One difficulty I could 21 characterize comes from getting possibly several submittals 22 in before we get a chance to send the SER out which 23 basically kind of reclocks the whole review back to the 24 beginning. In certain cases like that we never have gotten 25 an SER out. There are other reasons that definitely are HERITAGE REPORTING CORPORATION l
(202)628-4888 l 4
22
.1 attributable to our side as well. So I know there are a lot 2 of reviews that have occurred where an SER never was 3 produced.
4 But our basic thinking for how we delineated 5 plants on Table 1 had to do with whether or not there had 6 been a fairly recent meeting and a fairly recent resubmittal 7 for which we'd already started writing the SER.
8 If there was a situation where we had an SER but 9 it was several years old and there had been one or two major 10 resubmittals since then, we didn't put them on Table 1 11 because we basically would have been in the posture of 12 having to recommence the entire review.
13 I also have had it pointed out to me in a couple 14 of cases where there are some minor errors in this table.
15 There's an example of a plant on I think Table 2 that should 16 have been on Table 1. Those sorts of things have come up.
17 I think we've gotten all that sorted out through the PM's, 18 but I did want to make you aware of that.
19 Table 2 plants is also a grouping that's caused a I
20 little bit of confusion because some people have said I did 21 get an SER, why am I not on Table 2? Those were cases, to I k
22 the best of our knowledge, where the SER was still current. I 23 In other words, we hadn't received a major resubmittal since 24 the SER was issued for which we would have had to basically 25 redo the entire review. So we know there are cases where l
l l l HERITAGE REPORTING CORPORATION l
(202)628-4888 1
s, ,
o i-23 1- there were SER's issued'in plants that aren't'on Table 2.
2 If the plant isn't listed on Table 2, the decision was made 3 because the SER was no longer current.
4 As the generic letter says, Table 1 and Table 2 5 plants don't need to send in a confirmation letter on the 6 generic letter itself. But it is our intent, and I think 7 the generic letter makes this clear, that it's essential 8 that those plants review their procedures against the 9 generic letter and make sure that they are consistent with 10 what the generic letter is talking about.
11 We have some questions that have come up in cases 12 where a relief request was granted in an SER on Table 2 and 13 the relief request is not consistent with the generic 14 letter. We weren't really aware that that was going to be 15 an issue, but it came up pretty strongly at our last meeting 16 and we're going to have to adjust that in the meeting 17 summary. I don't think we're quite prepared to address that 18 particular aspect today.
19 On the previous slide I was indicating that the 20 SER is essentially what constitutes your approval of the 21 relief request. In the ;ase of plants that are not on Table 22 1 and Table 2, it's the mechanism of the generic letter that 23 constitutes the approval. There are certain steps, of 24 course, that we want you to go through in order for that 25 approval to be valid. What the generic letter is indicating HERITAGE REPORTING CORPORATION (202)628-4888
24 1 is'that you should review your programs against the attached 2 positions and make sure that they're in conformance with the 3 attached positions.
4 In those cases where there's a situation of 5 impracticality we've given an alternative mechanism for 6 addressing deviations from the code and that'has to do with 7 addressing the four points on page three that are basically 8 maintenance history related. We're asking that if you 9 decide to take that route, that you document it in the IST 10 program and submit that along with the confirmation letter.
11 Then we're also looking that in the confirmation 12 letter would be a commitment to make any modification that 13 you need to make for consistency with the generic letter.
14 That would be in the time frame that would be after the 15 confirmation letter.
16 We recognize that there's a certain amount of 17 short cutting that we're doing with this approach. The 18 short cutting basically has to do with all of those relief 19 requests that are submitted to the NRC as of the time of the 20 generic letter that aren't covered by the attachment one 21 positions. What we're basically saying through this generic 22 letter is that those relief requests aren't going to be 23 reviewed. They're to be considered as if they're approved.
24 So if you don't havy an SER or you're not going to get one, ]
25 in other words you're not a Table 1 or Table 2 plant, and HERITAGE REPORTING CORPORATION (202)628-4888 l
L_- _ - _
l .
L l l
l 25 L 1 you have relief requests that aren't covered by attachment 2 one in the generic letter, they' re approved through the l
3 mechanism of the generic letter.
4 MR. MARSH: Let me just clarify that. When the l 5 NRC wrote the generic letter we decided on the scope of the l
l 6 issues for which we were going to give technical guidance.
7 In other words, when we decided on those 11 that were going 8 to be included in the attachment, we decided to draw that 9 line to count those 11 based on how many relief requests 10 there are, what the safety significance of those whole set 11 are. We drew that line such that we would write guidance on 12 11.
13 There are other relief requests. There may be 14 more depending on your plant specific submittal. What we're 15 saying is if you have a program that's been submitted, 16 that's on the docket prior to April 3, 1989 when the generic 17 letter was dated, that program is approved in its entirety 18 as long as you change those relief requests associated with 19 the attachment one positions and conform with the attachment 20 one technical positions. Those that are outside the 21 attachment one area are approved. No further review will be 22 done. But what will be done is we've drawn the dividing 23 line based on knowledge of the scope and the extent of all 24 the relief requests, and we've said these 11 are the most 25 important areas that we need you to address.
HERITAGE REPORTING CORPORATION (202)628-4888
26 l' If you change,' if you rewrite your program based 2 on some plant modification or based on some rethinking of 3 your IST program or something like that, and you change in 4 the areas outside of attachment one, in other words, in 5 those areas for which we have not given guidance, the 6 generic letter says what to do in those cases. It says you 7 have to follow the' code, and if you can't follow the code 8 then you have to plan for relief in accordance with
- 9. regulation.
10 MR. SULLIVAN: As the generic letter indicates, 11 there is a mechanism for some alternatives that you can 12 take. When the positions in attachment one are impractical, 13 it is not our intent that we're going to review them and 14 issue SER's against them. We would look at them as part of 15 the course of an inspection, but it's our basic approach 16 that we're not going to respond to the way you document them 17 until we come to do the inspections.
18 We're looking for the IST progrnms to be sent into 19 us as a result of this generic letter. As you need to 20 revise your programs to conform to the generic letter, we're l 21 looking that you send that information in to us. Basically, 22 we don't want to change the mode that has been the practice 23 in this industry to date, which is that utilities send their 24 programs in or changes as they've made to them.
25 I'd just go back to the same point again a little HERITAGE REPORTING CORPORATION (202)628-4888 u _ ___ __ _ _
27 1 bit. When you send in a confirmation letter, it's not our 2 intent to go and do a detailed review of the program you l 3 sent in and the confirmation letter. If you do send in 1
4 changes to the IST program where.you're changing what's in l 5 the non-attachment one creas, in other words, you're adding 1
6 a relief request that's not covered by attachment one, then 7 that relief request will fall under the provisions of 8 50.55A(G) and will have to be reviewed.
9 What we'd be asking is that you clearly identify 10 in your program submittals those relief requests that aren't 11 approved through the mechanism of the generic letter.
12 So just to amplify a little bit, in the time frame 13 that you're responding to the generic letter, and assuming 14 that you' re not on Table 1 or Table 2, what would make it 15 most clear for us is if for each relief request you indicate 16 what the mechanism of approval is. Either it's a previously 17 submitted, non-attachment one relief request so it's 18 approved by the generic letter, or it's covered by 19 attachment one and conforms to attachment one so it's 20 approved by the generic letter, or it's outside the scope of 21 attachment one and it's new and it needs review by us. It 22 would help us to respond if the programs are delineated that 23 way, and it would also help you retain your conformance with 24 tech spec 405 if you can bring to our attention what needs 25 prompt action.
HERITAGE REPORTING CORPORATION (202)628-4888
28
-1 MR. MARSH: I think this is a very different mode l 2 that we're entering into. We're saying that for those 3 relief requests associated with attachment one, you've got 4 approval for those relief requests as long as you conform-5 with those positions, all the alternate tests as provided in 6 paragraph B.
7 Outside of attachment one, if it's submitted 8 already, was on the docket prior to April 3rd, it's 9 approved. If it's new or if it's changed, it's not 10 approved. You do not have approval for those relief 11 requests. You cannot implement that relief request without 12 prior approval.
13 (Q) When we write our new program, our program is 14 due to be submitted within the next year, on previous relief 15 requests that we've had in place, do we need to justify 16 these?
17 MR. MARSH: Are these relief requests in 18 attachment one? In the attachment one area?
19 (Q) No.
20 MR. MARSH: They're outside of attachment one?
21 (Q) Yes.
22 MR. MARSH: Have you received an SER on them?
23 (Q) No, we haven't.
24 MR. MARSH: Se you are a non-Table 1 or Table 2 25 plant?
HERITAGE REPORTING CORPORATION (202)628-4888
'29 I 1 (Q) Right.
l
'2 MR. MARSH: Are they new?
3 (Q) I guess my. question is if we submit the same 4 relief request again, will that --
L l 5' MR. MARSH: If it's the same as that which is on L
6 the docket, was on the docket as of April 3rd, it is an
. 7 approved relief request..You do.need to. resubmit it. Submit 8 it from the standpoint of this is part of your program, this 9 is a deviation from the code requirement,'this relief 10 requirement is approved by the generic letter. The generic 11 letter says if it was on the docket prior to April 3rd, it's e,
12 approved even though it's outside attachment one.
13 As Ted was saying, the best way to do that is just 14 t indicate from whence the appro al came for this relief 15 request.
16 MR. SULLIVAN: I would like'to amplify on that a 17 little bit, though. If as part of your update you, well as 18 part of your update I think you need to reevaluate each 19 relief request that's in your older program and make sure 20 with the current thinking at the plant that you still have l
l 21 the same situation, that the code requirement is still 22 impractical, it's something about which the system hasn't 23 changed, or instrumentation or whatever, and reverify that 24 the situation is the same. If the situation isn't the same, 25 then the previous relief request approval through the
'l HERITAGE REPORTING CORPORATION (202)628-4888
1 l
r 30 1 generic letter really can't hold any longer.
2 You've sort of led.into the next slide in the 3 sense that I've been trying to talk about the present 4 situation, but the generic letter was intended to be a 5 vehicle for the future as well. What we're indicating here 6 is that for program changes that are covered by the attached 7 positions, in other words the positions in attachment one, 8 what we're interested in is that you seek approval through 9 the same mechanism. In other words, use the attachment as 10 your guidance. If it's a relief request that conforms with 11 the attachment, then it gets its approval through the 12 generic letter. If it's a change, if it's a new relief 13 re que st,, it's going to have to come in for staff review.
14 It's really outside the scope of the generic letter.
15 The next basic bullet is in response to the 16 previous question.
17 We do intend to shift some of our resources from 18 program reviews over into assisting the regions on 19 inspections. We're going to be preparing a temporary t 20 instruction for this effort. We're going to try to go to 21 each utility in the country on a certain schedule. The 22 focus of the inspection will be the attached positions. As 23 we have time we'll look into other areas. As I tried to 24 indicate a few minutes ago, when we prepare for the 25 inspection we will spend our time at that point looking at HERITAGE REPORTING CORPORATION (202)628-4888
31 1 areas such as how you conformed with the generic letter.
2 What was behind your confirmation letter. That's when we 3 would spend our time reviewing the justifications that you 4 may want to use for deviating with the attached positions.
5 Before we get into the question and answer session 6 I wanted to check whether or not there's anyone here from 7 Arkansas Power. We know that's not this region, but it's 8 closer than California. If not we'll get on to some of the 9 other questions.
10 MR. MARSH: Before we get started on that, I have 11 to apologize. I did not introduce the rest of the people up 12 here at the table. That was my job and I didn't do that.
13 This is Horace Shaw. He's the Mechanical 14 Engineering branch. You've already met Ted Sullivan. Ted's 15 the Section Chief under me for IST. This is Herb Rockhold.
16 He's our contact at EG&G. This is Tom Scarbrough who is on 17 assignment to us from the appeal board.
18 Let's take a ten minute break and then come back 19 for questions.
20 (Whereupon, a brief recess was taken) 21 MR. MARSH: One of the things we're seeking from 22 these meetings is feedback from you on the generic letter.
23 For the past two meetings we've gotten some pretty strong 24 feedback that the six months time frames that are in the 25 generic letter are very very difficult to meet. What I'd HERITAGE REPORTING CORPORATION (202)628-4888 I
)
l l
l l
i
- - _ _ _ _ - _ - _ _ _ _ _ _ _ _ _ - _ ._. _. l
32 1 like to hear from you is the reasons you believe it's 2 difficult to meet, what are the logistics problems. Is it 3 something that's a p2.lority within the plant, or is it just l 4 resources not available within the industry? What is your 5 perception of the difficulty? Can I have some feedback?
6 (No response) 7 MR. MARSH: Then it must not be a problem.
l 8 (Laughter) i l
9 (A) My name is Joe - . Right now we're very 10 resource limited in my department to implement the change 11 that we're talking about here. That's why I feel six months 12 is a little bit too short a time to implement this.
13 MR. MARSH: Is the problem in program development?
14 Or is the problem in procedural writing once the. program 15 development has been done? Where is the problem? Can you 16 pin it down that way?
17 (A) Primarily it would be in program development 18 and reviewing the revision that we have currently submitted 19 with the generic letter, its impact to that revision and 20 relief request, the implementation of the modifications that 21 would need to be made then to our program and our 22 procedures, putting together the justifications in a good 23 enough format such that when you do get an inspection team 24 in, that we can readily show how we get from position A to 25 position B. That's a lot of background and a lot of HERITAGE REPORTING CORPORATION (202)628-4888 1
L 33 1 development work that we, just right now, are very resource 2 limited.
l 3 In my particular department we're currently 4 understaffed by about one open position that utility 5 management has at this time been holding open, hasn't 6 allowed me to fill. So I've got other jobs that people are 7 doing in the performance area that have to be done as well.
8 So for me, for my utility, it's a resource question.
9 MR. MARSH: Is this something that can be solved 10 if you went to a contractor? I'm not encouraging that, I'm 11 just looking for the resource perspectives to that.
12 (A) Oh yeah, we could go to a contractor, but 13 then you're talking resources there in terms of dollars, not 14 people. That's also limited.
15 MR. MARSH: We've also heard that there aren't l
16 very many contractors that are out there who do that kind of 17 work. I don't know whether that's true or false, but we've 18 heard that's a limited amount of people anyway.
19 Mark?
20 (A) I could give you some estimates of what we 21 did. We revised the whole IST program at Dresden Station.
22 To identify all the problems and revise the program it took 23 us two guys approximately six months including the fire 24 drills they came up with in inservice testing just to 25 program it. Implementation with modifications, depending on HERITAGE REPORTING CORPORATION (202)628-4888
34 1 when you can put the modifications on, could take anywhere 2 from a year to two years depending on your outage situation.
3 So revising the procedures to conform to your 4 program is a long and cumbersome road. Getting baseline 5 data for valve testing, setting reference limits, those all 6 have to be special procedures. We're about 85 percent 7 complete right now on our implementation, but we still have 8 two modifications that are going to be coming down the road 9 and that's been identified already as the next available 10 refueling outage, to put in gauges.
11 MR. MARSH: The generic letter gives you 18 months 12 or one outage after the confirmation letter for equipment 13 modification. So you're basically talking no less than two l
14 years from the date of generic letter through modifications.
15 (A) Yes. Depending on when you identify the l
16 modification and if it's necessary or not.
17 MR. MARSH: So it shouldn't be less than 18 18 months. It's the 18 months or the next refueling outage.
19 (A) Either unit, though. If you're in a 20 refueling outage on one unit right now, you might have two 21 years to put that modification together, identify it, and 22 then put in a modification. So you might have some extra 23 time to be added to that.
24 MR. MARSH: Thank you.
i 25 (A) I'm Gary Knap from Clod City Station. One )
HERITAGE REPORTING CORPORATION (202)628-4888 l
1
35 l
1 thing I also wanted to point out was that myself, I received 2 the generic letter through our corporate mail service the 1 3 first week in May, so I was one month shy on that end of it.
1 4 Plus my response is due back to my corporate one month prior 5 to the due date when it's due to you for their preparation 6 of the response. My response from the station has to be to 7 my reinsurance people approximately six weeks prior to your 8 deadline. So I have approximately working time of three and 9 a half months to do all of this. So the cut down, as you 10 can see, is...
11 MR. MARSH: There are a variety of reasons to cut l
12 into your time, I understand that. Some of which are 13 controlled by your own procedures and some of which are 14 controlled by others. I'm looking for resources that are 15 not available throughout the industry, or positions for 16 which you needed clarification, that you couldn't act on 17 until you had these meetings. I've heard that comment 18 before. If that is true, I need to hear that from you as 19 well.
20 There was a gentleman from NUMARC who was at the 21 first meeting in Philadelphia, Clyde Calloway, and he said 22 that utilities could contact him if they wanted to work 23 together towards something. I'm not saying yeah or nay, I'm 24 just offering that to you as something that was offered to 25 us. We said we would hear the comments and reflect them in HERITAGE REPORTING CORPORATION (202)628-4888 I
- - - , - - - . - - - - . - - - - . - . - . - - - - - - - . - - . - - - , . - . - ~
K'.
36 1 the meeting-minutos. We probably won't be- able to make any 2 kind of. decision in the meeting minutea. It's too big a 3, decision 1for mesting minutes. ,
4 But as Carl was saying, if the time ~ frames aren't 5 right, we' don't want to encourage poor programs, poor !
6 procedures, by setting up unreasonable times.- We really 7 don't want to do that. But we also want to make sure that 8 you' re putting the rigLt emphasis on this. We don't want to 9 say oh heck, take as much time as you need. We don't want 10 'o eay that. So we need to hear good, solid reasons. There 11 are some that are out there, and we'll reflect the comments 12 in the meeting minutes.
13 (A) My name is - , from Boston Ed. There is a
- 14 timeline in implementing this letter. People like to work 15 things in-parallel, but to implement this letter correctly, 16 you can't really'do that. Thero's a research and review 17 process.. They're going to want to make a plan, and they may 18 .have to write new relief requests. Granted, if they're
~
19 covered by the generic letter and they need it, it will be 20 accepted.
21 A relief requast can take a day to write 22 correctly. A little word that's unclear here or there can 23 change the whole intent. Also there's baseline data that l.
L 24 has to be collected at soms time. The acceptance criteria 25 has to be funneled up to the engineering organization, and HERITAGE REPORTING CORPORATION (202)628-4888
l 37
- 1. if there aren't resources for that they've got to be b
2 diverted. Diverting resources in a utility takes time.
3 Contractors. don't show up in one week.- Bids have-to be put 4 out.and that takes time.
5 If you start writing the procedures before you 6- have your acceptance criteria or your plan of attack, the 7 procedures may not meet their function. We have to have 1
8 this stuff almost in hand, not all of it, to write adequate 9 procedures. Sometimes it requires temporary procedures to 10 collect solid reference values to ensure that either you 11 don't have a suspect acceptance criteria uhat doesn't 12 adequately. monitor performance, or an acceptance criteria 13 that -- unnecessarily. An example is the MOV stroke time.
14 Trying to come up with solid MOV stroke times that are going 15 to, I've been playing around with acceptance criteria for 16 six months now, tnd I've tried them against equipment on a 17 trial basis, and in some cases I would have thrown equipment 18 into alert that was satisfactory and come very close to 19 inopping and putting a plant in an LTO, because if I had to 20 try to come out with an acceptance criteria too soon, it 21 would have caused the plant major hardship.
22 There is a time period for first time use in 23 refinement procedures.
24 MR. MARSH: Thanks. That's good stuff and that's 25 the kind of feedback we need.
HERITAGE REPORTING CORPORATION (202)628-4888
I 38 1 Another thing that Ted, you've been asking for in 2 the meetings are equipment changes.
3 MR. SULLIVAN: When we started off the set of 4 meetings last week I made the statement that it was our 5- intention to, or at least what we had envisioned was that 6 the equipment mods needed out of this generic letter 7 pertained to instrumentation only. No one contradicted that 8 in the first meeting. But in the second meeting there was 9 someone who indicated that wasn't going to be the case.
10 There were certain situations where a valve couldn't be flow 11 tested or disassembled. I think he was thinking of, certain 12 PWRs have these small' seal-welded valves and they really 13 can't be disassembled repeatedly.
14 I guess I would like to get your feedback on 15 whether there are other examples where modifications would 16 be needed to conform with the letter beyond the two that 17 I've just mentioned.
18 MR. MARSH: Think about that. We'd like to hear
.19 that kind of feedback as we go through the meeting. If 20 something comes to mind for equipment change beyond those 21 two, we'd like to hear from you.
22 The intent was that the equipment changes that 23 this generic letter requires are minimal. It would only be 24 flow instrument kind of stuff. We were surprised to hear 25 about the check valve changeout. We weren't sure how much HERITAGE REPORTING CORPORATION (202)628-4888
39 1 that would be too, how many valves would really have to be 2 changed out.
3 Why don't we go on to the questions and answers.
4 What we'd like to do is read questions and then give our 5 answer, those that have been provided to us already, and 6 then we'll go.to our cards.
7 MR. SULLIVAN: We have a set of questions from 8 Toledo Edison that we're going to start off with. A lot of 9 these questions are kind of inter-related, so we'll kind of 10 treat them as they relate to each other.
11 (Q) MR. SULLIVAN: The first one says, "Is it the 12 intent of position two of Generic Letter 89-04 that during 13 valve testing by disassembly that the valve be completely 14 disassembled and each internal valve part removed, if 15 possible, and 100 percent of the part visually inspected?
16 Or may only the valve bonnet be removed and the valve 17 internals inspected in place without the removal of the 18 internal valve parts unless elidence of discrepant 19 conditions are found, which would then require further 20 inspection and possible removal of the part?
21 (A) MR. SULLIVAN: Let me try to give a fairly 22 comprehensive answer rather than just a yes or no.
23 I'd like to start by saying that we don't really 24 consider disassembly of a valve to be a test. I know that's 25 sort of an interpretation that's coming through this HERITAGE REPORTING CORPORATION (202)628-4888
40 1 quest.'on and I think we may have written the generic letter
.2 unintentially in a way that leads people to believe that we 3 think that a disassembly is a test. It's really not a test.
4 It's a verification of the condition of the valve. It's not 5 really a. substitute for operability. I think what you 6 really need to do to verify operability is actually do a 7 flow test. So that's one point I want to mention.
8 Another point I want to mention is that it's kind 9 of indirectly embodied in this position, but we were trying 10 to indicate that any time you disassemble a valve because 11 you can't. meet position one, that you should do a partial 12 flow. That really is as close as you're going to get in 13 these situations to operability testing.
14 It isn't our intent through position two that 15 valves be broken down piece by piece and then completely.
16 rebuilt. . We think that is something that is a complete 17 maintenance issue that may be driven by other considerations 18 such as the recent problems that have been identified with l
l 19 certain types of Anchor / Darling swing checks. But for 20 purposes of position two, we're really looking that you do 21 minimal disassembly. Just take the bonnet off if that's all 22 you need to do, and reach in and manually move the disk, 23 inspect as well as you can the condition of the parts.
24 As this particular question indicates, if you see 25 discrepant conditions, then check into it further with that HERITAGE REPORTING CORPORATION (202)628-4888 l
41 1 valve and all the similar valves.
l 2 MR. MARSH: I'd like to add some more to that. I 3 just want to reiterate what Ted said about valve 4 disassembly, and that it is not a substitute for a good 5 operability test, although disassembly can tell you 6 something about the internals of a valve that a flow test 7 would not tell you necessarily, whether it's in good order 8 or not. But they ought to go hand in hand.
9 There ought to be a flow test and you periodically 10 disassemble to look for degradation of a valvc inside that 11 you can't see.
12 We really shouldn't be in the framework of doing a 13 disassembly test for an operability test. That's an 14 invasive test. That means you're having to break a 15 component down and by looking at it infer that when it was 16 assembled it would have worked for you. In some cases 17 that's a good inference to make. In other cases it's not.
18 I think we're in that situation because there is a lack i
19 within the industry of a good, definitive program of 20 substitute operability tests for valves that can't be 21 disassembled, valves that can't be readily flow tested. ,
22 For example, what would a good program be if you 23 can't disassemble the valve? If you can go put some flow 24 through it and yet you want to find out what the disposition 25 is, how do you do that? Is it radiography? Is it HERITAGE REPORTING CORPORATION l (202)628-4888 l
42" 1 acoustics? Is it a magnetic field? Is it some turbulence 2 techniques? what is the optimum way for a particular type 3 of valve, for a particular configuration, for a particular 4 flow? You can see there's almost an infinite set of 5 variables in that kind of a question, so it's not going to 6 be an easy program to develop.
7 I do want to say there is an industry group that's 8 been developed. It's called NIC. It stands for the Nuclear 9 Industry Check Valve program. It's a recently formed group.
10 It was headed up by Baltimore Gas & Electric who were 11 looking,for data to try to verify operability of check 12 valves. I don't know yet how far they're going to go or 13 what they're going to do, but I'm encouraged by the kinds of 14 things that they're looking at. They're looking at doing 15 flow tests, a whole section of valves, and seeing what is 16 the best way to find out the flow through that disposition 17 of that valve with that flow without having to invade a 18 valve. The name of the group is NIC, and I think they 19 report to the ASME subcommittee on performance testing every 20 meeting.
21 There needs to be, my opinion, a good concerted 22 effort in the industry to avoid having to disassemble 23 components that you really don't want to disassemble like 24 accumulator injection valves and PWRs. The primary pressure 25 may not be stopped. It's unfortunate that you have to be in HERITAGE REPORTING CORPORATION (202)628-4888
43
~1 the business of disassembling primary pressure boundary for 2 an operability check of a valve like that.
3 I'd just like to point out one other thing about 4 testing valves on a disassembly that Ted mentioned before 5 sort of in passing. In moving those parts that are movable, 6 in one case I could have prevented a reactor shutdown.
7 People had the parts out in their hand, did not move the 8 relative parts, the shaft through a sleeve, and that would 9 have shown that there was a resistance to motion there.
10 MR. MARSH: By all means check the operability of 11 the parts once they're disassembled, once they're available 12 to you.
13 By the same token, be careful, indeed to say this 14 because of the Salem event where the accumulator injection 15 valves were being tested after reassembly and through some 16 difficulties ended up injecting nitrogen in the system and a 17 loss of RHR. So you need to exercise care in putting the 18 system back together again as well, and testing 19 subsequently. This information notice that's coming on that 20 event, for those of you that haven't heard about it.
21 (Q) You mentioned about the movement of parts and 22 to make sure of free movement and there's no bowing, 23 etcetera. Are you advocating getting into a dimensional 24 taking and things like that?
25 MR. MARSH: No. If there is evidence in your l
l HERITAGE REPORTING CORPORATION (202)628-4888 1
44 1 opinion, after Cw.sansembly of the valve, checking 2 operability, it looks to you like there's something going on 3 that would warrant dimensional checks, by a11'means do it.
4 But not unless it's warranted.
5 M". . SULLIVAN: I want to add one more point that I 6 hope won't be confusing. We've talked about disassembly and 7 its limitations in terms of the code. As you're aware, the 8 way we've approached this in IST has to do with those 9 situations where a valve can't be so-called full flow 10 tested. But we do also want to put a pitch in for the kinds 11 of recommendations that EPRI and INPO have, come out with 12 INPO in the SOER 8603, and EPRI in their Applications 13 Guideline. That pitch is that all safety-related valves and 14 probably all valves in the plant, all check valves in the 15 plant for that matter, ought to be disassembled over and 16 above what the Section 11 requires on some frequency that's 17 appropriate for their service conditions. Because even if 18 .you can do full flow testing, you really can't tell what the 19 condition of the valve is. It's not a diagnostic test.
20 It's kind of a go, no go test. To determine what the test ,
21 condition. of the valves are we recommend that you follow 22 those guidelines in the EPRI document.
23 MR. MARSH: That didn't add too much confusion, I 24 hope.
)
25 (Q) MR. SULLIVAN: The next question says, "Do l
HERITAGE REPORTING CORPORATION (202)628-4888 l
)
45 1 personnel performing the visual inspections addressed in-2 position two have to be VT3 certified, ANSI 45.2.6 3 certified? Or may engineering-personnel competent in check 4 valve technical requirements perform this visual 5 inspection? 5' 6 (A) MR. SULLIVAN: I think it really isn't our 7 intent to try and add requirements right now that have to do 8 with certification. To the best of my knowledge, the 9 standards that are indicated in this question really 10 wouldn't qualify the inspector for check valve mechanical 11 condition. It really goes more to inservice inspection 12 requirements. So at this point in time we're looking for 13 utilities to use their procedures and their qualification.
14 guidelines under their own Appendix B program for this type 15 of disassembly.
16 (Q) MR. SULLIVAN: The next question gets into an 17 area that's come up in some of the other meetings. I may 18 have indirectly gotten at this but we'll get into it a 19 little bit more right now. It says, "Even if the check flow 20 valve testing can be performed as required by Section 11, 21 may the valve test be performed by disassembly as permitted 22 by position two in Generic Letter 89-04 when it is 23 considered by the utility that testing by disassembly will 24 provide the same or greater assurance that the valve will 25 function properly?"
HERITAGE REPORTING CORPORATION (202)628-4888
46 l; 1 (A) MR. SULLIVAN: We've gotten this question, I L 2 believe, because we didn't make it clear in the generic 3 letter that we felt there was a hierarchy here, and that we
, 4 felt that disassembly was only to be used in lieu of full 5 flow testing when full flow testing couldn't be done. -It's 1
[ 6 not our intent, as I indicated in addressing these questions 7 so far, that these two different positions are a one for one 8 substitute. We're looking for full flow testing as the 9 requirement that should be met. The position two we're 10 looking to be followed only when the full performance flow 11 test of the valve can't be performed. We're also looking 12 that when disassembly is used the partial stroke be used in 13 concert with it.
14 (Q) MR. SULLIVAN: The next question says, "If 15 the answer to Question IB is yes," and I think it's no, "can 16 the test frequency, etcetera, as described in the generic 17 letter position two be used in lieu of ASME Section 11 18 requirements even if the Section 11 test could be 19 performed?"
20 (A) MR. SULLIN AN: I don't think I need to 21 address that.
22 (Q) MR. SULLIVAN: The next question says, "If 23 the answer to Question 1B is yes, must a relief request be 24 processed or may this test by disassembly be noted in the 25 valve IST program submittal to the NRC7" l
HERITAGE REPORTING CORPORATION (202)628-4888
- - _ _- -___ - -________ _- __ _ l
47 1 (A) HR. SULLIVAN: I think since the answer is 2 basically no, that that dispenses with this question.
3 HR. MARSH: If we play this right we're going to 4 get done with this meeting real quick. (Laughter) 5 (Q) Mr. SULLIVAN: The next question says, "If 6 the answer to Question 1C is yes, must a relief request be 7 processed or may the frequency sample size, etcetera, be 8 noted in the valve IST program submittal to the NRC?"
9 (A) MR. MARSH: Let me go backwards to the 10 question that we didn't answer because the answer was no.
11 We're talking about a process point at this point. What 12 does the generic letter allow? The generic letter does say 13 on what provisions you go to position two. If you cannot do 14 position one, what do you do in position two? It says what 15 . you' re allowed to do and what you' re not allowed to do in a 16 position two. So the answer in you do not have to have the l
17 specific relief request approved for the implementation of 18 position two in attachment one. Is that clear to everyone?
- 19. If you're using the provisions under position two 20 of the attachment, that is the disassembly provision, that 21 you do not have to have a relief request explicitly reviewed 22 and approved? That you are approved by implementation of 23 that position? But you do need to document it because it is 24 a deviation from the code.
l 25 (Q) MR. MARSH: The next question says "If the HERITAGE REPORTING CORPORATION (202)628-4888 i
_ ~ - -
48 1 answer to this was yes, must a relief request be processed 2 or may the frequency sample size be noted in the valve IST 3 program submittal?"
4 (.A) MR. MARSH: The answer is you need to note 5 those things in the program itself, justify that change in 6 frequency.
7 (Q) When you say you need to document your 8 program, is that under a relief request but it just doesn't 9 have to be submitted before you implement it? Or is it just 10 a separate paragraph? How would you like to see it in the 11 program work?
12 MR. SULLIVAN: I think there's more than one way 13 to do this. Let me talk a little bit about what our 14 thinking on this particular point is.
15 When you use the terminology relief request, it 16 evokes the sentiment that it's going to get a response.
17 MR. MARSH: Because it's a request and therefore, 18 you have to have something backwards from the request.
19 MR. SULLIVAN: In these cases where essentially 20 the deviation from the code is approved through the generic 21 letter, that's for example position two, we do want you to 22 document the same type of information that you would put in 23 a relief request. You can call it a deviation approved per 24 position two of the generic letter as sort of a heading or 25 as a note, or you can call it a relief request approved HERITAGE REPORTING CORPORATION (202)628-4888
_ _--__-__ _ l
49 1 through position two in the generic letter. I-think either 2 way. But you need to make it clear.where your approval is 3 coming from.. If you want to change the terminology in these l
4 specific instances to deviation, that's fine. If you're 1
i 5 uncomfortable with that and you want to continue calling it 6 a relief request, that's fine too. But indicate that it's 7 already approved through the generic letter. It's sort of a 8 relief to the generic letter and the generic letter is your 9 response. Either way.
10 (Q) MR. SULLIVAN: The next question says, "May 11 the valve testing by disassembly / visual inspection 12 identified in position two of generic letter 89-04 be 13 applied to reverse flow testing of check valves?"
14 (A) MR. SULLIVAN: I think the same kind of 15 thinking is operative for both directions. If you can do a 16 reverse flow test, that's really what we're looking for.
17 That's actually a test. We would say if it's possible to do 18 it that way, that's what you should be doing. If it's not 19 possible because of different test type arrangements or 20 whatever, as long as we're not talking about inconvenience, 21 then the disassembly by taking the valve bonnet off and 22 checking the condition of the valve, the' position of the 23 disk and so forth, would be an alternative.
24 There does get to be a little bit of a problem 25 here though in cases where you have a valve where once you HERITAGE REPORTING CORPORATION (202)628-4888
50 1 take the disk off the internals come out with it. I would 2 say in that particular case disassembly doesn't fit the bill 3 for checking reverse flow.
4 (Pause) 5 MR. MARSH: We were just discussing whether in 6 fact position one and position two can be used for reverse 7 flow testing in check valves. When this generic letter was 8 written and when it was defended before the CRGR, it was not 9 a substitute for reverse flow testing. So I have to amend 10 what Ted was saying. The letter, I think I'm reading at 11 this point, is not a substitute for reverse flow testing.
12 (Q) That would be consistent with EPRI 13 guidelines. You can't do a disassembly to --
14 MR. MARSH: To prove the valve is closing tight?
15 I think it may get there at some point in concert with some 16 other type of test, but at this point I think we need to be 17 real careful on one and two. One is titled full flow 18 testing of check valves, and then two says as an alternative 19 to full flow testing of check valves. Those are the 20 applications that are approved.
21 (Q) A key point, that's correct because you 22 shouldn't reverse flow testing because many times you have 23 limit cycling on a valve, on a check valve. It will open up 24 and impact on its seat. It may do this ten cycles and you 25 warp the disk. This really happens. Thereby, if you really HERITAGE REE 3RTING CORPORATION (202)628-4888
51 1 flow test it, you'll pick this up. If yoi do a seep leakage 2 test you will not see that type of damage.
3 MR. MARSH: Let me make sure I understand what 4 you're saying. When you do a flow test on a valve you bang 5 against the open seat and --
6 (Q) That's correct.
7 MR. MARSH: And you damage the disk or the valve 8 by banging against --
9 (Q) No. It may go into limit cycle. You slow it 10 down or go into the seat and then it will lift off fivo or 11 six degrees, come back into the seat, lift off and come 12 back. That's what happens when you shut down systems. They 13 approved this, the modified valves in Canada. This is 14 published material. They call it limit cycling. So if you 15 do a flow test, a reverse flow test, you will actually pick 16 this type of damage up.
17 MR. MARSH: And you wouldn't pick it up by 18 disassembly of course.
19 (Q) No. Nor would you pick it up in the test 20 itself. It's sitting there, you wouldn't pick up the 21 physics of it. What we're after in operations -- I have a 22 paper that I'll give you on thet.
23 MR. MARSE: Great.
21 (Q) MR. SULLIVAN: The next question says, "Is it 25 the intent of Generic Letter 89-04 that the only reactor i i
HERITAGE REPORTING CORPORATION l (202)628-4888 l
1 52:
1 coolant system pressure' isolation valves to be included in .
l l 2 the ISt program are those listed in the tech specs and those !
l 3 which are event PIVs?"
4 (Pause) 5 MR. MARSH: Who wrote this-question? It's a good 6 ' question.
7 (Q) I think if you look at the third question on 8 down you'll see'what I'm getting to-by the next two 9 questions.
10 MR. MARSH: You're helping us so we don't paint 11 ourselves into a corner before we do-that.
12 (Q) I addressed that question for pre-1979 plants 13 and then I led up to what I was getting-at, that you guys 14 were eventually going to make us include all the PIVs, 15 RCS/PIVs, whether event fives or listed in the tech specs.
16 The older plants --
17 MR. MARSH: We know that.
18 (Q) That's what I'm leading up to.
19 MR. MARSH: I don't know where we're going long 20 term. I know there are a lot of concerns that are going on 1
21 right now so to speak, about Event V within NRR and testing 22 that's going on in PIVs and whether plants should get ne
-23 orders to include all the PIV Event V valves in the testing 24 program or what is the extent of this and that.
25 (Q) As I read the letter, the answer is yes to HERITAGE REPORTING CORPORATION (202)628-4888
l 53
~
1 that question. !
2 MR. MARSH: The letter was written to state 3 explicitly, at least for this letter within the tech spec 4 confines, because we knew we couldn't go beyond that at this 5 stage. That's not where we want to be, at least long term, l 6 but we knew that was specifically allowed for the short 7 term.
8 But that doesn't mean that PIVs should not be 9 tested. PIVs should be tested in accordance with the IST 10 scope statements. That's clear. Whether they should be in 11 their tech specs is another kind of an issue. That's a tech 12 spec issue. But IST, the scope statement of IST is 13 operative. We're not trying to redefine the scope statement 14 of IST. The generic letter did recognize that there could 15 be a difference between what the scope statement for IWV 16 says and what the tech specs say. So for this narrow cut, 17 we just said stick with your tech specs. It's clear there.
18 And by the way, make very sure that your 19 procedures pick up the testing of those valves too, because 20 we've seen some cases where procedures don't. Although the 21 programs say the valves are being tested, they're not by 22 procedure. So that's a corollary issue.
23 But as I was saying, we've got a scope statement 24 in IWV and it is operative too. We're not trying to relax 25 the scope statement of IWV. It's there.
HERITAGE REPORTING CORPORATION (202)628-4888
54 1 I've answered'p art of your question, I think.
2 (Q) I was looking'for a yes. (Laughter) This is 3 almost verbatim right out of the, if you take a look at that 4 position, position four, those words are, that's what you 5 say.
6 MR. MARSH: Yes, and we wrote it becuase frankly, 7 what happened, you guys can refresh me, when the Event V 8 orders came out in the '79 time frame they didn't cover all 9 'PIVs. We now know that. We didn't know that then. There 10 was sort of a narrow set of PIVs for which these orders 11 apply. And those subsequently within the tech specs. So 12 the tech specs now cover just those Event V orders, which 13 was a narrow part of all the valves that really do comprise 14 PIVs.
15 TER's were written by a contractor we had back l
l 16 then that said this was okay. So we realized that we 17 couldn't change the positions that we said then in this 18 generic letter. It was too big of an issue to bite off this 19 generic letter. We wouldn't get there. We'd get to the end 20 of this generic letter. All these other issues would be 21 slow if you had to justify changing technical positions that 22 shouldn't change, but you have to bite off the pieces we can 23 bite off on. So we said stick with your tech specs for the 24 mean time and make sure that your testing those valves.
25 Make sure.
HERITAGE REPORTING CORPORATION )
(202)628-4888 J I
J l.
L__ _ --- _ _ - - - o
55 1 As I say, there are other things going on, other 2 things happening. For those of you that are familiar with 3 the Event V precursor that happened at a foreign plant, 4 there's going to be inspections this summer into next fall 5 at six pilot plants. What's going on in PIVs at six pilot 6 PWR's. Let's make sure that they're actually testing them.
7 I'm not sure of the plants, but I think those that have been 8 selected know who they are.
9 I think I've given the answer to your question on 10 that.
11 (Q) MR. MARSH: 2C says, "Does the NRC anticipate 12 that some time in the futare requiring all RCS PIVs to be 13 included in the IST program?"
14 (A) MR. MARSH: I think that's a good goal. I 15 think we ought to steer ourselves that way. There are steps 16 that we' re thinking about taking that direction. So I don't 17 know how I can answer it any better than that.
18 (Q) MR. MARSH: Will Generic Letter 89-04, this 19 is the third question, "Will Gensric Letter 89-04 be updated 20 from time to time to provide additional positions on IST 21 programs in areas such as the following: the ASME Section 22 11 code does not require leak testing for valves where 23 leakage is continuously monitored. However, for PWR plants, 24 the NRC often requires leak testing for category A valves 25 such as the RCS accumulated core flood discharge checks HERITAGE REPORTING CORPORATION (202)628-4888
- _ _ - - _ _ _ _ _ - _ _ _ ___ .__ l
56 1 which are monitored continuously for seep leakage."
2 (A) MR. MARSH: But you understand, we're 3 required to forward flow test. That's not forward flow 4- test. We understand that backflow is monitored and there is 5 no specific tests on that amount of leakage because.of its 6 categorization, but the test we're talking about, position 7 one, are forward flow tests for those check valves. That's 8 where the problem arises.
l 9 (Q) MR. MARSn: "Do we intend to supplement the 10 Generic Letter 89-04 should it be needed?"
11 (A) MR. MARSH: We don't have any explicit plans 12 to supplement other than what I said in my opening remarks.
13 We're thinking about reg changes. We're thinking about a 14 subsequent generic letter on issues that go beyond this 15 generic letter. That's a pre-decisional kind of thing.
16 We're. working on that internally. We're working it through 17 the system.
18 Before we go into the other set of questions I 19 want to make sure you're all aware that there is another l 20 vehicle available to you for answering questions about the 21 code. That's, of course, the inquiry process or the code 22 case process. Then if you feel that the words that are in 23 the code itself aren't clear and you don't know what that l 1
24 means, like what does the torque value mean of the check j
)
25 valve? What is the basis for this 50 percent or 200 foot ;
l HERITAGE REPORTING CORPORATION <
(202)628-4888
i 57 1 ; pounds? Use the code process. Send an inquiry in, send a 2~- . code. case in, and ask the code to address that specific 3 issue._ What is the intent of that part of IWB that talks 4- about torque testing-for check valves?. They should be able
- 5. to answer that question. They wrote'the code and that's 1
6 what they're intended to be there for. They haven't been 7 getting a lot of inquiries. They've been getting some. But 8 make the process work.
9 The process really shouldn't be the NRC 10 interpreting all of the code. Of course that's a vehicle and 11 the regulation sets that up, but use the code too. That's 12 as good a process.
13 MR. SULLIVAN: The next set of questions is.from 14 Commonwealth Edison.
15 (Q) MR. SULLIVAN: "When the stroke time of a 16 power operated valve exceeds its reference or average stroke 17 time as established in accordance with position five of the-
'18 generic letter, but it's still within its plant technical 19 specification of FSAR stroke time limit, can performing an 20 evaluation which determines if the valve may remain operable 21 be used to sound select position five in lieu of making it 22 mandatory that the valve be declared inoperable?"
23 (A) MR. SULLIVAN: Position five is the one that 24 talks about limiting value of full stroke times for power 25 operated valves.
IRITAGE REPORTING CORPORATION (202)628-4888 i
f
58 1 I want'to clarify something that I was a little 2 bit confused about with this question. It starts out l
3 saying, "When the stroke time of a power operated valve 4 exceeds its reference for average stroke time." I'm a 5 little bit confused what you mean by reference or average 6 stroke time, when position five talks about limiting value 7 of full stroke time.
8 MR. MARSH: You mean the limiting value of full 9 stroke?
10 (Q) Yes. I mean a limiting value of stroke time 11 which is established from the average stroke time. A 12 certain percentage above that. That's what was meant.
13 MR. SULLIVAN: This is a question that's come up 14 in the previous two meetings. The answer hasn't been very 15 popular, but it's our intent with this position that you 16 establish a limiting value in full stroke time on a level 17 where you would seriously question the continued operability 18 of a valve. Where you would feel that there's a reasonable 19 chance that the next time you would have to call upon the 20 valve it may not work. For that reason, our philosophy or 21 intent, position, whatever, through this generic letter is 22 that when you exceed that limiting value the component be 23 declared inoperable. You enter the LCL, if that's what 24 declaring it inoperable forces you to do.
25 MR. MARSH: The only addition I need to make onto HERITAGE REPORTING CORPORATION (202)628-4888 j
59 1 that is the corollary concern about 24 hours2.777778e-4 days <br />0.00667 hours <br />3.968254e-5 weeks <br />9.132e-6 months <br />. The generic 2 letter does remove the 24 hour2.777778e-4 days <br />0.00667 hours <br />3.968254e-5 weeks <br />9.132e-6 months <br /> provision that does imply in 3 the code itself. The reason we took the 24 hour2.777778e-4 days <br />0.00667 hours <br />3.968254e-5 weeks <br />9.132e-6 months <br /> out is it 4 in many plants' technical specification basis, it says that 5 there is no increased period. When valves become 6 inoperable, there isn't an evaluation time period associated 7 with it. The 24 hour2.777778e-4 days <br />0.00667 hours <br />3.968254e-5 weeks <br />9.132e-6 months <br /> time frame should not be construed to 8 be an added time above the time when the valve exceeds its 9 limiting value of full stroke.
10 In answering questions about the 24 hours2.777778e-4 days <br />0.00667 hours <br />3.968254e-5 weeks <br />9.132e-6 months <br />, the 11 removal of the 24 hour2.777778e-4 days <br />0.00667 hours <br />3.968254e-5 weeks <br />9.132e-6 months <br /> provision, we asked licensees if 12 there were other technical specifications that gave grace 13 periods, and there aren't, to our knowledge. There isn't a 14 place that says if you go below so many GPM for a pump 15 you've got so much time to evaluate it. Or if you relay for 16 your safety injection system goes beyond 6.2 seconds, you've 17 got so much time to evaluate it.
18 So we feel like removal of the 24 hour2.777778e-4 days <br />0.00667 hours <br />3.968254e-5 weeks <br />9.132e-6 months <br /> provision l 19 is technically consistent with the technical specification 20 testing philosophy which IST is a part of. We feel the 24 l 21 hour2.430556e-4 days <br />0.00583 hours <br />3.472222e-5 weeks <br />7.9905e-6 months <br /> provision of the code really is a confusion factor.
22 Along with the 72 hour8.333333e-4 days <br />0.02 hours <br />1.190476e-4 weeks <br />2.7396e-5 months <br /> provision of the code, that's a 23 confusion factor too. When asking the code what is the l 24 basis for 24, what is the basis for 72, 96 hours0.00111 days <br />0.0267 hours <br />1.587302e-4 weeks <br />3.6528e-5 months <br />, I'm sorry.
25 Ninety-six hours for pumps. There isn't a good basis there l
HERITAGE REPORTING CORPORATION (202)628-4888 l
1 1
l l
l E________.__._
60-1 either.
2 So we've gone full route on the 24 hours2.777778e-4 days <br />0.00667 hours <br />3.968254e-5 weeks <br />9.132e-6 months <br /> and the 3 196 hours0.00227 days <br />0.0544 hours <br />3.240741e-4 weeks <br />7.4578e-5 months <br />. Where we land I guess is in the generic-letter 1 ~4 .itself.
I 5 Any comments you'd like to make on the 24 hour2.777778e-4 days <br />0.00667 hours <br />3.968254e-5 weeks <br />9.132e-6 months <br />?
6 (A) I thought that was taken away from a Region 7 .III letter, from our Region III letter, that we didn't have 8 any time to evaluate it once it got in the required action 9 range. That-was quite awhile ago.
10 MR. MARSH: We've worked with Region III and we 11 understand the Region III position. Our position.where we 12 .are. There are other regions that didn't get that position.
13 We needed to say it generically. We needed for everybody to 14 understand our basis for saying it so that it would be 15 clear. Because even though we said it here, Region III said 16 it, sne said it, it hasn't been promulgated.
17 (Q) When you go out there and test a pump and'say
'18 you test it and you get some odd ball number for 19 differential pressure and it 's outside your acceptance 20 criteria'and then the operator notices the head of the gauge 21 is broken, somebody must have hit it or something, do you 22 have to declare that inoperable? Or can you fix it?
23 There's a difference between a gauge being broken and a 24 gauge being --
25 MR. MARSH: You ought to have the gauge calibrated HERITAGE REPORTING CORPORATION (202)628-4888
L-:
l 61 1 and ready to' perform the test before you perform the test.
2 That's the optimum way of doing it.
3 (Q) I'm just saying, what happens when-they go 4 out there and.say there's maintenance going on and it gets 5 broken?
6 MR. MARSH: It's the same thing as if you're 7 operating and doing a test on the relays or you're doing a i
8 channel test on'the safety injection and you realize that l
9 one of the channels breaks on you. What do you do? That.
10 channel is inoperable.
11 (Q) .But those are different circumstances.
12 MR. MARSH: Why?
13 (Q) The pumps aren't going to fail, the failure 14 mechanism isn't going to, just because a gauge is broken.
15 Say you haven't met your critical date for performing the i
16 surveillance. I'm just saying it seems like a lot of red 17 tape and a lot of paperwork to declare it inop because you 18 know the gauge is broken and you can put a new gauge in 19 while you're running.
20 MR. MARSH: I understand that, and I'm trying to 21 make to you an argument that's from a consistency standpoint 22 with technical specifications. We're trying to treat them 23 all equally. There really shouldn't be an evaluation time, 24 per se. There really shouldn't be a time frame that you can 25 take data that may not quite be right and do something with HERITAGE REPORTING CORPORATION (202)628-4888
l 62 p.
1 it. .Because if procedures are clear, the guidelines are
~
-2 clear,.the people are trained, if the instrumentation is all l
- 3. ; calibrated and ready to go to perform the test. Ycu perform 4 the test and it shows-that equipment isn't proper, declare 5 it inoperable or do whatever you'need to do for technical 6 specification compliance. Fix it, and get out of your LCL.
7 That's not a popular answer. I understand it's 8 not a popular answer because that's the way that we've been-9 sort of doing things. It's not a good way. We've seen 10 cases where it's really been abused. Where the time periods 11 haven't been treated properly.
12 (Continued on following page) 13 14 15 16 17 18 19 20 21 HERITAGE REPORTING CORPORATION (202)628-4888 l
63 11 MR.'SULLIVAN: ' Generic Letter 89 of 4 statra in
~
2 _ paragraph D, "that when licensees are unable to comply with T 3 the positions of attachment 1, evaluation of alternate 4 testing should address four points: maintenance history of 5 individual components, maintenance history of related 6 components in similar environment, component vendor records 7 .and degradation of other facilities, records of degradation 8 of the same or like components from other utilities."
9 This is just a quote from the Generic Letter.
10 (Q) Is it mandatory for each instance to address 11 all four of the above items? In some instances or 12 situations, the above items may not apply, or only a portion 13 may apply.
14 MR, SULLIVAN: We realize, when we put together 15 these four criteria, that they were fairly tough to meet.
16 And'one of the reasons we went this route, as indicated 17 before, was to try and provide some alternatives that we 18 would be comfortable with in lieu of having to review 19 everything before you implemented deviation.
20 We deliberately chosa four fairly tough criteria 21 with the view in mind that if they weren't applicable, or if 22 you couldn't do them, then this particular vehicle isn't the 23 vehicle you should be following.
24 (Q) The next question says, "When evaluating an 25 alternate test to one of the position of attachment 1, HERITAGE REPORTING CORPORATION (202)628-4888
F- b i- , . 64 M '
, 11 Generic' Letter 8904, may the alternate test be implemented-l 2 without prior.NEC approval providing an evaluation as 3 performed and documented'and retained in the IST program?"
h 4 MR. SUaLIVAN: I want te just'go over this, even
'5 though we talked about this in my presentation.
15 Our view is that when you're evaluating an 7 alternate' test,.and using those criteriafI just read, our E
EL approach in the Generic Letter is that if you document-it 9 and' justify it fully'in the IST program, adjusting each of
- 10 the four points, that we're not going to evaluate it as a 11 _ relief request. And, hence, you can-treat it as ifLit's 12 ' approved by the Generic Letter, and to go ahead and 13 implement it.-
14 MR. HARSH: Yes, sir.
15 (Q) That full question really had in it sometimes
- 16 you can, maybe, be doing testing and you admit your 17 component is failing. So you're not really going to gain 18- anything by making its history, because you know that 19 former, but you haven't altered a test that you feel is 20 equivalent to one of the positions, or_you really want to-21 justify it on that basis, on the merit of the test itself.
22 That's really.the reason for the question. In other words, 23- in that case, some of that history would be not applicable 24 because you're not really -- or maybe it would be 25 applicable, the fact that you know have a component.
HERITAGE REPORTING CORPORATION (202)628-4888
.h3 65 l .
1 MR. MARSH: It's real tough foi us to evaluate the v -
L 2 particulars of that piece of. equipment, to say you got some 3 equipment that! indicates that 'it's going down hill, but yet, 4 you'got an alternate test, and through that test you believe 5 it's all rigLt. I don't want to give that kind of an 6 answer. Because that implies that at any time -- the L
7 minutes'could be read or anybody could think that anytime 8- - you got-a piece of failing equipment, you can come up with 9 an alternate test that could be all right. And I don't know 10 that to be the case.
11 All I want to say is, you are given some latitude 12 within the' provision of the Generic Letter. How well you 13 apply that latitude is going to be judged when we come and 14 take a look at how you do an alternate test.
15 The criteria are pretty tough, and it does -- how 16 you go through it explicitly, look at these four factors, 17 some of which you may not be able to do. And it wasn't 18 meant to be an impediment. It really wasn't to -- we 19 weren't trying to put things that are unachievable for you.
20 We were really asking ourselves when we evaluate relief 21 requests for you, what are the things that we look fort what 22 questions do we ask of you. We ask you these types of 23 questions.
24 So we're asking the utilities to docament the same 25 type of information that we would have you document to us HERITAGE REPORTING CORPORATION (202)628-4888
--. h_____._.,__----m_ - _ _ _ . . _ - - - - - - - - - - -----
NL 66 1 when we perform the review. So it's not new questions, it's 2' just asking you to evaluate them for yourselves.
3 Anyway, it's.real hard to be real definitive with 4 the answer. To say, yes,.it's okay, without knowing the 5 particulars with the pump or whatever it happens to be and 6 the particulars of the alternate test.
L 7 (Q) On that same question, say you do your 8 maintenance history and your review, would you have to 9 maintain Jome type of program to make sure that you're 10 always getting this feedback in to make sure that you're E 11 relieved or whatever your position is, is still justifiable?
12 I mean, do you have to continually monitor it, or do you 13 just look it over at the 10 year upkeep?
14 A. That's a good question. We hadn't thought about 15 that question in the course of writing these criteria.
16 MR. SULLIVAN: I would say -- I don't want to try 17 and answer specifically what kind of records and so forth 18 you need to keep, but it's incumbent on you to make sure 19 that the justification you put together remains valid.
20 Otherwise, the relief that you sort of grant yourself, 21 really wouldn't be appropriate.
22 MR. MARSH: That's probably true, although I've 23 never said it this way, in general -- If a valve cannot be 24 tested, if continuous spray check valve cannot be tested 25 because you do not want to spray the containment; therefore, HERITAGE REPORTING CORPORATION (202)628-4888
67 1- the check valve can't be' tested reasonably. And that relief 2 request, the Generic Letter through the subvehicle was 3 approved.
4- 'But at some point, you want to test that-pump, so 5 you put in a spool piece in the system, and another valve 6 which enables you to test that pump. That relief request 7 really is no longer operative. It should no longer remain 8 in effect, because now there is a method for testing that 9 check valve. Although we haven't said it in those terms, I
'10 think that's automatic in relief of the code.
11 MB. SULLIVAN: The last part of this m:estion
'12 says.:
13 (Q) "Thus, the documented, alternative test 14 evaluation in the IST program have to be formally submitted 15 to the NRC as an IST program revision, and if so, in what 16 time frame?"
17 MR. SULLIVAN: I think I talked about this before.
18- But just to take this particular example, this would be an 19 area that relates intimately with conformance letter. What 20 we're looking for is when you send out the conformance 21 letter, if you had to change your program, that you send in 22 the revised program in the same time frame.
23 MR. MARSH: We ought to also go on to say that if j 24 you need to change a program at any point, or if you do 25 change a program at any point, the revision, editorials, '
HERITAGE REPORTING CORPORATION (202)628-4888
o 68-1 whatever else,;that we have that program. I don't think we, i
2- want to be in the position where you have a program that's. '
3 different.than the one we have, that you've implemented 4 something_different.
5 So what we're saying is, anytime that you change 6 your program in the future for whatever reason, we want it.
7 We want to have on our shelves what you have in the plant.
8- So that when we do an inspection, we have right available to 9 us exactly what you've got, whether it's program changes 10 because of the Generic Letter or subsequent program changes 11 for whatever reason.
12 And it doesn't mean that we're going to get into 13 the review mode again. This is just for the purpose
.14 preparation for inspection or for looking through what types 15- of relief our plant's using, in general. It's to make the 16 act better, namely.
17 MR. SULLIVAN: Tad's remarks then directly answer, 18 coincidentally, the next question. So I'm just going to go 19 on to the following question.
20 (Q) Do the requirements to conform to the stated 21 positions of the Generic Letter, within six months of the 22 date of that letter, mean that all procedures have to be 23 revised and approved within the six-month? Or is it 24 acceptable to have procedures in the process of being 25 revised within the six-month period?
HERITAGE REPORTING CORPORATION (202)628-4888
s W y.
69 1 MR. SULLIVAN: I think we've given you an
{
2 indication from the conversation that we started off at the 3' beginning of this session, that we know the.six-month period 4 is a hardship, possibly in'some cases. And we've talked 5 about noting this kind of information in the summary. We
-6' haven't really said clearly how we're going to respond to 7 that. The reason is we're not sure.
8 I can just go back to the thinking of the Generic 9 Letter; that's the only way I can answer this question. It-10 apparently isn't clear, because this is about the third time
'll we've gotten this question in three meetings.
12 But our intent was that the program and procedures 13 would be wrapped up in the same type of -- as the 14- confirmation letter. In other words, six months.
15 (Q) You couldn't submit,.like an outline, and 16 make commitments on certain dates. Give you an outline like 17 you do for modifications. Would that be an option possibly?
18 MR. MARSH: Well, right now, the requirement is 19 the Generic Letter says, as we said it says, we're thinking 20 about alternatives. We're thinking about whether the 21 program review can be done within six months, and the 22 procedural review be done at some different time frame, 23 consistent with what -- We're not sure. If you have 24 something to offer, please do it. Or work through Clive 25 Calloway from NUMARC or whatever. We do want to set up HERITAGE REPORTING CORPORATION (202)628-4888
- - - - _ _ _ _ - _ - _ l
70 1 something that's achievable, that makes sure you do the best 2 . job you can do, but not just rolling on. We do want to get 3 on with this thing properly.
4 One question we had was what happens if we're f 5 going to do a new program within six months or so. And I'm 6 going to have new relief requests in that program that are 7 beyond what I have right now. Outside of attachment 1, so 8 they're not approved,'so they need to be approved, and it's 9 an update. So it's going to be a new program, new relief I a
10 requests, they're outside of attachment 1, the question was, 11 are they approved or not? -
12 The answer is no; they're not approved. The 13 question was, well, can I implement that? No, you can't 14 implement them.
15 If the relief requests are coming from additions 16 of valves to systems, in other words, I'm not increasing the 17 scope of my testing, I'm basically adding stuff, I can't do 18 the test that the code would have me do because it's inside 19 containment, and it bypasses around MOVs, the thermal 20 release around MOVs, but my test is good. .
21 I'm adding valves to the system. I'm not reducing 22 testing; I'm adding components that need to be tested. I 23 just can't do it on that kind of frequency. Is the relief 24 request approved? The answer is still, no, it's not 25 approved.
HERITAGE REPORTING CORPORATION l (202)628-4888 i
' i
-l 71 1 .But I think what we're driving at here is we are 2 going to need to know these types of relief requests 3 promptly. So that when we say the program, that you said' 4 conforms to the Generic Letter, is approved, and it's
.5 approved meaning it's done, it's finished, it has NRC 6 approval as well.
L 7 Perhaps that's the way to restructure the time 8 frames. So that relief requests that you need for program 9 approval are approved at the confirmation letter time.
10 Did you follow me?
11 (Q) Are you saying then that the Commission will l- 12 support, if we identify four or five relief requests? Like 13 in the next month were submitted, than when October comes 14 around, we'll have a response to those?
15 MR. MARSH: I'm going to try. The whole intent of 16 the Generic Letter was to just focus on those things that 17 are the most important. We really thought that we're 18 cutting down on relief requests. We've given you guidance 19 with.the things that you can implement. The idea was not to 20 end up with a whole bunch of more relief requests; the idea 21 was to help the situation, not hurt it. But we recognize 22 that there's going to be new relief requests. It's going to 23 be.
24 But we're not doing program reviews any more.
25 We're not looking at programs in the entirety. We still got HERITAGE REPORTING CORPORATION (202)628-4888
72 1 our contractors; we stil'1 have our staff. So my hope is to 2 be responsive. So that when you got a relief request that 3 you really need to have approved, that's the point in time 4 that it gets approved by that point in time. That's the 5 hope, and I think that's achievable.
l 6 What has to happen is when you send in a relief l
7 request through your project manager, you go to indicate to 8 the project manager: I need this. This is why I need this.
9 This is when I need this. So that he can come up to our 10 office and say, great, Arnold needs it. This is the date.
11 This is why. Please do it. It gives it the right priority.
12 So that it's a closed loop. It doesn't end up with stuff in 13 my in-box that you can't handle.
14 The whole thrust is to not only improve resources.
15 I mean, resources are a concern, but it's not what's driving 16 this thing. What's driving this thing is the lack of 17 guidance of us to you of what programs ought to be about.
18 What we think IST is about, and with the curt environment, 19 with the lack of a good ASME code, makes us have to provide 20 a lot of stuff. And without us telling you what's most 21 important to focus on, it ends up with just a huge volume of 22 in-box stuff. That's a resource that if I had a million 23 people I could put on it, but it still wouldn't solve the 24 problem. It still wouldn't tell you generically what we 25 think is the most important thing to do. That's what we're HERITAGE REPORTING CORPORATION (202)628-4888 l
I
_ _ _ _ _ _ _ - - - - . - - - I
73 1 trying to do.
- 2 (Q) What you are saying is there is going to be'a 3- short-term type thing where as long.as we submit our program
- 4. and the associated relief requests within the six-month time 5 frame of the Generic Letter, that once we submit it, that's 6 like in turn approval until you get a chance to look at it, 7 or we have to get something in writing back from you
- 8. saying --
9 MR. MARSH: I'll be frank with you. We haven't 10 thought this part of the circle through. We need to think 11 this part of the circle through. Right now the Generic 12 Letter says, get your confirmation letter in there right 13 away by six months with relief requests identified in that 14 confirmation letter. They.need to approved before they can 15 be -- that's the way the structure is right now. But we're 16 finding out that six months is achievable. A lot of people 17 are having a lot of troubles meeting six months.
18 So maybe there can be a restructure of the six 19 months. Maybe it says -- I don't know whether this is the 20 way we're going to go. It says if you need relief requests 21 approved, prior to your confirmation letter, get the relief 22' . request in by six months, so that when your confirmation 23 letter comes in, in some other time frame, it embodies NRC's 24 approval of those relief requests. So that the confirmation 25 letter says not only does this thing cover attachment 1 HERITAGE REPORTING CORPORATION (202)628-4888
. k.
74 1 positions, it also covers the reliefs that you approved l
2 already prior to this time.
L l
I'm not sure this is the way to )
l 3 go. Maybe that's the way to go.
4 (Q) There may be some impact on some utilities, 5 because there are certain things you just can't do. Of I 6 course, we're asking for relief in those case. And what I'm j
-7 hearing you say is you don't have approval for those. So 8 you have to go through a lot of abnormal gyrations to 9 accomplish that or'try to expedite that approval. And if 10 you get it in within the six nonths, that's still not 11 approval.
12- MR. MARSH: That's right. That's a key. The way' 13 the instruction right now, the outside attachment 1, new 14 relief is not approved. We recognize that. And we need to 15 say that to clear up the 405 issue. But we need to say, 16 then what about implementation of these things? What if I 17 just physically can't test these things? At what point are 18 you going to come knocking on my door to see whether I am 19 not testing things in accordance with the relief request 20 that I've given you? At what point do a need to say that?
21 I haven't said that yet. The Generic Letter 22 doesn't articulate that, and we ought to say that. That's 23 not as much as I can say, but I haven't thought more 24 carefully about it.
25 MR. SULLIVAN: The last question says -- the last HERITAGE REPORTING CORPORATION (202)628-4888
1 o
75~
l' 1 question on this list -'says:
2 A "Due'to' outage schedules and constraints, are 3 -there.any provisions-for not completing all equipment 1
i 4 modifications within the 18 months of the date of 5- confirmatory letter for the first scheduled refueling outage.
6 following confirmation letter?"
7' MR. SULLIVAN: I guess this is an area that's l
8 similar to the area we've been talking about with respect to 9 the six months. We really, in this particular case, l
10 tried -- maybe'even more so in'the sense of the six l
11 months -- to make sure that we're giving you an adequate 12 ' window. So I don't know exactly what circumstances you have 13 'in mind. And I really don't have a resolution for the
- 14. question.
15 MR. MARSH: Anytime the NRC says we want this'done 16 by 18 months from whatever else, this is what we want. I'm 17 not encouraging it, but licensees are always free to go to 18 their project managers.and say, "I can't do that. I need 19 relief from that." And if it is justifie.d, and if you can j 20 be given approval for provisions of the generic requirement, 21 it's always allowed by the regulations. It's always there.
22 So if there is a justifiable reason for be it relief from 23 the time frames that are there, it's a vehicle that's 24 available.
25 But is that a problem? We thought that the time HERITAGE REPORTING CORPORATION (202)628-4888 l
l i
76 1 frames that we were using 18 months to two years were 2 achievable time frames.
3 (Q) One example we have in mind is, as least one 4 of our plants, like immediately after the confirmation 5 letter, could be going into an outage --
6 MR. MARSH: But you got 18 months. It says, "18 7 months or the refueling outage, whichever is later."
8 (Q) Okay. Maybe I misunderstood that. So you 9 could use the 18 months.
10 MR. MARSH: Yes, sir, right. It says 18 months or 11 the next refueling outage, whichever is later.
12 (Q) And for some reason your outage was 19 months 13 later, that would probably be grounds to say, hey, the 14 immediate intent was -- I'm doing it in my next outage --
15 MR. MARSH: Yes. You have to be careful with 16 the --
17 (Q) Right. I understand that.
18 MR. MARSH: There has to be a good justified 19 reason for that. One month, one way or the other, is 20 perhaps the good reason for it.
21 (Q) The reason would be that an outage 22 immediately following a confirmation letter. We may not be 23 able to --
24 MR. MARSH: We want to avoid situations where the 25 confirmation letter is on such-and-such a date, and then IIERITAGE REPORTING CORPORATION (202)628-4888
l 77 1 three months later there's a refueling outage. Well, 2 obviously, we could wait'for the'18 months, but then, gee, 3 you need an outage to-do this modification. But you're on a l
4 two-year cycle. So now we're out to 27 mons.s from the date 5 of the confirmation letter, compared to 18 months. Well, 6 we're not real receptive to that kind-of logic.
l 7- There are shades of gray, but-we thought 18 months 8 in the refueling outage was a good time frame.
9 Let's see, we're at 1:15 -- 12:15.
10 '
Okay. This is from public service corporation.
11 Question number 1:
12 (Q) Currently, we only test the ICS pump suction 13 check valves, ICS 3AB. See the attached drawing to verify 14 that they open as part of the ICS pump test.
15 Originally, the only safety function recognized 16 was for the valves to be open to provide a water source, the 17 RWST for the ICS pumps. During an independent review of the 18 IC program, it was determined that these-valves may also 19 have a safety function to close when the pumps are taking 20 suction from the RER system. These valves, if they fail to 21 open, could provide another flow path to the RWST, besides 22 the normal flow path that contain it. This flow path would 23 also allow potentially contaminated water from the 24 containment sump into the RWST, which is obviously not 25 desireable.
HERITAGE REPORTING CORPORATION (202)628-4888
i 78 1 As part of our company's in-house safety system 2 functional inspection, it was determined that if these check 3 valves failed to open, adequate flow to the containment 4 would still be achieved.
5 We're also converting the manual valve, upstream ;
1 6 of ICS 3AB, into niotor operated valves in order to prevent I 7 sump water from getting into the RWST.
8 Our question is whether these check valves need to 9 be leak-tested?
10 MR. MARSH: If the safety system functional 11 test -- it sounds like it focused on adequate flow to the 12 system with these check valve stuck open, not on the 13 radiological part. So it sounds like part of the safety 14 function of that valve was not assessed in the safety system 15 functional test. You have to test the check valves, or any 16 kind of valve, in its safety function mode, whichever that 17 may be, for flowback or whatever. And if you determined 18 that the safety function is only in one direction and that 19 backflow is not a safety function, then it need not be leak 20 tested in that direction. But I don't know if that's where 21 you come down in this question.
22 It sounds like there's still the issue of backflow 23 through that test valve, back into the RWST, even though the 24 flow requirements may be back.
25 Where is the gentleman from --
HERITAGE REPORTING CORPORATION (202)628-4888
79 1 (Q) I didn't know there were going to be so many 2 questions on the Generic Letter. We didn't have many 3 because we've been at discussions with NRC on most of these 4 issues.
5 MR. MARSH: I think it is generally true 6 regardless of this particular question of the many types of 7 questions that exist, that valves need to be tested in their 8 safety direction. And that could be by low flowback or 9 whatever else. And if the flow requirements of the pump are 10 being met with the valves stuck open, the valve need not be 11 tested in the backflow direction. But all aspects of the 12 problem need to be addressed.
13 Let me give you an aspect of the problem that is 14 frequently not addressed. We've had problems on feedwater 15 systems on PWRs, check valves leaking from the AFW system 16 into the main feed system. Normally, there's one last check 17 before it enters the feedline.
18 Frequently that valve is not -- may not be 19 backflow tested at all, only forward flow tested. If it is 20 backflow tested, it may only be that the valve promptly 21 moved to its closed position, not with any leakage. Check 22 for it in the upstream direction.
23 We've had problems in a couple plants now where 24 those check valves leaked back. And you don't detect that 25 leakage unless there's something upstream of it leaking, HERITAGE REPORTING CORPORATION (202)628-4888 i
I
80 1 'another gasket, something's happening. When that has j 2 happened -- in fact, one of the recent situations was there 3 was a valve misalignment. And because of the leaking of the 4 .feedwater check valve, we ended up with mainfeed all the way 5 through the AFW system. And you got back into the storage 6 tank, because of some bypass, some miniflow lines that were 7 aligned. It depends on how many check valves you go in 8 line.
9 The problem is, first, pump operability, making 10 sure the feedwater pump is operable with the main feed in 11 that line.
12 But more than that was the piping system was not 13 designed for that kind of temperature. And the stresses and 14 the supports, and the potential for water hammer that 15 continued the penetration, all those types of questions 16 weren't addressed. So when a valve is categorized, and when 17 a valve is then leak tested, you have to also ensure that 18 the piping upstream of that, and all the associated 19 equipment upstream of that is designed consistent with the 20 categorization of the valve. If it's a B valve, leakage is 21 generally not important, which ought to mean then that a 22 small amount of leakage upstream isn't going to impact 23 piping operability or support operability.
24 But in these case, it seems to be impacting piping 25 operability and support operability. So we need to be HERITAGE REPORTING CORPORATION (202)628-4888
81 1 mindful of that type of issue when you categorize the test 2 check valves, too.
3 What, if.anything, is being done with the licensee 4 response to Generic Letter 8706?
5 I'm sorry, what is 8706? PIV's, okay.
6 Generic letter 8904 references PIVs in section 4.
7 However, it appears that there are no longer changes 8 required due to Generic Letter 8706. Is this'true?
9 (Q) Such as putting on tech specs. We' re an 10 older plant. We only have event B --
11 MR. MARSH: It comes back to the previous answer I 12 was giving. What happened in the '85 time frame, 1984-5 13 time frame, there was some thought given to relaxing leakage 14 requirements in the tech specs PIVs. Maybe this was in 15 8706; I'm not sure.
16 So the group that was sponsoring that proposal 17 went to the CRGR and said, "We want to relax the 18 leakage limits." The CRGR said, "Why are the leakage limits 19 there in the first place? Are you sure they should be 20 there? And what about other PIVs?" And it opened up all j 21 the issues surrounding PIVs being in the tech specs, the
.J 22 initial orders. l 23 So this generic letter went out as an information 24 gathering process. That's all it was. It said, "What's 25 there? How are you testing it? How many valves do you HERITAGE REPORTING CORPORATION (202)628-4888 i
I
>g.
L
[k 82 1 have? 'Where are the information gathering?
2 But the thrust was, we wanted to find out what the 3 industry is doing for PIVs. Beyond the tech specs, what are 4 you doing with PIVs? And it was aimed at supporting a new 5 Generic Letter, a new set of requirements for PIVs.
6 The information that came from that Generic Letter 7 is being inputted into a generic issue 105, which is PIV-8 testing. That information is all being input into this 9 latest concern of event B check valves and how they're being 10 tested because of this precursive --
11 So it's all a train of information and concerns 12 that's leading to where we are now. It is being pursued.
13 There are research generic issue, and there's ongoing NRR
, 14 programs to look at how valves are being tested.
15 There are no further requirements for you at this 16 point beyond responding to the Generic Letter 8706, and 17 beyond our Generic Letter at the code itself.
18 (Q) Do you think that there will always be some
'19 separation between event B PIVs and all the other PIVs?
20 MR. MARSH: I don't know. Separation meaning, 21 these are more important than these. Test these this way; 22 test this, this way. Stay tuned. I really don't know.
23 Because this may depend on what happens as a result of these 24 inspections --
25 MR. SULLIVAN: The next question says:
HERITAGE REPORTING CORPORATION (202)628-4888 l
b
, 83' e 1; (Q)' Section 3 of Generic Letter 8904 deals with 2 backflow testing of' check valves. It has a~ list of several 3- valves'that'NRC states provide a safety function. Some of
- <4 '
these' valves do-not appear to provide safety function, and-5 we would like to hear NRC's reasons classifying these valves 6 as safety related.
7 MR. SULLIVAN: I think in the letter we indicated 8 that these were examples. We weren't explicit. We were 9 .trying.ta say that in all' plants, they may not perform.a 10 safety function, but at least in some plants they do.
11 I guess what I'd like to do is ask the questioner
'12 Eto identify which of that list of six or seven he's 13 concerned about.
'14 MR. MARSH: This gentleman was some preconsulting.
15 We found that out.
16 (Q) I guess I wanted -- VCT outlet check. I 17 think maybe some plants it might be a containment isolation 18 valve; I'm not sure.-
19 HR. MARSH: No, no, no. VCT outlet valve is in 20 some plants very important depending upon whether the MOV 21 isolating the VCT is a safety created reliable valve which 22 automatically isolates on low VCT level, and you don't end l 23 up with problems.
l 24 For example, without mentioning plant names, we've 25 had problems with the MOV not closing on low level. Now, HERITAGE REPORTING CORPORATION (202)628-4888
E 84 if this is really not a check valve concern; this is a VCT
-2 level concern with the VCT not closing on low level, having 3 nitrogen gas find the charging pumps, which are safety grade 4 pumps at this plant. And if there were a problem with 5 backflow, if the MOV weren't closed properly, for example,.
6 if you couldn't rely on that valve to do its job and there 7 were recirculation concern or RWST concern, then that valve 8 provides the isolation for RWST water from being diverted up 9 by control tank.
10 It's not a containment isolation valve. It's 11 really separating the safety grade water supply source for 12 your safety inject 2on pumps from the nonsafety grade, 13 normal, uncontrol tank water supply for those pumps. It
, 14 depends upon the plant figuration and really the motor 15 operated valve.
16 I think I described to you that, now. Do you want' 17- some clarification on some of the others?
18 (Q) That was the only one, I guess. Thanks a 19 lot.
20 MR. MARSH: Do you have a safety grade MOV 21 isolation?
22 (Q) How about clarifying the mainstream return 23 valves? MR. MARSH: So you got an 24 MSIV, and it's a safety-grade MSIV. Is the configuration 25 such that you need both? Why is the nonturning valve there HERITAGE REPORTING CORPORATION (202)628-4888
M ~85
~1
.1' in the first place? Let me'ask you that from a design 2 perspective.
3 (Q) It',a a good question. It was outside of our
-4 ASME code, but it's there as a second valve to the SMIV.
5 MR. MARSH: It's'not there to mitigation of.a 6 single failure of the MSIv? You need tocask yourself that
?' 7 kind of a question. Because this configuration, where we
'8 mention it, was there really as a supplement for a single 9 failure for the MSIV. In other words, if you end up with a i 10 break inside a containment, you're trying to prevent with a 11 single failure double blowdowns, two steam or three steam 12 generators blowing down inside containment. That's what'it 13 was there for.
14 So it really depends upon the configuration that 15 you got, whether you need that valve, why it's there. And
'16 if it's there for backflow prevention for MSIV single 17 failure purposes, then it needs to be in the program, which 18 is what prompted this concern.
19 MR. SULLIVAN: The last question on this list 20 says:
21 (Q) We are only able to perform partial flow test 22 of the accumulative discharge check valves due to
- 23. limitations based on system configuration. Do we have to 24 supplement this test with disassembly of the check valve?
25 MR. SULLIVAN: We talked about this quite a bit in HERITAGE REPORTING CORPORATION (202)628-4888 1
86 1 the previous set of questions. And I think we were fairly 2 clear that these two things go together in this situation.
3 But I would like to amplify one point I don't think we hit 4 on earlier. That is sort of an expansion of some of the 5 remarks Tad was making, that we are concerned about taking 6 these valves apart more often that they need to be, which 7 following this position literally could lead you to do.
8 We would encourage, in this particular situation, 9 that you take advantage of the last couple of paragraphs in 10 position two, to construct the kind of information you need 11 to justify a different sampling frequency if the data really 12 does justify it, in which case the test that you would be 13 performing regularly would be the partial flow, supplemented 14 by disassembly on an appropriate inspection frequency.
15 MR. MARSH: Okay, gents, it's 12:30. We've got 16 about five questions on cards. We've gone through all of 17 the resubmitted questions. What I'd like to do, with your 18 okay, is take a lunch break. And we don't need that half-19 hour caucus that we've provided for. Why don't we just come 20 back. Please use the hour break for writing down any more 21 questions that you have. We got some more cards up.here, 22 too.
23 Let's reconvene at 1:30 Central time. l l
24 (Whereupon, at 12:30 p.m., the hearing was l l
25 recessed, to reconvene at 1:30 p.m., this same day, Tuesday, HERITAGE REPORTING CORPORATION (202)628-4888 l l
I
87 1 . June 13, 1989.)
2 3
4 5
HERITAGE REPORTING CORPORATION (202)628-4888 1
j
88 1 AFTERNOON SESSION 2 1:30 3 MR. MARSH: Well, get right'at it.
. 1 4 Let me start with another set of printed 5 questions. This is from Point Beach:
6 .(Q) First, how are extensions of the October 30, 7 1989,-deadline viewed? What factors are considered on such 8 requests?
9 MR. MARSH: I think we've talked a lot about that.
10 What I need from you, and I think I'm getting from you, is 11 feedback on why this schedule is retractable, what types of 12 competing schedules there are when things need to be done, 13 what you think would be the right type of sequence, and what 14 a reasonable due date would be.
15 Wait until you read, please, the meeting minutes.
16 We'll give whatever guidance we can on those meeting 17 minutes, what could be viewed, what can't be viewed. So 18 don't send anything for your project manager, or anything 19 else, until you see the meeting minutes.
H2O (Q) Two: Is relief required for items per the 21 Generic Letter 8904 which differ from the ASME code? what 22 is the long-term status of the relief system?
23 MR. MARSH: I think you understand what we're 24 saying about relief in terms of the ASME code per the 25 Generic Letter. Is there something more you would like me HERITAGE REPORTING CORPORATION (202)628-4888
l 89 1 to say, or something that's not clear? Where's the Point 2 Beach gentleman? Mr. Swensen, Warren Swensen, is he here?
3 (Q) I t.. ink that's pretty clear.
4 MR. MARSH: Okay, all right. We talked a lot 5 about it, probably since we got the question.
6 (Q) It sounds like in the long-term, if we do 7 something different than what's in the Generic Letter, then 8 we submit a relief request.
9 MR. MARSH: Yes, yes. And if it is also different 10 from that which is already approved; keeping in mind that if 11 you're adding valves or components to the program and you're 12 in compliance with the code, that's not relief. In other 13 words, you can always comply with the code. Or, if you're 14 changing a relief request such that you're adding compliance 15 to the code. For example, we got a relief request now 16 that's in your pregram. It's submitted and therefore it's 17 approved by the Generic Letter, and upon review of it, you 18 decide that you can meet the code; and so, therefore, you're 19 going to rollback the relief that's requested. That's not 20 something that has to be acted on, because you're meeting 21 the code. It just needs to be highlighted as a change and 22 it's approved.
l 23 (Q) Many of the alternatives given seem vague and )
24 subject to interpretation. Who declares adequacy and what 25 are the ramifications of differences between licensees and j i
i HERITAGE REPORTING CORPORATION (202)628-4888
90 1 'the'NRC? -
! '2 MR. MALSH: We try not'to be vague. We understand L
3 that there's:always going to be interpretation of the 4 . wording when you write generic words. How they really get 5 read and implemented is going to be plant specific. It's 6 going to be independent. That's why we're having'these 7 meetings to help you in understanding what the thinking at B least was.
'9 If you have a question beyond the meeting, beyond
'10- what's in the Generic Letter, get in touch with us. Let's l
11 get it clarified. Go through your Project Manager; and, i
12 please, that's what we will do.
13 (Q) What are the ramifications of differences 14 between the licensees and NRC?
15 MR. MARSH: I think that depends on the 16 difference. That depends on the clarify of the guidance 17 with which you disagree. It depends on many factors.
18 That's an inspection area at this point. We're going from 19 program review, relief request approval, to inspection to a 20 different mode. I really can't comment beyond that. It 1 21 really depends on the particulars; it depends on the i
22 significance of the issue, too. '
23 (Q) What is the NRC's opinion per Generic Letter 24 8904 of nonquantifiable demonstrations of performance? For 25 example, a solenoid valve has no position indication that HERITAGE REPORTING CORPORATION (202)628-4888 i
i L . _ _ _ - _ _ - - - - - - . - . - ~ . - - - - - - - - - - - --- -
91 l' can be observed or timed, but varied temperatures showed no 2 overheating.
3 MR. MARSH: We're striving for quantification of 4 such things as valve performance, stroke time of pump 1
5 performance, et cetera. We had a question at another 6 meeting where someone was asking, what is a suitable
.7 alternative for some of these things. We answered to the 8 extent that's possible to a quantifying task, measure 9 parameters and report and show the suitability of that test 10 by measurement of information, not by qualitative tests.
11 Steer clear of qualitative tests.
12 For example, flow-way through a check valve may 13 not know the flow of the line itself, but you can quantify 14 its flow by looking at tank changes, boron concentration 15 changes. There are ways of getting back the basic flow 16 parameter.
17 MR. SULLIVAN: I'd like to ask the gentleman from 18 Point Beach to elaborate on what you have specifically in 19 mind with question 4. What system, what type of components?
20 (Q) Question 4 is flux feedpump turbine bearing 21 coolant. And its to a supply valve to the cooling water 22 supply valve.
23 MR. SULLIVAN: I'm sorry. I can't hear you.
24 (Q) It's to cooling water supply to the bearings.
25 What we're using, and it's a relief request in our existing I i
HERITAGE REPORTING CORPORATION i (202) C28-4 8 8 8
92 w ,
l l' system, is, we' measure bearing' temperature in each run. And 2 if you don't have coeling to the bearing, obviously our 3' temperature will go up.- So if our bearing temperature 4 stays -- specified level,.then --
5 MR. SULLIVAN: I don't want'co try to give a 6 definitive'yes or no because I don't know the details. But 7 it sounds like you're going in the right direction. You're 1
8 measuring something quantitative. I don't know what your (
9 acceptance criteria might look like, but I think that's one E
10 of our concerns with qualitative approaches, is that the 11 acceptance criteria then become very subjective.
12 (Q) I think another case is -- were looking a 13 pressure change ~before'and after opening a valve. Zero l 14 before, we say you have a pressure increase; after, we open 15 a valve.
{
16 MR. MARSH: And you would'use that'as a 17 quantification of backflow or something?
18 (Q) No. That's just valve position. I don't 19 believe that's a' check valve. That's just an automatic 20 valve.
21 MR. MARSH: As a demonstration that the valve 22 opened? )
23 (Q) Right.
24 MR. MARSH: Okay. That may satisfy some parts of 25 the requirement, but still, stroke time would have to be HERITAGS REPORTING CORPORATION (202)628-4888
f
.s.
~
93 1 measured for-that valve."
2 (Q) Just to answer his question, that's a good
- 3. way of doing it. .It.seems to me it's pretty. accurate for 4 the valve. If you can get pressure across it,.you're pretty 5 close.- That's a nice way of testing it. It's a valid test, 6 if you look at Instrument Society of America standards, 7 because they have many standards. I'm just agreeing with 8 the gentleman, that's all.
9 MR. MARSH: Okay. I don't know all'the' details of 10 it, though. I'm not sure exactly what it is.
11 (Q) I understood what he's trying to do. It's 12 something that perhaps your personnel would like to look at,
- 13. too.
14 MR. MARSH: If the test is a demonstration that 15 the valve opens, and all you're doing is looking at down 16 stream pressure and upstream pressure, and you're measuring 17 differential pressure --
18 I think that's when he was doing something 19 different, though. He was just looking at the downstream 20 pressure and observing that it went up, and that would be a 21 demonstration that the valve opened. Did I misunderstand?
22 (Q) No. That's pretty much what we did. I'll 23 have to review that test and see exactly what we're doing.
24 MR. MARSH: What system is it and what component?
25 (Q) It's the aux feedwater coolant. {
i I
HERITAGE REPORTING CORPORATION l (202)628-4888 l
94
- l. MR. MARSH: .And this is a check valve in line, or
? this is just a' manual. operated valve.in line?
3 (Q) It's a solenoid operated --
4 MR. MARSH: A solenoid operated valve inline?
5 You're checking only that it opens.
6 . (Q) Right.
7 MR. MARSH: Only that it opens. Now, that's not 8 stroke time,-only that it is open.
9 (Q) Because we don't have any other indication 10 that shows it's going to open.
11 MR. SULLIVAN: And you're using these two sets of 12 parameters in concert with each other.
13 MR. MARSH: We had questions of that sort from the f 14 others. You can quantify with a test what the flow is 15- through that system, if you do a heat balance across the 16 heat exchanger, with some precision. I'm not sure how much.
17 It-just depends on how much you can quantify the heat 18 performance -- the heat transfer coefficient in the heat-19 exchanger for one thing. That's not a very easy thing to 20 do. But tests can be performed so that you can then go back 21 and repeat the test to make sure the valve is still 22 similarly performing.
23 MR. SULLIVAN: We'll start working our way through 24 the cards next.
25 This next question is very plant specific, and HERITAGE REPORTING CORPORATION (202)628-4888 1
i
)
i
- - _ _ _ _ _ _ _ - - - - - - - - _ _ - - . - - - - - - - - - - _ - 1
95 1 maybe it would be' appropriate to just read it and give a
'2 brief answer and then' talk about it later with the 4 1
3 questioner,.since it's very plant unique. I don't think
,4 it's too much of general interest.
5 (Q) Should the Sow County station be on table 2 6- of the Generic Letter? If not, why?
7 -MR. SULLIVAN: Basically, the reason we didn't 8 put -- let me just back up a little bit. I'm not familiar 9 with all the dates. But I know we did issue an SER on the 10 Sow County station. It's not been too long. It's been 11 probably within the past year to year-and-a-half.
12 (Q) August 26.
13 MR. SULLIVAN: So it's been within the past year.
14 The reason we didn't put it on the Generic Letter 15 is because on the Generic Letter table 2, is because you all 16 made a program submittal that was so extensive we felt we 17 would have to basically recommence an entire review. I 18 guess it-could have been when we put on there, but we 19 decided not to in the interest of getting SERs done on 20 plants that haven't received any SER. We can talk about it 21 more later. But that was my basic thinking.
22 MR. SULLIVAN: The next question says:
23 (Q) Our plant is on table 1. We have revised the 24 program to identify Generic Letter 8904 as a reference and 25 made some minor changes consistent with 8904. Do we need to HERITAGE REPORTING CORPORATION (202)628-4888
_ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ l
96 i l
I resubmit the program?
2 MR. MARSH: So it's one that we are about ready to 3 issue an SER. But that which we're reviewing is not what 4 you have now. You've changed your program beyond the 5 submittal. Right?
6 MR. SULLIVAN: This is from --
i 7 (Q) I guess the question is our reading oi' the 8 Generic Letter is nothing that talks about an additional 9 submittal to the NRC for us; and yet, in this meeting that's 10 been identified as a requirement. It seems to be 11 inconsistent. We told our management, based on the reading 12 of the Generic Letter, that there's no need to submit 13 anything further as 1c.g as we're consistent with the 14 Generic Letter.
15 MR. MARSH: What troubles me is, the Generic 16 Letter said we have your program right here. We have it; 17 it's the same. What you have at the plant is the same as 18 what we have right here. And we're going to issue an SER on 19 it. You're a table . plant, you don't have to do anything.
20 So sit on your hands for awhile.
21 But I think you said in the question that you 22 changed your program.
23 (Q) Well, we changed our program primarily to 24 say, we have within our program a list of references in 25 there. We reviewed our program using the Generic Letter.
HERITAGE REPORTING CORPORATION (202)628-4888
97 1- One.of the major changes'in the program is to say, by the 2 way, this program complies with Generic Letter 8904. But I 3 think what I heard this morning is now as a result of --
4 ~ change, we now locked into --
5 MR. MARSH: That's not a new requirement, by the 6- way. The fact that the NRC should have in its hands the 7 program that you have in the plant, that's not new. That's 8 just something I think is clarified. We're not asking for 9 something that'a poing to result in extensive review. It's 10 something we want to have just for QC standards, just 11 because Appendix B -- we want to have what you have. So 12 that when we do an inspection it's the same thing that you 13 have in the plant. That's the reason we want it. It's not 14 because it needs a submittal that's going to be reviewed and l
15 approved. It's because what's on our shelves was 16 implemented in the plant.
17 But the thrust of the Generic Letter for a table 1 18 plant was that you don't have to do anything. You're about 19 ready to get the SER. You may have to change your program, 20 base on what the SER says, but you're in front now.
21 I heard you say that you just changed the 22 reference. No changes in the program; is that whdt you l
23 mean? No changes really, it's just something that says that l' 24 it's improved --
25 (Q) Let me revise the question. For utilities HERITAGE REPORTING CORPORATION (202)628-4888
98 I that are in table 1, or aftec you get your SER, if you 2 revise your program, _does that program require a resubmittal 3 to the NRC7 4 MR. MARSH: Yes. Yes, it does.
5 (Q) The letter doesn't say that, I don't believe; 6 MR. SULLIVAN: No , it doesn't. But as I said, 7 that's not new.
8 (Q) No. I'm not saying it's new.
9 MR. SULLIVAN: That's a clarification.
10 (Q) The Generic Letter does not address that. As 11 a matter of fact, it just simply says you don't need to send 12 it.
13 MR. SULLIVAN: No, I don't think so. Where does 14 it say that?
15 MR. MARSH: It says you don't have to issue a 16 confirmation letter.
17 (Q) Section D.
18 MR. MARSH: Section D. I'll just read it for 19 everybody else who doesn't have it.
20 "D is program updates and revisions if the 21 licensees modify their ISD program beyond that currently 22 submitted to the NRC. They should review those 23 modifications against the positions found in Attachment 1.
24 For ISD program changes for which specific positions are 25 provided in Attachment 1, licenscas should follow the HERITAGE REPORTING CORPORATION (202)628-4888
V 1
1 !
l 99 1 guidance in section B above. For ISD program changes in 2 areas not covered by Attachment 1, the provisions of 10CFR50 3 55AG should be_followed. A modified program should comply 4 with the disposition of relief request in any applicable SER 5 based on a previously submitted ISD program."
6 That's not the t' action you want me to read; is it?
l 7 (Q) Yes. That's the section I was looking at, 8 sir. I don't see anything in there that says I need to send 9 in a new program, a revised program. What we're saying 10 here, we need to do that. Right?
11 MR. SULLIVAN: Yes. That's what we're saying.
12 We , I guess, agree with you that the Generic Letter was an 13 explicit one from this particular point. I'm not sure. I'd 14 have to re-review the whole thing.
15 MR. MARSH: Well, I think even if it says for ISD 16 program changes in areas not covered by Attachment 1, the 17 provisions of 10CFR50.55AG should be followed, that means 18 that if relief requests are there, they need to be reviewed 19 and approved prior to them being implemented.
20 This is a question of you having a new program 21 that still complies. All we're saying is we want that, too.
22 And it's not because we're going to do a big job on it, it's 23 because we want to have on our shelves what you have. I 24 think that's your intent as well.
25 (Q) Yes. That was my question. Do we have to HERITAGE REPORTING CORPORATION (202)628-4888
100 1 send in a revised program to the NRC? Reading the Generic.
2 Letter, it appeared we didn't have to, and then here we have 3 some program comments that, yes, you issue a revised program 4 on your utility, send it in to the NRC because we want to 5 keep abreast of the changes.
6 MR. SULLIVAN: Yes. that is true. The Generic 7 Letter doesn't say that.
8 (Q) Yes. It's not specific.
9 MR. MARSH: It's not specific.
10 (Q) Okay, good. That's all we wanted to know.
11 MR. MARSH: Okay, good.
12 But I want to come back to -- there has been a 13 program change beyond what we have. Right? What you say is 14 pretty narrow. It's no testing changes, no scope changes, 15 no relief request changes. Okay.
16 Wait until you get the SER. You may have to make 17 another program change; I don't know. And then once you're 18 done with that, make sure that comes to us. And we'll say 19 that in the meeting minutes, too. That's a clarification.
20 The regulations requires --
l 21 (Q) We would appreciate if you could put in the {
i 22 meeting minutes to say that having revised your program, you 23 need to send it in to the NRC. I i
24 MR. MARSH: Okay. It will be in there.
25 (Q) It's not clear in the Generic Letter, i
l HERITAGE REPORTING CORPORATION (202)628-4888 1 l
l
101 l' MR. MARSH: Okay,. you got it.- Thank you.
1 2 (Q). Thank you.
1 13 MR. MARSH: You bet. I 1
4 MR. SULLIVAN: I don't want to overwork this 5 question, but I'm curious. Do you have-a program that 6 you're going to send'in imminently? Or, is it just sort of 7 in the works? l 8 (Q) No , it's done.
9 MR. SULLIVAN: Have you sent it in already?
10 -(Q) No.
11 DR. SULLIVAN: You're just sort of waiting for 12 direction from this meeting?
13 (Q) Well, we weren't waiting. Based on our 14 reading of the Generic Letter, we assumed that since we were 15 on 16 table 1 --
117 MR. SULLIVAN: I guess I just wanted to add one 18 point. As I read this card literally, it sounds like if you 19- send it in, it could resolve some items that otherwise might 20 be open'in your SER. Because it says that you sent in some 21 changes, or you prepared some changes to make the program 22 consistent with the Generic Letter.
23 (Q) More of the changes in the program were 24 referencing the Generic Letter to ensure that our program 25 complied with the Generic Letter. There weren't wholesale HERITAGE REPORTING CORPORATION (202)628-4888
102
_ changes in the ISD program.
L 1-p 2 MR.. MARSH: No' programs, no relief request, no L 3- scope-changes, just reference to the Generic Letter, i
- 4. (Q) Well, quite frankly, other than the reference L 5 to the' Generic Letter and the changes made, we went through
- 6. a comprehensive review of all the valves in our plant, and
, 7 = identified some changes in direction. We thought some 8 things should be either added or deleted in terms of their 9 direction, their_ safety function. There are some changes of 10 that type, a few, not a lot.
11 MR. MARSH: In the future,.please make sure that 12 what you have is what we have. What troubles me is we've 13 got a program ut, der review of an issue in NCR, and it's_not 14 going to refer to your program now. You may have added some
-15 valves and added some relief requests, maybe different 16 than --
17 (Q) No relief requests. I don't add relief 18 requests.
19 MR. SULLIVAN: Okay. The next question says 20 (Q) Once OM6 and OM10 are approved, will it be 21 required to implement them immediately (within six months) 22 or will they be implemented at the next program update?
23 HR. SULLIVAN: Tad was talking in his opening 24 remarks about OM6 and 10 being in the process of 25 incorporation by reference in 10CFR50.55A. And we indicated HERITAGE REPORTING CORPORATION (202)628-4888
.f
g l2 .103
.1 ;we'weren't exactly sure when that rule was going to work-N 21 'it's way through to completion. But I.can say that, as it's O
rf 3 currently envisioned, .OM6 and 10 would be incorporated by.
'N 4. ' reference as previous addenda were in the past. And the 5 provisions for updating aren't.being changed in'this 6- particular rule making.
7 So the answer really is to follow the rule here.
8 Will they be implemented at the next program update? That's 9 when you would get into it, because that aspect of the rule 10 'hasn't been changed -- or isn't being changed.
11 MR. MARSH: I don't want to discourage its'use.
12 If.you've got a program revision that you decide to make 13 early, you want to pick up OM6 and 10. There's nothing 14- wrong with that.
15 Yes, ma' am.
16 (Q) Is there any plan to speed up the updating 17 process for OM6 and 10 -- 10CFR50?
18 MR. MARSH: If you mean make it more frequently 19 than every 10 years?
20 (Q) Is there any plan to require the 21 implementation on OM6 and 10, other than the normal update 22 procedures that's now in effect?
23 MR. MARSH: There is some thoughts to doing that, 24 within making regulatory change, regulation changes. There 25 is some thoughts to that. It's real premature at this ;
HERITAGE REPORTING CORPORATION 1 (202)628-4888 i
m__________________________ j
Atm E 104 L
- 1 . point, but we're getting --
2 MR. SULLIVAN: Theinext question is similar to the a
3 one'we were talking about from the Point Beach list of 4 questions.
5 .(Q) For' check. valves where design accident flow 6 is not specified, what' guidance can you give'for full-flow
- 7. testing?
8 MR. SULLIVAN: I assume, Don, what you mean.by
-9 that are valves in supporting systems where the concept of.
10 an accident flow really doesn't fit the way it would for an 11 ECCS system, like an air start system for diesel generators 12 .or something like'that.
13 (Q) Yes, that would'be my example.
14 MR. SULLIVAN: I think it is a similar answer to 15 what we were talking about from Point Beach. We know there 16 are many valves like this in plant. We don't think it's 17 feasible. We're not recommending that people wholesale 18 through the plant insta11' flow gauges, trying to figure out 19 what flow is required, a specific number. But we are 20 interested in this type of test being more than a 21 qualitative evaluation.
22 MR. MARSH: Are you having trouble determining up 23 front what the flow rate should be through that valve? Or 24 are you having trouble at the backend, testing to make sure 25 that it needs that flow?
HEF TAGE REPORTING CORPORATION (202)628-4888
= _ - - _ - _ _ _ _ - _ _ _ _ _ _ - _ _ _ _ _ _ _ _ _ _ _ - _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ - _ _ _ _ _
105
' a i' >
. 1 ".- (Q)' . More up front. ..We-really have a hard time t.
2 'specifying what the limits,should be. ,There might'not be 3 .what'we call a design safety flow, like an air-startLvalve,. 3
~4~ .for. example.. Another example, where FC cooling water. check 5 valves -- it might not be written'anywhere what that design' i
'6 flow:would be,-although you probably have'an idea of what-
.7 the check valve has to pass to get to be; operable. In other 8 words,.could we just do the testing'while the check' valve 9 really performed its function --
10 MR. MARSH: _It may not be a safety analysis. It 11 may not be a Chapter XXV safety analysis. But there should 12 be some way'of backing out what the flow rate should be i
13 through that component -- for probability purposes. !
14 (Q) You can calculate.that. Anytime you have a 15 process,.you can always calculate the flow.
-16 MR. MARSH: Sure.
17 (Q) We have the pressures; that's no problem.
18 MR. MARSH: Sure. The piping was designed, the 19 valve was purchased, the compressors were sized. The 20 accumulators were sized with some thought of what the flow 21 rate was going to be when it started the diesel generator or 22 when supply cooling water with heat exchanger. It may not 23 be readily available. You may not have the calculation.
24' But we really should know up front what it's supposed to be.
25 Right, Mark?
HERITAGE REPORTING CORPORATION (202)628-4888
F.
a-l- U N '
6 ls- 106-
' ~
1 (Q) Right.- In some cases we have to dig out.or
.2- .do a-reference value test, or do a calculation, like Janice 3' says._ Eut a lot of these things that we have are not li readily available for safety.
5 MR. MARSH: But if you' don't know what it's 6 supposed to pass, then how are you going to check to see 7 that it's passing then?
8 (Q) Sometimes your system requires a parameter to 9' function correctly. You can do a flow calculation. You may 10 have it within 100 percent. 7 c's not a very accurate flow 11 calculation -- trying to back down from the parameters to a 12 flow calculation, and then verify that flow calculation.
13 What you really want to verify for the system safety 14 function is the parameter to start with. So that's part of 15 the dilemma there. You can do a hell of a lot of 16 calculation, number -- to try to get flow number. And the 17 number may be really what's going to help you.
18 MR. MARSH: Because you're trying to control 19 temperature, as opposed to trying to control flow.
20 (Q) If you monitor a temperature that is lower 21 than, say, what you need to operate the equipment 22 satisfactory, you know that that flow is conservative.
23 MR. MARSH: I understand the problem.
24 MR. SHAW: You must have an acceptance criteria.
25 And the acceptance criteria itself is derived from your HERITAGE REPORTING CORPORATION (202)628-4388
107 1 requirement.
2 (Q) That is correct.
3 MR. SHAW: -It may take a different form, but so 4 long as you have this acceptance criteria clearly stated, 5 then you are away from judgment, which is what we are trying 6 not to get into. So long as you can justify your acceptance 7 criteria, I would say it's acceptable.
8 MR. MARSH: That's good. That'means, for example, 9 if you were going to try to quantify the operability of a 10 check valve in a heat exchanger, for which you could resdly 11 measure flow, but you could construct a test with a design 12 heatload, with some kind of heatload, no heatload on that 13 heat exchanger, and with no river water temperature or 14 something, with the pump performing in its known good 15 condition, that the heat exchanger was performing operably.
16 And it seems like that's a quantifiable test, too. That's 17 not qualitative, that's a quantification of the 18 functionality of that whole system of the check valve, and 19 of any other system that's there.
20 Now, you have to make sure that you can measure 21 degradation of those components, too. That's the ticklish 22 part. Because heat transfer coefficients are very tough to 23 measure if the river water is not in its design condition in 24 accordance with the FSR. It's colder than average. Then, 25 how are you going to be rble detect that the valve isn't HERITAGE REPORTING CORPORATION (202)628-4888
j..
108 1 performing like it shoul-d be.
2 (Q) Sometimes there's various ways to do that.
l 3 Now, if you're talking a big heat exchanger.with 10-inch 1
4 pipes, that obviously -- we're going to do some thinking, 5 hard thinking. Sometimes you're talking about, for 6 instance, the PWR and NCR -- support system which draws
- 7 water out from the intermediate pump and runs it through a -
8 ~ seal condenser or something like that, or a lube oil 9 cooler, and you're talking very small lines.
10' You know that component needs to get a set flow --
11 with a guarantee in oil temperature, say. What'you can do, 12 you can protect, say, a maximum oil temperature in an 13 accident is a 160 degrees. You can calculate and pick a 14 lower temperature which is conservative, which will cover
-15 these unknowns, say, 110 degrees. If you know during your 16 test that you meet that 110 degree parameter, then you'll be 17 conservative for these flowing variables. To try to pin 18 that down to a set flow rate is very, very difficult.
19 MR. MARSH: It may be a temperature as a function 20 of river water temperature. Lube oil temperature is a j 21 function of river water temperature.
22 (Q) In this case, it's -- storage temperature. i l
23 MR. MARSH: Yes, sir.
24 (Q) Can we take that one step further. We have a 25 question there on the same subject. Ours is really trouble j i
HERITAGE REPORTING CORPORATION (202)628-4888 !
l l
I
~ _. . _ - _ _ _ _ - _ _ - _ _ _ - _-_ _ _ -
f 109
=i 11 with check valves..'We'have some flow operated valves that 2! are containment isolation valves. And we have a check valve l g, 3 which bypas'ses around that. And-it's purpose is to' relieve 4 .the pressure' building up between the inboard and outboard 5 containment isolation valve. There's no real way.then to 6 close the that valve.
7 But.we quantify that with a pressure between the, 8' lines that this' valve will relieve 1that pressure. You 9: isolate the pressure and fuse the lines and we verify that 10 the pressure is relived.
11 MR. MARSH: So you isolate the valves, pressurize 12' .the line, and check the pressure decays, indicating that' the 13 valve is passing some kind of flow.
14 (Q) It's actually the containment isolation valve in 15 this case,.and we do a performance task, downstream of that 16 line. Then,.of course,'the check valve opens up to' relieve 17 lthe pressure.
18 MR. MARSH: It's a thermal relieve. So it's got a 19 design purpose. And you can find out what the design 20 capacity for that valve should be by calculation or by
' 21 design documents of some sort.
22 (Q) Well, the design -- to relieve the pressure.
23 MR. MARSH: Without quantification. I mean it was 24 sized in a certain way. I I
. 25 (Q) Well, it was put in there because of the HERITAGE REPORTING CORPORATION (202)628-4888
110 l
1 relieve valve. You could'not order a relieve valve, than i I
2 went from a low class, or class 2, to a class 1 system. l 3 Instead of dumping it - put a check valve in there so that 4 you can have safety classification.
5 MR. MARSH: Okay. It's, what, a three quarter 6 inch valve of something?
7 (Q) Very small. You also have along that line, 8 second grades. You can always make an issue of -- breakers 9 as opposed to relieve vacuum as a -- It doesn't have to go 10 with check valves, simple lift check valves. It just comes 11 off. How high does it have to come off? We have no concern 12 over that as long as we don't get a vacuum.
13 MR. MARSH: You could construct some type of 14 scenario where the system is isolated. You got the adjacent 15 system at temperature, what type of maximum thermal 16 expansion there could be, what kind of fluid flow should 17 there be through this valve in order to relieve that 18 pressure. So you could do an easy back-of-the-envelope 19 calculation on capacity through that valve, what it should 20 be.
21 (Q) Most of these check valves are there when we 22 isolate a lube, but the -- temperature goes up. It's a 23 long, drawn-out process.
24 MR. MARSH: But there's thermal expansion that's 25 relieving, so there's a heat process going on where that HERITAGE REPORTING CORPORATION (202)628-4888
111 1 isolated system is heating up, trying to relieve that 2 expansion through this relieve valve. It's a thermal 3 relief; you're just trying to relieve thermal expansion.
4 (Q) That's the process.
5 MR. MARSH: So you can do a simp 3e calculation for 6 what type of flow there should be minimum through that 7 valve. And it would be small flow rates.
8 MR. ROCKHOLD: You can't measure it.
9 MR. MARSH: You can measure flow through that 10 valve, but you're pressurizing the lineup as a test, and 11 checking to see that the pressure decays.
12 MR. ROCKHOLD: No, you're not even pressuring it.
13 I don't think you attempt to pressure it and verify that you 14 can't end this pressure.
15 (Q) Most of these are in test return lines. So 16 as you isolate the test valve to CST, you isolate the first 17 one that pumps the running, then we shut the second one 18 isolating the falling of the pressure. And we look at a 19 test gaugo, that's where pressure drops off. Then to 20 actually relieve that pressure, we're talking -- We identify 21 these valves, but we have no concern, say, we want to 22 measure the flow rate to these valves.
23 MR. MARSH: I'm trying to construct a way for you 24 to quantify what the flow is to that valve. You can't 25 measure it directly, but you can measure it by pressuring HERITAGE REPORTING CORPORATION (202)628-4888 l
L---._----_-_-..-- .
t.
112 1 it. You know what the. volume is inside that pipe. You know
- . 2 .what it's pressure, and you can back out some. type.of flow-
-3 rate through that valve.
4' (Q)- The trouble is you're going against another 5 . valve, the boundary valve. And it's not really a true-r 6 pressure -- open a downstream valve, the pressure that's 7 fouled up is one step -- the downstream valve is open. That 8 fouled up pressure is going to flow through the check valve 9 and --
10 MR. MARSH: Okay. I'm afraid we're spending too 11 much time trying to give you a real good answer. And I 12' don't understand enough about the process.to do that at this 13 point.
14 (Q) It's not a generic, general-type question for 15 all the check valves. A lot of them can only test like 16 that ,.
17 MR. MARSH: What we're trying to do is construct a 18 general answer that says quantified to the extent that you 19 can. If you can measure a process, if you pressure and 20 pressure decay, back calculate -- use of that system. If 21 you know something about that valve, and it's CCV, you can 22 do that. And the whole thrust of this has been quantified 23 to the extent that you can. It's not qualitative assess 24 matter, quantification.
25 If you don't have the data to find out what the HERITAGE REPORTING CORPORATION (202)628-4888
113 11 flow rate is, it should be through that valve, figure out 2 what it should be through that valve in order to meet its 3 design conditions. If you don't have that kind of job-4- )
limitations to the extent that you can, if you can't do !
5 that, then come up with some system test.which assures 6 system functionality by maintaining a process temperature.
7- So we've all been talking about trying to quantify I 8 to the extent that you can, without trying to rely on a 9 qualitative assessn.ent. So it's generic; and I'm trying to 10 give you a generic answer which is quantified. Use 11 quantified tests, quantified bases for acceptance criteria 12 for these valves. If you can't do that, then process with a 13 quantification on the test parameters.
14 I think we really need to go into some more 15 details on this particular valve. I need to think about it 16 some more, even beyond that. I don't mean to diminish it's 17 importance, I just don't have enough at this point to answer 18 that.
19 MR. SULLIVAN: The next question pertains to 20 position 11. It says:
21 (Q) Do safety related compone1 s outside of Class 22 I, II and III need to be tested in accordance with the code 23 and included in the IST program? Or is it the intent to 24 have some form of testing to demonstrate operability?
25 MR. MARSH: I'd like to answer this. If you saw HERITAGE REPORTING CORPORATION (202)628-4888
114 1 the first version of the Generic Letter, position 1, it said 2 all safety-related components -- are you Vince Treaty (ph),
3 Point Beach?
l 4 (Q) Yes.
5 MR. MARSH: You're Vince? Okay.
6 You saw the first Generic Letter. You saw 7 position 1 that said all safety-related components that were 8 there. We got a lot of feedback from a lot of industry 9 groups, from the Code Committee, that said, "Uh-uh. Look at 10 the regulations and look at the ASME code. It's Class I, II 11 and III components.
12 And we went back and looked very carefully through 13 the regulations, through the regulatory history of this 14 thing. And the way the code regulation was originally 15 framed, it said safety grade stuff. That's where it 16 started. But it got changed. It got wiggled.
17 I think that happened because folks thought that 18 safety grade equals Class I, II and III. You can find the 19 point when it changed. Of course, you don't know what the 20 thinking was, why it changed. But we think that's why it 21 got changed.
22 We changed the position in Position 1. We changed 23 it to its current wording in Position 11. Position 11 says, 24 Class I, II and III components, which have to have a new 25 program, because that's consistent with the code, that's HERITAGE REPORTING CORPORATION (202)628-4888
115' 1 consistent with the regulations.
2 But it went on to say, in the last four lines of 3 that position, it says, "However, we believe that Appendix B
- 4. says that you should be testing these components by some 5 type of program to make sure they're operable." And that's 6- ~ basically what we mean. Just because you've got a safety-
.7 grade equipment that's not code classed, it doesn't mean it
.8 shouldn't be tested. There should be some operability check 9 on that component.
10 Now let me just tell you that good plants, with 11 good IST programs, have all safety grade stuff down there.
12 All safety-related equipment is in their program. So if 13 you're looking to elevate IST up to the place where it 14 should be, that's the way to do it.
15 You understand exactly what the regulations now 16 say, what they mean. But we're saying that when an 17 inspector comes and he says, "How are you making sure that 18 that thing is operable?" You're going to have to answer 19 that question.
20 One way to answer that question is to have it in 21 the IST program. Another way to answer that question could 22 be another testing program, some other way to make sure that 23 it's operable.
24 Yes, sir.
25 (Q) Would it be acceptable to say that tech spec HERITAGE REPORTING CORPORATION (202)628-4888
_ _ . _ _ . _ ~ _ _ _ . _ - _ . _ _ _ _ .
116 1: had an operability t'est'on it, that you could use that if.
2- it's a noncode valve, and say this is how I'm going to test 2L .it , and we'll just use that?
4 MR. MARSH: Use the system' functionality?
- 5. (Q) Identified that that's a safety-related-6 valve, and this is how I'm testing it. Is that what you're 7 looking for in one~ program --
8 MR. MARSH: Tech specs 9
10 11 12 i 13 14 l
15 16 l'
l 17 18 19 20 21 22 23 24 25 i
HERITAGE REPORTING CORPORATION (202)628-4888 i
Lg , * '
117 c1 :HR. MARSH: The': tech spec has the valve in it'and-t 2 , .it's got=a. surveillance. test'on it'in the tech'apec.- I 3' would think that's reasonable. I'm-hard pressed'to find ~a' 4 tech spec that --
5 (Q) It will even give you frequency.
6 MR. MARSH: Excuse me?
7 -(Q) It will even give_you frequency. And I' m 8 looking at the diesel specifically_for the air start system.
9 It tells you to test it one way.
- 10 MR. MARSH: Well, that's'a system test. That is 11 not necessarily a valve test. That's a. system test. You 12- may start the diesel up, check both parallel valves, both 13 sets--of parallel valves.- You have'four valves, two in
' 14 series or two valves, each one in series. Each one in 1
15 parallel, I'm sorry.
16 (Q) Two parallel lines.-
17 _ MR . MARSH: Each line has one valve in it?
18- (Q) Yes.
_ 19 MR. MARSH: Are you testing each valve?
20 (Q) Specifically?
21 MR. MARSH: Yes.
22 (Q) No.
23 RMR. MARSH: Well, you're testing that the diesel 24 can start and that's what the tech spec has you do. Check 25' the diesel can start. But how are you making sure that both HERITAGE REPORTING CORPORATION (202)628-4888 l
118 1 of those valves are doing their job? Each train are doing 2 its job?
3 (Q) They check it whenever they replace the motor 4 which is like after it starts.
5 MR. MARSH: They test each valve?
6 (Q) Yes. They test it specifically to make sure 7 that each line will start.
8 MR. MARSH: When they do the surveillance test in 9 the tech spec.
10 (Q) Right.
.11 MR. MARSH: That may be al1 right.
12 (Q) I mean it satisfies the tech spec operability 13 concerns.
14 MR. MARSH: You can't say that system operability 15 in the tech spec satisfies component operability in general.
16 You can't say that. In this case, if the tech specs or your 17 procedures say you test each valve individually, you 18 alternate back and forth in accordance with the tech spec 19 test for a diesel generator startup, they may be adequate --
20 (Q) That's what I'm saying. It's non-code and 21 that's how we would describe it.
22 MR. MARSH: I am leaning towards yes.
23 (Laughter.)
24 MR. MARSH: I've got to hedge a little.
25 Sir?
HERITAGE REPORTING CORPORATION (202)628-4888
119 j
1 (Q) But you are still looking to meet the code of i i
i 2 frequencies.
3 MR. MARSH: Well, you can't -- by the intent of 4 the question, by the intent of the code, we can't require 5 you to meet the code frequencies for non-code components. I 6 If you do, that's adequate. Clearly adequate.
7 If you do not, we're going to ask you how are you 8 making sure that component is operable. Now, we get into 9 subjective areas. If you do it every time you start the 10 diesel generator, alternate back and forth, that may be 11 okay. If you have another component on some other 12 non-code class equipment that you don't test, except after 13 you replace it or something that's every two years, that is 14 not. It gets very subjective.
15 That's why having it in your program, meeting the 16 code requirements is the clearest way of meeting the 17 requirements. But we understand what we can and cannot 18 require.
19 (Q) Every 50 starts we tested each train --
20 MR. MARSH: It's every 50 starts.
21 (Q) You totally rebuild them.
22 MR. MARSH: I see. I think we would want to look 23 carefully at that.
24 Yes, sir?
25 (Q) Does non-code equipment have to comply with HERITAGE REPORTING CORPORATION (202)628-4888 i
g.
J ,; 1201
'l._ (tech spec 405?
l 2 MR. MARSH: .Okay, the question was d'ea k :3 n6n-code equipment have to comply with tech spec 405. I m
4 think the answer ~is no. Non-code equipment.does not have to 5 comply with-tech spec 405.
6 Yes, sir?
p-N 7 (Q) Just as a matter of clarification. When you 8 say non-code, you mean.non-category A, B, C within your ISI' 9 boundary, right?
'10 . MR . MARSH: Class 1,-2, and 3.
31- (Q) Equivalent to Class 1, 2, and 3 for Section
.12 11.
13 MR. SULLIVAN: This' question is coming possibly.
14 out of the design that preceded Section 3.
15.- (Q) For example, in our plant, we-criginally 16 classified the safety guide 126. Now, our second 10 year 17- submittal has to be Reg Guide-125 and NuREG 0800. So, we la have in our'second 10 year submittal extended a lot of the 19 ISI programs beyond the present as we call it safety-related 20 boundary in the original design.
21 But when you say " code" you mean like 86, Section 22 11 says " Category A, B, C/ type 1, 2, 3. You are referring 23 to Class 1, 2, 3, as Section 11 code.
24' MR. SULLIVAN: Yes.
1 25 (Q) MR. SULLIVAN: Okay, this question is from HERITAGE REPORTING CORPORATION (202)628-4888 n __-_m._.-.---_..--n -
l 121 1 Roger -- I'm not sure what your last name is. It says, 2 "What additional NRC guidance can be provided on testing 3 Skid mounted pumps and valves, i.e., diesel generator lube 4 oil pumps and valves, internal engine cooling, RCI system, 5 vacuum pumps with only once source of power, et cetera?"
6 "Most of these pumps and values do not have the l
7 necessary test instrumentation to support ASME 11 testing 6 and do not fall within the scope statements of IWPMB. Will 9 modifications need to be performed?"
10 (A) MR. SULLIVAN: In the last two questions we 11 talked about position 11 and I wonder if we have answered 12 your question or do you still think we need to talk about 73 this more? Basically, what you have talked about here are 14 safety-related support systems that are not built according 15 to the code. And then we also talked about being --
16 (Q) Well, they may meet the code like for a motor 17 or something, but they don't have an alternate power supply, 18 so, they don't meet the criteria for putting it in. You 19 know, it's safety-related but it only has a single source of 20 power. So, the code says you need two sources.
21 MR. MARSH: Now, the issue is number of power 22 supplies to meet the scope requirements,.not as a code 23 classification?
24 (Q) Right.
25 MR. MARSH: It is a code class component.
HERITAGE REPORTING CORPORATION (202)628-4888
122 1 (Q) I mean that's the guidance. If you go to the 2 code, it tells you you are only required to have --
3 MR. MARSH: In'the IST program, it has got two 4 power supplies, emergency power.
5 (Q) That's a standard.
6 MR. MARSH: Give me an example.
7 (Q) If you take a diesel generator, the' fuel pump 8 is' mechanically driven on diesel. There's those type of 9 equipment. Then there's the auxiliaries, vacuum pump. It 10 just pulls the~ vacuum on the condenser, the barometric-11 condenser. 'Or you have a condensate pump. Those pumps do 12 not have an alternative power supply emergency source.
13 .HR. MARSH: I want to think on that. Okay? I 14 want to give a'little thought to that. It's a different 15 twist to the same type of question.
16 (Q) Right.
17 MR. MARSH: But we've never heard it in the 18 context of pumps and power supply to it.
19 (Q) It's the same thing with the valves in our 20 systems.
21 MR. MARSH: Yes. I just want to think about it.
22 Apparently they are considered safety-related components, 23 but they don't have the power supply provision in the scope 24 statement. The question is should they be in the program?
25 Are they in the program now? Does anybody have those types HERITAGE REPORTING CORPORATION (202)628-4888 i
123 1 of components in their program?
2 (Q) Do you have your Section 11 up there in front 3 of you?
4 MR. MARSH: Yes.
5 (Q) If you look at the exemptions of IWP, I think 6 you will find those direct drive on diesel generators 7 specifically are exempted from the programs altogether.
8 MR. MARSH: Okay.
9 (Q) I'm just saying when you get around to saying 10 you should include all safety-related pumps and valves, I 11 mean there are some of those things that aren't designed to 12 be tested like that.
13 MR. MARSH: Well, if we say how do you assure 14 operability of that component? What type of provisions do 15 you do to test it to make sure that it's not degrading 16 itself by Appendix B requirement? How can you answer that?
17 (Q) We do a quarterly pump check throughout the 18 whole system. And you verify it like the vacuum, yos pull 19 it back in for the vacuum pump.
20 MR. MARSH: What are you looking at on that pump?
21 Are you looking at anything on that pump specifically?
22 (Q) As long as it pulls the vacuum out of the 23 tank.
24 MR. MARSH: It's a system pump. It's not the pump 25 itself.
HERITAGE RFPORTING CORPORATION (202)628-4888 m_______.. _ _ .
124 1 (Q) You n9y not even need the pump because you 2 have ruptured --
3 MR. MARSH: You're confusing system functionality 4 with component functionality. The code is a component code.
5 It is not a system code. It is not meant to test the 6 diesel.
7 (Q) That's what I'm saying. When you apply it to 8 the diesels, it doesn't work. See, that's what the question 9 is supposed to get around to is some of those things are 10 made to test by the code component specific. It's a diesel 11 generator package. That's all the question is.
12 MR. SULLIVAN: They weren't designed tu the code, 13 were they?
14 MR. ROCKHOLD: No. They are designed to IEEE 15 requirements not Section 11 requirements.
16 (Q) Right.
17 MR. ROCKHOLD: Which never the two shall meet.
18 (Q) They do not fall in our code boundary.
19 MR. SULLIVAN: Okay, they're not ASME Code Class 20 1, 2, or 3 in any of these examples you' re talking about.
"1 (Q) But I'm still saying you still have that 22 requirement in the code: to exempt them. Or that it 23 doesn't fall within the scope statement.
24 MR. ROCKHOLD: I looked at the code and IWP 1200 25 does not say that these components are exempted. It says HERITAGE REPORTING CORPORATION (202)628-4888
-_ -- - - )
b g 125 1 that drivers:are exempted except where they --
2 (Q) Not the exemption requirements.
3 (Q) What he is saying is when you go through and 4 set'up your Class 1, 2, 3 boundaries these are outside those 5 boundaries.
6 MR. MARSH: Oh, you're talking because it doesn't 7 have power supply.-
18 (Q) Right.
9 MR. MARSH: Not because it's exempted because of 10- the provision of exemption there.
11 (Q) Right.
12 MR. ROCKHOLD: For the diesel,-that's true. But 13 that is not.necessarily true for'these little support pumps, 14' say, on the high pressure coolant injection pump for the L 15 BWRs. Is that true? This' condensate pump on the HIPSY and 16 RIXY pumps and the vacuum pump. Were they purchased as code 17 class components? Were they designed as code class 18 components?
19 (Q) The *7XY pump is, yes.
20 MR. ROCKHOLD: The pump was and all the supporting 21 systems provided on that skid were purchased as code class 22 components. You're saying that's not true?
23 (Q) No, I mean you can look at your' design 24 documents.
25 MR. ROCKHOLD: I'm not saying whether they're HERITAGE REPORTING CORPORATION (202)628-4888
l I
i 126 L -1 exempt or~not exempt.. I'm saying'are they purchased-as-code L 2 class components of the support equipment on, say, the HIPSY 3 skid? I 4 (Q) If you look at our PNIDs, it does show those 5 codes.
6 MR. MARSH: So, they're not code nor are they 1.
7 powersd by emergency power. So, you've got two balls.
8 (Q) Well, still yo'u'said you treat all systems as 9 safety-related.
10 (Q) There is non-ASME Section 3 code items-within 11- your ISI boundaries.. That's why I was saying when you say L 12 non-code, .do you mean non-ASMC Class 1, 2, 3 classified?
13 Because Class 1, 2, 3 encompasses more than this ASME' code.
14 MR. MARSH: You had to make a call, didn't.you?
15~ At some point you had to translate B31-1 into Class 1, 2, 16 and 3. So, you may have B31-1 stuff in there, but you still Tl know whether it's code Class 1, 2, and 3. .So, there should 18 be no question about B31-1 stuff. You have already made 19 that transposition.
20 Okay. Where are we on this gentleman's question 21 about the pumps?
22 (Q) Maybe it was just meant to stimulate 23 discussion.
24 MR. MARSH: It did that, but I want to give you 25 the answer. The question was not retracted.
HERITAGE REPORTING CORPORATION (202)628-4888
L , I L l L i 127 1 (Laughter.)
2 (Q) HR. SULLIVAN: -This question has come up at 3 the other two meetings.- It says: In the first 10 year ->
i 4 submittal, an SER approved or relief request which is'not 5 consistent with the alternative positions in Generic Letter- J 1..
E 6 89-04. Does Generic Letter 89-04 void previously approved ,
1 o 7 relief requests via an SER or may these alternative relief 8 requests not consistent with Generic Letter 89-04 still be 9- considered valid and so documented in the IST program?
10 (A) MR._SULLIVAN: Let me just try to clarify one 11' point. This is from Toledo Edison. The SER that you're 12 referring to, you are still considering_ operative, aren't L 13 :you?
14 (Q) Yes. We are one of the -- we fall in the 15 category 2 of an approved SER.
16 MR. MARSH: Have you changed your program beyond 17 the version that the SER was based on?
18 (Q) Well, we submitted it early in the '70s, 19 updated it in '82. And now we're at the time we are getting 20, ready for our second 10 year submittal. So, the question is 21 we may have some relief requests -- for the first 10-year 22 submittal have been approved. They may not necessarily be 23 consistent with the Generic Letter. So, in preparation of 24 our second 10-year submittal, we would like -- the question 25 is can we still use those old relief requests provided HERITAGE REPORTING CORPORATION (202)628-4888
128 a
1 configuratf.on and stuff'like that.is still applicable. Or 2- do we have to go to the Generic Letter. There may be an 3 instance where we would rather go with the relief request as
'4 opposed to what's in the Generic Letter.
5 I'm not in a position to give you examples at this 6 time.
7 MR. SULLIVAN: We have had a lot of discussion 8 about this and I think it's the kind of question that makes 9 us uncomfortable. I think one of the issues that we're 10 talking'about is what takes precedence? The SER or the 11 Generic Letter?
12 And I think from a strictly legal point of view, 13 your relief request was approved and the Generic Letter 14 wasn't written in such a way to undo those approvals. And I 15 guess in the same kind of looking at it legalistically, when-16 you go into your next update, if there is no aspect of that 17 relief request that is changing or that circumstances around 18 which it was originally written --
19 MR. MARSH: Like the code. The code is the same.
l 20 It hasn't changed.
21 (Q) The code requirement hasn't changed.
22 MR. SULLIVAN: Right. Then we would say that from 23 a strictly legal point of view, the relief request continues 24 to be valid. We would encourage you to take the high road, 25 at least what we consider to be the high road and follow the HERITAGE REPORTING CORPORATION (202)628-4888 i
129 1 Generic Letter.
2 I guess we would also say that at sometime in'the ;
3 future we could look at those relief requests and reopen 4 some discussions. We would in those discussions probably 5 try and re-evaluate why we gave relief in the first place 6 and try to convince you, if we thought it was appropriate, 7 to reconsider. And we would just have to see whether we 8 thought arguments were convincing enough to persuade you to 9 follow the Generic Letter.
10 I think what we're saying 4.s that from the legal
'll point of view, the Generic Letter doesn't take away those 12 approvals. That is not going to stop ur from reopening 13 discussions with you on areas where we don't really agree 14 with what was done in the past.
15 MR. MARSH: The Generic Letter reflects the 16 current best thinking we've got. That is where we want to 17 go at this point. And we know there are SERs that are out 18 there that are les.s than where we want to go. So you need 19 to understand that. That's where we are going to be coming 20 from. The technical bases for those SERs may be a mistake.
21 And we're going to try as best we can to get us to the best 22 state we can, not a lessened state.
23 (Q) I have a question. Currently, inspectors on 24 sites and the issue of manual valves is being raised.
25 Certain valves that are manual valves are safet '-related HERITAGE REPORTING CORPORATION (202)628-4888
130 1 valves and must be exercised. And what basis are we using l 2 to say that a manual valve must be exercised? What throws l .3 them in that category?
4 If it is a manual valve, we're saying that it 5 could be a mile away and we have an accident, you know, we 6 can't get an operator out there. It might not be involved.
7 And all of a sudden, somebody says, _"This manual valve 8 appears to have a safety function." What is the basis used 9 for manual valves?
10 MR. MARSH: Well, I don't know the particulars of
- 11. your situation. I don't know the inspectors' concerns, 12 whatever. But let me just say in general I'm not satisfied 13 with the code wholesale exempting manual valves. I think 14 that's a mistake.
15 MR. ROCKHOLD: The code does not specifically 16 exempt manual valves. If you take credit for the 17 operability of a manual valve, i.e., you have to change the 18 position of the valve sometime in any safety analyses, then 19 you have to periodically demonstrate you have the capability 20 to change the position of that valve. Typically, they are 21 not addressed in any safety analysis because of the time 22 constraints that are typically put on these valves.
23 (Q) I guess that's what we're trying to get down 24' to is art wp worried about valves now because of emergency 25 procedures?
HERITAGE REPORTING CORPORATION (202)628-4888
131 1 MR. ROCKHOLD: *Sure.
2 (Q) Ultimately, those are the ones you are 3 concerned with?
4 MR. MARSH: Well, I don't know what the system is, 5 whht the specifics were, what the concern was. I don't know 6 whether the inspector's concern was driven by emergency 7 procedure or whether it was driven by a code requirement or 8 something.
9 (Q) I mean recently -- manual valves were 10 excluded. There used to be a Category E.
11 MR. ROCKHOLD: No. Category E valves were valves 12 that were locked in position. Locked or sealed.in position.
13 (Q) Right.
14 MR. ROCKHOLD: And manual valves -- as a matter of 15 fact, Category A and B manual valves are not that atypical.
16 They exist. And operability can be demonstrated by an 17 operator changing the position of that valve. I think maybe 18 a popular misconception is that typically when you talk 19 about valves that have to change positions that are not 20 check valves, you are talking about power-operated valves.
21 And then the acceptance criteria is fairly well defined on 22 what stroke time requirements there are and all that, but 23 very little is said about any manually operated valve.
24 So, we have never viewed manual valves as being 25 excluded from Section 11 testing. It's that the acceptance HERITAGE REPORTING CORPORATION (202)628-4888 I l
l
132 1 criteria might be a little less vague. An operator has the 2 ability to change the position of the valve if you take 3 credit for being able to change the position of the valve.
4 (Q) So, we're actually narrowing down the valves 5 that serve a safety function in emergency conditions. And 6 you would say we need an operator to operate that valve.
7 (Q) I think the question is in order to satisfy 8 the IST program, you may have x-number of valves in there 9 that are manual. And you use them as part of your emergency 10 plan.
11 MR. MARSH: Right.
12 (Q) But, you know, as a last-ditch effort, you 13 want to try to get water to that vessel and you may include 14 other valves. Where do you stop?
15 MR. MARSH: We're not expanding. I don t believe 16 that this question involves an expansion of the scope of 17 IST. I think the scope of the IST is a legal boundary for 18 the program. And I don't think anything that Herb has saio 19 or I have said limits the scope.
20 But if you are talking about valve-to plant valves 21 that are beyond the code class, that's not within the scope 22 of IST. It's not there.
23 We know that procedures involve all kinds of ways l 24 of getting water into the vessel. And if you want to do 25 that, that's is probably a prudent thing to do, but we're HERITAGE REPORTING CORPORATION (202)628-4888 i
, 133 it i not requiring that at this point.
12- (Q) So, _ we are basically saying that Section 11 3 valves that are required --
4 MR. MARSd: Code class stuff that has a specified 5 function of shutting down the reactor as needed to mitigate 6 the consequences of an accident, et cetera, the wording is 7 in the scope statement for you. I don't know what - you 8 know, it depends on what code version you're on is your 9 operative' scope statement.
10 I would like to know the particular valve and what 11 was the issue of the particular valve.
12 (Q) We have a swale line. And what happens is in 13 alternate path, if we lose our path or emergency service 14 route --
15 (Q) If a barge runs into your discharge --
16 (Q) And you can actually shut the valve --
17 It was considered that you had to go out and shut 18 that valve. And that's all well and good, but we can't do 19 that feasibly in the plan when it's running because we'll 20 blow away some of our buildings.
21 (Q) MR. SULLIVAN: The next question is rather 22 long. It says: On page 2 of Ted Sullivan's presentation he 23 indicated NRC will not issue SERs on Attachment 1 items and 24 justified alternatives. First question: Are the justified 25 alternatives as indicated on that slide the four points on i
HERITAGE REPORTING CORPORATION (202)628-4888 i
134 1 past component history?'
2' (A) MR. SULLIVAN: That was what I meant by that 3 slide. By the justified alternatives, I meant addressing 4 the four points that are on page 3 of the Generic Letter.
5 The second part of this question says: Can I use 6 these four alternatives to justify a deviation from the 7 Attachment 1 positions?
8 That was the intent of Section B of the Generic 9 Letter on page 3 that talks about maintenance history, et 10 cetera, was to be used just for justifying alternatives from 11 Attachment I which leads into the third question or fourth 12 questions, I'm not sure which.
13 (Q) MR. SULLIVAN: If so, are these then approved 14 by the Generic Letter.
15 (A) MR. SULLIVAN: ' he idea of the Generic Letter 16 was that they would be approvec through the Generic Letter 17 in that if you meet -- if you address all four items and 18 justify your alternative esting on the basis of the 19 maintenance history identified in those four items, then 20 they are in effect approved by the Generic Letter.
1 21 (Q) MR. SULLIVAN: After issuing a confirmatory 22 letter, can I go through the above process to get automatic I
23 or pre-approval of Attachment 1 exemptions in the future? '
24 MR. MARSH: What does that mean?
25 (Q) The problem I saw there was if you go over to HERITAGE REPORTING CORPORATION (202)628-4888 L
135 1- page 3 of the same presentation that Ted had, that 2 . alternative item column disappears and you go to 50.55A(G) 3 that would be "all others". You don't satisfy the attached 4- positions, i.e., here I have -- say I've added a new gizmo 1
5 or something or I am re-examining existing components and I 6 find that for some reason I can no longer satisfy your )
7 . Attachment 1 positions. I've already issued a confirmatory I B letter -- let's say I'm 10 years down the road and I'm 9 re-doing my program. Can I go through that same process 10 that I did with my confirmatory letter, get this 11 pre-approval and justify it on those grounds or am I being 12' channeled'right back in to 50.55 if I don't satisfy to the 13 ' word your attached position.
14 MR. SULLIVAN: Okay. Let me try and respond to 15 that. I think there is a potential conflict in the Generic 16 Letter. It sort of says it in one sentence and it doesn't 17 say it in another.
- s. -
18 For example, it says -- I'm reading from Part D.
19 Program Updates and Revisions to which this question is l
20 being addressed.
p 21 "For IST program changes for which specific 22 positions are provided in Attachment 1, licensees should 23 follow the guidance in Section B above."
1..
L 24 So, by that, we intend to indicate that if you 25 could sometime in the future take advantage of that part of HERITAGE REPORTING CORPORATION (202)628-4888
I 136 1 Section B that talks about justifying alternatives, you are 2 still allowed to do that. That was the intent of this 3 particular sentence.
4 I realize the next sentence may confuse the point.
5 The next sentence says: "For IST program changes in areas,"
6 no, I guess it doean't, "not covered by Attachment 1 the 7 provisions of 50.55A(G) should be followed."
8 No, I don't think that confuse the --
9 (Q) Would that read better if it read:
10 " Attachment 1 or Section B above."
11 MR. SULLIVAN: No , because in your question 4, I 12 am going to address applying those alternatives to positions 13 not covered by Attachment 1.
14 (Q) That's question 5. Right. That's the 15 follow-up to the question 4.
16 (Q) MR. SULLIVAN: Okay. Question 5 says: "Can 17 the four points be used for non-Attachment 1 items following 18 a similar process?"
19 (A) MR. SULLIVAN: This idea that we put in Part B 20 of justifying alternatives through maintenance history was 21 kind of a new one. As we said this morning, we tried to 22 come up with something that we feel pretty comfortable with.
23 But we also want to limit its use until we have some 24 opportunity to evaluate how well it was working. So, it 25 wasn't our intent to open it wide open to all areas covered HERITAGE REPORTING CORPORATION (202)628-4888
137 1 by the IST program. So,'in short, I would say, no, it 2 wasn't our intent that those.four points could be used for 3 non-Attachment 1 areas.
4 (Q) Let's go back to item 4, then. Can I then 10 5 years from now -- say I have a check valve that I can't do 6 anything with to meet your approved variations -- can I use 7 the history standpoint in the alternative testing 8 justification process to get that pre-approval? That's 9 going back to something we recognize as being a problem area 10 where we've had shortcomings or other concerns. Not 11 something that falls outside of the scope of Attachment 1. I 12 MR. SULLIVAN: Okay. I don't want to pinned down 13 to what's going to happen 10 years from now because I'm not 14 sure how many of us will be involved in IST, but I think-15 unless we withdraw this in some future generic 16- communication, unless we withdraw this option, it would be 17 our intent that you could continue to use it in the future 13 for Attachment 1 areas.
19 Yes?
20 (Q) When it references back to Attachment 1 21 areas, okay? For instance, the check valve testing. If you 22 would like to use an alternate type of test for that and you 23 would follow your four points for a justified alternate 24 test, does that require a relief request for the original 25 process or can you just incorporate and document and add to HERITAGE REPORTING CORPORATION (202)628-4888
138 1 your program.
2 MR. MARSH: Okay. -Remember in position 1, it says 3 not only for full-flow testing, it also says there are other 4 ways of detecting full flow through this valve other than 5 having flow instruments. And it gives some other things 6 that you should look at, too.
7 If you have gone through all of that and you 8 cannot meet that, then you go to position 2. Position 2 9 says as an alternative to full-flow testing, with flow 10 measurement, you can use these things. Then you go to 11 position 2, which is a last ditch effort, which is 12 disassembly.
13 Then at says, failing all that, you go to the 14 maintenance history, but you have got to go through all of 15 those things. After you have gone through all those steps 16 and documented all those steps in your program, then you 17 have approval. You don't have to submit and get relief.
18 Those specifics have to be all documented within your plant 19 records, that you have gone through and justified for 20 yourself, you can't meet position 1, you can't meet position 21 2. Okay?
22 Yes, sir?
23 (Q) We are a Table 1 plant and you currently have 24 our 4 on our IST program. And we would like to make one 25 change using this alternative method and which you currently HERITAGE REPORTING CORPORATION (202)628-4888 l
t
_ - _ _ - _ _ - _ _ _ __ l
139 1 have. This is something new on our end. Can we put this in r
2' writing and'get it to you by the six month deadline type 3 date?
4 MR. MARSH: I don't know. What plant, sir?
5 (0) Okay, this is Braidwood.
6 MR. MARSH: I don't know what our schedule is for '
7 Braidwood for the SER. Is it just about done? It's in the 8 briefcase. Sorry.
9 .(Laughter.)
10 Really, if you're about ready to prove what you've 11 got, take the SER and do what the SER says to do. And then 12 you have a program approval. If you want to change beyond 13 that which is approved by the SER, then follow the Generic 14 Letter. It says how to do that.
I 15 (Q) Okay, thank you.
l <6 (Q) HR. SULLIVAN: The next question is from Gary 's 17 Rosner of Callaway. I think we have answered this, but I 1
18 just will read it and make sure,
)
1 19 "When will it be known what the staff position is l j
20 on SER approved relief requests that contradict Generic )
21 Letter 89-04 dictated testing?"
22 (A) MR. SULLIVAN: We talked about that a few 23 ninutes ago since lunch time. It was a similar question 24 from Toledo Edison. Did you get a satisfactory answer?
25 Okay, good.
HERITAGE REPORTING CORPORATION (202)628-4888 i
n~----_-____- - _ _ _ - _ _ - _ _ - _ _ - . _ - _ - - - - - i
140 1 (Q) MR. SULLIVAN: The second question:
2 "Regarding full-flow testing of check valves, why is 3 knowledge of total flow through parallel flow lines
'4 unacceptable? This seems to challenge conservative tech 5 spec requirements for flow balance.
6 (A) MR. MARSH: The intent was not to challenge 7 any of the' flow-balancing requirements. The intent was that 8 you construct tests that make sure that the flow that's 9 required to be passed through individual lines is in fact 10 being checked.
11 Total header flow won't tell you whether you have 12 an imbalance going on within the individual injection lines, 13- themselves. If you have a check valve that is degrading and
. 14 it has too much flow resistance in one individual line, the
.15 first thing that will happen is that some of the flow will 16 be_ distributed to the other lines and the header flow may 17 stay the same. So, you won't see any degradation in total 18 . matter flow.
19 We have had this question a lot. One thing that 20 we always suggest is looking to make sure that you can't 21 measure individual header flow by the use of the, generally, 22 the flow valves or setting devices. When those valves were 23 set up initially they had flow measuring devices on the 24 headers to make sure that they in fact were balanced. And 25 they are capped and sometimes not used anymore, but they're HERITAGE REPORTING CORPORATION (202)628-4888
141 1 'available depending upon'your plant, itself. .But'you do 2 need to know individua1' header flow -- excuse me.
3 Individual flow line flow to make sure that that valve is
'4 performing properly. Okay?
5 (Q) It is your opinion, then that a degraded 6 situation would not be reflected then --
8 (Q) -- in total header flow.
9 MR. MARSH: That's true. It would not. You have 10 got a lot of degradation before you see header flow changes 11 and you can do some little calculations and show that with 12 the flow, with decay starting to change a little bit in that-13 valve, a little bit more that the header flow doesn't change 14 much. It really does not. And that you are looking for 15 individual flow line changes..
16 Now, there are ways to do that. Again, this NIC 17 ~ group is looking at various ways of looking for a check 18 valve disk pos.i bion as a function of flow wi't h magnetometers 19 or with with other devices on it. So, there are ways to do 20 that. Okay? I sense you are not satisfied with that 21 answer.
22 This is a difficult area because most lines do not 23 have flow devices on them.
24 (Q) Even if they do, you've got ALARA 25 considerations with respect to crawling around in back of HERITAGE REPORTING CORPORATION (202)628-4888
__-._________----_--_.m.- - - - - - - . - - - - - - - - - - - - - - - - - - - - - - - -
142 1 the bio-wall next to the radioactively hot areas. And I 2 don't'have a'2eal good feeling of sending mechanics in and 3 doing that if I can prove that it's not necessary. I 4 MR. MARSH: Well, it is up to you to prove that, 5 of' course. . You are going to have to prove that any 6 degradation in that valve performance is going to be.
7 detectable and that's a very tough test to prove. We have 8 tried this on plant and tried that and it -- we have always >
9 ended up with inability for that to be shown. Okay?
10 Degradation in valve performance in those four parallel )
11 lines takes a pretty big degradation in an individual check 12 valve for you to see it in the header flow: virtual 13 blockage in one line. And that is really what it's meant to 14 be. It is really meant to balance of individual valvo ficu !
15 like that so you get safety injection into the other lines.
16 (Q) The tech spec regarding ECCS flow bounds were 17 for the plan, the flow bounds procedure specifies going 18 through three loops, flow verified to three loops as opposed 19 to four. So, there once again that is an indicator that 20 total flow is looked at as opposed to a four-loop 21 verification.
22 MR. MARSH: I'm not sure I understand what you 23 mean. Do you mean that you have to verify that you have 24 flow into three loops instead of four loops? ,
25 (Q) Total flow. We can talk about it afterwards.
HERITAGE REPORTING CORPORATION (202)628-4888 l
J I
l
143 1 MR. MARSH: Yes, let's do that, please. I'm not 2 sure I've heard that argument before.
3 Don?
4 (Q) I was going to say, you're right. It seems 5 to me that if the disk is 50 percent open and if you can 6 raise it all the way up, it's not going to change the flow.
7 That is the key point. Perhaps you should write a tutorial 8 to this to help out the user. Sincerely, people aren't 9 aware of this.
10 MR. MARSH: Well, we've heard this question a lot.
11 (Q) That's the reason. People are confused on 12 how it works. I have some test data on actual valves I can 13 share with you.
14 MR. MARSH: Okay, super. We need to find out 15 about that tech spec, though.
16 (Q) MR. SULLIVAN: The next question says: If a 17 plant with an SER on its IST program has a 10-year review 18 upcoming, how should that be handled? Resubmittal?
19 (A) MR. SULLIVAN: I'm not exactly sure what your 20 question is. I think we have addressed this.
21 (Q) Yes.
22 MR. SULLIVAN: Okay, you're indicating we've 23 addressed this.
24 (Q) MR. SULLIVAN: The fourth part of this 25 question says: Regarding disassembly of check valves, HERITAGE REPORTING CORPORATION (202)628-4888 i
4
144 1 please define extreme hardship speaking with regards to 2 extension of disassembly interval.
3 (A) MR. SULLIVAN: You have really put me on the 4 spot with that question. I think what we had in mind when 5 we wrote that question were the kinds of check valves that 6 are close to the loop where there's high radiation exposure, 7 where you may have to offload a core because the valve may 8 be below the loop.
9 MR. MARSH: The prime example that I cite was one 10 that you gave was where you got to offload the core in order 11 to disassemble that valve. That's the clearest example that 12 there is. Even ALARA -- of course, ALARA is always a 13 concern, but ALARA needs to be an element in an assessment.
.14 It can't be the only element that drives you towards not 15 disassembling a valve. Okay? There is a competing risk and 16 that is that the check valve is doing its job properly.
17 We have been asked numerous times about how should 18 ALARA be factored into proposed changes to the plan or to 19 proposed maintenance practices, whatever else. The NRC has 20 given some guidance on ALARA, the use of ALARA. There is a 21 NUREG available. I don't have its number, but it does 22 discuss ALARA and how ALARA should be used. And, as I say, 23 it is only one element in any kind of assessment for 24 disassembly. Okay? That's about all the guidance I can 25 give.
HERITAGE REPORTING CORPORATION (202)628-4888
145 1 (Q) Could I ask one more question on that last 2 hardship?
3 MR. MARSH: Sorry. No more gyestions.
4 (Laughter.) ;
5 (Q) Sorry about that. Besides draining down for l 6 reload for check valves, what about ones that would be where 7 you would have to operate at mid-loop. With all this mid-8 loop concern and RHR systems, those would be -- we have some l 9 check valves at our plant that are -- that we'd have to 10 drain out the mid-loop valve to disassemble. Would that be a 11 hardship or not?
12 MR. MARSH: Good question. There is a lot of 13 concern about mid-loop operation. That is a definite safety 14 problem that is going on in operating plants. If that's 15 part of the concern, then you should ascess that as part of 16 a basis. In cases where you should avoid mid-loop operation 17 -- for valves that have known to be good performers - you 18 have to balance this out against known good performers where 19 you have done tests, you've done inspections. There is a 20 basis for -- in addition to the concern of the mid-loop, 21 that there is an indication the valve is going to be all 22 right. You have to balance all these things out together.
23 So, yes, I would answer that it should be part of l
24 an assessment, certainly not the sole basis because you have 25 the other competing concern which is make sure the valve i HERITAGE REPORTING CORPORATION l
(202)628-4888
- - - _ - i
146 1 works properly. But it'is a safety concern.
2 (Q) MR. MARSH: Okay. This is from Brett, the 3 Illinois Department of Nuclear Safety. Where's Brett.
4 Brett Metro? The question is: What is the NRC's basis for
- 5. stating approvals by virtue of the Generic Letter for 6 previously submitted relief requests when such relief 7 requests could be outside the scope of the positions in the 8 Generic Letter and have not undergone NRC review?
9 (A) MR. MARSH: I think what you'are asking for, 10 Brett, is how can the NRC say that your program is approved 11 when some of the relief requests are outside of Attachment 12' 1, right? Some of them are outside that.
13 Well, as I said, we are considering where to draw 14 the line in terms of guidance to provide the positions that 15 licensees must meet. This was done with knowledge of the 26 body of relief requests that were there. We drew the line 17 in such a way that we hit what we believed to be the 11 most 18 significant technical issues out there that we could address 19 in that kind of a frame work, knowing that we weren't 20 getting all of them.
21 Then we went back and spot checked. We went back 22 and looked at some programs to see what were the ones we 23 were giving up on. And how safety-significant are they?
24 After that check, we convinced ourselves that 25 those that we left approved with the idea of explicitly I
HERITAGE REPORTING CORPORATION (202)628-4888 l
C.__-___._____-__-_-____ ._
'147 1 reviewed were not : that safety signiff. cant.
n 2 'So, on balance, you rere petting this much safety 3 significance, this much and give up that much. That's how 4 'we considered that.
5 (Q) MR. SULLIVAN: I don't understand this next 6 question, but I'll read it out and see if I can get you to 7 clarify it.
8 It says: Generic Letter position on power-9 operated valves with stroke time of greater than 10 seconds 10 is to place the valve in at an increasing frequency and if 11 stroke is greater than 25 percent of baseline stroke time.
12 Could you clarify that question?
13 (Q) If you follow the code, it's 25 percent -- I 14 want to know what the position is. You just mentioned that 15 you have to have a baseline and acceptance rate.
16 MR. SULLIVAN: I think what we were trying to do 17 in the first paragraph of position 6 was basically sort of 18 set the stage for the rest of the position by citing code 19 requirements. We weren't trying to create something new.
20 Since we are talking in position 6 about rapid acting 21 valves, we didn't talk about the whole background of code 22 testing. We were just trying to sort of set the stage.
23 So, we basically were talking about valves whose 24 stroke time is less than 10 seconds and we just really 25 weren't trying to address the whole scope of the problem.
HERITAGE REPORTING CORPORATION (202)628-4888
( f 148
'l And I think-that the second paragraph then goes on 2 to say: With' reference to valves whose stroke time is less 3 than 10 seconds, we were saying that we would be receptive 4 to an approach where you basically determine changes from a 5 reference time as opposed to a previous stroke time..
6 So, I think what you may be asking is why didn't 7 we also addrer,s that aspect for the longer stroke time 8 valves.
9 LWe just didn't think to. This position was trying i
10 to set the stage for the rapid acting valves and we weren't 11 trying to address the full scope of the problem.
12 This question came up at our meeting in Atlanta 13 and possibly if we had thcoght the whole thing through to 14 this type of question, we might have treated the-subject 15 differently but the Generic Letter is written only to 16' address -- is written in the way it is with respect to 17 valves under 10 seconds and we're not in a position through 18 this meeting to try.and give --
19 MR. MARSH: To extend the scope of the Generic 20 Letter.
21 2R. SULLIVAN: -- to extend the scope of the 22 Generic Letter.
23 Yes, sir?
l l 24 (Q) Since OM-10 uses a reference value instead of 25 previous stroke time for all valves and the statement about HERITAGE REPORTING CORPORATION (202)628-4888 l
I.
L
m 169 1 encouraging the use of OM-10, I would think that would be a.
2 problem.
3' MR. MARSH: You're right. It's just that now 1
l '4 we're not in the OM-19 frame work. We're in the IWP frame 5 work. And if you want to pick up on 10, we're in a relief 6 request kind of a frame work. Not that I want to discourage 7 the use of OM-10 because I think that's better in this 8 regard. But we're into a process' question. Okay?
9 The Generic Letter doesn't give a vehicle, Larry, 10 for updating yourself to OM-10 either. All it says is how 11 to address the Attachment 1 positions.
12 You-can tell these meetings get a little longer 13 and a little longer towards the end of the day. We're 14 holding up here.
15 (Q) MR. MARSH: The question is: Back-flow 16 testing the check valves, what is the position of the 17 Generic Letter on the word " verified by other means"?
18 MR. SULLIVAN: I think this sentence or a large
'19 part of this sentence is lifted directly out of the code.
20 (Q) Inaudible.
21 MR. MARSH: Not for back-flow testing. This is 22 forward flow testing. We're in the back-flow testing 23 position. We've talked about disassembly as a means of 24 verifying full flow not as a means of verifying back flow.
25 Disassembly is not one of those techniques that you can use HERITAGE REPORTING CORPORATION (202) 62B-4 8 80
t t
.f-
^
r.
150 L1 - to say,.'yes, Lit would-have.been leak tight had'it been
- 2 called upon to' work.
'3- You can^t'necessarily say that.
4 (Q) The back-flow and' leak tight --
5 (Q) Yes, we don't mean leak cight.
6 .(Q) 'Because if it is leak tight, it has'a leakage-7 criteria. And you measure that criteria through a media..
8 Back flow, the valve goes reversing directions.
9 MR.-MARSH: Okay. Hold.on.
10 (Discussion off the record. )
11 MR. MARSH: Jesse, we want to gi<e you the right 12 answer for this one. Okay? We don't want to confuse you.
13- So, let's take a few-minutes break and come back. We're 14 going to caucus about it.
11 5 (Whereupon, a brief recess was taken.)
16' MR. MARSH: We caucused and had some discussions 17 about what would be a suitable alternative for back-flow 18 testing for Category C check valves, whether disassembly 19' would suffice? That is whether position 2 would apply.
20 We all agreed that the current Generic Letter
- 21. doesn't give you guidance. It does not apply to those 22 conditions. So, the question is does the code allow you to 23 do it. And if the code does not allow you to do it, would 24 the NRC entertain relief requests such that disassembly 25 would satisfy that.
HERITAGE REPORTING CORPORATION (202)628-4888
l 151 1 _And we.could not definitely say one way or the 2 other amongst ourselves. We didn't reach that point where 3 we could tell you clearly which way.- So, what we will do is 4 we want to think on it some more and we will'give you a 5 position'in the meeting minutes, because we think we're~at 6 least that far along. But you won't hear it today, but 7 you'll read it in the minutes. Okay?
8 Yes, sir?
9 (Q) So the question that was asked earlier today: l 10 Does position 2 apply to reverse flow test -- I raean 11 position B apply to reverse flow test. And you said no to 12 start with.
13 MR. MARSH: I said, no, it does not. As it reads 14 right now, it doesn't. Just the heading of it, it says, no, 15 it doesn't.- So, what we are giving you is another code 16 interpretation beyond the Generic Letter is what we would be 17 giving you.
18 Position 1 and 2 were not presented, reviewed and 19 justified in light of back-flow function. Th<sy were really 20 forward-flow function. That is what their heading is.
21 (Q) Are you leaning toward yes?
22 (Laughter.)
23 MR. MARSH: We have already said in the Generic 24 Letter that for forward flow, for forward flow, checking the 25 forward flow capability of a check valve, disassembly can be HERITAGE REPORTING CORPORATION (202)628-4888
152 ii. approved, is approved if you have gone through position 1 1'
2 -and verified for yourself that you cannot' forward-flow testL 3L Lit yourself. You can then take the bonnet off and make sure 4' <that it works. And there is grouping techniques that are in 5 there for the forward-flow direction.
6 (Q) Now, I'm looking at - you remove the bonnet,
~7 ycb do forward flow by removing the bonnet?
8 MR. MARSH: Right. That satisfies the code, 9 that's true. And what the code goes on to give you some 10 words about torque. Okay? It goes on to give you words 11 about torque measurement and that type of thing. And we 12 think that sf lifting the valve and checking for its 13 operability that satisfies the torque requirements. You
~
'14 don't have to measure torque explicitly.
15 (Q) MR. SULLIVAN: There is a question here, I 16- don't have a writer, but it says: Is temporary flow 17 instrumentation, i.e., portable flow meter permitted in lieu 18 of modification to install permanent flow instrumentation?
19 If so, is relief requested -- I think it means " required".
20 MR. MARSH: Is relief request required is that 21 what you mean?
22 MR. SULLIVAN: I am kind of curious to know what 23 is behind this question, if anything. If I should just take 24 it at face value or if there is a little bit more to the 25 question?
HERITAGE REPORTING CORPORATION (202)628-4888
- ~ ~
~ ='~~~ - ~ - - - - - ' ~ - ' ~ - -~ ~ ~ ~ ~ ~ ~ - - ~ - ~ ~ - - - ~ ---
1 4 4 m4, .,
a s /: :153-il:
~(Q) Well,1there'are numerous systems that.e use yA '2 exteraal flow measurement devices.. And I have seen s'ome 3 -plants get'that-'as okay. I mean'you can use it.' But the di : code says "in". *
., 5 ' ';MR. SULLIVAN: But the, code says what?
3 6L (Q) "In" the' system.
i 7 (Q) "In the circuit."
{
8 MR. SULLIVAN: What do you mean "in"?
9 -(Q) That's a normal granted we can use external
~ 10 flow measure devices. That's all we're asking.
. 11 Di MR. SULLIVAN:. I don't know if the code was 12 written with a view in mind to exclude portable flow meters, 13 but I don't think it was written at a time when this was 14 particularly thought of.
4, 15 My' understanding is that the issue involved with 16 portable flow meters is their accuracy. And that's why I:
.17 was asking if there was a little more to that?
o 18- (Q) That is another issue.
19 (Laughter.)
20 (Q) The code requirement, the prragraph that ;
21 states that the last sentence gets you ou' of that.
It will 22 allow;you to use other devices as long as --
23 (Q) They are plus or minus 10 percent. t 4 24 (Q) Well, yes, but also you can qualify the 25 method that you're using for the application. I think the HERITAGE REPORTING CORPORATION (202)628-4888 i.
r o
- - - - - - - - - - - - - - - - ~ - -
' ^ g' .
153 1 (Q) Well, there are numerous systems that e use 2 external flow measurement devices. And I have seen some 3 plants get that as okay. I mean you can use it. But the 4 code says "in".
5 MR. SULLIVAN: But the code says what?
6 (Q) "In" the system.
7 (Q) "In the circuit."
8 MR. SULLIVAN: What do you mean "in"?
9 (Q) That's a normal granted we can use external 10 flow measure devices. That's all we're asking.
11 MR. SULLIVAN: I don't know if the code was 12 written with a view in mind to exclude portable flow meters, 13 but I don't think it was written at a time when this was 14 particularly thougli cf.
15 hy understanding is that the issue involved with 16 portable flow meters is the!.r accuracy. And that's why I 17 was asking if there was a little more to that?
18 (Q) That .8 s ar.other issue.
19 (Laughter.)
20 (Q) The code requirement, the paragraph that 21 states that the last sentence gets you out of that. It will 22 allow you to use other devices as long as --
23 (Q) They are plus or minus 10 percent.
24 (Q) Well, yes, but also you can qualify the l
25 method that you're using for the application. I think the HERITAGE REPORTING CORPORATION l (2021628-4888 l
4
_n________.___ _
'154 1- last sentence of_that code paragraph --
2 MR. SULLIVAN: We didn't think that the code 3 restricted, prevented the use of portable flow measurement 4 devices. We know a case where it's used.
. But again I 51 thought it was the flow accuracy problem.
6 (Q) Well,-I guess you were saying that if you 7 have 18 months to perform modifications of instrumentation 8 and we were just -- this is not a concern, that we can use 9 external that we don't have to modify our system.
10 MR. SULLIVAN: That's true.
11 (Q) What instruments are we worried about 12 modifying in 18 months.
13 MR. SHAW1 But there are conditions where you just 14 cannot put in any flow meter in the system and make it work.
15 Therefore, you may need a modification.
16 MR. ROCKHOLD: There is also systems, where 17 ultrasonic flow instruments won't work, but a permanently 18 installed Orpheus type of flow instrument would work, 19 concrete-lined pipes, overlying pipes, something like that 20 where ultrasonics just won't work, but you still have the 21 straight run of pipe available for an Orpheus type flow 22 instrument. So, that might be a credible alternative.
23 MR. MARSH: See, the miniflow is a low -- .
24 generally, these are mini: low kinds of lines. And those are 25 low flow rates. I think accuracy problems, cavitation HERITAGE REPORTING CORPORATION (202)628-4888 mm _ _ _ _ _ . _ _ . _ _ _ _ _ __ _ _ _ _ _ - _ _ _ _ . . _ _ _ _ _ _ _ . _
_t 155
- 1. accuracy ~ problems on that type of line.--
p 2 (Q). Miniflow? Are you talking about recire 3 ' systems?
-4. MR. tiARSH: Right.
l 5' (Q) There's tremendous cavitation. 1 6 MR. MARSH: Right. That's the reason that 7 technique may not be useful on those lines because of 8 cavitation on those lines.
9 (Q)' You may.be able to, but you should check it
- 10. first because that is really dangerous.
11 MR. SHAW: Exactly.
12 (Q) MR. SULLIVAN: The next question says:
13 HDo utilities have to contact the project managers to 14 schedule immediately a meeting to resolve any requested 15 relief request outside the Generic Letter prior to require d 1G test frequency to obtain approval and avoid violation after 17 submittal or will there be a grace period?
18 I don't understand that.
19 MR. MARSH: I do. You are saying you've got 20 relief requests that are outside Attachment 1. They are 21 being implemented now. And you want to know so that by the 22 Generic Letter they are not approved. They are not approved 23 right now, but yet they are implemented right now. And the 24 Generic Letter says that they should not be implemented 25 without approval. Should you be contacting the project HERITAGE REPORTING CORPORATION (202)628-4888
e i
156 1 manager in terms of getting them approved so that you are' 2- legal in' terms of those relief requests.
3 (Q) You're not in Table 1 and you're not in 4- Table 2.
t 5 MR. MARSHi ,Okay._ These are relief' requests that 6 are not submitted? Or these are. submitted relief requests?
7 Already submitted or not?
8 (Q) Tnere are some that are already submitted.
9 And then there is some that as a result of complying with --
10 ycu know, reviewing 89-04 you see there may be some 11 deviations that you can't always comply with the 12 Attachment 1.
13 MR. MARSH: So, you're just going to have new
- 14 relief requests.
15 (Q) Right.
16 MR. MARSH: But Attachment 1 gives you a vehicle, 17 but you can't even make those vehicles? Those four part 18 tests, you cannot do those?
19 (Q) There are one or two cases where they may not 20 be able to.
21 MR. MARSH: Okay. Well, first, it ought to be 22 clear that are already submitted. If it is submitted, if 23 the NRC has said that those are outside Attachment 1 are i
24 approved because we know generally what the characterization 25 are. If they are new relief requests, the idea we had was HERITAGE REPORTING CORPORATION (202)628-4888
1 157 1 that they would come in with the confirmation letter.
- 2 (Q) Yes.
3 MR. MARSH: With the confirmation letter, not-4 before the confirmation letter. *a. there wasn't going to 5 .be any_NRC approval prior to'that point in time.
6 We thought you needed that time to get everything 7 in order so that when the confirmation letter came, it would 8 have a set of relief requests that needed to be approved.
9 (Q) That's the grace period.
10 MR. MARSH: That's the grace period.
11 Now, you are still not in total conformance. When 12 the confirmation letter comes in, it says, "We conform with 13 -Attachment 1." .Okay? There we are.
14 But, "Here are some relief requests you need to 15 act on, NRC. Here you are."
16 And for you to be approved, we've got to act on '
17 those things. We know that. -You have to tell us the 18 immediacy of those things and when you are going to 19 implement them so that we can act on them as quickly as l
20 possible so that we can be responsive to you.
21 So, yes, you are not in compliance at the timo of
?? the confirmation letter. You needn't g3t with the project 23 manager now.
24 (Q) Do not'have to.
25 MR. MARSH: You needn't. That's right. You need HERITAGE REPORTING CORPORATION (202)628-4888
l 158 1 not de that at this point.
2 (Q) MR. SULLIVAN: This question says: Subject, 3 check valves which are never required.to fully open, i.e.,
4 thermal expansion or second breakers.
5 (0) We have already discussed that.
6 MR. MARSH: I think we did talk about that. We 7 were talking about just'to the maximum extent'that you can B quantify what the flow rate should be, quantify what the 9 flow rate through that system is. If you can do it by 10 pressurization tests and checking for the pressure decay and 11 make sure that the flow rate is what it should be, that's a 12 good quantification.
13 (Q) MR. SULLIVAN: This question is from Northern 14 States Power Monticello: Position 1 implies that the only 15 method acceptable to the NRC for full-flow stroke exercising 16 is a full-flow test. No mention is made of check valves 17 with external features which can be used for full-stroke 18 exercising.
-19 Do the six criteria presented have to be addressed 20 in the IST program to justify an external operator?
21 (A) MR. MARSH: Pat, I hope that position 1 22 doesn't imply that that is the only approved method for
- 23. forward flow testing and check valves.
24 It says -- there are words there, the last sort of 25 sentence -- that talks about other code-approved technigpes.
HERITAGE REPORTING CORPORATION (202)628-4888
j.
-159-lL . I_am right, I hope. -
-2 Pat, where are you?
3 (Q) Right here.
4 MR. MARSH: Okay. We are not trying to imply that 5 the only approved method is forward flow testing because the 6 code gives you options as well, external operators. And we 7 are not taking those options away. This is meant to be an B elaboration of full flow test which what you should do, 9 we're trying to set up a hierarchy that code requirements 10 are also there. We didn't say explicitly there are other 11 code requirements. There were some words about other 12 techniques.
13 The six criteria do not have to be addressed if 14 you use an external operator because the external operator 15 is a code requirement that meets the' code. And this Generic 16 Letter is not going beyond the code in the sense of forward 17 flow testing.
18 (Q) MR. SULLIVAN: Okay, the next question is also 19 from Monticello: Position 2 goes into the scheduling of 20 disassembly inspection in a very detailed manner. Are other 21 scheduling schemes acceptable as long as they have each i
22 valve disassembled and inspected within six years? Would I 23 approval of an alternate schedule have to be informed of an l 24 SER or acceptance of details provided in confirmation I 25 letter.
HERITAGE REPORTING CORPORATION i (202)628-4888 !
160
- 1 Then it says in parentheses: ' Existing schedule 2 for disassembly and inspection agreed upon in IST program i -3~ review' meeting with NRC, but SER never issued.
l 4 (A) MR. SULLIVAN: I am a little' reluctant to give 5 a general answer. I could think'of an example for a PWR 6 that one might refer to. But I don't know enough about 7 this. So, what I would like to ask you to do is elaborate a 8 little. '
9 (Q) Okay. When we were reviewing our second 10 10 year submittal, we came down to check valves that we 11 couldn' t full flow test. And the agreement that came out 12 was that we would do an disassembly and inspection in the 13 first refueling outage and if they all looked good, we could 14 exterd out and inspect them every five years thereafter.
15 And that's the program we're on right now for our check 16 valves that we couldn't do a full flow test.
'17 But in getting to this where you do specific group 18 of four check valves, we do one each refueling outage. We 19 are in the program where we are doing all of them every five 20 years.
21 MR. SULLIVAN: Okay. Are you doing them in 22 sequences or are you doing them all at the same time?
23 (Q) We're doing all at the same time.
24 MR. SULLIVAN: I think the position in position 2 I
25 is indicating that -- is written around the concept of j HERITAGE REPORTING CORPORATION (202)628-4888 1
_ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ . - . _ _ _ J
161 1 taking advantage of sampling techniques. It is directing 2 that you take them apart one at a time so that you are sort 3 of getting a snapshot each refueling outage as to what's 4 going on. And that is a concept that is used elsewhere in 5 testing. For example, in SNUB or functional testing, you 6 take a certain sample so that you sort of get a picture of 7 what's going on with the passage of time. And with the 8 approach that you are taking you basically lose the benefits 9 of sampling.
10 I would prefer to see that you do the -- that you 11 take advantage of that kind of concept. There are some 12 cases where, for example, in PWR where you might have to 13 offload the core to get at a group of valves, you would have 14 a real driving reason for not doing each valve every 15 refueling outage, in which case, I would say it is something 16 that we're aware of and would be possibly receptive to a 17 change in the approach provided that when you address the 18 points at the end of position 2, you address whether or not 19 the history is good enough that you feel comfortable or that 20 you can justify losing the benefits of the sampling process.
21 So, I think that within the context of position 2, 22 you can approach what you' re doing, but I think you have the 23 extra burden of justifying in terms of the benefits of 24 sampling.
25 MR. MARSH: Herb and I were just discussing it.
HERITAGE REPORTING CORPORATION (202)628-4888
_____ _ - I
162 I 1 Apparently, we were not~ enamored to the disassembly 2 frequency that you have. And it wasn't approved and.that's 3 why there wasn't an SER issued on it, because it wasn't 4 approved. 'So, please don't think that we just didn't get s 5 around'to it. I think we had difficulty with it.
6 (Q) .Okay. ,
7 (Q) MR. MARSH: Can components be deleted without i 8 prior NRC approval?
9 (A) MR. MARSH: It depends on what you mean by 10 deleted. If you mean deleted from relief requests, that is:
11 Can you now meet the code on those components? Yes. You 12 can. Absolutely. No problem.
13 If that means taking them out of the program, 14 itself, you've still got the scope issue and scope is 15 operative. Is that what you mean? You are physically 16 removing equipment from the plant and, therefore, you can't 17 . test it if it is not there anymore? Is that what you mean?
18 MR. SULLIVAN: We're not sure who wrote this one.
- 19. MR. MARSH: He's gone? Okay, end of question.
l 20 (Laughter.)
21 MR. MARSH: You've got to be careful what the l
22 question is. Scope may require relief as well. If there is l
l 23 something that should be using the code, but you believe it l
E 24 can't be used in a code for whatever justified reasons that 25 you think you have, then that may require a relief request HERITAGE REPORTING CORPORATION (202)628-4888
- _ - _ _ - _ _ _ - _ = _ _
i
'l 163 l i
1 and approval. :
2 If you are physically taking stuff out of the 3 plant and, therefore, it is just not there any more to test, 4 that is a notification' issue. That's an update issue. If 5 you are meeting the code requirements -- as I say, you don't 6 have to notify us of changes in relief requests when 7 equipment is meeting the code. That's another notification 8 issue as opposed to an approval process question.
9 (Q) I.can't think of any case, but while we're on 10 this question, I'd like to kick it around a little bit.
11 We're in the business of always adding components. But say 12 engineering comes down and just determines that a component 13 is not safety-related. That would be a cese where you would 14 want to delete it from the program. And would we be 15 required -- would approval be required then? It's 16 hypothetical.
17 MR. SULLIVAN: Let me try and address that point.
18 That's the point of thought I had in mind with whoever wrote 19- this question. I think it's your obligation to meet the 20 code. That's what 50.55A(G) says. 50. 55A (G) doesn't really l 21 require that you get an NRC approval for each and every 22 component in your plant, whether it's in or out.
23 That doesn't mean that when we come and do 24 inspections we don't questien why some components may not be 25 in the program or why a component may have been removed. I HERITAGE REPORTING CORPORATION (202)628-4888
164 1 think'the burden is on you.-
2~ But strictly speaking, the way 50.50A(G) is 3 constructed, it doesn't require NRC approval if you.
4 reevaluate.and decide that you have a good basis for taking 5 the. component out. I 6 MR. MARSH: So amend my last. That means you
- 7. don't need to get specific written-approval for~ additions or 8 subtractions from programs.
9 MR. SULLIVAN: That's not what the Reg requires.
10 MR. MARSH: Your scope has to be justified. Keep 4
11 in mind, though, as we said earlier,.we do want to make sure 12 that it's clear in the meeting minutes: Whatever program 13 you've got implemented in the plant is what we would like to 14 have as well just for QC purposes.
15 That means if you do take things out of scope, let 16 us have a copy of the most up-to-date one. We and the 17 region. So we all have the same copy so we are all on the
- i. 18 same one. Not for review purposes, but just for preparation 19 purposes.
20 (Q) MR. SHAW: This question concerns training 21 requirement. It says that UPRVs have relief from IWV 3427B.
l 22 Item 10 on Attachment 1 only discusses CIVs.
23 (A) MR. SRAM: I think in general 3427B is a 24 trending requirement. We really believe that all trending 25 requirements are significant. And the -- mission is true to HERITAGE REPORTING CORPORATION (202)628-4888 l
1
h 165 l' the requirement or perform the test in such a way that the 2 data gathered becomes meaningless.
3 I say this is because we all recognize that.
4 trending is meaningful only when it goes to a repeated 5 weight. In other words, you got the data in such a way that 6 it is repeatable. .Then trending becomes minimum.
.7 But in most cases I think our people because of 8 many reasons, the data you collected are really not 9 repeatable. And in that case requires a scheduling. If the 10 data is scheduling, then you can see that it really doesn't 11 make any sense -- this time you come in there, it's here.
12 The next time, it is perhaps one part lower. The third 13 time, same equipment may come to half a point up again.
14 Therefore, when you start to trend this kind of thing, it 15 doesn't make any sense. And we do not feel that kind of 16 requirement is meaningful. And especially for CIVs because 17 the limits, the official limits are so low and therefore we 18 do not at this moment put the requirement in for the CIVs.
19 But we really think that people should put in more 20 emphasize if they believe that their program should be used 21 together with other people, such as maintenance. Then 22 trending would really give you a very significant point that 23 you know your equipment is deteriorating or degrading and 24 therefore that you can catch it beforehand.
25 Essentially, what we are saying 3s that we still HERITAGE REPORTING CORPOR7 TION (202)628-4888
m 166 1l 'believe trending is very important and we think that:the 2 right thing to do is that when you perform a test, you want 3 to look into the testing, itself, see whether you violated 4~ the so-called repeatability problem, that you ought to use.
- 5. consistency to perform a test. And in that case, once you
-6 set a low range, anything beyond that low range becomes
.trendable and useful. And I think what-~we aro'saying there 8 is we really believe that trending is still important.
9 (Q) We.still are granted for CIVs, we identify 10 the valves.
11 MR. SHAW: Right. I think this is'because the 12 LOCA leak rate, you do have problems of determining what is
- 13. the limit. And for that reason we thing that usually we 14 come in there with very low limits. And to say that 15 trending is useful and meaningful, I think it stretch the-16 point. That's why we do not particularly emphasize that 17 part.
18 (Q) Would you also be receptive is trending is 19 not usable on stroke time for certain valves, to the extent 20 that we have data that 48-inch air-operated valves closing 21 in 4 seconds limit, stroke time is 3 and we get in these 22 action requirements, that this data is not meaningful r
23 because we would be on a multi-frequency with any deviation.
24 MR. SHAW: In general, we do not believe in 25 performing any effort which is not meaningful. And I think l
HERITAGE REPORTING CORPORATION (202)628-4888 l
l
__ _--_--__ - - - - - - - - - - - . - 1
J l
167 'l
[ 1- when.you point out that you collected data, we think a 2 certain time limit. But'yet human errors comes in there..
u i 3 You've got'a schedule way beyond your trending range. And !
4 in-that' case, I agree it is meaningless.
5 (Q) I just want to say, in his case, your motor j 6~ ; operated valve, the way.they're designed -- if-you get the 7 stroke time on it and you come down later and it~ changes and 8 .then you definitely have some type of problem. There is no 9 other way.
10 Perhaps a packing froze up on you or something.
-11 (Q) We're talking about big air-operated valves.
12 (Q) Well, like if you got a 4-second stroke and-13- you see 10 or 12 seconds --
.14 (Q) No. Our acceptance criteria is 4' seconds.
15 Ne're saying that there are deviations that could throw us 16 in an increased frequency because you only allow 2 seconds 17 and you say we have to do trending data and meet the 18 requirements. And we have a problem there.
19 MR. SHAW: I think you notice when we talk about 20 this,'this is a generic position there. And I think each 21 individual case, you ought to review that and see whether 22 human errors cuts into the margin. And I think anytime 23 you've got some data which is dubious or doubtful, you want 24~ to' identify it. Whether that was caused by - you want to 25 find the root cause.
HERITAGE REPORTING CORPORATION l (202) b28-4 888 !
1 1
_ _ _ _ _ _ _ _ _ _ _ . . . _ _ _ _ . i
l' ,
168
- L -(Q) Right.
2 ~MR. SHAW: And if the root cause shows that.this 3 is because of a human error or other things, and you 4 convince yourself that you are in the scheduling regs, then 5 I think that you have a good reason for not trending.
6 .Otherwise, trending is still very important.
7 (Q) Thank you.
8 (Q) MR. SULLIVAN: The next question says: For a 9 multi-unit site if one unit has an approved SER which grants.
10 relief on items which do not meet all the criteria of the 11 ' Generic Letter, can the approved SER provide a basis for the 12 other unit to go ahead and implement the relief request -
13 prior to NRC re-review. Note: Design differences do not 14 exist between the two units.
15 (A) MR. SULLIVAN: I think if we take the. case of 16 a single unit site, I've already talked about the fact that 17 we don't think that relying on a relief request tht.t differs 18 with the Generic Letter is necessarily the right way to go.
19 We are sort of in a posture of saying that it may 20 be that legally -- from a legal point of view, we may not 21 have --
22 MR. MARSH: Go ahead and say it.
23 MR. SULLIVAN: We may not be able to get you to 24 change things because we could be into a backfit situation 25 and.we would have to evaluate how seriously we want to HERITAGE REPORTING CORPORATION (202)628-4888 )
l
)
l
169
'11 -, consider the issue from the point of view of pouring 2 resources into fighting through a backfit. But as I 3 indicated earlier,. we would definitely in cases that we're-4 aware of on a singe unit site try to convince you to 5 withdraw the relief request and file the Generic Letter.
6 For a multi-unit site where there isn't a legal 7 situation involved, I would say definitely you can't use an 8 approved relief request for one site and apply it to the q 9 other where the relief request is different from the Generic 10 Letter.
11 And I think in general from - going back to a 12- legal perspective -- I don't think that regardless of 13 whether or not we would approve it based on the precedents 14 'weLhave set, I don't think that you can apply a relief 15 request from one unit to another. The way 50.55A(G) works, 16 as I understand it, each relief request has *; ue addressed 17 on a unit-specific basis.
18 (Discussion held off the record.)
19 MR. SULLIVAN: Okay. Herb is noting one other 20 point. If there is a relief request submitted on a -i 21 multi-unit site and it's identified as being a relief 22 request applicable to both units, then we can write an SER j i
23 And it may all be rolled that grants relief for both units.
24 up in one.
25 But if there is a relief request that specifically HERITAGE REPORTING CORPORATION (202)628-4888 1
, 170 1 is written for one unit and-specifically approved.for
.21 'another unit -- I mean it is specifically approved'for that 3 unit, then it is only approved for that unit. It is not
- 4. ' approved for both units, even if'there are no design.
5 differences.
Y 6 (Q) MR. SULLIVAN: The next question says: If I-7- understand correctly,.a relief request submitted prior to 8 . April 3, 1989 but not discussed on any SER and is.not a 9 subject of the Generic Letter Attachment 1 is approved for 10 use without'any further utility reviews. I understand this 11 to be true.
12- (A) MR. MARSH: A relief request submitted prior 13 to April 3rd,'1989, but not discussed in any SER and is not 14 'the subject of Generic Letter Attachment 1 is approved for 15 use without any further utility review.
16 I assume that you are a non-Table 1 or Table 2 17 plant, first. That means that you -- then you say SER, 18 which implies there is an SER. Is that what you mean?
19 (Q) On that one there's an SER, review of the 20- first two revisions of our IST program. There is a 21 subsequent submittal on the IST program, two additional 22 revisions that went in before the Generic Letter was issued.
-23 MR. MARSH: That resets you in terms of the SER 24 and what it said. In other words, you are a non-Table 1 or 25 Table 2 plant, you're staring from zero, okay.
HERITAGE REPORTING CORPORATION (202)628-4888 u____________ _ - 1
L 171 1 (Q) That's correct.
2 MR. MARSH: In that sense, you are correct.
3 Relief requests which were submitted and were on the docket 4 at the time of the Generic Letter are approved as they were 5 submitted and all you need to do is to address -- we need to 6 address the Attachment 1 issues in your confirmation letter.
7 You need to modify your program to address the Attachment 1 8 issues. You need to address your procedures in all respect, 9 inside Attachment 1/outside Attachment 1. That's a key to 10 make sure that things are being tested individually, for 11 example.
12 So, yes, there is further utility review with 13 respect to outside Attachment 1 issues and that's for 14 procedures, making sure your procedures are adequate.
15 All this -- what I have just said is true as long 16 as what you have at the plant right now is the same as what 17 is submitted and on the docket. If you have gone beyond 18 that, then we are in a different frame work. Okay?
19 (Q) MR. SULLIVAN: The last question is kind of a 20 tough one. We have gotten it before. It says: What 21 methods are considered acceptable for establishing a i
22 limiting value for full stroke time for power-operated 23 valves as given in position No. 5 of Generic Letter 89-947 24 (A) MR. SULLIVAN: I don't think that we have l
25 criteria that we can give you. We are not in a position to l l
HERITAGE REPORTING CORPORATION I (202)628-4888
172 1 give you any sort of quantitative guidelines. We have tried 2 to set down a philosophy. We think that it requires a 3 pretty considerable knowledge on your part about different 4 valves, how they behave, where they start to be considered 5 questionable in terms of their performance. And I think that 6 it is something that you basically have to come up with on 7 your own.
B When we identified this issue, we basically 9 identified it because we would come out and do either 10 inspections or conduct review meetings and see that the 11 limiting values of full stroke times were corresponding to 12 what was in the tech spec. And there may have been a valve 13 that would normally stroke in a few seconds and the limiting 14 value was coming out of the system requirements maybe it's 15 at 60 seconds. So, that's how we identified the issue.
16 We are not in a position on the other hand to try 17 to prescribe what we think the values should be. It is a 18 fairly big job.
19 MR. MARSH: It's too component-specific, too 20 application-specific. It would be wrong for us to specify 21 in your particular plant or your particular valve with its 1
1 22 own history what the proper limiting value should be. It i 23 really depends upon your own individual plant-specific 24 information.
25 We have been asked to give a multiple. What is a j HERITAGE REPORTING CORPORATION (202)628-4888
173 1 proper multiple of normal stroke time to be used? We can't 2 even do that because that again depends upon a variation in 3 the value of normal stroke for the valve..
4 We have been asked whether licensees should use a 5 statistical analysis and use three standard deviations.
'6 We're'not even sure that is the right way to go. It depends' 7 upon whether the valve, whether'there's been maintenance on 8 a valve recently which would throw that statistical analysis 9 off.
10 These are the kinds of questions you need to ask
~
11 yourself. If there hasn't been maintenance on the valve and 12 if it's been a good performer, it's been pretty steady, that 13 there is a good mean value, perhaps statistics is okay for 14 this particular valve for this particular application. But 15 if it has got a widely varying value, then maybe you want to 16 be very careful about how you set it to pick.up -- not 17 declare it inoperable incorrectly, but to pick up variations 18 that are outside the norm.
19 Yes, sir?
20 (Q) Would you agree that that limiting value 21 would be some above your alert limit?
22 MR. MARSH: The alert limit?
23 (Q) Yes. 1.25 to 1.57 24 MR. MARSH: Probably.
25 MR. ROCKHOLD: As long as it did not go beyond any 1
HERITAGE REPORTING CORPORATION (202)628-4888 l'
m . ,.. -
174 .)
1, l 1 system or tech spec identified' criteria, yes.
1 2 (Q) You don't have a problem with the.elert 1
L 3 limits in the code of.25 percent or 50 percent?
l .
4 MR. ROCKHOLD: Except for ACMOVs.
l-h 5- MR. MARSH: Right. At that point, you have 6 already burned the motor out. And you've got a problem with 7 -the valve at that point.
8 (Q). So, is that what you're saying not for those 9 valves? You don't have to do the code requirements if we
'10- set lower limits than that 25 or 50 percent.
11 Do you know what I'm saying? The code requires 12 .you to make a comparison.
13 MR. MARSH: No. We're not saying you don't have 14 to do those tests for those valves or you don't have to do 15 those comparisons.
16 MR. SULLIVAN: Are you saying, for example, if you 17 used -- are you saying, for example, if you had an ACMOD and 18 you set your limiting value full stroke time to be 110 19 percent of the normal time, I think the alert time becomes 20 moot. You would always declare the valve inoperable before 21 you ever hit alert.
22 (Q) But what I'm saying, even trending in that 23 point is probably meaningless.
24 MR. SHAW: Again, you are talking about things 25 very component-dependable. You are talking about a HERITAGE REPORTING CORPORATION (202)628-4888
175 1 different kind of valves. 'You are talking about the AC 2 driven valve which would be altogether different from a DC 3 driven or air powered operating. If it's air or hydraulic 4 cctuator valve.
5 I don't think we want to give you a generic 6 guidance number-wise.
7 (Q) I'm just saying, if you said 110 percent and 8 the code gives you an alert limit of 125 percent, then you 9 still have to --
10 MR. MARSH: Follow the 125 percent /150 percent.
11 (Q) Do we need a relief request again? To me it 12 doesn't even make sense to trend some of those valves in 13 that case.
14 MR. SHAW: You are right in saying that certain 15 types of applications, trending does not make too much 16 sense, yes, we agree to that.
17 But essentially, I think again we are talking 18 about we don't want to make a generic statement here. You 19 are saying that because of AC driven motor, AC motor driven 20 valves, the variation is so small, trending makes very 21 little sense.
22 Now, that cannot be exter 'ed to cover other 23 things. And I think again we must be -- we must look at the j 24 problem. And if you do have a problem by itself then maybe 25 the thing that we would -- we certainly want to look at HERITAGE REPORTING CORPORATION (202)628-4888
176 1 that.
2 And, unfortunately, the code gives you a very 3 generic percentage, say, 10-25 percent. In some cases, that 4 10 or 25 percent is not adequate. We are saying that we 5 don't want -- I think essentially if you have equipment, you 6 should know that equipment much better than we do. We are 7 trying to give you generic guidance. We say that, "Use 8 common sense." And, thererore, you want to do certain 9 thir.gs to help yourse3 f.
10 Dut if we are pinned down to say that would 5 l 11 percent be all right? Would 3 percent be all right? It is 12 difficult when we don't know all the details. I think you 13 can appreciate that situation there.
14 MR. MARSH: We tried. I'm not sure we gave you a 15 clear-cut yes or no. <
16 (Q) I think I know where you are on the trending.
17 You are saying if - you're assuming that you're trending to 18 determine whether the valve goes on alert to increase the i 19 frequency. I 20 (Q) We're trending to go along with the code. j 21 (Q) There's another side to that. If you lower 22 the alert limit and the action required on the valve, your 23 trending becomes even more important because now instead of l 24 increasing the frequency, if the trend goes up, it goes up a 25 little bit, you're going to inop the valve. But you still HER2TAGE REPORTING CORPORATION (202)628-4888 l
,,.-au.-a
177 l' have.to' trend to determi'ne that degradation. The point is 2 it is just going to come sooner.
l 3 (Q). Okay. So then you come up with a reference L 4 value. 'Okay? .And you can say, "This is my reference value l 5 and I'm going to set a' fixed alert." Okay? A fixed alert 6 now, instead of just doing what the code says I shou)3 look 7- at 25 percent of last stroke. That's what the code says.
8 What I am saying is if you just set a flat stroke-9 time, okay? It's a fixed value.
10 MR. MARSH: I know you're trying. We're trying, il too, here.
12 MR. SULLIVAN: What is the trent.ing requirement 13- 'that you are specifically referring to?
14 1G1. SHAW: Yes, give us example, now. I think we 15 are talking about something that I don't have any feeling on
'16 at this moment.
17 (Q) I am just saying that if you are reducing a 18 limit to below what tre code is telling you for alert 19 values.
20 MR. SHAW: Again, you are talking about 21 components. 1 22 (Q) Here it is. I've got it.
23 The code tells you to compare your stroke time.
24 You stroke it today, you compare it with the last time it 25 stroked.
HERITAGE REPORTING CORPORATION (202)628-4888 l I
(:
178 1 What I am saying is, why can't -- if you determine 2 an average value or reference stroke for that valve, okay, 3 you set your limit on that. Okay? And not the last stroke.
4 You see what I'm saying? To me it deviates from the code a 5 little bit. And I am just wondering if you need a relief 6 request or anything that says that because if you are 7 comparing it to a reference stroke time versus the J 2st time 8 it stroked, it could be different than what the code is 9 saying. Because it always tells me to compare it with the 10 last value.
11 MR. SHAW: You're just taking the trending benefit 12 away. In that case, I'm not saying that you are doing 13 things wrong, but then there is no guarantee -- well, you 14 can set a limit value there. And this time it becomes a 15 go go-go situation. And it could be very close to that and 16 you still believe it is functionable. And there is no 17 guarantee that between your tests the thing would go wrong.
18 It all depends on how limits were assessed. It is 19 very difficult for me to say generically we agree to it or 20 not.
21 (Q) Well, can you set a reference stroke time 22 based on your past history? When you have five years of 23 data and you pick a number, okay? Now, you say, "That's my 24 reference value and I will put in a procedure of fixed alert 25 limit." Okay?
HERITAGE REPORTING CORPORATION (202)628-4888
179 1 MR. SHAW: Fine.
2 (Q) The code tells you to compare it with the 3 last limit to determine if you went above that 25 percent.
4 MR. SULLIVAN: Let me just try. If it is outside 5 of what we wrote in I think it's one of the positions in the 6 Generic Letter. We got into that a little bit saying use of 7 reference times in a certain example was acceptable.
8 (Q) Okay.
9 MR. SULLIVAN: But it is outside of that. It's in 10 position 6, the second paragraph. We talk about reference 11 time very briefly for valves that were stroking in less than 12 10 seconds. We talked about that being an acceptable 13 approach.
14 We didn't address valves with longer stroke times 15 than that. In eit'.ter case, I think you need to read the 16 Generic Letter. !.f you can use it to justify an 17 alternative, then do that. Otherwise, I think you have to 18 follow 50.55A(G) and submit a relief request.
19 We can talk about what we would be receptive to 20 and I think we really need details to give you any better l
21 answer.
22 (Q) I don't have any problem with trending.
l 23 Maybe I've got a little confusion there because I asked a 1
24 couple of questions in that area.
l
! 25 I'm just talking about when the code tells you to i
HERITAGE REPORTING CORPORATION (202)628-4888
180
, 1 compare your value stroke with your last stroke. Okay?
^ 2 And you're saying that if you compare it with the 3 reference stroke -- it doesn't say that?
4 MR. SULLIVAN: No, that's not really true.
5 MR. ROCKHOLD: It says that, but we don't see that 6 as being trending.
7 (Q) Okay. Let's get trending out of the picture.
8 I am just looking at that requirement.
9 MR. ROCKHOLD: Okay.
10 (Q) Now, if I compare my actual stroke time today 11 with the reference value where I have a fixed alert range 12 because I based that on the history.
13 MR. SHAW: That's fine.
14 (Q) I'm not violating the code?
15 (Q) You are doing something better than the code.
16 The code allows you to do that.
17 (Q) You're saying that your position here allows 18 me to do that.
19 MR. MARSH: In fact, that's what 10 does. Doesn't 20 it, Larry? 10 does that explicitly.
21 MR. SULLIVAN: I think basically you need to see 22 whether it is addressed in position 6. If it is not 23 addressed in there, in your particular case, then I think 24 you need to submit a relief request, even though it is 25 better than the code, it is not the same as what's in the HERITAGE REPORTING CORPORATION (202)628-4888 i
m_______. ___ ._
v,'
181 1 code. It is different from the code. It is not addressed 2 in the Generic Letter. I am afraid you are going to have to 3 submit a relief request.
4 (Q) Well, could we get it in the minutes a
.5 position that you fe.el?
6 MR. SULLIVAN: We talked about that before and we 7 decided that the minutes really aren't th,e right vehicle for 8 generically extending the letter. We would need to send the 9 minutes back through CRGR.
10 (Q) Well, how about a clarification of that?
11 MR. MARSH: We can clarify code requirements.
12 MR. SULLIVAN: We can clarify what we just said.
13 MR. MARSH: We can clarify the Generic Letter, 14 clarify what you need to do in order to see whether you are 15 within the Generic Letter and what you need to do. But not 16 to state new positions.
17 Okay, any more questions: Okay, that's it.
18 (Laughter.)
19 MR. MARSH: Ceriously, any more questions?
20 (No response.)
21 MR. MARSH: Thanks very much, gents. It's been a 22 useful meeting for us and we hope it has been for you, too.
l 23 (Whereupon, at 4:35 p.m., the hearing was l 24 adjourned.)
25 HERITAGE REPORTING CORPORATION (202)628-4888 l
p;,
m -- ,
~
[>.' -i ,
is
<, 9 :1 CERTIFICATE
.o
~
3 'This is to; certify.that the~ attached ~ proceedings before the 4 United.. States Nuclear. Regulatory Commission in the-matter of .
5 ,
Name:: Inservice 'asting, Generic Ietter 89-04
[y 6:
7' Docket Number: 342-710 8 Place: Chicago, Illinoic 9 Date: June-13, 1989 10 were held as:herein. appear., and'that this.is the original p[-_
11 transcript thereof for.the file of the United States Nuclear ri 12' Regulatory Commission taken stenographically by me and, 13 thereafter reduced to typewrit.ing by me or under the-direction
- 14 of the. court reporting company, and that the transcript is'a-
- 15 true.and accurate re 'rQ) I or oing proceedings.
16- /s4._
~
/Ard s 17' (Signature' typed):
r )
Timo M pyne 18 Official Rep er 19 Heritage Reporting Corporation
- . 20'
.21 22 23 L- 24 25 Heritage Reporting Corporation (202) 628-4888 11.u.a. _ .. . . _ - _ .- - -