ML20236X501
| ML20236X501 | |
| Person / Time | |
|---|---|
| Site: | Sequoyah  |
| Issue date: | 11/24/1987 |
| From: | NRC COMMISSION (OCM) |
| To: | |
| Shared Package | |
| ML20236X475 | List: |
| References | |
| NUDOCS 8712090247 | |
| Download: ML20236X501 (180) | |
Text
{{#Wiki_filter:Ti} Q %j$$@f3y79" i El~'~ ~'i #{"~ ~ ~~ " S EE'9'$ ENCLOSURE 1' ~ ~ L 1 UMIBD STATES ~ ~ NTJCLEAR REGULATORY COMMISSION IN THE MATTER OF ) DOCKET NO: 50-327 I ) 50-328 ' DISCUSSION ON SILICONE RUBBER ) ) l INSULATED CABLE AT SEQUOYAH ) ? a ! ^ LOCATIONt Bethesda, Maryland PAGES: 1 through 122 DATE: November 24,. 1987 Heritage Reporting Corporation 8712090247 871202 M A W W2 ) PDR ADOCK 05000327 122S L Streer. N.W. T PDR Wasaanstoa. D.C. 2000$ (202) 623 4488 1 l I
.. ~.....
- n,UIgo,'
c : c pa UINITED STATES NUCLEAR REGULATORY COMMISSION ~ IN THE MATTER OF ) DOCKET NO: 50-327 ) 50-328 DISCUSSION ON SILICONE RUBBER ) ) INSULATED CABLE AT SEQUOYAH ) LOCATION: Bethesda, Maryland PAGES: 1 through 122 DATE: November 24, 1987 ......................................e Heritage Reporting Corporation O(11cial Reportert 1:20 L Street. N.W. Wasaangton. D.C. 20005 ( 02) 628 4888
~ ~ ..o i UNITED-STATES NUCLEAR REGULATORY COMMISSION i 1 2 3 In the Matter of: ) DISCUSSION ON SILICONE RUBBER ) q -41 ) Docket No: 50-327 5. ) 50-328 INSUIlATED CABLE AT SEQUOYAH ) 6- ) y: 8 Tuesday 9 -November 24, 1987 -10 Nuclear Regulatory Commission Room.'550
- 11 4350 East-West Highway Bethesda, Maryland 12 The above-entitled matter came on for hearing, 13 pursuant to notice, at 14 1
- 00 p.m.
15 APPEARANCES: 16 On behalf of the Nuclear Regulatory Commission: 17 GARY.ZECH, Assistant Director for Projects I Office of Special Projects 18-SMUL Projects Division 19 B.D. LIAW, Assistant Director of Technical Programs TVA Projects Division 20 EILEEN MCKENNA, Office of Special ProjectsProjects Manager for Sequoyah 21 22 JANE AXELRAD, Deputy Director Office of Special Projects c 23-24' (Continued on next page.) 25 Heritage Reporting Corporation (202) 628-4888
j 2 1 APPEARANCES: (Continued). 2 On behalf of the Nuclear Regulatory Commissions 3 STEWART'EBNETER, Acting Director Office of Special. Projects 4 ' STEVEN' RICHARDSON, Acting Director-5 TVA Projects-Division. L 6 7 On behalf of the Tennessee Valley Authority: 8 CHARLES FOX, Deputy Manager Office of Nuclear Power 9 RICHARD GRIDLEY,. Director 10 Nuclear Licensing 11 ROBERT CANTRELL. Deputy Director Division of Nuclear Engineering -12 MARK BURZYNSKI, Suquoyah Licensing Supervisor 13 KENT BROWN, Senior Electrical Engineer 14 Division of Nuclear Engineering 15 ANGELO MARINOS, Chief, Reactor Operations Branch TVA Projects Division 16 Office of Special Projects 17-18 19 j ] 20 21-J 22 23 24 25 Heritage Reporting Corporation 1 (202) 628-4888 l i I a
f-7 ) i 'O l f. 3 1 PROCEEDINGS 2 MR. ZECH: I guess.that we are ready.to get started' 3 now. I am Gary Zech, the Assistar.t Director. for Projects in 4 the-TVA Projects Division of the Office of Special' Projects. 5 The purpose of this meeting is to provide the opportunity for 6 TVA to present some additional information on the silicone 7 rubber insulated cables at Sequoyah. 8 This meeting is being transcribed. For those who would be speaking for TVA or for the NRC, I would ask that you 10-speak'into a mike, state your name and spell your name.as well 11 if you would, so that the transcriber can get the correct 12' 3pelling of your name. Also there is an attendance' sheet being 13 passed around, and I would ask that you sign that with your. '14 name and organization as well. 15 Before we get started, why do we not-go ahead at 16 least around the. table, and whoever else you would like to 17 introduce, Charlie, and we will go from there. Gary Zech, and 18 do you want to go ahead. 19 MR. LIAW: I am B.D. Liaw, Assistant Director of 20 Technical Programs at the Division of TVA Projects. 21 MR. MARINAS: I am Angelo Marinos, Chief of the 22 Reactor Operations Branch of the TVA Projects Division. 23 MR. RICHARDSON: I am Steve Richardson, the Acting E .24 Director of the TVA Projects Division. 25 MS. AXELRAD: I am Jane Axelrad, Deputy Director, t Heritage Reporting Corporation (202) 628-4888
. r. ) 4 1 Office of_Special Projects. 2 MR. EBNETER: I am Stewart Ebneter, Acting Director 3 of OSP.- '4 MR. CANTRELL:. I am Bob Cantrell, Deputy Director of. 5 the Division of Nuclear Engineering, TVA. 6 MR. FOX: Charlie Fox,. Deputy Manager, Office of 7 Nuclear Power,'TVA. 8 MR. GRIDLEY: Dick Gridley, Director of Division of 9 Nuclear' Licensing, TVA. 10 MR. BURZYUSKI: Mark Burzyuski, Sequoyah Licensing. 11 MR. BROWN: Kent Brown, Division of Nuclear 12 Engineering, TVA. 13 MS. MCKENNA: Eileen McKenna, NRC, Project' Manager 14 for Sequoyah, Special Projects. 15 MR..MAVRO. O.J. Mavro, and'I am an advisor to 16 Charlie Fox. i 17 MR. RAUGHLEY: Bill Raughley, Chief Electrical 118 Engineer, TVA. '19 MR. SHEA:. Tim Shea representing TVA's Electrical 20-Engineering Branch. 21 MR. GILL: Paul Gill, NRR, Siectrical Systems Branch. I 22 MR. GUILLEN: Jaime Guillen, NRC,'NRR 23 Communications Branch. 24 MR. GOODWIN: Ed Goodwin, Office of Special' Projects, 25 Technical Systems. l .i U Heritage Reporting Corporation 1 (202) 628-4888 l i u 1 L
V 5 1 MR. Z'sCd t Okay, thank you very much. 2 Stu, do you h' ave anything before we get started? 3 MR. EBNETER: No. 4 MR. ZECH: Charlie, do you want to go ahead? 5 MR. FOX: Yes? TVA has completed an' extensive test 6 program at Sequoyah that has addressed the original TER 7 concerns and the evidence of damage to cables of pull-bys, 8 jamming, and unsupported vertical cable. Now these original 9 concerns have been successfully dispositioned. No anomalies or 10 evidence of cable insulation damage has been found. And this 11 program shows that TVA's cable insulation practices at Sequoyah 12 were acceptable.. As we speak, I just found out from Angelo 13 that we have formally docketed that program, and it covers the 14 original TER concerns. 15 As.a pirt of our unsupported vertical cable testing 16 at Sequoyah, we did notice anomalies initially in AIW cable. 17 This is a silicone rubber cable. 18 MR. LIAW: Excuse me, Charlie. 19 MR. FOX: Sure. 20 MR. LIAW: I am sorry to interrupt. I think that I 21 would like the record to show that the TER concerns are not 22 limited to those three concerns. 23 MR. FOX: I said the original TER concerns. 24 MR. LIAW: What is the difference between the t l 25 original? I Heritage Reporting Corporation (202) 628-4888 I l
6 1 MR. FOX: Well, the' original concerns were damage due 2 to three' factors, B.D., pull-bys, jamming, and unsupported 3 vertical cable. 4 MR. LIAW: Not side wall pressure? .5 MR. FOX: That was. net a part of the. Original,TER. 6 ..MR.. BROWN: This is Kent Brown. I.think that the 7. concern.there that B.D. is trying to express is'that with the 8 TER that certain things were identified as restart concerns. 9 MR. FOX: Yes, excuse me. I am speaking strictly ta) 10 the restart considerations in the TER. Excuse me, I stand' 11 corrected. 12 MR. LIAW: I was stating something that was a concern 13 . identified in the TER. 14 MR. FOX: Yes, absolutely. .15 MR. LIAW: They were identified to be pre-restart 16 . items.- 17 MR. FOX: Well, right now our plate is full, and we .18 are trying to focus on those items that we are facing to get 19 our plant back on the line. And those tend to be the things 20 that we focus on the must there. 21 As I mentioned, we did find that we could' easily 22 damage silicone rubber cables, be it Anaconda, Rockbestos, or 23 AIW cable. We demonstrated this in a laboratory test program. 24 And'this susceptibility that we discovered particularly to 25' minor impact damage led to us submitting a potential Part 21. j Heritage Reporting Corporation (202) 628-4888 i ) b 3' j
j 7 1 1 Today we are going to go through and emphasize what 2 we have done on silicone rubber cables. And you are going to 3 hear a good bit of new information that has not been out on the-4 table previously. 5 Specifically, you are aware that we did post-mortems 6 on the four silicone rubber cablos, AIW silicone rubber cables, 7 that failed as a part of the unsupported vertical cable testing 8 inside containment at Sequoyah. 9 We have also now done post-mortems on the five 10 additional silicone rubber failures that we have seen. And we 11 have shown that the test program that we will spend a good deal 12 of time on today at Wyle has enveloped, more than enveloped, 13 the worst case that we have-seen in the plant. 14 So what I am telling you that is new today is that we 15 have gone ahead and done post-mortems on the failed cables, 16 those which fail the voltage testing at Sequoyah. And we will 17 go through the details of that in just a short period of time. 18 We use that minimum wall thickness on the Yukon 19 reports to set up a special test program at Wyle Labs. We are 20 here to present that program to you today, and we are also 21 going to docket that program today. We have that test report 22 with us, and a memo from our Director of Licensing, Dich 23 Gridley, to provide to you. So we will officially be putting 24 that on the docket today. 25 The Wyle test program is of particular significance Heritage Reporting Corporation (202) 628-4888
l 8 1 to TVA, because it demonstrates that silicone rubber cable with 2 substantially reduced wall thickness will in fact pass a LOCA 3 and continue to function properly. 4 TVA took this information from Wyle, the preliminary 5 information and preliminary report, to the ICC meeting of the 6 IEEE November 9th through lith. Mr. Marinos, your staff was 7 present at that meeting. We presented the program and results 8 to the cable manufacturers, to representatives of the utility 9 industry, and to a number of industry experts. 10 In fact, I would like to introduce two people who are 11 here today. They have mentioned their names as we went around 12 the room. One is Ted Balaska, the present Chairman of the IEEE 13 ICC Committee, who is with us today. His term expires the end 14 of Dec, ember. And Reinhold Luther is the incoming Chairman, and i 15 he is also with us today. 16 We also have Bob Gehm and Mike Mennone here of 17 Rockbestos. I think that Dr. Ling has not shown up yet. Dr. 18 Ling is with Cabelink and is a'former Anaconda manager. He is 19 not here yet, I presume. 20 We bounced this off the committee, and there were a 21 number of items of good feedback. For example, we got an eye l 22 opener on the polarization index. Larry Ke'lly of Olcenite 23 pointed out to us that maybe our results were indeterminate on 24 the polarization index, because the rate of voltage input into 25 the cables, and also the length of the cables, and a number of Heritage Reporting Corporation (202) 628-4888 i
9 1 other parameters need to be tak'en~into account.before you draw 2 any conclusions about polarization indexes, be they less.than 3 one, greater than one or whatever. 4 We also'since that time have become aware of an ASTN 5 document. It is ASTN D-149-81 and D-3755. This document deals 6 with dielectric strengths for cabling solution. In fact, there 7 are a number of things in this standard that I think can go'a 8-long way toward explaining some of the voltage anomalies.that ( 9 we saw both in Sequoyah'and at.Wyle, both pre and post-testing, j .10 Specifically, ASTN D-149 shows that dielectric l 11 strength is inversely proportional to the thickness. It is .12 also very sensitive to.the rate input of the voltara. It is -13 also sensitive to wet versus dry. We will discuss those in 14 some detail with you today. That goes a long way towards 15 explaining some of the.other anomalies. ] -16 MR. ZECH: Charlie, what are the dates of these 17 standards, when were they issued? 1 18 MR. FOX: 1981. We will give you a copy of both the 19 ASTN standards. We have those with us today. They are 1981 20 standards. 21 MR. EBNETER: Do they talk about bulk characteristics 22 or cable characteristics? 1 23 MR. FOX: They talk about cable characteristics. We 24 will provide the staff a copy of these today. 25 In addition, we feel that our cables based on the Heritage Reporting Corporation (202) 628-4888 I l i
10 1 information that we have at hand now are good. We believe that 2 the insulation practices at Sequoyah were okay as evidenced by 3 the reeults by pull-bys, jamming, and unsupported vertical 4 cable. 5 MR. MARINOS: Are you certain that these cable 6 characteristics are not bulk characteristics? 7 MR. FOX: I am told that it is cable characteristics. 8 We will get into that when Kent Brown does his presentation 9 just a little later. 10 MR. BROWN: If you like, I can clarify that real 11 quickly. 12 MR. MARINOS: If you can clarify that. 13 MR. BROWN: It is a material standard being an ASTN 14 standard. However, they do talk about that the testing was a 15 variety of different electrodes and a variety of.different 16 specimen configurations, if you would, so that the subject is 17 really a dielectric testing phenomenon. 18 MR. MARINOS: But not even a form of a cable. 1 4 19 MR. BROWN: I would say that the phenomena that are 20 under discussion would be sensitive sometimes to the j 21 orientation such as the wet versus dry which we will discuss a j 22 little later. But they would be directly a'pplicable to the 23 kind of testing that we are talking about. 24 MR. FOX: We can have Ted Laskos speak to that a 25 little later in the presentation, if you would like, on the l l Heritage Reporting Corporation (202) 628-4888 i l I
11 1 applicability. 2 Specifically, as I said, we have shown that our basic 3 cable insulation practices have been shown to be adequate as it 4 relates to the restart issues that were identified in the TER. 5 We have found some silicone rubber failures at high voltage. 6 We submitted a potential Part 21 due to ease of damage. 7 Based on the information that we have today and 8 primarily the Wyle results, we feel like that Part 21 is t 9 probably no longer valid. We have also done post-mortems on 10 the silicone rubber cables that have failed. We have shown 11 that we had a minimum of 8 mils of wall thickness on all of 12 those cables. That was the minimum that we saw. We have done i 13 tests at Wyle that were substantially below that and a wall 14 thickness that passed the LOCA. 15 We feel like that we can now begin to e.?olain some of 16 the voltage anomalies that we have seen based on some of the 17 more recent standards that we have been made aware of. And we l 18 are here today to formally present to you the information which 19 we feel like is necessary to close this iscoe. 20 MR. ZECK: Charlie, you said sonething about the Part l 21 21, are you going to cover why you think chat is as well? 22 MR. FOX: Yes. In'other words, at the time that we 23 put the Part 21 in is when we saw that you could significantly 24 damage the silicon rubber cables with minor impact. We have 25 Heritage Reporting Corporation (202) 628-4888
12 1 since bounded. And in the failures that we have seen, we have 2 had a minimum of 8 mils of wall thickness. In fact, I guess 3 that the first population of cables that we sent up to Yukon 4 had 8, 10, 12, and 20, 5 (Continued on next page.) 6 7 8 9 i 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 I Heritage Reporting Corporation (202) 628-4888
1 13 l 4 1 MR. FOX: And then this last set of post-mortems 2 'that we.did at TVA showed 21 mils minimum remaining wall 3 thickness. So we tested cables at Wyle, specifically 4 Rockbestos at 6 mils in both Anaconda and AIW, got down to 4 5 mils, and put those cables through a full-blown LOCA and they 6 not only passed the LOCA, they continued to operate 7 successfully well after the LOCA. 8 MR. ZECH: But those cables were not from installed 9 systems. They were off a cable reel, is that correct? 10 MR. FOX: Yes, that's correct 11 MR. ZECH: Okay. 12 MR. FOX: We took and artificially reduce the wall 13 thickness down to a level which was conservative relative to 14 the minimum we saw in the plant. That's correct. 15 MR. MARINOS: But a post-mortem of the last five that 16 you did was -- 17 MR. FOX: Those were from the plant. 18 MR. MARINOS: -- was'that failed. 19 MR. FOX: Yes. 20 MR. MARINOS: And you have determined that the ones 21 that filed at 7,000, whatever the voltage was, did have 21 mils 22 of insulation on them. 23 MR. FOX: That's correct. 24 MR. ZECH: Let's go through the program. Thia is 25 going to be covered? i Heritage Reporting Corpor6 lon (202) 628-4888 j
14 1 MR. FOX: Yes, I'm giving you a summary of what--- 2 MR. ZECH: Letts go through that. 3 MR. FOX: There you are. The speaker is Kent Brown. 4 from the electrical engineering branch 13f D&E. 5 MR. BROWN: Okay, good afternoon. 6 As Mr. Fox said,.my name.is Kent Brown. I'm with the 7 Division of Nuclear Engineering at the Tennessee valley [ 8 Authority. I am in the electrical engineering branch. 9 .(Slide) 10 Before I get into this, I want to correct one thing-11 that I heard out in the hall. A comment was made as I was 12 coming in that I was going to come and present a Thanksgiving 13 turkey. ILdon't think that's characteristic at all. I think' 14 what'we're coming to do is to give you a good program, to show' 15 you what you've done. Maybe I can rephrase it and say that-16 maybe it's a holiday goose. l'7 MR.' FOX: Let me say if we have to eat a little crow 18 today, it will make the turkey taste better Thursday. 19 MR. EBNETER: Well, I hope the thing gets a little 20 more serious as we go along. 21 MR. BROWN: Certainly. 22 I do want to present to you the results of the effort 23 -that we have undertaken at TVA, the work that we have done. As 24 Mr. Fox said,- it involves both a lot of effort on our part in j 25 insitu tests, effort on our part at Wyle Labs and their Heritage Reporting Corporation (202) 628-4888
f 15 1-cooperation doing some rather unique tests there. We have done 2 a lot of things in-house in our own laboratories both at 3 Singleton and in our central labs at Chickamanga. We have 4 engaged some of the experts. Some have been introduced 5 already. Some of the best experts in the field that we think 6 - can contribute to really trying to break some new ground with 7 this, and to try to understand the things _that we have seen, 8 and to try to help us in focus where we need to be putting our 9 attention. Is this a serious problem, and if it is, how to 10 unravel it. 11 Let me give you an overview then of where we would 12 like to go in this meeting, the kinds of things that we would 13 like to touch on. -14 (Slide) 15 Our objective, if you will, will be to do several 16 things. 17 one, to look at the background: how did we get where 18 we are today; how did we come from over the last few months of 19 one set of proposalo, and then a revised set of proposals, and 20 now we're coming back in with some new information. And I want 21 to put it in a context where you'll understand that a lot of i 22 what we're doing is based now on hindsight. That if we had all 23 been as smart six months ago as we are today, this meeting 24 probably wouldn't occur. Thing would be very much different. 25 But the actions that were taken in March, April, May, k l Heritage Reporting Corporation (202) 628-4888 I I
3 16 l' June,.whatever,'are indicative of the point.that we were at.six 2. months ago'. 1They.are indicative of the point I think that the 3 ' industry was at in their inability to. answer some-of the very 4 pointed questions that came from both the NRC and their con'ultants' that we' simply did not have the answers for,.nor 5 s 6 did the experts in the industries have;the answers for at that 7
- time, q
8 Third,.we'want to talk a little bit about the 9 soecific resolution of the silicone rubber insulated cable 10 concerns. What approach did we take and why did we take it, I 11 what were the results that we found, and of course very ] g . 12 importantly, what conclusions have we. drawn from those'results 13-and how do we intend to apply them. ~ j ' 14 Lastly, we want to give you a bit of'a feel for-15 what's going;on in'the industry at this time. I think'the 16 thing'that we've all wrestled with is the fact that.there are 17 no clear-cut standards 1that say, " Yeah, verily, in'this 18 situation, go.and do X, Y or Z." 19 So as we approach new ground, as we blaze a new 20 trail, if you will, we have involved the industry. I think 21 that what I'm going to present is that there was a problem 22 there, TVA saw the problem, and in cooperation with the NRC, as 23 we stumbled sometimes blindly through it, but we hit the 24 problem head-on. We've done' things that some of the people who 25 are experts.in the room I think would tell you that they Heritage Reporting Corporation (202) 628-4888
17 1 probably wouldn't have risked doing. 2 We have done things, and some of the data I think 3 that I will show you, I'll tell you when I get there, that if I 4 would have known what I was doing at the time that I did it, I 5 would have probably been a whole lot more conservative. 6 So I think what I'm going to present is a program 7 where we took a problem, we tackled the problem, we hit it 8 head-on. We did some difficult things; some things that no one 9 else has done, and what I want to do with you is share not only 10 our results, but share with you what the industry has done as 11 we've involved the industry and said, help us out in trying to 12 resolve-this particular problem. 13 (Slide)- 14 Our objective, then, is to tell you how we have 15 concluded our review, most particularly, the key point, I'll 16 give you the background, but we have concluded our review on 17 the silicone rubber insulated cable concern, and we want to 18 provide and explain to you what our final position is: llow it 19 is that we hsve arrived at the point where we have a level of 20 confidence where I can look anybody in the room in the eye and 21 tell you that the cable that we've got out there not only will 22 work for its intended function on a day-to-day basis, but that 23 that cable will work throughout a harsh environment condition, 24 and it will work after that harsh environment condition to 25 bring the plant to its intended safe shutdown condition. Heritage Reporting Corporation o I (202) 628-4888 4 1 1 --______z_ _J
U}i ' C ~18 1-And'I think' safety is the heart-of the concern,. .; 2 - Lwhether you(are TVA,' Lor whether you are the NRC.- And.'if we can- ~ 13' ' agree'_that'those safety concerns are. met,-and that's my_ job up 4 here.today,:isLto present to'you that basis, if we can~ agree 5 thatLthose, safety concerns have'beentfulfilled, then'I think we 6 'will bPh be happy. -7 We'want-to review the Wyle test program,fbecause it 8 is part and parcel _to everything I'm going to tell you. It is
- 9
- right at the heart of_our conclusions and-supports them very
.10 well.- 11 We want to review the significance of some of those-12' 'dielectricLatrength testing phenomenon that Mr. Fox alluded to,; 131 and relate that to what our experience has been.- There have '14-ibeen a lotuof times when I've given data, either.in a meeting 15 or-by the: phone or in personal contact with~some.of the people-16 here.at the table, and:quite honestly, will look at the data 17 and will say, now, how can that'be, how can that be. 18 And I think what we have.got today we don't.have tne 19 gem that came out a standard-that says, whenever TVA tests 20 cable, it will be this way or that way. What we hate are 21: basic, fundamental materials principles. As engineers in the- ] 22 utility industry, we tend to think in terms of~ application, so -23' we go to'the available standards guidance, whether it's IEEE or 24 ICEA. _They are applications oriented. 25 What we have done is step back one step, and said, I' Heritage Reporting Corporation zo (202) 628-4888
i f' 19-1_ instead'of going to the application. standards, we were led-to 2 .the materials standards themselves-that talk about the basic 3 fundamental principles of the testing' phenomenon, and then a4-let's relate that to what we've done. Maybe'it will' help us to 5 understand. And if'we can understand, maybe some of ths 6 uneasiness that we felt regarding the d.ata will be reduced. 7 We want to review those current industry activities. 8 I think we say there is a problem, and are not sometimes 100 9 percent sure how to approach it. We don't pretend to have all -10 the expertise in the world. We have hired most of the 11 consultants. You have hired the rest of them, 1 think. That 12 we' don't pretend to know everything,that there is to be known. 13 But'we have engaged the industry then to work with us so there 14 is a program going on. We're not always going-to be left then 15 saying what do we do when we have a cable problem. But we 16 intend to be in business a long time, and inevitably there are '17 going to be other cable problems. 18 Also, we want to correct a misconception regarding 19 insitu test breakdown levels. Now, I'll personally tell you 20 that that misconception -- I'll get to it later -- stems from 21 data that we passed on, I passed on in a September 10th 22 meeting, based on something we'got over the phone, and told you 23 the breakdown levels were in the range of 5,000 volts or so. 24-That was said in a transcribed meeting. And what we did upon 25 further searching is we found out that that was not indeed Heritage heporting Corporation (202) 628-4888 = _ _
20 1 true.- We'll get into the' details on that.- I want to clear 2; that record as we meet here today. 3 One more thing as far as an objective-and Mr.. Fox 4 touched on this partially,'I think, is that, yes, there-is a 5. Part-21 out there, a potential Part 21 that we wrote.
- And, 6
yes, there was a notice which came out.as a result of that. 7 .And I think you will see that some of the data that we present. 8 today will be very critical, and very helpful in offering ways 9 or proposing ways that we can come to a resolution of that Part 10 21. 11 (Slide). 12 Okay, let me give you an overview. This is.the 13 background part. I want to give.you an overview of what ve 14L have done, set the stage. basically for the problem that we 15 encountered, and let you know then how it is that we approached ~ 16 that-problem. 17 As a result of concerns that stem 'from really the 18 ,atts Bar employee concerns program, which were then reviewed '19 against Sequoyah Nuclear Plant, because the two plants shared .0 the same construction spec during that period of time. The NRC 21 decided that there could be problems -- the NRC and its 22 consultants decided that there could be problems most 23 specifically in three areas that needed to be reviewed and 24 resolved prior to restart of the Sequoyah unit. L 25 These involve the potential for cable jamming within i Heritage Reporting Corporation (202) 628-4888
l .J ) 21 1 a conduit run. That is, when three cables of roughly the same 2 diameter are within a conduit where the ratio of the conduit i 3 inter-diameter to the outer-diameter of a single cable was 4 roughly 3 to 1. And the concern is that when you go around the 5 corner in that conduit system, though you may have started out 6 in a triangular configuration, that in the process of going l l 7 around that bend, the cables will pop out of the triangular l 8 configuration, and go three abreast and lock up or jam within 9 that conduit system. 10 That was considered a serious enough concern that it 11 was tabled as restart. -12 Second was the issue of cable pullbys. When you have 13 cables resident within the conduit system and you go to install 14 additional cables within that same conduit system, the new ones 15 aro then pulled in over top of the former cables. The concern 16 there is, again, when you go around a corner where side wall 17 pressures are reasonably high that due to that side wall j 18 pressure there is a possibility that the cables which are being 19 pulled in could saw or cut into the cables which are resident 20 within that conduit system. 21 Third, there was a concern for lack of support in l l 22 vertical conduit runs, or more particularly in the case where 23 our vertical runs exceeded the allowable and the national 24 electric code, and no additional, or no separate cable support I 1 25 had been supplied. That the conduit system itself became the Heritage Reporting Corporation (202) 628-4808
m 1. b. 22 f if . supporting mechanism. c2: .Most particular,-as1we review that'particular 3 ' problem,-the concerns centered ~upon silicone rubber because of .4 the: softness of that insulation system. And,1further, centered 5 upon runs where immediately.at the top of.this long vertical 6 drop.there.was.a 90-degree condulet. 7' We did our-testing at'.240 volts'per mil. Initially, 8' .this was: applied against the. nominal: insulation thickness; that: 9 'is, if youricable had-45 mils nominally, you applied it against 1'O that. criteria. 11. ' Subsequently, we revised our program, and reduced. 12 that particular criteria to where the voltage was applied 13 against'the' minimum qualified thickness. The thought there ~ 14-being that. cable insulation thicknesses are often! driven by4 15-many things other than their environmental' qualification 16 requirements..They are driven perhaps-by manufacturing. -17 re'quirements wherein if you have got a cable that is very ~ 18
- 1arge, it's nearlyl impossible'from a processing standpoint to 19; extrude very, very thin insulations on it.
20-So if'you go and look at the standards, you will see 21 that as conductor. size grows, commensurate with that growth 22 there is also a growth in the insulation thickness as well. '23 A 600-volt cable, whether it's the size of your 24' littleffinger, or, you know, an. inch in diameter, still needs 25 insulation just to hold 600 volts from an electrical Heritage Reporting Corporation (202) 628-4888 __u__
( 23 1 standpoint. But for mechanical and processing reasons, 2 insulation thicknesses do vary. 3' The program was structured so that we did our 4 evaluations against worst-case configurations. With the issue 5 of jamming and the issue of pullbys since the damage woulc' be l 6 resident in the conduit system in an area that it cannot be 7 physically viewed and evaluated, we did a program ~there whera i 8 we ranked our installations according to an agreed upon set of 9 criteria. And then we tested 15 out of the worst case 20. ] 10 The idea being there that since we couldn't verify 11 that the problem did exist, or didn't exist anywhere within the q 12 conduit run, that we would establish a statistical basis saying 13 that it could be out there in that worst case group somewhere. ~ 14 That the approach then was that we would' establish by multiple 15 tests whether the problem really had occurred or not, even 16 though the possibility was there. 17 In the case of number three here, the lack of support 18 in vertical conduit runs, since it was evident that you could i 19 go and take the cover off the condulet and you could see ] I 20 whether that particular installation was undertrained, it j l l 21 wasn't undertrained, it was easy to categorize by what was 22 exactly your worst case, only a single test was done. And 23 that's in accordance with both the letter of January 15th from 24 the Nuclear Regulatory Commission, and in accordance with the i 25 TER. i Heritage Reporting Corporation (202) 628-4888 I l
) 24 l' The acceptance criteria that we used for this test h 2 program is derived in part from IEEE 141. That particular 3-standard -- I might say, of course, part of the acceptance 4 criteria was that during the course of a withstand test that 5 you do indeed withstand, all right? 6 The second part of the criteria, the portion which 7 was derived from IEEE 141, is that which Charlie has -- Mr. Fox-8 has referred to already, that a polarization index. The 9 criteria that we applied said that if you had a polarization 10 index which was 1.0 or greater, that that cable was acceptable 11 for use, and that if the polarization index was lessJthan-1.0, 12 we considered that a failure. 13 I have some additional comments, as Mr. Fox 14 Indicated, about that-later in our program. 15 (Slide) 16 Okay, a little bit of background here just to set the 17 stage for how we've applied cables, what our design philosophy 18 is. 19 Inside containment at Sequoyah, all of the low 20 voltage power and control cables are single conductor, silicone 21 rubber insulated with asbestos braided jackets. There are 22 approximately 950 conductors that are 10 CFR 50.49 required
- 3 cables.
They were supplied from three different vendors, and 24 our design philosophy and application has been such that these -25 which are in control applications, of course, have no OMIC Heritage Reporting Corporation (202) 628-4888
e s 25 1- ' heating, no self-heating, very low current. And those which 2 are in power applications, review has shown that those cables 3 do not operate until the onset of an accident event. So they 4 basically set their at ambient condition until some accident 5-event occurs. 6 Outside containment design philosophy is a little bit 7 different. Our low-voltage power. cables are.both single-8 conductor and multi-conductor, depending upon the size. They 9 have cross-link polyethylene or ethylene propylene rubber 10 insulation systems with -- generally with hypalon jacket. 11 There are some which also have_PVC jackets. 12 The control cables are multi-conductor, all'aulti-13 conductor. They are either cross-lined polyethylene and EPR 14 with hypalon jackets, or they are: thermal plastic polyethylene 15 insulated with PVC jackets. 16 MR. LIAW: Excuse me, Kent. 17 MR. BROWN: Yes. 18 MR. LIAW: Are you saying that you don't have silicon 19 rubber cable outside containment required to meet the 20 environment, or harsh environment? l 21 MR. BROWN: There are no 50.49 silicon rubber cables 22 outside containment except when used as pigtails like in the 23 case of a target rock silinoid valve that has a high 24 compartment temperature, we use on the last few feet perhaps a j 25_ pigtail, an application such as that. Heritage Reporting Corporation (202) 628-4888 1
O 4 26 1 There are some 1E cables that are -- I think it's 2 Jimmy Huston, is that 11 outside -- 3 MR. HUSTON: Approximately 11 cables. 4 MR. BROWN: Yeah, 11 cables outside containment which 5 are 1E but they are 50.49 required. 6 MR. LIAW: What do you mean by that? 7 MR. BROWN: All right. The simplest way would be, 8 for instance, if you had a 1E cable but it was not located in a 9 harsh environment. Then it doesn't fall under the boundaries 10 of 10 CFR 50.49. 11 MR. LIAW: Concerns master breaks condition? 12 MR. BROWN: Yes. Another way wouAd be if it was 13 classified as a 1E cable, but it was what we call a Category C. 14 Its operation is not required, nor is it a problem if it fails 15 upon the onset of an accident. 16 MR. LIAW: It's somewhat inconsistent with the method 17 we received previously. I guess we're going to have to 18 clarify, you know,-or pigtail or all along, are those cable -- 19 my understanding were under MAC 24. 20 MR. BROWN: Again, I believe that the number is 4 21 around 11, and we can confirm that. There are a very low 22 number, and as I said, they could be 1E but not in a harsh 23 environment which would take them outside the boundary of j 1 24 50.49, or they could be what we call a Category C, and not only j i 25 not required to function, but they also could fail, whose i Heritage Reporting Corporation j (202) 628-4888 j i
1, ] 27 f 1 failure has been shown not to -- 2 MR. LIAW: What is Category C now? 3 MR. BROWN: Category C, it doesn't have to operate 4 and it doesn't matter if it fails, and that doesn't harm -- 5 it's a fall-safe piece of equipment, for instance, and the 6 -cable can fail and the equipment still does its function. 7 MR. EBNETER: Kent, excuse me. Are there any 1E 8 cables inside containment other than 50.49s? 9 MR. BROWN Yes. 10 MR. EBNETER: Of silicone rubber? This says 50.49 11 cables inside containment. I'm using your argument on the last 12-statement now on reverse in containment. Is there some it 13 cables in containment that are not 50.49, but are silicone 14 rubber cables? 15 MR. BROWN: Under the scope of 50.49, at least as I 16 defined it, I'm thinking in terms of A and B, I'm confident 17 there are many 50.49 cables that are Category C which, of 18 course, are silicone rubber if they are power control. But if 19 they are Category C, obviously they are not 50.49 required. 20 They are not required for a 50.49 event. It gets to be a 21 circuitous argument. 22 Is that something you would like further 23 clarification on? 24 MR. EBNETER: No, I'm just curious about it. 1 25 MR. BROWN: Okay. Heritage Reporting Corporation (202) 628-4888
28 1-MR. LIAW: Back to Category C again. You've got an Mi 2 internal definition or something that you have committed stated 3 in FSAR? All different'way of categorizing your system 4 components, I was jusc -- 5 MR. BROWN: Yeah. As far as whether that's internal 6 or whether that's common industry practice to categorize A, B 7 and C, I really:can't tell you. 8 However, I do know that this is fully a part of our 9 EQ program'and has been. reviewed, and categorization was 10 reviewed as well as qualification. So when we categorized that 11 process of categorization, also fell within the scope of the 12 review. 13' (Slide) n. 14 Okay, on to the results then. i 15 In the jamming test program, again we picked 15 out l 16 of our worst case 20 to do the test. There were in those 15 i 17 cable 45. conductors in also 15 conduits. All of those cables i 18 fell outside containment due to the selection criteria. l l 19 Jamming is not a problem until you get into larger conductor 20 sizes-We specifically, by our agreed up criteria, were 21 looking for conductor sizes No. 10 or later, which is basically 22 the bulk of, you know, your power cables outside containment. 23 Out of the 15 conduits'that were tested, 10 of these l .24 were dry during the test, five were wet during the hipot. All 25 conductors passed the 7200 volt exposure. L l Heritage Reporting Corporation (202) 628-4888
f. .e-29 1 Just to give you some background on the specifics of 2 the result, 44 of those conductors had a polarization index of. 3 1.25 or greater, and one.had a polarization index between 1 and 4 1.25. 5 What we feel like we showed there conclusively was that having exposed our worst case configurations to this test 6 7 we conclusively showed that cable jamming was not a problem.at 8 Sequoyah Nuclear Plant. 9 (Slide) 10 All right, similar story on pullby test results. And 11 I guess the thing that impresses me as much when I look at this 12 every time is the scope of that program. When we went looking 13 for cables in a pullby situation, the very fact that you have-14- 'got pullbys means that you have got multiple cables to contend 15 with. 16 And the way our criteria was set up, we were looking 17 for situations where there were at least seven cables in a 18 conduit, where there had been at least two pullby situations, 19 so that the number of cables grew very, very rapidly. There 20 were 878 conductors in all, and 298 cables that were exposed to 21 this hipot test, a tremendous program in scope when you look at 22 it. 23 All the conduits were outside containment.
- Again, 24 because of the structure of the screening criteria that were 25 applied as we agreed, we were looking for situations where Heritage Reporting Corporation (202) 628-4888
30 1 there were thermal plastic jackets involved. Given that we 2 were looking for thermal plastic jackets, if you remember a 3 couple of slides back wl.en I said here is our design 4 philosophy, the control cables outside containment, those which 5 are multi-conductor, some of them have-polyethylene, thermal 6 plastic polyethylene with a PVC jacket. And others would have 7 cross-linked polyethylene with a PVC jacket. 8 So immediately, as we're looking for the PVC or the 9 thermal plastic jacketed cables, and as you can understanding, 10 this is if you were in a sawing application pulling one cable 11 across another, a thermal plastic would tend to be violated 12 first. It would tend to coften, and the saw thrnugh wnuld 13 occur, and then you would be sbrading a primarily insulation 14 itself. 15 Twelve of these conduits were tested in the dry 16 configuration. Three of the conduits were flooded water prior 17 to the hipot itself. All conductors passed the hipot. Now 18 there were various voltages involved. Again, we're applying 19 our criteria against the minimum qualified thickness. So some 20 were tested at 4800 volta DC, and others at 7200 volts DC. 21 Out of the 878 conductors, 874 had a polarization 22 index of 1.25 or greater. And four conductors had a 23 polarization index between 1 and 1.25. 24 MS. AXELRAD: Can I ask you, are you sayi.; that you 25 didn't have any situations inside containment in which you had Heritage Reporting Corporation (202) 628-4888 j
31 1 at least seven cables in the conduit, and therefore had at 2 least two pu11bys? 3 MR. BROWN: No, I'm not saying that. I'm saying that 4 the criteria that we agreed upon was structured that we were 5 looking for thermal plastic jackets, because the very nature of q 6 them being thermal plastic with the friction that is created 7 when you are pulling around a corner pulling one set of cables 8 across the other, heat will be generated, and it will be the 9 thermal plastic cable that will tend to be violated first. 10 One of the other criteria was that the cable which 11 was being pulled by needed to be on that had a thermal set l 12 jacket, a rubber-like material. Again, they are much less 13 slick, if you will, than the thermal plastic jackets. So they 14 will tend to be higher in nature in friction, and tend to start 15 that abrading process earlier in time sequence. 16 MR. FOX: These criteria defining worst case for l 17 pullbys and jamming and the unsupported vertical were worked 18 out between NRC staff, the NRC consultants and the TV technical 19 people. These were agreed to. 20 MR. MARINOS: We accepted the criteria. 21 MR. FOX: Pre-test, and they were accepted, yes. 22 MR. BROWN: Yeah, there were many consideration. For 23 instance, we had switched pulling lubricants at a certain point 24 in time. And we went to one that everybody agreed stayed on l 25 the cables better. So we established that as a cutoff date and l Heritage Reporting Corporation (202) 628-4888 l \\
i-32 1 said, look, for times prior to that, pullbys'that were prior to 2-that when you would have had a less sophisticated pulling 3 ' lubricant'and one'that would have-been pulled off-perhaps as 4-soon as the cable got into the conduit. 5 So there were a number of factors that drove towards 6 this particular cable situation. . as it modified some point in time? Might W 7 MR. LIAW: 8. be Jim was correct. In one of your submittals, he indicated' 9 that, that the minimal number of cable conduit. Seven sounds 10 like -- 11 MR. FOX: For the minimum number -- 12 MR. MARINOS:
- Yes, 13 MR. FOX:
Yes, that was one of the criteria. 14 'MR. LIAW: -- seven of the criteria. 15 MR.. FOX: I don't remember specifically. There was a 16 number that defined a minimal number we were looking for. 17 MR. BROWN: Yes, seven was-always the criteria.for 18 pullbys. That's the initial criteria. 19 MR. FOX: My staff says seven wasn't -- 20 MR. LIAW: No, seven. Now Kent say initial criteria. 21 MR. BROWN Initial-meaning that the first thing that -22 we did, the first process that we did was we went and asked the 23_ computer system spit out every conduit that's got more than 24 seven cables in it. That was the first step. 25 MR. FOX: That was our initial -- Heritage Reporting Corporation (202) 628-4888
r 33 1 MR. BROWN: First step in the screening process. 2 That's what I mean by initial. 3 It was always seven. It still is seven. Seven is 4 what we apply. 5 MR. LIAW: So she was correct then. Jane was 6 correct. 7 MS. AXELRAD: That's what they said. I was just 8 checking. 9 MR. LIAW: Yeah, okay. 10 MR. EBNETER: I would hope we all understand and 11 agree on this at this point in time. 12 MR. FOX: Absolutely. 13 MR. LIAW: We will accept the program and bless the-14 results. 15 MR. FOX: One of the reasons that we're covering this 16 in a little more detail than we would normally if we were just 17 meeting with the Office of Special Projects is that we do have 18 NRR people present who have not been intimately involved in the 19 details of this disposition of pullbys, jamming and unsupported 20 vertical cable. So.if you will bear with us, we'll hurry on 21 through, but we wanted to cover it for their benefit, so Paul 22 Gill, et al., would have familiarity with what we have done. EndT2 23 Thank you. 24 (Continued on the next page.) 25 Heritage Reporting Corporation (202) 628-4888
') 34 1 (Slide) 2 MR. BROWN: In the national support arena, the 3 criteria that we applied was to look for conduits that 4 contained silicone rubber insulated cables, again, because of 5 the softness of the silicone rubber, we were looking for 6 applications where these were inside containment, because of 7 the exposure to a harsh environment that they wo.31d see. 8 The criteria set there was where there was a minimum 9 of five cables to be contained in each of those conduits. We 10 were looking for a minimum of 20 percent fill. 11 The thought there was that with a high number of 12 cables and a high percentage fill, we're going to work towards 13 an. application that will find the high bearing pressure in that 14 condulet. 15 The next criteria was that there needed to be a 16 90-degree condulet at the top of the run, again, with the very 17 small radius in the condulet, the fairly high load as a result 18 of percent of fill and number of cables, that the bearing 19 pressure developed at that small radius in the condulet might 20 be sufficient so that there would be a conductor creep through 21 the fairly soft silicone rubber insulation. 22 Of course, the last criteria, which was the heart of 23 the whole matter, was that the drop beneath that condulet had 24 to exceed NEC Article 300-19. I { I 25 That criteria basically says that if you exceed 25 Heritage Reporting Corporation (202) 628-4888
35 1 percent of the drop, 25 percent of the length allowed in that 2 particular table, that an additional cable support should be 3 included. And that was something that we had not done. 4 (Slide) 5 MR. BROWN: Okay. The test that we did in the 6 vertical conduit application initially, as we ranked 7 our -- 8 MR. MARINOS: We apparently have a slide that's not 9 in your hand. 10 MS. AXELRAD: A Xerox. 11 MR. FOX: Bear with us. We can furnish it. 12 MR. BROWN: We can correct that after the meeting. 13 All right. So our screening process, we came to 14 define a particular conduit that we defined as our worst case. 15 In that conduit, there were six cables, 16 conductors, and the 16 drop exceeded the National Electric Code. All these were 17 silicone rubber insulated with asbestos braided jackets. All 18 of them were single conductor configurations. 19 The conduit was located in containment. The conduit, 20 as I explained earlier, was tested dry because of the immediate 21 proximity of the point of concern to a known ground plane. 22 As I noted earlier, these tests were done applying 23 the 240 volt per mil criteria against the nominal insulation 24 thickness. 240 volts times 45, 10,800 volts. 25 There were no failures at the support point. We did Heritage Reporting Corporation (202) 628-4888
36 1 encounter three breakdowns. One at 7,500 volts as we were 2 bringing the voltage up, one at 10,00 volts again, during the 3 increase of the voltage, and the third one at 10,800 volts DC 4 after one minute. 5 Of the remaining conductors, one conductor had a 6 polarization index less than one. Eight conductors had a 7 polarization index between, greater than one, greater than or 8 equal to one, but less than 1.25. And four conductors had a 9 polarization index of 1.25 or greater. 10 As you know, we took the cables out, took out the 11 faulted sections after we had done an isolation process to 12 determine whether the faults had occurred at the support point 13 o'r away from the support point. 14 When we had satisfied both ourselves and your 15 resident inspectors that indeed we had isolated them away from 16 the support point, those cables were removed, sent to the 17 University of Connecticut, to their Electrical Insulation 18 Research Center, and subjected them to a battery of tests. 19 MR. LIAW: And these cables are made by AIW? 20 MR. BROWN: Yes. All 16 of those conductors were 21 manufactured by American Insulated Wire. Just as a bit of 22 review, and I think this has to be one of the oldest slides in 23 creation at this point, one of the most used, is a picture of 24 that particular conduit run. 25 What we have is a junction box at the lower extent of Heritage Reporting Corporation (202) 628-4888
37 1. that run, maybe.a' ten or 15 foot horizontal run, a condulet in-2 a fitting', then~a vertical drop which exceeded the code. 3 At Point B it just slightly exceeded the Code. At 4 Point D where another condulet was, and we had a branching off 5 of additional conduit to other equipment, it was another three 6 or four feet in excess of the code. 7 The failure points however,.were not here.at this 90 8 degree condulet, as was I guess anticipated, perhaps,.if you 9 felt that the mechanism was correct. They were not, however,, 10 at that point, nor at Point D, which greater drop, greater 11 bearing pressure. 12 .They were well away from what we show as a Point C 13 and yet another point here where we say Cable Number 2. 14 I would say this is perhaps five feet, okay, and this 15 perhaps ten feet to the other failure location. 16 As mentioned, we sent the cables to the University of 17 Connecticut,,to their Electrical Insulation Research Center. 18 They did a number of tests, including tensile and elongation. 19 We did some insulation resistance, some dielectric breakdown, 20 infrared spectroscopy, X-ray spectroscopy, quite a number of 21 examinations, optical, of course. 22 What we found was that there were no contaminants. 23 One of our first feelings was gee whiz, if these cables had 24 failed, at these levels and when we had taken the jackets off, 25 we didn't see anything of any significance at all. Our first Heritage Reporting Corporation (202) 628-4880
i 381 I 1 feeling was there's got to be some sort of manufacturing 2 problem.. So we went looking for contaminants, and we found 3 none. 4 We looked to see-if there was any age-related 5 degradation, perhaps, that the cables were just aging faster 6 than our documentation would have led us to believe. There was 7 no local heating, there was no problem from excess radiation. 8 None. 9 We found that the cables were within their spec. I 10 sa} OD here. Really what I mean to say is that the insulation 11 thickness met the spec. requirements.- 12 If I took and cut the cable at any point, end, 13 middle, whatever, what I found is that the~ nominal wall, in all 14 cases, met the insulation thickness. 15 We did find damage, localized. .16 (Slide) 17 MR. BROWN: What we found was a result of what we 18 believed to be impact damage..And this is a fairly crude 19 graphic representation of what was uncovered at the. University 20 of Connecticut. 21 The phenomenon, as we believe it occurred, and we 22 have duplicated it, as that through some kind of impact, as the 23 insulation system is compressed onto the strands beneath, they 24 literally cut up into that insulation, into that fairly soft 25 silicone rubber. Heritage Reporting Corporation (202) 628-4888 ___________E
39 1 Now, I've shown it as a large opening here, but that 2 is really more graphic than anything. Literally what it does, 3 it cuts and basically closes back up. When we were at the 4 University of Connecticut, we measured the remaining intact 5 insulation wall. 6 This area that here that I've shown is reduced wall. 7 And on those which we measured, the smallest or thinnest intact 8 insulation wall was eight mils. There were a number of pockets 9 that we measured eight, one for eight, a couple for ten, one 10 for 12, another at 20. 11 We found this on two cab 1'es, out of the four 12 conductors. I say cables. We found this on two conductors 13 out of the four conductors which were sent to the University of 14 Connecticut for examination. And really, as I think we all 15 know, it has been the concern then for the reduced insulation 16 wall thickness that has been the sticking point throughout our 17 discussions. j 1 18 MR. LIAW: Kent, tell-me how you made those areas 19 already electrically broken down, and how you come back to 20 measure the reduced thickness. 21 MR. BROWN: Okay. You'll remember on, let's take ] 22 the, probably the easiest one to discuss is Cable Number 4. In 1 23 the U. Conn. Report you'll see it listed as Cable Number 4. 24 There was no dielectric breakdown during the test 25 itself. It was the cable that had a polarization index less Heritage Reporting Corporation (202) 628-4888
40 1 than one, as a conductor that we had removed and then in 2 sending it to the lab, what we did was, using a Tesla coil 3 arrangement where -- 4 MR. MARINOS: But Kent, you're confusing us. You are 5 talking about the U. Conn. tests, aren't you? 6 MR. BROWN: Yes. 7 MR. MARINOS: You didn't use polarization index then. 8 The polarization index arrived later in your revision to the 9 cable test program. 10 MR.. BROWN: Specifically, polarization index was not 11 an issue at all, I guess, until the July 31st semester. 12 However, what we did do is, in evaluating leakage current, and 13 without the use of the word " polarization index," we had a 14 situation where the leakage current increased throughout the 15 course of that five-minute withstand test. 16 Now, we called that one susceptible, from the 17 beginning. 18 MR. MARINOS: You said you didn't allow it to break 19 down, you interrupted the test? 20 MR. BROWN: No. No, at the end of its five-minute 21 test, the prescribed five-minute withstand test, it had not 22 broken down whatsoever. Okay? But, we had observed increasing 23 leakage current throughout the duration of the test, so we 24 categorized that one as susceptible and removed it as well. 25 MR. MARINOS: So this is the cable at 10.8 KV that Heritage Reporting Corporation (202) 628-4888
41 1 you declared that it was, didn't pass the test. That's the one 2 that we're talking about? 3 MR. BROWN: That's -- 4 MR. MARINOS: There was a bunch of them that failed 5 at lower voltage than that. 6 MR. BROWN: Yes. A couple slides back I had one that 7 said that there was a breakdown at 7.5 -- 8 MR. MARINOS: 7,200. 9 MR. BROWN: -- 10 KV, and at 10.8 KV after one 10 minute. None of those are the one I am talking about now. 11 This is the one that shows on that particular slide as a 12 polarization index less than one. No breakdown. 13 MR. MARINOS: The other ones had higher thickness 14 insulation, the ones that failed? 15 MR. BROWN: Let me explain, okay? I'll get there. 16 On this cable, which one we called Number 4, because 17 there was no specific discharge channel, therefore I can make 18 very conclusive statements about what we did and didn't find. 19 There was no breakdown. 20 We used, as I said, a Tesla coil to find the weak 21 spots in the insulation, if you will, and there was, according, 22 as you'll see in the U. Conn. report, one three-inch section on 23 this cable, in which there were a number of punctures as a 24 result of exposure to the Tesla coil itself. 25 Now, it leaves a tiny, tiny, microscopic puncture in Heritage Reporting Corporation (202) 628-4888
l 42 1 the cable as-opposed to a hipot test set which leaves maybe a 2 one millimeter. hole. So they're readily distinguished one from 3 another. 4 What we did is we took, and because there was some 5 surface crazing where the braid had impinged upon the 6 insulation, you could tell which side the impact came from. 7 All right? You would see surface crazing on this side. We 8 then went to the opposite side, 180 degrees around, cut the 9 cable longitudinally and unrolled that insulation. And then 10 you can see where the voids are. 11 Now, in order to measure the thickness, what we did, 12 remember now I have, I've cut like so. Once I've said all 13 right, all the voids are in this area, what we did is we sliced 14 the cable in half. All the voids are on one side. 15 We took a heated bath of dye, put that piece of white 16 insulation into the heated dye, left it for some extended 17 period.of time, and I don't remember the time period at the 18 moment, to allow the dye to penetrate as far as the void would 19 allow it to go, took the cable then out of that dye bath and 20 began to cut into it, shave into it would be a better word, 21 looking for those -- well, we knew where the voids or cuts 22 were -- it was simply a matter of shaving in a hair at a time 23 until we found the thinnest remaining intact wall. 24 That is the cable I referred to. 1 25 MR. MARINOS: So you were cutting it axially, right? Heritage Reporting Corporation (202) 628-4888 l \\
) 1 .I, 43 1 I mean you were slicing. 2 MR. BROWN: Radially. In other words, I was cutting 3 this way. 4 MR. MARINOS: I mean radially. I meant radially when 5 I said axially. 6 MR. BROWN: Okay. That is Cable Number 4. Only one 7 other of these conductors had a pocket that we located. All of 8 them were examined, but only one other one had a pocket and it 9 was something like 19 or 20 mils. At least double. Whether 10 it was 19 or 20 I don't remember exactly which. 11 Now, Conductor, I believe it is, 1, did not have a 12 pocket. And I forget now if Conductor 2 also did not have any 13 pockets that we could locate. 14 There, we had a one millimeter discharge channel, but 15 with no indication of a pocket that we could find. And as you 16 know, in our earlier presentations, we said that aside from the 17 looseness of the tube and aside from some marking on the 18 surface, we found nothing. 19 MR. MARINOS: So you say you found nothing. The 20 thickness was as you predict, 45? 21 MR. BROWN: Spec. thickness, right. 22 MR. MARINOS: To the spec. thickness. 23 (Slide) 24 MR. BROWN: Very quickly, then, in order to confirm 25 our mechanis.n as being one of impact, we took and used an l I Heritage Reporting Corporation j (202) 628-4888 ] l 1
f i l 44 1 impact test set with the sketch I have here and subjected the 2 cables to varying impact levels. 3 What we were trying to establish was at what impact 4 level some damage occurred, and how much damage occurred, and 5 what we found was that the different manufacturers exhibited 6 different degrees of resistance to the impact forces. 7 AIW -- I'm going to put all these in terms of number 8 -of pounds dropped from a foot. AIW was something like 3.6 9 pounds dropped from a foot. Anaconda, I believe, was ten, help 10 me out here, five and a half pounds dropped from a foot, I 11 think. And Rockbestos was nine something pounds dropped from a 12 foot. 9.9, I believe. 13 Okay. Following that initial set of tests, we 14 expanded our sample because we couldn't definitively say what 15 event caused the damage that we saw. 16 We expanded the sample and we expanded to other 17 manufacturers as well, because at this point in time we had 18 seen that there was some susceptibility of silicone rubber to 19 impact damage. l 20 The slide here shows the total number of conductors j 21 installed in Sequoyah Unit 2, in 5049 applications. The total i 22 number that we tested, we tested some 91 conductors. We i 23 exposed them to either 7,200 or 9,600 volts according to 24 whatever their minimum qualified thickness was at that point. 25 You'll see we had a total of seven breakdowns. These Heritage Reporting Corporation i (202) 628-4888
4 1 45-II include the original group of 16,.by the way.,-We had aLtotal'
- 2 ofL eeven breakdowns and then three we categorized.-
That 3~ includes the initial one. Three, then, we categorized also as .4 = anomalies, on the basis of the polarization index being'less 5 than one. 6 (Slide) 7 MR. BROWN: I guess I might just mention 8 that -- well, let'me go ahead with this. 9 You will recall, early, I guess it'was in'the 10-objective slide, I said that we had made statements in'our 11 September 10 meeting that we had breakdowns on the order of 12 5,200 and 5,800 volts. I want to correct that at this. point 13 in. time.- 14 After we got the data to review and reviewed 'hati t 15 data and discussed with the test people involved, we found that' 16 the lowest value of dielectric breakdown of any cable silicone 17-rubber that we tested in our program was 7,000. volts. 18 The situation was there that the people conducting 19 the tests, they had the end meter set on a certain scale. As 20 they ran the current up, or ran the voltage up, rather, they 21 pegged the current on that scale and instead.of simply 22 switching to the next higher scale and continuing the test, 23 they said, meter peg scale, and they stopped the test. And 24 they called it a fail and stopped. 25-Now, when we saw what they'd done, we challenged Heritage Reporting Corporation (202) 628-4888
46 1 them. We sent them back out there. They repeated the test. 2 And what we found: on those was that they may, I think one of 3 the conductors actually had an acceptable reading. Its 4 polarization index was well above one. They simply had the 5 meter on the wrong scale. I believe the other one had a 6 polarization index of less than one. 7 But there were no breakdowns, and understand this 8 now, there were no breakdowns less than 7,000 volts. 9 MR. MARINOS: So the number is reduced from three to 10 two, then? You said one, that you redid the test; you found 11 that the polarization index was acceptable? So the three that 12 you had indicated at first as test failures, have been reduced 13 down to two? 14 MR. BROWN: Yes, that's correct. The third one was 15 indeed a breakdown at 7,000 volts DC. 16 MR. MARINOS: Okay. 17 MR. BROWN Now, the reason I mention this is because 18 in a letter that we received dated November 13, you all took 19 note of the breakdowns that we had reported, so we understand 20 where you got it. 21 MR. MARINOS: Those were figures we had to deal with. 22 MR. BROWN But you had reported.that there were 23 breakdowns between 5,000 and 7,000 volts and that this resulted 24 in a heightened level of uncertainty, because you felt like 25 there was a high percentage that had broken down in these l Heritage Reporting Corporation (202) 628-4888 l l
[, o 47 1 levels. 2. What we want to do is, we want to correct that 3 .information, and we hope that that will alleviate the 4 heightened level of uncertainty because there were no failures 5 below 7,000 volts whatsoever. 6 I guess you would say that at that point we were in a 7 quandary. We had a number of breakdowns at this point, 7,000 8 volts all the way up to 10,800 volts, and we had tested 91 9 conductors, a fairly it. tensive program, and we had several 10 options as to what we could do. We could continue testing, you 11 know, embark on a test and replace program. We could go out 12 and do a wholesale replace program, yet not knowing perhaps 13 what the cause for our problems were. 14 We knew it was impact, but yet in a sense hadn't 15 really solved what the problems were. We could change cable 16 types completely. There were a number of options that we could 17 pursue. I 18 What we decided to pursue was since we had found a 19 cable with an insulation thickness as small as eight mils, and i 20 since the point that we had stuck on in our July 13 21 presentation was the specific pointblank question was asked, a 22 very valid question. Are you saying that eight mils is 23 qualified? Silence from the TVA side of the room. We couldn't 24 tell you. Nobody knew. So wi sat out to know. 25 (Slide) Heritage Reporting Corporation (202) 628-4888
e 48 1 MR. BROWN: We decided that we would attempt to 2 qualify' cable with reduced insulation wall. We decided that 3 this would do one of two things for us. It would either prove 4 that a reduced wall in this particular kind of cable in this 5 particular kind of application, was of no concern, or, if we 6 were continuing a testing program, perhaps it would allow for 7 reduced test voltages, since test voltages were derived for 8 minimum EQ thickness. 9 In studying the problem we said that there were 10 several possible ways that we could reduce the cable wall. One 11 was, we could contract with manufacturers in order to get them 12 to extrude cables with reduced walls. We could somehow shave 13 the cables down to some reduced thickness. We could buff the. 14 cables down mechanically, or we could impact them. i 15 Now, of course, up to this point in time we had been 16 doing some impacting, as I had shown. The problem with that is 17 the repeatability of it. At that threshold level it seems 18 very, very particular as to exactly maybe how the strands are 19 oriented or how you hit the cable or the temperature on a given 20 day or the height above ground or sea level, or something. I 21 don't know what it's particular on, but it's hard to get a 22 repeatability. ] 23 We didn't want to go into a loca test situation where 24 we didn't know whether we had eight mils or 28 mils or 40 mils. 25 So we finally decided that what we would pursue was shaving the i Heritage Reporting Corporation (202) 628-4888
b o 1 49 1 cable insulation down to some reduced thickness. 2 Target' thicknesses for this test were eight, 15, 20,- 1 3 25 and for one manufacturer,30 mils. For two manufacturers, we 4 already had cables that were qualified as low as 30 mils and 5 that was the criteria that we were applying. 6 For one vendor, his thinnest, qualified silicone 7 rubber insulation system was 40 mils. 8 So we opted to put one in for 30 mils there. As you 9 see, we were basically hedging our bet, not to put all of our 10 eggs in one basket because we didn't know at that point in time 11 what the answer might be. 12 Our target qualification was set for ten years. The 13 intent of this was of course to facilitate Sequoyah restart. 14 We weren't really intending to embark upon a research program. I 15 We got a plant that we want to operate and we want to operate 16 it now. 17 MS. AXELRAD: But under 50.49 you are required to 18 qualify them for the life of the plant, which is 40 years, 19 right? 20 MR. BROWN: Right. Well, we're required to qualify, 21 to be qualified, whenever we want to operate the plant. So if I 22 I want to change our cables every year, I could come up and 2L operate for a year and then I have to shut down because of EQ 24 and change them out. 25 So we chose a time that would provide a sufficient i Heritage Reporting Corporation 1 (202) 628-4888 l l l l
v 96 1; 7200, 9600 or-10,800. And'it works out to roughly, oh, let's 2 say 120-volta per second. So I've done a slow rise test which 3 remember is more severe than the quick rise,.and look at the 4-levels that I broke down. S They are well above, far enough above that I would. 6. believe that one could sustain a five-minute test at the lower 7 ' levels if all I had to go on was faith in looking at these 8 7500. 9 MR. MARINOS: If you look at the square root of the 10 . distances to these voltages, you will come up with certain 11. thickness which is way beyond the 4 mil that you established 12 that you really need. Is this the aim of these numbers here? l 13 MR. BROWN: No. The aim of the numbers is, first of-14 all, to show you if I follow any given recognized method of 15 establishing what a voltage should be, 383, new cable, 16 whatever, that we have done more than.that. 17 MR. MARINOS: The premise here, as Charlie has 18 indicated, you will demonstrate that with the tests you have 19 conducted you have adequate insulation level in your cables. 1 d 20 You have no failure, okay. They break down at 7,500 and 7,000 21 volts, and that's indicative of a successful test based on 22-something, and the something I can only relate to this ratio, 23 to this definition of thickness. The ratio went over this 24 total of the distance, of the thickness. q 25 MR. CANTRELL: I guess, Angelo -- Bob Cantrell. I i' Heritage Reporting Corporation (202) 628-4888 i
e i s .97 1 guess the point he's making is that if we were standing back 2 and had never done any test and had gone to Wyle Lab and said, 3 you know, let ae design a perfect cable, and then devise a test 4 for it to do the electrical needs of that, installation, I would 5 have designed a cable at 4, 4 and 6 mile that would have met 6 the functional service needs of that application. 7 Now if I went to the industry standards and developed j i 8 an insitu test program to give me confidence in that design, 9 the EQ levels and the production levels -- 10 MR. MARINOS: But our question is that you have an 11 insulation level. 12 MR. CANTRELL: Okay. 13 MR. MARINOS: That's what we are trying to 14 MR. CANTRELL: But if we were testing for that in a 15 new -- you know, if we bought new 4-mil cable, they would f 16 finish it and test it at those levels he had up there, 3600. i 17 We were saying our objective was not to develop a destructive { 18 test, but to develop a test to give us confidence. 19 And I contend that we would have come in here and I 20 probably reached agreement based on industry standards of test 21 voltages for that theoretical 4-mil cable at 3600 or whatever 22 those numbers were. We tested considerably.above that and 23 broke down, which gives us a pain in'that we have to understand j 24' those anomalies of breakdown. 25 But if we looked at that and explained those l Heritage Reporting Corporation (202) 628-4888 i
50 1 qualification level to get us up, allow us to set a methodical 2 rath towards maybe long-term qualification and decide what we 3 want to do. Do we want to make a change in cable types, do we 4 want to stick with the same thing, do we want to requalify? We 5 thought this gave us the greatest number of options. 6 MR. FOX: Ms. Axelrad, your point it well taken. We 7 have not qualified this cable for beyond ten years. If we want 8 to run the plant, say we've got a minimum thickness of eight. 9 miles, then we have to do some additional qualification tests. 10 We did it for ten, and I'll be maybe a little more 11
- frank, wa also wanted to get this test work done in time to 12 minimize the down time in the sound.
13 It just looked like a pacing item. And to go ahead 14 and run the additional 30 years worth of life or so would have 15 taken more time. We just wanted to get to the bottom line and 16 show that we were safe for restarts. That's why we terminated 17 at ten years. 18 MR. GRIDLEY: Kent? 19 MR. BROWN Yes. 20 MR. GRIDLEY: If you're making a transition, we ought 21 to take a short break. 22 MR. BROWN: Yes, this would be a. good time, because 23 starting here I'll begin to talk about the Wyle test program 24 itself and get a little more detail on the results. 25 MR. ZECK: Let's try to get back in ten, 15 minutes Heritage Reporting Corporation (202) 628-4888
51 'I at.the most. 2-(Whereupon, a brief. recess was taken.) 3 MR. ZECK: Why-don't we go in and resume. Stew will-4 . join us I'm sure as soon as he can l 5: (Slide) 6 MR. BROWN: Where we left off, of course, was that 7 TVA h ad decided that we were going-to intentionally shave down 8-cable walls and expose them to qualification tests. 9 ~After much trial and error, we did work out a method 10 by which we could shave the cables to what we thought was a 11 reasonably consistent remaining wall thickness. .I'11'tell you 12 right off that we recognized it was a difficult thing:to do. 13. When you start shaving into.something, it's hard to-14 say how much is left. You can say how much you took off, but 15 if you're.not,100 percent sure that you started with say -16 exactly 45 mils, when you've taken off 10 or 20, you're not 17 sure how much is left. 18 There is a tolerance in the manufacturing process. 19 You may specify 45 and get 45 to 50 or 52. So when you've 20 shaved off 20 mils, maybe you've got a four mil window there. 21 So we did work out a method that we thought was 22' fairly consistent. So we prepared the samples at out -23 Chickamauga labs. We engaged Wyle Laboratories to work with us 24 in this qualification program. 25 The cables being silicone rubber insulation have a Heritage Reporting Corporation (202) 628-4888 x _
L 52 l 1 temperature rating of 125 degrees heat. As I mentioned, their 2 application is such that there is on ohmic heating. They are l 3 using controlled circuit applications where there is no current 4 flowing essentially, and they are used in low voltage power 5 applications that do not see service until after the onset of 6 some accident event. 7 So for the course of the normal service life until l 8 the day that the accident occurs, they are sitting there 9 basically at ambient temperature. 10 Given that ambient, within our containment structure 11 as Sequoyah, is a maximum of about 49 degrees C, and given that 12 the cables are rate for 125 degrees C, if you plod through the 13 erroneous calculations, you will find that ten years worth of. 14 normal service at 60 degrees C does effectively zero aging to 15 the cable insulation systems. 16 Consequently, no thermal aging was applied prior to 17 the accident itself. We did, of course, apply all of the 18 requisite post-loca aging on both the control and the power 19 cable applications. 20 We did, however, apply the full, ten year normal dose 21 of radiation along with the accident dose and the stipulated 22 accident radiation dose margin so that the. total dose that we 23 applied to encompass our ten year goal was slightly less than 24 one times ten to the eighth rads. 25 We did a ten day loca exposure. We were simulating a Heritage Reporting Corporation (202) 628-4888
q l l 53 1 100 day loca event, but in order to get that effort over i 1 2 quickly, we did an accepted practice, we accelerated the event 3 and those of you who are familiar with the qualification 4 process know that this is very much the norm, to accelerate 1 5 that event. 6 There is a risk involved to the user, because you're ) 7 going to be driving that equipment at higher and higher 8 temperatures for this period of time, but we regarded that for 9 the materials that we were testing, for the ten-day excursion 10 that we proposed, that this was acceptable risk on our part. 11 We exposed the cables to chem spray. I might mention 12 that in normal service, all of these cables are in conduit. I 13 wouldn't by any means say that the conduits are 100 percent 14 leak tight, but certainly the conduits inhibit the egress 15 either of water or chem spray. In the course of doing these 16 tests, though, we put the cables in the chamber on a flat tray 17 with direct chem spray exposure on the insulation. Some of 18 these, in order to do the shaving, we had to remove the jacket, 19 so the jacket was out of the way, and the chem spray was 20 directly on the insulation itself. 21 The acceptance criteria that we established was to 22 hold rated voltage and current throughout the loca excursion. 23 We wanted to prove that the cables could perform their intended 24 function, that they could do what they're supposed to do. We 25 took insulation resistance measurements wet. We took them dry. Heritage Reporting Corporation (202) 628-4888
) 54 1 We took them before the loca, we took them after the loca. 2 Those of you agcin who are fmmiliar with EQ testing 3 will know that that is a little bit beyond the norm. 4 We did post-loca hipot tests both wet and dry. 5 Normally, you do them wet. As I am going to show you in a 6 minute, we did one horrible program to the cables, very, very 7 conservative in terms of the way that we did our post-loca 8 hipot, as we were trying to demonstrate the degree of margin 9 that we had remaining, even fs11owing all the abuse. And 10 finally, we took the cables to a wet DC breakdown. 11 (Slide) 12 MR. BROWN: Just as a quickie, there is a slide in 13 your package there that shows the profile excursion for the l 14 ten-day event. We picked out at 342 degrees F. Again, I will l 15 remind you that we were loading the cables through the 16 excursion so that the internal heating was provided by the 17 current itself. 18 (Slide) 19 MR. BROWN: After the tests, in recognition of the i 20 fact that there was a bit of, if you want to call it research, 21 that's going on, that we were trying to really establish, and j 22 remember one of the two priorities we had set, we said perhaps L l 23 we could establish a new minimum qualification thickness. l 1 24 Perhaps that could be used in establishing what a reasonable 25 test voltage might be in an application like this. Heritage Reporting Corporation (202) 628-4888 i i --_--__-----_J
55 1 The other prime consideration was to establish 2 whether or not these reduced walls were of any significance. 3 We established our post-loca hipots in a step 4 fashion.- At any point in that profile, going from 1,000 to 1 5 4,700 volts, at any point, if we. encountered a voltage that we 6-could correlate, either to an industry standard or to a 7 commonly accepted dielectric exposure, if you will, we stayed And I'll e T ain that as I go into this just l r for five minuten. 9 a bit. 10 But in the course of going from one plateau that I 11 say relates to some industry standard to another plateau, if 12 there were a large gap or wide band in between, we played it 13 safe and we stopped at an intermediate plateau. 14 Let's put some numbers in all this or I'm going to 15 confuse everybody. 16 For instance, in going from 500 volts DC, which is a 17 typical voltage that you would use in a megger application, we. 18 stayed there for five minutes. But in going to our next 19 five-minute plateau, and I'll explain what's behind that 20 shortly, which was 3,00 volts, said there was a wide band in 21 there. And we want to gain as much information as we can. So 22 we don't want to just be increasing the voltage and presto, we 23 break down at 2,600 volts, and the question is, could you have 24 stayed at 1,000, could you have stayed at 2,000? We're left 25 with an answer we can't provide. l Heritage Reporting Corporation (202) 628-4888 i.__ - -.___.. - - _ _ _ _. -. - _. _.. - _ -. __._.-....__-.-___.O
56 1 So we established a couple of plateaus on the line. 2 We said that we will take the cable and go to 500 volts CC, .3 correlates to.a megger. Then we go to 1,000 volts, stay there 4 for a minute. No particular correlation there. Go to 2,000 5 volts, stay there for a minute. No particular correlation. At 6 3,000 volts, if you remember our July 13 presentation, we were 7 working on coming up with test voltages that we believed were a 8 function of the' application'of that particular cable. 9 We said that it matters how you should test a cable, 10 according to what its service is. In the case of control 11 circuit applications, back in July we d'erived a test voltage 12 which was a function of twice that rated voltage of that 13 system, plus 1,000, time's three -- the three is a multiplier 14 conversion from AC to DC -- times.8 to correlate with the fact 15 that we're doing field testing rather than factory testing. 16 I think all of you who have been in engineering for a 17 while are familiar with the old rule of thumb on dielectric 18 testing of twice rated plus 1,000. 19 So we.had taken that approach and said for control 20 circuit applications, we will test those cables at-3,000 volts, 21 and because'twice rated plus 1,000 was a recognized normal i 22 practice, we'll stay there for five minutes. 23 Another intermediate step then was one minute at 1 24 4,000 volts. 25 4,700 volts is a similar derivation of the test Heritage Reporting Corporation (202) 628-4888 \\
k 57 l' voltage based on the application for cables that are in 480 2 volt control circuit usage. And so we stayed'there for five 3 minutes as well. 4 If you are not familiar with EQ testing, what you are 5 going to see if you go to anybody's environmental qualification 6 file and you look at any test program that I have ever seen 7 that has been completed, if it has been tested under recent 8 standards, you will see a single, five-minute test at 240 volts 9 per mil, or 80 volts per mil IC, depending on whether it was AC 10 or DC. 11 So we included an additional step at 240 volts per 12 mil DC. We based that on the so-called measured insulation 13 thickness, and we remained at this level for five minutes. So 14 we encompassed general industry practice. That's twice rated 15 plus 1,000, and we encompassed a derivation from 383. 16 The way that that might work, and I'll jump the gun 17 just a little bit, if we measured cable and said we had four 18 mils, 240 tim four is 960. So we would stay five minutes at 19 500 volts DC and then rather than doing a step at 960 and a 1 20 step at 1,000, we did a single step at the 240 volts per mil 21 times four, or 960, and then we went to 2,000 for a minute, 22 3,000 for five minutes, et cetera, et cetera. l 23 So instead of the usual five-minute exposure, we got j 24 18 minutes as a minimum of dielectric exposure. Now, that's 25 wet. Heritage Reporting Corporation (202) 628-4888
q l l 58 1 Then we turned-around, took the cables out of the 2. water. We dried them. We actually.put them in an oven,-dried! 3 them,' and with the cables on a mandrel, we'd done the mandrel 4 rebend, with the cables on a mandrel, then we repeated the same 5 program, in order to get more information on the wet versus' 6. dry. 7 MR. LIAW: Kent? 8 MR. BROWN: Yes. 9 MR. LIAW: Each step along the way, did you maintain 10 the tungsten brand rate in termsLof voltage? 11. MR. BROWN: Yes, we did. As.we increased from 3,000 12 to 4,0~00 volts we went up at 500 volts per second~. 13 MR. LIAW: All the way? l 14 MR. BROWN: Yes, sir. l l \\ 15 MR. LIAW: 500 volts per second. 16 MR. BROWN: Yes, sir. L 17 (Slide) 18 MR. BROWN: As you remember, I said that we had shot 19 for nominal thicknesses of eight, 15, 20, 25 and in one case, 20 30 mils. Between the time that we did our sample preparation 21-and the end of the loca, our laboratories in Chattanooga had 22 developed an eddy current technique in order to actually 23 measure what we ended up with. Because, as I said, there was a 1 24 tolerance that was a result of the very tolerance on the 25 insulation itself, we never knew, or said we wouldn't, if we l Heritage Reporting Corporation (202) 628-4888 1
/ ] 59 were targeting eight mils, whether we had ten or whether we had-l 2 six. 3 So we. developed an eddy current technique which we-4 could pass along that shaved surface of the cable, and 5-determine-by the' readout of that particular device what the 6 actual, minimum remaining thickness was. 7 'I just might' note in passing that this same 8 technology is used:for measuring paint thicknesses or coating 9 thicknesses, on any conductive surface. 10 (Slide) 11 MR. BROWN: Here are the specimens that we put into 12 this exposure,.and we can correlate these in~a minute.then with-13 the test voltages that we saw. 14 .With AIW, where we were targeting for.eight mils, we 15 actually ended up with two mils. Now, you'll notice there's a 16. tolerance on.that last column that says plus-minus two mils. -17 The. number that I'm giving you here is.what we think it is, and 18 we can apply the tolerance to that. k. 19 where we were shooting or targeting for 15, we ended 20 up with four. And you can see on down the chart there. 21 Now, this w&s quite a surprise to us at first. And-22 what we found upon challenging our Central Labs people was that 23 in the process of cutting the silicone rubber, if you have i4 handled the stuff,.you'll know that it gives very easily. l 25 Consequently, as you try to cut it or shave it, it also tends Heritage Reporting Corporation (202) 628-4888
60 1 to bunch up or tear as you try to cut it out, so that in the 2 course of doing any of the shaving, you could always identify 3 low spots that were significantly low. 4 When we exposed them to this eddy current measuring 5 technique, what we found was that we had significantly undercut 6 that which we had anticipated. 7 Quite honestly, in retrospect, had we known that we 8 were going to get insulation walls as thin as these, we would 9 probably box the samples up, send them back to our lab and tell 10 them to try again. 11 So what I'm going to present to you is just what 12 happened when we did actually test what we had already run 13 through the loca. 14 One clarification. Anaconda sample one. Our target 15 was for eight mils. When that sample, I should note that on 16 any given sample which was ten to 15 feet long, even though all 17 we found in the plant was highly localized, basically point 18 damage, we shaved three sections on each cable approximately 19 three inches l'ong in each section, down to the estimates 20 reduced wall. 21 Our target was to intensely degrade three separate 22 areas down to that wall. 23 On Anaconda sample number one, when the lab came and 24 measured that sample, they could basically tell us that we just I 25 had something greater than zero. The insulation was so thin it j l Heritage Reporting Corporation (202) 628-4888 s
f 9 61 1 was not within the resolution of their particular measuring i 2 technique. 3 After we did the loca, when we did the wet hipot, 4 this sample which had functioned satisfactorily throughout its 5 loca exposure had carried rated current, rated voltage 6 throughout that loca. When we did the wet hipot, we had 7 essentially zero insulation resistance at that point in time. 8 That was a sample that turned out to be basically 9 greater than zero mils. 10 I'm going to pass around a picture that will show you 11 Anaconda sample one. It will be the sample at the top of the 12 picture. If you will take a look i t, you will see that you 13 can practically see the conductor. The insulation is very, 14 very thin. And I'll let that go around while I talk. 15 ('Ficture passed to audience) 16 MR. BROWN: Following the wet insulation resistance i 17 tests, where we had zero insulation resistance, we looked at 18 that particular segment that has probably a mil or lesa. We 19 found a very small pinhole. We cut that particular section out I 20 and continued the test then with the remaining, still six 21 inches worth of remaining damage with minimum insulation 22 thickness on each of those segments of approximately two mils. I 23 (Slide) I 24 MR. BROWN: Now, each and every one of those samples ] 25 that we had, having now removed the segment that had something f Heritage Reporting Corporation (202) 628-4888 l
c 62 1 less than a mil, each and every one of those segments was 2 exposed to the full post-loca dielectric program, both wet and 3 dry. 4 Five minutes at 500, 1,000 volts for one minute, 5-2,000 volts for a minute, 3,000 volts for five minutes, 4,000 6 volts for one minute, 4,700 volts for five minutes, both wet 7 and dry, following a mandrel rebend test. And not one of them 8 failed. 9 I think this has got to be a significant point in 10 that it has to tell you a world about this particular 11 insulation system and its durability and its resistance to age. 12 MR. BROWN: What we are showing on this particular 13 slide is the results of the breakdown test that we did. We 14 exposed the cables to a fairly quick-rise test, 500 volts DC 15 per second, when the cables were submerged in water, until 16 breakdown. 17 And as you'll see, the breakdowns ranged as low as 18 12,000 volts, and there were several, those four samples that 19 are indicated there with the caret, that did not break down due 20 to the limitations of Wyle's test equipment. Approximately 60 21 KV was as high as they could go. 22 After the first sample or two, we were rather 23 incredulous. We sent the stuff back to the cal. lab and told 24 them to recalibrates it. We didn't quite honestly believe what 25 we were seeing. Heritage Reporti'ng Corporation (202) 628-4888 Ja--._n-_--_
t, t I 63 1 But the equipment checked out as being perfectly 2 within calibration, so we are confident and Wyle is confident 3 and.willing to stand behind us on the results we see. 4 Furthermore, even though it causes confusion 1 5 initially, here shortly I will get into a bit of the 6 explanation on why we're seeing dielectric levels as high as we 7 are. ,I think this indeed gives a degree of confidence that 8 we've not had before, that you can take a cable and expose it 9 to the kind of program that we have done, with shaving them 10 down to levele less than a tenth of what you intended to have 11 on that' cable, and that can go through the entire loca exposure 12 and still not only pass the loca, but be able to withstand 13 voltages for five minutes in accordance with the program that 14 I've told you, and then go to breakdown levels such as these. 15 Let's talk about breakdowns for a minute, now. 16 (Continued on next page.) 17 18 19 20 21 22 23 24 25 Heritage Reporting Corporation (202) 628-4888
l 1-MR.' BROWN: Probably the most confusing point =for '21 the'last few months'has been that of discussion over what is an l L3-acceptable test and what level should we apply, and should it h 4 be'300Lvolts per mil or should it 240 volts per mil,.or should 5 it'be some' function of nominal mil, or should it be some 6- -function of a. reduced insulation thickness according to a 7 minimum EQ thickness. You know, a lot of' discussion in that 8-regard. 4 9 We have talked intrinsic dielectric strength, and we 10 have talked typical..We have. shot the whole gamut'in 11 discussing this arena. Just recently, it was brought to our 12 attention that there was an ASTM or actually multiple ASTM 13 standards which deal wi'th the subject of dielectric testing in 14 general. 15 Typically, we had gone back to the application 16. standards, those of IEEE and IECA,-and looked at the guidance 17 that they provided there. But they do not really talk in a 18 tutorial fashion. They do not really give you the background 19 of what is behind your dielectric testing. And when we went to 20 -the materials standards in the ASTMs, I think that they gave us 21 some information which will help us to understand the great 22 variety that=we were seeing when we were testing. Sometimes in 23 water and sometimes in air, and when we were testing very thin 24 insulations and very thick insulations, we were seeing what 125 appeared to be a whole scatter of data, and I hope that this Heritage Reporting Corporation (202) 628-4888
65 ) I will help. 2 (Slide.) 3 The standards that I am referring to are 4 ASTM D-149 which deals with AC breakdown tests, and 5 ASTM D-3755. The significant parts for our discussion today 6 are the appendices of both of these standards. Even though we 7 did not do AC testing, there is a lot of information in the 8 appendix of the D-149 which is applicable to DC testing. In 9 fact, the 3755 standard refers you back to the D-149 for 10 basically a tutorial portion of that text within Appendix 11 X-1. 12 If you go to Appendix X-1, what you see is that there 13 are three postulated mechanisms of dielectric breakdown, which 14 you have got to consider when you talk about breakdown of any 15 particular material. The standard says tnat breakdown could be 16 caused by electrical discharges, partial discharges or corona. 17 Second, the standard points out that breakdown could 18 be the result of basically a thermal instability or a thermal 19 run-away. As current passed through a concentrated area within 20 the insulation when you are under a dielectric test, as that 21 current passes, the insulation at that point begins to heat up. 22 And if it heats up rapidly enough, you go to a thermal run-away 23 condition, and you will have dielectric breakdown. 24 The third value which has been kicked around quite a 25 bit I guess is that of intrinsic breakdown, and it is the Heritage Reporting Corporation (202) 628-4888
66 4 1 highest of all three values, and it is the value with which we 2 have often been compared. Intrinsic breakdo'wn would be if you 3 could eliminate I guess the first two effects completely. If 4 you had a nice pure material and you had no anomalous field 5 effects, and no stress concentrations, and no thermal 6 breakdown, and no corona whatsoever, and no surface discharges, 7 you would find the intrinsic or base value, if you will, for 8 the material itself. 9 As I noted, the testing that we have been doing often 10 has been compared to the intrinsic breakdown values that you 11 might find published. 12 MR. MARINOS: It is because we are doing DC tests, 13 right, and the other two values will not contribute 14 significantly to that? 15 MR. BROWN: The standard though it does_most 16 specifically reference those phenomena, it will tell you that 17 they are of a lesser effect for DC than they would be for AC. 18 Next we say that dielectric strength -- 19 MR. LIAW: Ken, let me read something to you. 20 D-3755, Appendix Section N-1.214. "Since thermal breakdown and 21 breakdown caused by electrical discharges are less important in 22 direct voltage breakdown than alternating voltage breakdown, 23 the direct voltage breakdown is more like to approach the 24 theoretical maximum breakdown voltage." 25 I suppose that is intrinsic breakdown voltage, right? Heritage Reporting Corporation (202) 628-4888
67 1 MR. BROWN: Right. 2 MR. LIAW: I interpret what this is'saying is.that 3 for DC tests that'the first two phenomena are really not a 4 -factor. 5 MR. BROWN: I think that you just agreed with what I 6. said, that they are certainly of reduced effect in DC breakdown 7 testing. I do not think that it is quite accurate or fair to 8 say that they are of no effect. Otherwise, we would not see 9 them in the standard at all. 10 MR. LIAW: I am not saying that. 11 MR. BROWN: We agree that they are of reduced effect 12 within the DC test. 13 MR. LIAW: Let me pursue another point based on what l l 14 you presented earlier. Charlie in his opening remarks talked 15 about the functional relationship between the thickness and the l 16 breakdown voltage. And by looking at the number that you just l-17 presented, I do not ses any functional relationship there. 18 That is why I ran my calculator during lunchtime, and I did not 1 19 see anything. l l 20 I am not saying therefore there is no such thing. I 21 am not saying that. I am simply saying that either the minimum 22 major thickness has such a large error band.that makes all of I the numbers that you have there difficult to interpret. l 23 i .24 MR. BROWN: Okay. The first thing that I would 25 like -- Heritage Report ing Corporation (202) 628-4888 l
68 1-MR. LIAW: Let me pose another question. 2 Have you attempted to plot this curve to see whether 3 or not there is a certain functional relationship as indicated 4 in the D-149? 5 MR. BROWN: No, I have not. ,6 MR. LIAW: Why not; the major point that you_are 7 trying to make, why did not try to do that? 8 MR. BROWN: Well, first of aII, I think that you have 9 got to realize that you are dealing with a very, very limited 10 data set. 11 MR. LIAW: Okay. We are dealing with a very, very 12 limited data set, and I think that everybody ought to bear that 13 in mind. 14 MR. BROWN: What we are talking about is on the one 15 hand we have theoretical knowledge which is now the industry 16 accepted understanding of how insulation systems work when 'l 17 subjected to high potential DC breakdown. This is the best of 4 18 the best understanding that the phenomenon occurs in that 19 fashion. 20 Now what I will also tell you is that there are a lot i 21 of factors that may influence dielectric breakdown of any 22-particular device, any particular slab of insulation. If I 23 take a slab of insulation and it is uniformly 4 mils thick. .t 24 Well, let's make it even more difficult. If I take it and it 25 is 4 mils and I step up to 6 mils, I am not going to guarantee Heritage Reporting Corporation (202) 628-4888
69 i l' you that every time that the 4 mils part will break down. 2 Because there are going to be areas within the 6 mils area that l 3 have a contaminant or they are non-homogeneous in some respect. 4 So if you are only looking at four data points, I am j 3 really hesitant to draw a great conclusion from three or four 6 data-points. 7 MR. LIAW: That is right. 8 MR. BROWN: What we have decided to do is rather than 9 drawing upon three or four data points is say let's go back to 10 fundamentals and let's engineer this thing. To be honest with j 11 you, we probably' plotted anything that you can think of along i 12 the way just to see if we can find relationships. 13 But here I want to say forget four data points and 14 trying to make great theory out of that, and let's go with the 15 theory that already exists within the science, and let's see if 16 we can gain some insight into how this worked. What I am a 17 trying to do is that I know that there is a lot of confusion. 18 MR. LIAW: You can stop there, Ken. 19 MR. BROWN: Yes. 20 MR. LIAW: I guess that what I am looking for is I 1 21 understand the basic phenomena, and I understand unat you are 22 trying to say. And in a sense maybe I am jumping the gun, but 23 I guess that I am looking for to you to see whether or not 24 with all of this qualitative argument or the theory, the 25 Petitioner's theory, whether or not you can establish some kind Heritage Reporting Corporation 4 (202) 628-4888 l 1
J l [ 70 1-of functional relationship upon which we can make a logical 2 ' conclusion, and that is what I-am looking for. You can tell me-l 3 now or you can tell me later. 4' MR. FOX:.That may-be a subject of a R&D program and 5 a rather lengthy one, B.D. What we are saying is that based on ~ 6 the standards that.we can make qualitative arguments to explain 7 some of the phenomena that we have seen. 8 MR. LIAW: Are you saying, Charlie, that you are 9. asking us to make a licensing judgment based on some kind of 10 qualitative data? 11 MR. FOX: What I want you to make your judgment:on is 12 pure and simple. We have beyond a shadow of a doubt shown that 13 cables with very small wall thicknesses will in fact pass a 14-LOCA. We can do fa'n dances forever around these voltage 15 anomalies and maybe never explain them. But what is important. 16 is that the cables that we have tested with wall thicknesses 17 down to 4 mils will in fact pass a LOCA. That is the 18 background of the EQ testing period. 19 MR. LIAW: To go cne step further, how are you going 20 to demonstrate to me that the cable is at this 4 mils 21 thickness? 22 MR. FOX: Every anomaly that we have observed in our 23 test program to date, we have clearly shown that we have a 24 minimum of 4 mils. In fact, we have not seen an indication of 25 a minimum wall thickness of less than 8 mils, right? Heritage Reporting Corporation (202) 628-4888
't ' k 71 1 .MR. LIAWz I am asking this question here. I am a i 2 questioning.the accuracy of the measurements.- Because you are ] 3 going to use that to say:that you have a certain number of mils ~ 4' or a certain amount of thickness, and I guess that I have to 5 see that. 6 MR '. FOX: Okay. Well, part of that is contained in 7 the Yukon report,.and part of it is' contained in'the 8 information that we will present to-you in just a few moments. 9 That is still coming in the. presentation on the post-mortems of 10 the cables that we-have just recently done, the five additional 11-cables. 12 -MR. LIAW: The post-mortem tests of the cables. 13 MR. FOX: The post-mortems of failed cables. 14 MR. LIAW: That you removed? 15 MR. FOX: Yes, and that is still to come in'this 16-presentation.. 17 MR. LIAW: Okay. 18-MR. ZECK: I do not think that you have told us the i 19 final position yet that you intend to state here, have you; I 20 do not think that you just paraphrased it, did you, or is there-21 more to it? 22 MR. FOX: No. 23 MR. LIAW: That is what I say. I may have jumped the 24
- gun, 25 MR. FOX:
Let's move on. 1 Heritage Reporting Corporation (202) 628-4888
a 72 1 MR. BROWN: Okay. All right. So number one, we 2 recognize that looking at the intrinsic breakdown strength 3 alone may not be the appropriate thing to do. Second, we found 4 in looking at the standards that dielectric strength is greatly 5 dependent upon specimen thickness. I think that as we all know 6 that the program has been driven by the statement in IEEE 383 7 that says that when you are doing a qualification test that you 8 apply 240 volts per mil. 9 So we were making a linear assumption of the 10 relationship between insulation thickness or dielectric 11 integrity or breakdown. And we had been applying this 12 uniformly. You had 45 mils, and it was 240 times 45. You had 13 30 and so on. 14 What this does is that it places a penalty or a 15 stress on thicker insulation systems. Because in fact the 16 standard says that it is not a linear function, but rather it 17 varies inversely as a fractional power of the specimen 18 thickness. Let me be more specific. For solid insulation 19 systems, it varies inversely as the square of the thickness. 20 MR. FOX: The square root. 21 MR. BROWN: Excuse me, thank you. The square root of 22 the thickness. 23 MR. LIAW: One thing, Charlie. The fraction, the 24 material, it is the square root -- 25 MR. FOX: One over the square root. Heritage Reporting Corporation (202) 628-4888 a
1 'j. - 4 73 1: MR.' BROWN: 'One over the square root.- Let me put 2 -some numbers.to it,.so that you understand what I am saying.. .3-MR. LIAW: 'No,-again let me read that-to you. He is: 4' correct. :!It11s'not the square root, but it-is.the fraction, 5 therinverse1 proportion ofLthe. fraction. 6 MR.' FOX:. One over.the square root. 7 MR. LIAW - You say(for a'certain type =of material: 8 that it is clos'e'to one over the square root. 9 MR. FOX: Yes. - 10 MR. LIAW:. Okay.- 11' MR. BROWN: Yes..And I believe that if.you look at.- . 12 fthe words that-it specifically says for solid dielectrics. And-13 of.-course,.that is what~wd-are' dealing.with. We are not 14-dealing with paper,-or oil, or whatever. Now let's put some 15 numbers to that. 16 MR. MARINOS: Have you made a determination that-17 silicone rubber-falls.in'that category of that definition here?~ '18 MR. BROWN: It is definitely 1a solid and not an oil, 19 or'a laminate, or'a gas. 20 MR. MARINOS: It is a solid, you say? 21 MR. BROWN: Right. 22 MR. FOX: 'Would anyone from the cable manufacturers 23-like to comment?. 24
- MR. BROWN
You will have to stand.up, please. - 25 MR. FOX: This is Bob Gehm, and you need to come to a Heritage Reporting Corporation (202) 628-4888 -.u
e-I 74 1 microphone, from Rockbestos. 2 MR. GEHM: Silicone rubber can always be considered 3 as a solid dielectric as characterized by polyethylene, 4 ethylene, propylene, and so on. 5 MR. MARINOS: Are you familiar with this standard 6 that Kent is presenting, the ASTM standard? 7 MR. GERM: Not that particular standard, but there 8 are some other factors that he has not even gotten into which 9 make that even more conservative. 10 MR. BROWN: Okay, thank you. Let's put some numbers 11 towards this, and just see how this thing follows. What we are 12 saying is that if I double the insulation thickness, or let's 13 go a little bit higher, if I quadrnple the insulation 14 thickness, I go up four times as thick, and my dielectric 15 strength is not four times as great, but it is twice as great. 16 Now on a per mil basis, then you say per mil that 17 each has not four times the dielectric strength per mil, but 18 only half the dielectric strength per mil. So if I go up, by 19 quadrupling the insulation thickness, but if I say that the 20 dielectric strength is half as much per mil, I end up with an 21 overall dielectric strength of twice, not four times. 22 Similarly if I go up by a factor of nine, I do not 23 end up, just because the cable is nine times thicker, I do not 24 have nine times the dielectric strength, but I have three times 25 the dielectric strength. And the insulation strength per mil Heritage Reporting Corporation (202) 628-4888 i
75 1 is a third for that thicker one as it is for the thinner one. 2 So again let me say that this would then, if you were 3 following a program that assumed linearity of dielectric 4 strength with insulation thickness, you would place a greater 5 stress on thicker insulation systems such as silicone rubber in 6 the way that we buy silicone rubber as nominally with 45. 7 MR. MARINOS: This relationship again is predicated 8 on the three factors that you identified at the beginning of 9 your slide, meaning the discharge breakdown, the thermal 10 breakdown, and the intrinsic breakdown contributing 11 synergistically, is that correct, in arriving at this ratio of 12 one over the square root. 13 MR. BROWN: I believe that the correct answer to that 14 is no. What I am saying is that if I have an insulation 15 system, and let's talk just intrinsic breakdown and forget the 16 others for a moment. 17 MR. MARINOS: You are saying that intrinsic would 18 also exhibit the same properties? 19 MR. BROWN: It is going to be a function of the 20 thickness of the material, okay, in accordance with the inverse 21 reciprocity. j l 22 MR. LIAW: Maybe you have a better memory. As I 23 recall, when we discussed 240 per mil DC back in March, I do 24 not believe that anybody thought about whether or not anybody j 25 assumed anything. As I recall, it was something that you l l Heritage Reporting Corporation (202) 628-4888 i l l (
76 1 proposed on the 383 as an appropriate level to test. 2 MR. BROWN: I will let someone who is on those March 3 meetings reply to that. 4 MR. FOX: On the basis of the pool of knowledge that 5 we had at that point in time, you are absolutely right. We 6' were not aware of the one over the square root of X 7 relationship. In fact, I became aware of it last Friday, okay, i 8 So we were assuming the linear 240 volt per mil number here. 9 MR. LIAW: We are here trying to make an important 10 licensing decision based on knowledge that you and I just 11 acquired last Friday, I guess. 12 MR. FOX: We should always be strong enough to accept 13 the latest technical information, yes. I would like Ted 14 Balaska to come and make a short statement on the applicability 15 of the ASTM standards on dielectric strength to our cable. Ted 16 is the curront Chairman of the IEEE insulated cable conductors 17 committee. 18 MR. BALASKA: I think that for the record to i 19 introduce myself that I will present to the Commission a little 20 short biography. You will notice that that is an excerpt from 21 a program on high voltage DC testing which I have coordinated 22 for field testing of power cables in the field. So I will not i 23 go into the details of the background. It is there for the 24 record. 25 But I think that it is appropriate that we talk about L Heritage Reporting Corporation l (202) 628-4888 w_.
h Lo-4 77 1 -intrinsic breakdown strength, the-breakdown' strength.of the. 2 materials. Very commonly, people look at breakdown strength of 3-the material, and they'are' going-to take a'look at material 4 " material". And one quick way of doing it is to follow an ASTM 5 . procedure, and what is under discussion right now is 3755-79, 6 . which is dielectric breakdown' voltage'and dielectric strength 7 of solid electrical insulating materials under direct voltage 8' stress. 9 What is important that.I think that.we should-10 recognize is what is'in the appendix indicating what:the-11' limitations are of this' test method. There-are many factors 12 involved. One of them is electrodes, the electrode area, the 13 electrode geometry, and the electrode material. 14 Generally speaking, decreasing ~ volts per mil unit 15 breakd'own strength comes along with increasing the area being 16 subjected to this stress. The electrode geometry. Are you ) i 17 talking about-flat plate electrodes or angular plate 18 electrodes. And you run into this problem of trying to relate 19 these unit breakdown strengths when you get into a piece of 20 cable, because-in a piece of cable we do not'have the same 21 electrode configuration as we have in ASTM. We have coaxial 22 electrode configurations. 23. The other factor that we take a look at is specimen 24 thickness, and this is where apparently there is a hang-up -25 occurring here. And it is generally accepted that unit Heritage Reporting Corporation (202) 628-4888 j i 1
78 1 breakdown strength, volts per mil, is inversely proportional to -2 specimen thickness. 3 To put it in twenty-five cent words, thicker material 4 has lower volts per mil than thinner material. So therefore, 5 it is= inversely proportional to a reciprocal of the square root 6 of the thickness. 7 MR..LIAW: Excuse me, sir. 8 MR. BALASKA: Yes. 9 MR. LIAW: Are you talking about an average or I 10. talking about a total? 11 MR. BALASKA: I am talking about the ASTM which is an 12 average stress, because it.is generally parallel plate-13 electrodes. 14 MR. EBNETER: B.D., he is talking about an 15 incremental' voltage breakdown, which is less incrementally for 16 a large block, but the large-block has more increments. So the 17 total breakdown of voltage will still be greater. 18 MR. LIAW: Yes, yes. That is the reason that I 19 asking whether we are obviously talking about an average per 20 unit thickness. 21 MR. BALASKA: I am talking about unit breakdown 22 strength, volts per all, okay. 23-MR. LIAW: Right. 24 MR. BALASKA: And this is why I keep referring to 25 unit breakdown strength and volts per mil and not total Heritage Reporting Corporation (202) 628-4888
79 1 breakdown, okay. The other factor of influence is temperature 2 of the specimen under test and the surrounding medium. It is 3 generally accepted and has been proven that decreased volts per 4 mil or decreased unit breakdown strength comes along with 5 increased temperature. 6 Another factor is time. The rate of rise of the 7 applied voltage. The faster the rate of rise, the greater the 8 unit breakdown strength as opposed to a slower rate of rise. 9 MR. MARINOS: Is that what affects the temperature 10 factor? l 11 MR. BALASKA: That will affect temperature, correct, 12 how quickly you bring it up; yes, sir. I 13 MR. GOODWIN: Ed Goodwin of the NRC staff. Going l l 14 .back just a bit to what you were discussing earlier, which is l 15 the inverse relationship with thickness. Clearly, such an I 16 effect exists. It is recognized in the standard I L 17 specifications, the ICEA and the EEI. But if you examine those 18 closely, it appears to be recognized as only important for 19 thicknesses of insulation in excess of 75 or 80. And for the 20 sorts of insulation that we are talking about here that the 21 greatest thickness of which is 60 and most of which are 30, the l 22 ICEA standards reflect a standard 300 volts.per mil DC and make 23 no allowance for the effect that you are talking about. They 24 seem to treat as trivial for thinner cables. Now I presume 25 that you were on those ICEA committees. Heritage Reporting Corporation (202) 628-4888
1 80 1 And if the effect is so important, why is it ignored 2 for thinner insulations? 3 MR. BALASKA: The 300 volta per mil is not intended i 4 to be a test for breakdown strength. It is intended to be a 5 go-no go test to verify the adequacy of the construction for 6 its intended purpose. 7 MR. LIAW: I do not know whether you are aware. I i 8 thought that was the intent of the test. 9 Are you aware of that? 10 MR. BALASKA: That is not a design test, the 300 11 volts per mil? 12 MR. LIAW: I am talking about the original purpose 13 that TVA wants to do, to try to prove that. Because there were 14 concerns raised regarding the insulation practice. And as a 15 result of that, TVA proposed to have this test to demonstrate 16 that those cables were properly installed. 17 Now are you aware of this fact? 18 MR. BALASKA: Yes, I am. 19 MR. LIAW: Okay. 20 MR. BALASKA: But what I am addressing are the 21 factors that affect breakdown strength. 22 MR. LIAW: So that is pushing it back to the 320. 23 MR. BALASKA: Right. 24 MR. LIAW: That you were talking about, the 25 go-no go. In fact, that was the original criterion used and ~ Heritage Reporting Corporation (202) 628-4888
L 81 ,1 accepted byythe NRC.in the beginning. 2 MR.-BALASKA: But in essence,-that 300 volts per mil 3 .that'was-just discussed -- 4 MR. EBNETER: Well, let's stop here. Let's make an 5 assumption that'everything that you have said is valid at this 6 point, and let's'get through the rest of the presentation. .7 MR. BALASKA: Okay. 8 MR. EBNETER: Let's see what the bottom line is, and l 9' -then we will come back and argue as to what we think. 10 MR. BALASKA: Okay. The other factor that influences 11 breakdown strength and is insignificant in this discussion is 12 wave form. We are talking about a unit of directional voltage 13 here with essentially very little ripple. 14 Also the other factor is surrounding' medium.of the 15 test, the specimen, because this influences heat transfer which 16 was touched upon very briefly. Surface discharges which can l 17 erode material, and heat uniformity and heat dissipation. l 18' Another aspect is relative humidity. In other words, what 19 moisture absorption characteristics does the material have. 20 That influences its dielectric loss and its surface 21 resistivity. 22 And now we come to the difficult part, how do you 23 take a materials test and evaluate it for a service performance 24 where your electrode configuration is entirely different. And 25 .this is the horns of the dilemma that we are all on right now. Heritage Reporting Corporation (202) 628-4888 E
I ) l l l 82 l 1 Because you have to recognize that test electrode geometry is 2-different and the material is different. However, a hign 3 direct voltage test is a convenient test to conduct mainly 4 because of some limitations dealing with alternating current 5 tests. 6 Also you get into this thickness problem that you 7 have been discussing. You have to recognize that there are 8 deteriorating influences during service produced by heat, and 9 mechanical stresses, and partial discharges. 10 And finally, and this is spelled out in the ASTM 11 document that we are talking about, that the final 12 consideration is actual service performance. So you cannot 13 hang your hat on one peg of this situation. There are a lot of 14 factors involved. You cannot take a direct relationship unless 15 things' are identical between. 16 (Continued on next page.) 17 18 19 20 21 22 23 24 1 25 Heritage Reporting Corporation (202) 628-4888 h-_____._-__
.1 ) .c 83 l' MR. BALASKA: To help you with my spelling, young l' 2 ' lady, I'll give you my card, okay. 3 MR. BROWN Okay, thank you, Ted. 4 I think.the key thing _that Ted probably brought out 5 is that;there is no one point in any of this that I'm going to 6 read to you, or he's going to read to you, or anybody else is,: 7 and you're going to go, ah, that's the answer to everything. 8 What you've got to do and what we're trying to do with this, 9- .and I think'I've saidLit already is that there was a lot of l 10 . confusion regarding how to interpret the dielectric breakdown 11-data that we had accumulated, some of it wet, some of it dry, 12' some of it with reduced walls, some of it with ful* walls, and 13-what did the data mean? How could you interpret it? How could L 14 you interpret the apparent' increase, if.you will, apparent, I l l 15 say, increase in dielectric. strength of these reduced walls l 16' .after the LOCA, how does that make sense? How can you 17 interpret that? L 18 What I'm trying to give you now is an. overview of 19 some of the technology, or the theory that's behind the 20 technology that we're applying so that you can understand that 21 what we have seen has not been anomalous in any fashion; that-22' there is reason, that there is established theory even though 23-we've not been aware if, there is established theory to 24 understand'what we've seen, and that there is a rationale for 25 it. L l Heritage Reporting Corporation (202) 628-4888
84 1 Furthermore, that there is a rationale that says that 2 some of the stresses that'we have applied have been in excess 3-of what any of us,probably intended. One I touched on just 4 before Ted came up, was the factor of insulation thickness, and 5 how it varies, how dielectric strength varies with insulation- ,6 thickness, and.the penalty of some degree, and I can't quantity-7 that degree, that was put on us because we assumed linearity. 8 Something else there that Ted touched on was. 9 increasing temperatures. The tests that we have done insitu in 10-containment were done at somewhat, perhaps 20 or 25 degrees 11 Fahrenheit-higher temperatures than were those that have been 12 done in the labs, or have been done in water baths. How much 13 of an influence is that? I don't.know, I can't quantify it. 14 Again, I'm building a series of steps. 15 Breakdown voltage will tend to increase with the. '16 increasing rate of voltage application. Again, Mr. Balaska '17 touched on this as well. We have been comparing, if you will,. 18 apples and oranges. We have been looking at withstand tests, 19 five-minute withstand tests. We've been looking at breakdown ~20 values there, and we've been comparing those to intrinsic 21 breakdown values which are based on some prescribed rate of 22 rise, say 500 volts per second, and trying to draw comparisons 23 between the two. 24 MR. MARINOS: You're really not looking at breakdown 25 values. You are looking at their reduced level of testing from-Heritage Reporting Corporation (202) 628-4888
) 4-h 85 1 the breakdown. The breakdown values for the nominal thickness 2 as you stated in your presented here and before, there are very 3_ high voltages. Their orders of magnitude are higher than what 4 you're testing for. So it's really breakdown voltage' 5 ' theoretically to a very minimum thickness, not to the actual-6 thickness of the insulation. 7-So you know, don't call it breakdown values. It's-8 not. It's just a value which is way below the breakdown. 9 MR. BROWN: Yes, what I'm trying to say there are a 10 variety of factors that we were trying to understand,'and there 11 are a variety of known factors which influence them. 12 MR. MARINOS: If you're defining breakdown, then your 13 view is correct and consistent, but it is not breakdown 14 voltage, because we' assume there is a reasonable amount of 15 insulation in that cable, much more than what you have 16 demonstrated you really need, and the voltage you apply it's 17 much less than the breakdown voltage overall. So it's just add 18 voltage you apply, and we're trying to explain it. It is not a 19 breakdown voltage. 20 MR. BROWN: Yeah, very clearly, with some of the 21 voltage that we were originally applying in tests'-- 22 MR. MARINOS: Never was a breakdown voltage. 23 MR. BROWN: -- they were certainly not geared towards i 24 the insulation that we're now saying are qualified, certainly 25 they were not. They were geared towards some nominal i Heritage Reporting Corporation (202)'628-4888
J 86 1 insulation thickness. 2 MR. MARINOS: It's just semantics though using 3 breakdown, because breakdown is an entirely different thing for 4 that cable. 5 MR. BROWN: Well, as you are aware, we did a lot of 6 breakdown examinations in our own laboratories. 7 MR. MARINOS: Yes, and demonstrating you have very 8 high voltage that withstand with these cables. 9 MR. EBNETER: Oh, let's move on here. 10 MR. FOX: But there are some that wouldn't take the 11 high voltage though also in the laboratories. 12 MR. BROWN: All right. I think Ted has got the rest 13 of that. 14 (Slide) 15 Something else Mr. Balaska touched on was the fact 16 that the dielectric breakdown voltage increases when the test 17 specimens are merged in a liquid dielectric as opposed to being 18 in air. So again at one point in time, either internally or 19 together, we have tried to draw comparisons between dry 20 breakdowns and wet breakdowns. What we can say and what the 21 standard says is that the results in one medium cannot be 22 compared with those in a different medium. 23 Furthermore, I think the conclusion that I want to 24 make is that the lack of a surrounding liquid dielectric may 25 have resulted in surface discharges and reduced the withstand 1 1 1 ) i Heritage Reporting Corporation (202) 628-4888 l l
87 1 voltages. Because of the phenomena of what I'm going to call ~ 2 stress concentrations since you don't have a uniform ground 3 plane surrounding that cable, a wet test can in fact be more 4 severe -- excuse me. A dry test can in fact be more severe 5 than a wet test. 6 MR. LIAW: Explain why. 7 MR. BROWN: What I'd like -- 8 MR. LIAW: Give a reasonable explanation. 9 MR. BROWN: Yeah, what I'd like to do, I'd like to 10 ask if Mr. Bob Gehm of Rockbestos would like to address that 11 point. 12 In fact, in conversations some time back, it was Mr. 13 Gehm, I guess, who first warned us and said,'no, this is not '14 something that you want to do in testing these cables dry. 15 MR. GEHM: There are at least two factors that apply 16 here in dielectric testing of a coaxial insulation. One is the 17 increase in stress due to the fact that the field is not 18 parallel, but is concentrated at the conductor. Therefore, the 19 stress is increased logarithymically at the conductor. 20 The second thing that applies is that the conductors 21 are stranded. They are not smooth, cylindrical surface. So 22 you have a further increase in stress at that point due to the 23 radius of the individual strands. 24 Finally, in a dry test, you may have the outside 25 surface of the insulation over'a sharp point. You therefore Heritage Reporting Corporation (202) 628-4888
s 88 1 even concentrate the stress ever4 more at that point, and you 2 can have str, esses which may exceed the parallel plane stress by 3 factors of 3, 4 and 5 under those conditions. 4 MR. LIAW: The last factor is influenced by whether 5 it is wet or dry, right? 6 MR. GEHM: If the insulation is wet, you now have a ground 7 plane, a uniform ground plane entirely surrounding the 8 insulation. Therefore, the stress factors from the center 9 conductor proceed radially from each point. 10 If the insulation is in a dry medium, it is in 11 contact with a sharp point at one side. 12 MR. LIAW: Sharp point or in a conduit it is spot 13 contact. That can be a location of the contingent stress. 14 MR. GEHM: Wherever -- it could be from layicg 15 against the side of the conduit. It could be from laying over 16 a nipple in the end of the conduit. It could even be from a 17 burr in the conduit from a thread. But the dry dielectric test 18 can be quite severe and can be destructive under those 19 conditions which is one of the reasons why even in high voltage 20 testing of non-shielded and high voltage cables the recommended 21 field tests are never anywhere near the factory test voltages. 22 MR. LIAW: Let me ask you opinion. Do you think that 23 was the reason why in 383 -- actually for 383 in which they 24 specified be tested wet so that to get a uniform distribution 25 of the stresses? Heritage Reporting Corporation (202) 628-4888
l H: 89 1 MR.-GEHMs :That's part of it, but.that's not the 2 entire story, I don't believe. I mean, you are talking about-3' the final post-LOCA. test. l.# 4' MR. LIAW: Yeah, tested.
- 54 MR. GEHM Right.
My understanding'is that is part
- 6 of' flexing' test under which the insulation is then straightened-7.
out, reflexed around it's original' mandrel diameter and then 8 . tested in water. 9 MR. LIAW: In water. 10 MR. GEHM: Right. 11' MR. LIAW: 'To ensure some kind.of uniform. 12 distribution of the -- 13 MR. GERM: To ensure uniform. distribution of stress, 14 yes. 15 MR. BROWN: Yeah, and something more than that, too. 16 If you have cracks, that water is going to penetrate those 17 cracks, so.there is multiple purposes in that rebend test. 18 MR. FOX: Isn't the principle theory though'behind. ~ 19 383-1974 that when you bend it back around if you somehow 20 embrittle the cable, you are going to crack it, the water is il going to go in, and in fact cause you to have a short. It's. 22 'not only reducing stress; it's to find cracks. That's the H23 purpose for it as it's been related to me. .24' MR. EBNETER: I think that's the correct - 25 qualifications. Heritage Reporting Corporation (202) 628-4888-
i
- )
90-1 MR. BROWN: Okay. k 2 MR. FOX: Kent, your objective is to finish by ten to 3 four. 4 MR. BROWN: Okay. All right, so finally -- 5 MR. FOX: No, that's him to finish. 6 MR. BROWN: Again, I put'a lot of steps together on 7 this, so what am I going to conclude from these ASTM standards 8 and from bringing them before you. Two things that Ted has 9 touched on somewhat is that the dielectric breakdown voltage 10 test represents a convenient " preliminary test". 11 And second, but the final consideration must always 12 be the performance of the material in actual service. So we 13 can talk about breakdowns all day long, but what we've really-14 got to consider for ourselves is will the cable.do its intended 15 function in the service for which it's being exposed. 16 Now with that in mind, I want to remind you of what 17 we have done at the Wyle program. The Wyle program I think has 18 demonstrated not only will our cables that are in place perform 19 their intended functions throughout their normal life, but also 20 under accident conditions, but cables which have been severely 21 reduced well beyond that which we have seen in any of the post-22 mortems, and they still, too, will perform their intended 23 functions as well. 24 So that those Wyle tests demonstrate that the 25 dielectric strength of the silicone rubber does not decrease ~ Heritage Reporting Corporation (202) 628-4888 l __________a
L 91 1 within a.specified aging parameters, and we say that that 2 reduces the need for high potential test as a measure of future 3 adequacy. We want to look at future adequacy based on having 4 performed through their intended function, and well beyond, 5 complete with a quite extensive and comprehensive series of 6 post-LOCA dielectric examinations withstand. 7 All right, further, just as a point of note there. 8 We're talking about adequacy for service. Let's just not that-9 we have experienced no silicone rubber insulated cable failures 10 in service to date. 11 (Slide) 12 I guess just as a footnote on that whole discussion 13 of the ASTMs. 14 MR. EBNETER: Could I interrupt you one minute? 15 MR. BROWN: Certainly. 16 MR. EBNETER: On your last point, TVA has experienced 17 no silicone rubber cable failures in service. There is a 18 gentleman here from AIW or Rockbestos? 19 MR. GERMS Rockbestos. 20 MR. EBNETER: What's your experience industry wide on 21 the failures on your cables? 22 MR. GEHM: I've been involved in both high voltage 1 23 cables and low voltage control cable, telephone cables and the 24 'like. It's been my experience that we have seen in-use 25 failures in high voltage type cables. But I have never seen a Heritage Reporting Corporation (202) 628-4888
92 1 failure in a low voltage cable which resulted from years of 2-service. 3 The cables were installed. They either operate 4 satisfactorily when they are first turned on, or there is some .5 reason for it, usually damage in installation. Once the cables 6 have been installed and are operating, I have never seen a 7 situation where three, four, five years down the road one of 8 these cables has failed. 9 MR. RAUGHLEY: If I could add to that. 10 MR. FOX: This is Bill Raughley, TVA's chief 11 electrical. 12 MR. RAUGHLEY: If I could add to that, TMI had 13 predominantly Anaconda silicone rubber cables in their 14 containment. 15 MR. EBNETER: So what? 16 MR. RAUGHLEY: It was able to shut the plant down. 17 MR. BROWN: Okay. 18 MR. LIAW: I wish you didn't say that. Here you sit 19 with TMI II. I 20 MR. BROWN: All right, now, one of the comments that 21 I made regarded the presence of a liquid medium surrounding the 22 cable that was under test for the dielectric that was under i 23 test, and we've had a lot of discussions about wet versus dry, 24 and wet versus dry. 25 The conclusion that I drew was that it could be more i Heritage Reporting Corporation (202) 628-4888
i I k L.. 93 ^ 1 severe to test-a cable'in a dry condition because of all the 2 things'that Bob said.than'in a wet condition. We looked back-3 through the dielectric breakdown data that we had accumulated, 4 not accumulated knowing that such a standard exists, but simply 5 accumulated in the course of trying to understand the problem. 6 What I want to look at are the first group of data-1 7 points, because I have them both wet and dry there from AIN.- 8 We took new cables,'some of them we took'in a dry condition, 9 put foil around the jacket material, and took them to 10 breakdown. 11 Another set of cable we took them and put them in 12 wator, and took them at the same rate of rising to breakdown. 13 This is,not great loads of data, and I simply want to point at 14 what appears to be a trend which will confirm the established-15 theory that I just said, is that all-those that were tested in 16 a dry configuration broke down earlier than those which were 17 tested wet. 18 MR. FOX : - Well, is there not also a difference in the 19 rate of voltage application for the two? Weren't there two 20 ' parameters? 21 MR. BROWN No, sir, not on those. These tests were 22 done the same. 23 MR. FOX: Okay. 24 (Slide) 25 MR. BROWN: We've dielectric now for a few minutes, Heritage Reporting Corporation (202) 628-4888
i i 94 1 and let'be see if I can tie a ribbon around some of this and tell yoe v3at.I want to conclude. Remember, I said the ASTMs 2 t '3 weren't everything, but that we had to look at the. picture as a 4 whole the tremendous amount of work we have done.to this point. 5 If I look, and across the top there I have 6' manufacturer, minimum qualified thickness according Wyle 7 program. I have added in the tolerances. So previously I was 8 reporting two and two mils and four mils as being.the.various 9 minimum thicknesses. I plugged in the tolerances here so I'm 10 going to be,a good, conservative guy. 11 If I follow 383 standard accepted practice for EQ, I' 12 would apply 240 volts per mil for five minutes, do it once,.do 13 it wet,.and boom, I'm done, okay. And I draw a conclusion 14' based on that. And that would in the case of the 4, 4 and 6 15 mil insulation systems be as you see, 969, 960.and 1440. But I 16 didn't just do that. I've done more. 17 If I were looking-at new cable, brand new cable.and 18 take my 300 volts per mil. That's 1981. And if.I did a 500-19 minute test -- excuse me -- five-minute test, if I were 20 producing new cable, what levels would I pick for this 4 mil, 4 -21 mil and 6 mill cables. It would be 1200, 1200 and 1800. But 22 I didn't do'that. I did more. 23 How about if I were looking in the design stage of a-24 cable? Let's take from one of our silicone rubber product data j 25 sheets a 600-volt per mil dielectric strength as its reported. I Heritage Reporting Corporation i (202) 628-4888 ) l ) i
- )
95-1 If I used those-numbers against the same' set of. insulation 2 thicknesses, I would do a-single'five-minute test,. 2400, 2400 -3 'and 3600, but I didn't do that. I did more. I did more. 4 (Slide) 5' In the course, if you will remember in the course'of 6-my testing, my LOCA test on these reduced wall samples, I 7 didn't do a single-test. I did multiple tests._ I didn't do 8 them just wet. I did them wet and I did them dry. And I 9 didn't do it just for five minutes. I did it for 18 minutes, 10 shut it down, put it in the oven, took it out,.and went through 11 another 18-minute program as high as 4700 volts. 12 So we're not_ talking a small program in that sense. 13 (Slide) I 14' Finally,'let's drop down and look at that very last 15 entry on there. I've separated it physically, because I-want '16 you_to. separate it somewhat in your mind. I don't want to be 17 guilty of telling you not to do what I just told you not to do'. 18 or doing what I told you just not to do. 19 That is a breakdown. That is a slow rise' breakdown 20 that we've experienced out in the field. The lowest one has 21 been 7500 for AIW, and 7000 for Rockbestos. I've had no-22 failures for Anaronda whatsoever. I've done a slow rise 23 breakd.en. I specified that you come up to rated voltage 24 within the space of a minute. Rate of voltage being in this 25 case whatever the voltage for that test was, whether it was Heritage Reporting Corporation (202) 628-4888 i
o 98 o t 1 anomalies and relate those back to the actual mils that were 2 relating to the anomaly and bracket those, then we have a 3 significant test program of 90-something conductors that have 4 been completed with an acceptance value when we adjust the 5 acceptance value on the basis of the Wyle test. And it's not~ 6 to demonstrate that we had a breakdown voltage. l 7 (Slide) 8 MR. BROWN: Let me go back for just a second. You 9 know, what we're looking at, Angelo, is this level we've said 10 is a good test for 4 mils, and that's a good test for 4 mils. 11 MR. MARINOS: It's a good test for 4 mils. It does 12 not confirm that you have 4 mils. 13 MR. FOX: For. production. 14 MR. MARINOS: It's a good test for this hypothetical 15 case. 16 MR. BROWN: Now, I'm wanting to show that I've got a 17 cable that' is suitable for its intended service. 18 MR. MARINOS: Right. 19 MR. BROWN: Now that's what I'm required to do as I 20 see it, is I have to show that I've got a -- 21 MR. MARINOS: You have established with the Wyle test 22 that 4 mils is good enough. Okay, now you've got to jump from 23 there to demonstrate through some test that you do have. But 24 are you going to use calibers to measure the thickness, or are 25 _you going to use some -- Heritage Reporting Corporation (202) 628-4888 j i i i l _--_---_--__-_____--,-)
99 1 MR.-RAUGHLEY: If we go back to the original '2 manufacturing standard and saying.if we've done what you said 3' you have -- 4 THE REPORTER: I can't_ hear you. 5 MR. FOX. Bill'Raughley. Bill, get to a mike. 6 MR. RAUGHLEY: If we go back to the original 7 manufacturing standard, that standard would say test the 4 mils 8 to 960. If we go back to the original qualification standard,. 9 that standard would say test that to 1200. 10 MR. MARINOS: But the qualification standard for the 11 manufacturer is not attempting to establish that you have the' 12 thickness. 13 MR. RAUGHLEY: No, he's trying to establish if its 14 . adequate to ship it and be good for 40 years of service. 15 MR. MARINOS: Right. It does not establish the 16 thickness though. 17 MR. BROWN: It establishes'the level of confidence in 18 a -- 19 MR. RAUGHLEY: That's correct. 20 MR. BROWN: -- dielectric integrity. 21 MR. RAUGHLEY: But it's a high level of confidence is 22 the standard that's been used throughout the industry. It's 23 the standard that all cables in the industry ar's shipped to. 24 MR. LIAW: To me the first line there really does not 25 mean anything. i l I Heritage Reporting Corporation (202) 628-4888 j l _------_-u
l 100 1 MR. RAUGHLEY: No, what we're trying to say -- no, it 2 does. 3 MR. LIAW. You really don't have such a thing as 4 4 mil cable installation. Nobody is going to manufacture the 5 thing. Okay, and based on what, you know, the tests say about [ 6 the nonrelationship, I mean, you know, no one is going to 7 accept that. 8 MR. RAUGHLEY: But let's -- you know, originally you 9 guys asked me will 8 mils qualify. All right, the answer is 10 yes. And what we're trying to show you here is that if those 11 top three numbers are 4 mils, isn't the bottom one 4 mils? 12 Doesn't your engineering judgment say that's 4 mils? 13' Therefore, we have at least more than 4 mils in the cables we. 14 tested. If those top three are 4 mils,. that bottom one'is 4 15 mils. 16 MR. MARINOS: I.was saying that if there was 2 mils 17 of insulation, it would not take the 960 volts? Would 2 mils 18 of insulation take 960 volts? 19 MR. CANTRELL: We were looking at insitu -- 20 (Simultaneous conversation.) 21 MR. FOX: One speaker. 22 MR. CANTRELL: We were looking at.insitu testing, not 23 to do destructive testing. 24 MR. MARINOS: I understand. 25 MR. CANTRELL: Okay, we were looking at a test Heritage Reporting Corporation (202) 628-4888 l _--_.m_____h
) 101 1-program to give us confidence in that installation..You know,- 2-if,I go back to a'6 KV cable, you know, and I look at those 3 margins, I don't do'an insitu test of breakdown to. gain 4 confidence. I'm scaling this down. We're trying to Jook at -- 5 we have -- by.the time that cable is in place, it's taken its 6 ' abuse of being pulled into place and whatever, which all cable 7 goes through. And we're trying'to say is there reasonable 8-confidence that a' test program will give us confidence.. l 9 MR. MARINOS: But we have established minimums here. 10 We're not talking about nominal thicknesses and lots of margin. 11 You have elected to demonstrated the minimum required. number 12 cf minimum thickness of insulation that you can possibly meet' 13 your intended' function with. 14 And now the problem we have.is to bridge that 15 chemistry that you got there by doing a qualification test 16 that's 960 volts. The 1-mil of insulation can pass it, too. I 17 So by telling me you are passing it with these tests, you're 18 not telling me you have 4 mils.. You may have 50 mils, you may 19 have little as 1 mil. I 20. We have established that you need 4 mils. So now how 21 do we bridge that? 22 MR. FOX: Okay, let me tell you, we can argue until 23 hell freezes -- 24 MR. LIAW: One of the charts say, new cable, high 25 potential breakdown test results, three manufacturers, foil, Heritage Reporting Corporation (202) 628-4888
- +=__ -
i
.o / l 102 l I water and nominal insulation thickness mil, 45-mil. 2 Rockbestos, you have three tests there, 2,100 volts i 3 per mil, 2222, water 2222, not just take it to 2,000 volt mil. 4 Take 7,000, so you have about 2,000. That tells me -- you 5 demonstrate you have 3.5 mil thickness there. Am I wrong? 6. Based on what you say presented earlier. Rockbestos, over 7 2,000 volt per mil. So I take that -- round it off to 2,000. 8 Take 7,000, you about 2,000. So what have you demonstrated? 9 MR. CANTRELL: Okay, but we have a shaving test which 10 shows what that minimum insulation thickness was. 11 MR. LIAW: Well, that's beside the point. We are 12 deal with a bunch of conflicting invention, some of them 13 qualitative, some.of the quantitative. I just don't know how - 14 - you help me, okay, how to reconcile the differences. 15 Here within the same presentation. This is something 16 you are showing me, okay? 17 AIW, 1,111, the lowest one. Highest is 1933. Let's 18 take the highest one, 1933. You have about 7500. So how many 19 mil thickneso you have demonstrated? I don't know. Maybe you 20 can explain to me or help me to understand that. 21 MR. FOX: I'm telling you, in my opinion, there is a 22 lot of extraneous information on the table on this voltage 23 stuff that has no bearing on the real problems. 24 I want to reiterate again the important thing is that 25 we went in and did post-mortems on our failures. We shows that Heritage Reporting Corporation (202) 628-4888
103- .1 .the minimum thickness that we observed on any of those failures 2 was-8 mils. Most of them, in fact all of them were in excess 3 of that. 4 We then went to the test labs and conservatively 5 'showed that 4 mils for Anaconda and 4 mils for AIW and 6 mils 6-for Rockbestos will in fact pass the test, will.take a LOCA. 7 MR. MARINOS: That makes more sense. 8 MR. FOX 4 Yeah, sne need to forget all this voltage 9 stuff. 10 MR. MARINOS:.If you want to argue it in terms of how 11 much thickness you have by a physical examination, that's one 12 set of -- 13 MR. FOX: ~ Absolutely, yes. ~ 14 MR. MARINOS:. -- argument. Don't involve the vo1tage 15 'because -- 16 MR. F.OX: Yes. 17 MR. MARINOS: -- that breaks down. I don't 18 understand the relationship. 19 MR. FOX: The voltage is -- 20 MR. MARINOS: I cannot understand it, and I will not 21 be able to -- 22 MR. FOX: I don't think we'll ever understand it. 23 MR. BROWN: I think -- 24 MR. FOX: It's a diversion. 25 MR. MARINOS: If you want look at 100 of the cables Heritage Reporting Corporation (202) 628-4888
i ) 104 1 and measure the thickness physically,-each one of them, and 2 make'a statistical, you know, prediction of that, an 3 extrapolation, that's another story. This is, you know, 4 ' consistent. 5 MR. BROWN: I think we missed a very key point that 6 Mr. Gehm made earlier is that when.you are testing in a dry 7 conduit configuration, you are subjecting yourself to the 8 possibility of stress concentrations that can result in a 9 significantly lowered -- 10 MR. CANTRELL: Let's get on through this thing. 11 MR.-FOX: Just a minute, just a minute. 12~ Let's just set the record straight on one point. We 13 introduced the standards today because qualitatively they offer L 14' .some limited explanation for some of these phenomena that we-1 15 have seen in the plant:versus in the test lab. 16 MR. LIAW: I agree. 17 MR. FOX: No way can we quantitatively come up with 18 clear, crisp answers. Again, I want to focus you on the 19 important things. We did post-mortems. We established what 20 the minimum. wall thickness was on all the failed cable. Then 21 we went and in the lab showed that that minimum thickness, even 22 much less than that, would in fact take a LOCA. 23 We can dance forever on voltages. I wish like hell 24 that I had told the people you will not do anything except put 25 rated currents on that cables after you finish the test program l Heritage Reporting Corporation (202) 628-4888
) 105 1 at Wyle. It wasn't germane. We passed the LOCA. 2 MR. MARINOS: I told you so. 3 MR. LIAW: Well, somebody could ask you why don't you 4 use UT to measure thickness. 5 MR. BROWN: Why don't we get off voltages and get to 6 what -- 7 MR. BROWN Yeah, you know, I had the same problem 8 that everybody else is struggling with. 9 (Slide)So what I want to present as the key 10 observation is not dividing this value by that value, et 11 cetera, but the fact that we could take and' reduce insulations' 12 thicknesses down to the level that you see on here. You can 13 see the conductor, you can count the number of strands, et 14 cetera, at arms length. 15 Angelo has seen these in person, and I think 16 everybody's comment was is there insulation there. Now is 17 that not true, you know, amongst the people? But yet look what 18 happened. There were no age-related -- no age-related cracking 19 occurred, no LOCA-induced cracking or splitting occurred, no 20 cracking or splitting occurred during the mandrel rebend, you 21 know, the flexing of the cable, and no significant dimensional 22 changes were observed following the LOCA exposure. 23 (Continued on the next page.) 24 25 Heritage Reporting Corporation (202) 628-4888
, f., -h r 106 1 MR. BROWN: So, the material, I'm saying, is a very, 2. vary stable compound:for both aging, accident and the' .3 . mechanical" stresses'that wereLinvolved.- Now,'we have taken and 4' shaved the cables about as far as one: can-physically do it. I-5, .think~we have gone to the. limit in that regard. The'significant 16-thing there is if the-insulation is'there.now, the insulation: 7-will be there after LOCA. 8 What level of insulation did Sne go down to? I-say 21 9 mils. I put'on the tolerance for convenience and say.that, 10. 'maybe, as much as 4 mils. So, I think,'again, I am back to the 11 . point where I am:saying, " Forget trying to divide numbers. 12 What we can say.is that the cable has been demonstrated to be 13 satisfactory.for the service that is required." 14 LMR. LIAW: Kent, it is easy for you to. say that,- but 15 how can I explain to people what those numbers'mean? 16 MR. BROWN: I think what we want to explain to the 17 . people is that we:have at TVA subjected ourselves to the'most 18 extensive comprehensive dielectric. test program that has ever 19-been undertaken in any utility. Anybody who knows any-20 different,-talking low voltage non-shield and cable, speak now. '21 Silence. All right. 22 Also, I am going to say that we have undertaken the 23 single most severe LOCA test ever conducted and I am talking 24 severe in the sense not of temperature, but in terms.of what. 25 you put into the LOCA and what you did to it both before and Heritage Reporting Corporation i _(202) 628-4888
107 ~ l after the LOCA. And I am telling you that we got results that 2 are great from a performance standpoint. 3 Now, do you know of any tests that are more severe 4 than that? 5 MR. LIAW: I am not disputing what you are saying, 6 but help me to explain all those numbers. 7 (Slide.) 8 MR. BROWN: All right, let's go on to the conclusion 9 and we will be glad to work with you on the stupid number, as 10 you say. 11 MR. LIAW: Those numbers were generated by you, by 12 TVA on docket. 13 MR. BROWN: The numbers, and that is why we added the 14 ASCM. l 15 MR. FOX: And, again, I want to point out: The 16 numbers are not germane. What is germane is cables with 17 minimum wall thickness passed the LOCA. I regret that we 18 docketed those numbers. 19 MR. LIAW: As long as you say you regret you put out 20 in docket, fine. 21 MR. EBNETER: Let's get to the bottom line. Let's 22 keep moving here. 23 MR. BROWN: All right. My conclusion is, speaking 24 now as the one who is technically knowledgeable of the cables 25 and what we have done to them both in the plant and in the Heritage Reporting Corporation (202) 628-4888
.) 108 1 labs, is that for the purposes.of restart, it is my. opinion 2 'that no further work is required based on the degree of r 3 ' exposure to which we have subjected these cables, whether it be 4 dielectric or qualification. 5 MR. EBNETER: That is your. opinion. Is that Admiral 6 White's opinion? .7 MR. BROWN I'll let Mr. Fox answer that. 8 MR. FOX: Yes. 9 MR. BROWN: 'Okay. 10 MR. FOX: He is speaking for TVA. 11 MR. BROWN: Second,-I say that.we have conclusive l'2 evidence in the nature of our program that substantiates that 13 the cables will perform their intended safety.functi'on. And - 14 that is what we are after. What I want to know is not how much .15 the cables have' changed since yesterday. What I want to know '16 is: Will they do what we have intended them to do? 17 My' opinion and TVA's opinion in reading that test 18 document is that: Yes, they will. And I think we have 19 established by the virtue of all those tests a greater. degree 20 of confidence in the cables' that are in those plants, in that 21 plant right now than the confidence that you should have with 22 anybody else's cables out there because the. test that they'have 23 .done, they don't demonstrate any particular insulation 24 thickness. What we have proven is we have proven that you can 25 take a cable and severely degrade it and it is still going to Heritage Reporting Corporation (202) 628-4888
] 109 1 do l'ts job. 2 All'right. So, again, these are things I have 3 already hit. The.in situ tests, the highest voltage of anybody 4' I'know, they are most. conservative in terms of acceptance 5 criteria. We have what has to be the largest sample ever 6 subjected'of low voltage cables to these dielectric-tests. We 7 have well exceeded the industry test for a field examination of 8 cables. We have exceeded any standards that we are aware of aus 9 well as regulatory guidance in this light. 10 WeLthink that the points that we brought out about 11 the ASTM standards demonstrate a level of conservatism. No, I 12 cannot quantify each and every point; but, overall, I say I 13 have demonstrated a level of conservatism. We have extensive. 14 laboratory tests and evaluations and they looked at all of the 15 parameters. Not just dielectric. We looked at chemical. We 16 looked at environmental. We looked at mechanical. And all of 17 those, we came out on the plus' side. 18 (Slide.) i i 19 We demonstrated by the application of minimum EQ l 20 thickness to original dielectric strength requirements that the 21 cables are as good as the original manufacturing and i 22 qualification requirements. In other words., we bought 23 something that was supposed to function for its intended 24 service. What I offer to you today is a set of cables that 25 will per' form their intended service function. Heritage Reporting Corporation l (202) 628-4888 j 1 L_i __ -_
11D 1 We demonstrated by tests that 100 percent of that 2 sample will hold far in excess of rated service voltage.
- And, 3
as we all know, that is what the cable is going to actually be 4 exposed to. Anything else is a measure of margin that the 5 cable may have. 100 percent of the sample exceed the minimum 6 EQ thickness. That is we have done post-mortems on the cables. 7 It is well documented what we saw in U. Conn.: 8 mils, 10 8 mils, 12 mils, 20 mils. 9 What I am telling you is that new data that we just 10 have from our central labs -- 11 MR. MARINOS: When you say it exceeds the minimum EQ 12 thickness, the new one that you have established now? 13 MR. FOX: The new ones, yes. 14 MR. BROWN: Right. Absolutely. Thank you. The 15 cabies that we did a post mortem on from that second group 16 where we had failures in some AIW in one rockbestos, that post 17 mortem showed a minimum, minimum now, remaining wall of 21 18 mils. Well in excess of what we had seen previously. We 19 believe that the problem has been bounded, that if we go 20 looking further, we are looking for what currently doesn't 1 21 exist and I have reason at this point in time to believe that 22 it exists. 23 The program that has been put together exceeds all 24 licensing precedent that we are aware of and should give added l 1 25 assurance for the work that we have done. l l l Heritage Reporting Corporation (202) 628-4888 I i
,) i L e i 111 L 1 ' We. believe that the probleras that we have encountered 2 thus far are not significant in terms of performance of the 3-cable.in that we have qualified as low as 4 mils out of a-4- nominal 45 mil wall. 5
- In other words, yes, we have encountered dielectric 6
breakdowns, but are they significant? No, in light of the fact 7 that we didn't need-45 mils and we were shooting for a 8 . dielectric test based on that. 9-No. 6, we exceeded the original NRC request. You 10 guys said, " Pull out one and do X, Y, and Z." We have removed 11' a number of cables now. -We'have done extensive testing. We 12 have gone off to the best labs.:We have engaged the best 13 consultants that we could find. And we believe that we have 14 exceeded that original request. We believe we have'done a good 15 job of investigating and pursuing this problem. 16 MR. FOX: There is one point I want to make right 17 here is on the silicon rubber tests that we did, we were also 18 looking at worst case, based on random type events which led us 19 to look at the longest runs of silicon rubber cable we had from 20 the three manufacturers. So, we were looking at worst case. 21 That is another confidence level. We have looked at worst 22 case. We have had very few failures. 23 MR. LIAW: No failures. 24 MR. FOX: None, yes, at this point. 25 MR. BROWN: Finally, the extent of the testing that Heritage Reporting Co'rporation (202) 628-4866 l l
L'a. .s 112 L 1 we have done should assure us that.we have pursued and are 2 proposing-a safe alternative. In other words, we should have 3-great confidence in those cables because we.know that they are 4 very. age-stab'le. They are very stable through an accident 5 environment and they are there now. They have proved themselves 6 in service. And we believe and we believe we have shown that 7 they will stay there and continue to perform their service.. 8 And, No. 8, just as a point of correction on 9 ourselves, we hope that we have cleared up the heightened level 10 of uncertainty that we induced by erroneously reporting 11 breakdowns at 5200 and 5800 volts. We hope you all go away 12 from here knowing that what we have seen, the lowest breakdown 13 voltage has been 7000 volts. 14 With that, I say, thank you for your patience in 15 hearing m out and I guess I will turn it back to Mr. Fox at 16 this point. 17 MR. FOX: I just want to add one final remark. We 18 are confident that our cables et Sequoyah are suitable for 19 -service. We are confident that we can safely run Unit 2. And-20 we believe + bat the information that we have presented today 21 should reasonabis to expect you to allow us to restart the 22 plant with the basis of the information you have heard today. 23 MR. MARINOS: Let me get another clarification. On 24 the basis of the tests you conducted, of which was something 25 like 91 conductors. Heritage Reporting Corporation (202) 628-4888
113 1 MR. FOX: That's correct. 2 MR. MARINOS: And you have broken down just nine of 3 them. 4 MR. FOX: No. Seven of them broke down. There were 5 three leakage -- well, I won't say " leakage current." That was 6 the original factor. But there were a combination of leakage 7 current and polarization index. So, there were neven 8 breakdowns. 9 MR. MARINOS: Those seven were pulled out and 10 examined for thickness and you have established certain 11 thickness-12 MR. FOX: Yes, Fir. 13 MR. MARINOS: The lowest was 8 mils and the best was 14 21 mils. 15 MR. FOX: Yes, sir. 16 MR. MARINOS: And something in between. Those are 17 the only ones you examined. Did you examine the ones that 18 passed the test? 19 MR. FOX: No, sir. 20 MR. MARINOS: How do you know you have sufficient 21 insulation on those? Just because you passed that test, you 22 have already determined that the voltage test does not 23 determine this kind of a thing, particularly since you did the 24 Wyle test and you applied breakdown voltages of 2 mils 25 thickness and demonstrated that they can take up to 16,000 Heritage Reporting Corporation (202) 628-4888
p ] 3 1 l 114 l-1 volts or 18,000 volts. So, therefore, the ones that passed the 2 9,000 volts or 10,000 volts test that we had all established J 3 does not really tell you what kind of insulation thickness you 4 have got; does it? MR. FOX: Again, I think you are gettin'g hung up on 5 l 6 voltages. The bottom line is that cables failed. The failure 7 points were at areas of reduced insulation thickness. 8 MR. MARINOS: But if we don't hang up on voltages, 9 then you really looked at seven cables, only. Seven conductors 10 in the measurement -- 11 MR. FOX: No. We looked at the failures. 'And those 12 failures manifest themselves as, " reduced wall thickness." 13 with a couple-of exceptions. We couldn't figure out why those 14 cables failed. 15 We then took that minimum thickness associated with 16 those failures and then went and qualified a thickness even 17 lower that. 18 MR. MARINOS: I understand. But you only looked at 19 seven or nine, whatever it was. That's all you looked at 20 physically. l 21 MR. CANTRELL: I think, Angelo, we gain confidence in j i 22 the remaining part of the test. And, maybe, in hindsight, with l l 23 the data from Wyle, I think that confidence is justified for l l 24 them passing a 7500 or 9600 volt test for five minutes. You 25 know, we would say that there was not reason for us to lose l Heritage Reporting Corporation (202) 628-4888
-+ 115 1 confidence in those installations. 2 MR. MARINOS: I have low confidence in the 3 methodology by which we -- his presentation, which was very 4 good, demonstrated to me at least at this juncture that maybe 5 they are not legitimate tests. So, therefore, what other 6 mechanism are we going to use as NRC staff to determine, to l 7 agree with your confidence. Not the seven cables that you 8 physically examined. It has to be something more or something 9 better. 10 MR. FOX: We have looked at well over 1,000 cables. 11 We were looking for gross installation damage in the plant. 12 That was what postulated. That wac what was alleged. We have 13 not found gross installation damage in Sequoyah on Unit 2. 14 MR. MARINOS: And it seems t'o me with the testing 15 that we have all agreed, you would have done it anyway. 16 MR. FOX: We may or may not have, but.we were in the 17 boat together. 18 MR. MARINOS: I think this is what was demonstrated 19 today by his presentation, Kent's presentation. 20 MR. FOX: I don't agree with that at all. 21 MR. EBNETER: I don't under the discussion -- where 22 are we going? What are you looking for? 23 MR. BROWN: I think the root of the problem is -- 24 MR. MARINOS: To share their contidence. 25 MR. EBNETER: I think you have heard enough comments. Heritage Reporting Corporation (202) 628-4888
i s l 116 1 I don't think we will solve anything, Charlie, by taking a 2' caucus and then coming back. I think we need to sit down and 3 look at the new material you have put on the docket, this 4 report. Is there a report that you were going to give us? 5 MR. FOX: This report, the Wyle report is signed by. 6 Gridley and -- 7 MR. EBNETER: Have we'had that previously, do you 8 know? 9 MR. FOX: No. 10 MR. EBNETER: There is enough concern yet that I 11 think Angelo and the staff, B.D., we need to discuss this at 12 length. i 13 MR. LIAW: May I ask Mr. Balaska a question? I 14 MR. EBNETER: Sure. 15 MR. LIAW: I have a couple of questions for you. 16 Earlier, you mentioned that there are so many factors that can-17 effect the result of a test. Thereby one could make a judgment 18 on the validity of the test. Have you looked at how those 19 tests both in situ and ex parte were conducted for them in 20 terms of electric area, in terms of the specimen thickness, 21 that kind of thing and to make a judgment yourself with regard ] 22 to the validity of the test and whether or,not -- I hate to use 23 the word -- you are willing to stake'your professional 24 reputation on line. You have to sign it under oath, but you 25 examined those and you agree and those are the appropriate test i i Heritage Reporting Corporation (202) 628-4888
117 1 and one should place high confidence on those results? 2 MR. BALASKA: I have not looked specifically at area. '3 Okay? Because I don't know the exact lengths that were tested. 4 So, I can't come up with areas. Okay? I need to know that. 5 MR. LIAW: I am saying that both the electra apply at 6 this site when they come up with whatever? And also at.the 7 Wyle lab, how they conducted the post LOCA voltage tect? 8 MR. BALASKA: I was not at the Wyle lab. Okay? 9 MR. LIAW: Whether or not you have examined those you 10 have mentioned. 11 MR. BALASKA: I have not examined. No. 12 MR. LIAW: So, it is really unfair to ask you whether 13 or not you are willing to put your professional reputation on 14 the line? 15 MR. BALASKA: That's right because 1 can't answer 16 your question. 17 MR. LIAW: I understand that. Thank you. 18 MR. FOX: I want to just put one more thing back on 19 the record. Angelo asked the question -- 20 MR. GILL: If I could just follow up? I am Paul Gill 21 from NRR staff. If I could ask you a question. Would you 22 comment if the validity of 149 applies to installed conductors? 23 MR. BALASKA: Now, 149 is a ma?.erials test. 24 Installed conductors, we have to get into a product test. And 25 149 is not a product test. It is a materials test. Heritage Reporting Corporation .(202) 628-4888 L.
I 118 1 MR. GILL: So,.the answer is, no. 2 MR. BALASKA: That's right. The answer is, no. 3 Right. 4 MR. GILL: Thank you. 5 MR. LIAW: 149 was for AC test. Here, we are talking 6 about ADC test here. Therefore, do you think that D-3755 is 7 applicable to the same question that Paul Gill just asked you? 8 MR. BALASKA: It is applicable with the same 9 limitations as the 149 is applicable with the same limitations 10 that I outlined and I discussed before as far as 3755 is 11 concerned. 3755 is not directly applicable to cable. Okay? 12 MR. LIAW: Yes. 13 MR. BALASKA: It's materials. Now, when you 14 recognize the factors that I had mentioned before, you can draw 15 engineering judgments. And you alluded to that when you asked 16 me about area, as an example. 1 17 MR. LIAW: Normally, when we have plenty of margins, 18 I am willing to say, yes, in your own judgment. But here we i 19 are pushing something to 2 or 3 mil right at the management 20 error band. It is difficult for me to rely solely on your 21 judgment. That does concern me. .) 1 22 MR. BALASKA: I can understand your concern, but when ] 23 I take a look at what has transpired, in essence, I think TVA 24 has proven that 4 mils, if they have 4 mils, will pass -- l 25 MR. MARINOS: If they have 4 mils. i Heritage Reporting Corporation (202) 628-4888 i L
119 1 MR. BALASKA: That's right. And, in essence they have 2 indicated that they know they have 4 mils. Okay? From the 3 tests. 4 MR. LIAW: They have indicated that. These are all 5 qualifications you make with regard to the application of 6 electra and other factors. 7 MR. BALASKA: Right. Sure. 8 MR. EBNETER: I would like to comment on that. These 9 material specs, I think we have heard enough on. We have got 10 to test on the actual cables; right? It came out of Wyle. 11 Let's look at it. If you want to discount whatever they said 12 on the ASTM standards, fina. But the most -- the best data we 13 have is that cable data. 14 MR. MARINOS: Right. 15 MR. LIAW: I think ASTM is good information. 16 MR. EBNETER: It is good, but -- that standard has 17 the same relationship to the cable that the ASTM material specs 18 have to pipe. The same thing. Just look at it in that light. 19 MR. FOX: I would like to -- 20 MR. EBNETER: Let me stop you, Charlie. 21 MR. FOX: Oh, okay. 22 MR. EBNETER: Does staff have any,other questicns '23 other than ASTM? 24 MR. FOX: Just a minute. I'm going to make a comment. 25 I want to answer Angelo's question. I didn't answer it Heritage Reporting Corporation (202) 628-4888
j 120 1 earlier. On the testing we had done previously, we had 2 mutually agreed on pass / fail criteria, polarization index and 1 3 voltage level. We failed 10 cables out of well over 1,000, 4 total test program based on that. And we did an analysis of I 5 those failures. { I 6 We then -- a combination of both laboratory tests and j 7 post-mortems on the actual plant cables that failed to 8 establish the basis for an EQ test which we went and conducted ] i 9 at Wyle. { 10 We did test to qualify the remaining wall thickness, 11 but we had a basis for accepting the initial passes, which I 12 don't think we ought to go back and question. In systems 13 engineering, you do a failure modes and effects analysis when. ) 14 you have a failure, when you have a problem, or when you are 15 trying to anticipate a problem. We have done that. 16 We have homed in on what we think causes that type of 17 damage. We have homed in on what that damage manifests itself 18 as, such as its reduced wall thickness. We have then gone and 19 qualified that. 20 Of theses failures that we had, we only had seven i l 21 high voltage failures. Six of those seven were AIWs. There was 22 only one rockbestos voltage failure. 23 I want to tell you about where we are in the plant 24 right now. About three or four weeks ago, before we even had 25 these Wyle results, we decided that we would prioritize AIW 1 Heritage Reporting Corporation l (202) 628-4888 1 1 l L-
=,, - e: s 121 1 cable and start replacing. The fact is we have stripped out. 2 prior to even getting these final results 70 percent of the AIW 3. cable. We are-replacing that. The critical path for Unit 2 of 4 Sequoyah'for heat-up right now are hanger modifications.- We 5 thought it-would.be' prudent:to prioritize the AIW cable worst 6 to best-and start-replacing it'on a prudent basis.. We 7-identified about 70 percent of that that was substantial in 8 nature. We. stripped that out. It goes without saying we are ~ 9 going to replace 70 percent of the AIW cable with a new hypalon 10 cable.- Hypalon-jacketed cable. That's brand Rex. So, we are 11 well on our way to doing that. 12 The critical path has us finishing these hanger mods 13 by December the lith. At that point in time, we will.go back-14 into valve alignments and we will hopefully be in heat-up 15 sometime around the end of December or the first of the year. 16 What we are doing is replacing A 1/2 cable consistent with 17 doing those hanger mods. And we will have approximately 70 18 percent of that AIW cable replaced, the cable that we have q 19 really seen the problem in, by the 4th of December, the 3rd of i 20 December or thereabouts. So, I wanted you to be aware of where 21 we are at the plant. 22 MR. RICHARDSON: Look, Charlie, as I understand, one 23 of the problems is making an inference on the large population j 24 of cables that are in the plant based on an examination of only 25 seven in terms of how much wall thickness really is there. But Heritage Reporting Corporation (202) 628-4888 i
c 122 1 I hear you -- 2 MR. FOX: But you have got to realize that a large 3 fraction passed. 4 MR. RICHARDSON:- Hear me out. You have stripped out 5 a lot of cable. It seems to me, now, we have got a lot of data 6 if we go-look at the cable you have stripped out? 7 MR. FOX: No. If we had carefully taken it out, it l 1 8 would have been worth looking at, but when you strip cable ] 9 out -- 10 MR. RICHARDSON: So, some of it has been damaged in 11 the removal process? j 12 MR. FOX: I am not telling Angelo anything he doesn't j 13 know about electricals. When you. strip it out, you subject it l i 14 to damage. We didn't -- I mean it was yanked out. 15 MR. ZECH: I think that is common with other cable. 16 MR. EBNETER: Does staff have anything else? If not, 17 you all have a nice Thanksgiving. Mr. Zech? 18 MR. ZECH: Anything else from anybody? 19 (No response.) 20 Thank you very much. We will close.- 21 (Whereupon, at 4:12 p.m., the meeting was concluded.) 22 23 24 25 Heritage Reporting Corporation (202) 628-4888 t__ _ -_
1 CERTIFICATE-2: 3' This is to certify that the attached proceedings before the 4 United States Nuclear Regulatory Commission in the matter of: 5 Names Discussion'on Silicone Rubber Cable at Sequoyah 7 Docket Number: 50-327,'50-328 8 Place: 450 East West Highway, Room 550, Bethesda, Md. 9 Date November 24, 1987 i 10 were held as herein appears, and that this is the original l 11 transcript thereof for the file of the United States Nuclear 12 Regulatory Commission taken stenographically by me and, 13 thereafter reduced to typewriting by me or under the direction 14 of the court reporting company, and that the transcript is a .l 15 true and accurate record /of the.fofegoing proceedings. 16 /S/ lLu 0 AVtA^ J i I l j 17 (Signature typed): j 18 Official Reporter 19 Heritage Reporting Corporation 20 21 4 22 23 24 25 Heritage Reporting Corporation (202) 628-4888 P
i ENCLOSURE 2 NRC/TVA Meeting on Cable Testing November 24, 1987 Name Affiliation Eileen McKenna NRC/0SP Richard L. Gridley TVA/ Licensing Charles Fox TVA/0NP R. W. Cantrell TVA/DNE S. D. Ebneter NRC/OSP J.-A. Axelrad NRC/OSP S. D; Richardson NRC/OSP Angelo-Marinos NRC/0SP B. D. Liaw NRC/OSP .G. G. Zech NRC/OSP Kent W. Brown TVA/DNE Mark J. Burzynski TVA/ SON Licensing O. J. Mavro Stone & Webster ENG T. A.'Ippolito TVA/ Licensing i W. S. Raughley TVA/DNE ] Timothy M. Shea Stone & Webster Eng Corp Hukam Garg NRC/OSP Pau1= Gill NRC/NRR Jamie Guillen NRC/NRR E. F. Goodwin NRC/OSP F. M. Sittason Wyle Laboratories Jim Glenson WYLE Melinda Malloy NRC/OSP 1 Frank Ashe NRC/OSP/CPPD J. E. Lyons NRC/OSP/CPPD W. S. Fanner NRC/RES J. J. LaMarca TU Electric K.'A. Petty Stone & Webster Eng. Corp R. C. Pierson NRC/OSP G. T. Hubbard NRC/OSP J. N. Donohew-NRC/OSP G. E. Gears NRC/OSP J. F. Stang NRC/0SP J. Clifford NRC/OSP B. Zaleman. NRC/OSP Jack Redding TU Electric Arthur R. Fitzpatrick Stone & Webster Eng Corp R. Luther Consultant /TVA T. A. Balaska Ins. Pwr. Cable Sources-Consultant to SWEC/TVA Carol Ayers TVA/0NP Michael Mennone Rockbestos Robert J. Gehm Rockbestos Douglas Nichols TVA/0GC Peter V. Judd Cousultant to TVA Jim Hutson TVA Phil Polk TVA-Bethesda J
$ yg - ENCLOSURE 3 -- y + i ' TENNESSEE VALLEY AUTHORITY. SEQUOYAH NUCLEAR PLANT-(SQN) UNIT 2 RESOLUTION OF CONCERNS.0N. r . SILICONE RUBBERLINSULATED CABLES PRESENTED..T0 THE U. S. NUCLEAR REGULATORY COMMISSION BETHESDA, MARYLAND NOVEMBER 24, 1987 /DNE4 -:1018W i-
f AGENDA MEETING BETWEEN TVA AND NRC S11 TCONE RilRBER f Nstil ATFD CABLES I.. MEETING OBJECTIVE II. ' REVIEW OF BACKGROUND' III.- RESOLUTION OF' SILICONE RUBBER' INSULATED CABLE CONCERNS
- 1. ' APPROACH 2.
RESULTS .3. CONCLUSIONS IV. CURRENT INDUSTRY ACTIVITY. m l l L. ) .t l i l I
b__-.___________.
d 8 s i MEETING OBJECTIVE Sil TCONE RURRFR INSill_ATED C ARI FS TVA HAS CONCLUDED ITS REVIEW 0F THE CONCERNS O i INSULATED CABLE AND WILL. PROVIDE'AND EXPLAIN TVA'S FINAL POSITION REVIEW THE WYLE TEST PROGRAM AND ITS RESULTS REVIEW THE SIGNIFICANCE OF DIELECTRIC STRENGT RELATES TO TVA'S TEST PROGRAM REVIEW CURRENT INDUSTRY ACTIVITIES CORRECT MISCONCEPTION REGARDING INSITU BREAKDOWN LEVELS i -_m__._____m_
1 J i-Ei i i i l SEQUOYAH CABLE' TESTING l 1 .NEED FOR-TESTS' STEMMED FROM INSTALLATION CONCERNS 1. JAMMING J 1 s 2. PULLBYS 3. LACK OF SUPPORT IN VERTICAL i CONDUIT RUNS ' CABLES TESTED AT 240 V/ mil DC " MINIMUM QUALIFIED THICKNESS" USED TO ' DETERMINE-MAGNITUDE. 1 WORST CASE CONFIGURATIONS. TESTED ' ACCEPTANCE CRITERIA PER IEEE 141 ' PARAGRAPH 11.11.4. J l i a
) ggGUOYAH' DESIGN PHILOSOPHY b INSIDE CONTAINMENT ALL LV POWER AND CONTROL CABLES ARE 1/C SILICONE RUBBER INSULATED WITH ASBESTOS BR A I DED : JACKETS l APPROXIMATELY 250 10CFR50.49 i CABLES CAPPROX. 950 CONDUCTORS) I THREE VENDORS NO - SELF-HEATING UNDER NORMAL OPERATION OUTSIDE CONTAINMENT LV POWER CABLES ARE 1/C AND M/C XLPE OR EPR-WITH CSPE JACKETS CONTROL CABLES ARE M/C XLPE OR EPR WITH CSPE JACKETS OR THERMO-PLASTIC-POLYETHYLENE INSULATION WITH PVC JACKETS 4 4
'7AMMING TEST R E S U L_ T S 45-CONDUCTORS < 15 - CABLES > -TESTED-IN 15 CONDUITS ALL C A B LESC OUTSIDE CONTAINMENT 1O CONDUITS DRY DURING TESTS 5 CONDUITS WET DURING HIPOT. ) ALL CONDUCTORS PASSED ~7200 VDC -44 CONDUCTORS HAD PI 1.25 OR GREATER. 1 CONDUCTOR HAD PI 31.O. BUT 1 25 i e
-] l 6 1 PULLBY TEST RESULTS s J S78~ CONDUCTORS C298 CABLES) TESTED . IN 15. CONDUITS l ? r. ALL CONDUITS. OUTSIDE CONTAINMENT 12 CONDUITS' DRY DURING TESTS 3 CONDUITS WET DURING HIPOT ALL CONDUCTORS PASSED HIPOT -VARIOUS VOLTAGES (4800, 7200 VDC) S74 CONDUCTORS HAD PI 1.25 OR GREATER 4 CONDUCTORS HAD PI 1.0 BUT 1.25 O ' ^ - "--------m.
c ) I 1 l VERTICAL SUPPORT SELECTION CRITERIA I { l I J l CONDUIT CONTAINS SILICONE RUBBER-INSULATED-CABLES ) t LOCATED-IN CONTAINMENT l l ) MINIMUM OF 5 CABLES t-MINIMUM O F:~ 20 PERCENT fr I L L 90 ' DEGREE CONDULET AT TCP O f= RUN l J DROP EXCEEDS NEC ARTICLE 300-19 1
i TEST OF CABLES IN VERTICAL CONDUIT l
- 1 6 CONDUCTORS (6
CABLES) IN 4 ONE CONDUIT. ALL ~ CABLES SILICONE RUBBER,: 1/C. CONDUIT LOCATED IN CONTAINMENT CONDUIT TESTED DRY SABLES TESTED AT 10,S00 VDC I '40 FAILURES AT SUPPORT POINT 3 BREAKDOWNS (7500, 10,000, AND .0 900 VDC AFTER 1 MINUTE) 9 CONDUCTOR HAD P I. 1.0 e DECONDUCTORS HAD PI 1.25 BUT g 1.0 h CONDUCTORS HAD PI 1.25 OR GREATER OST LIKELY CAUSE DETERMINED O BE IMPACT f i I ^ ~ ~
..a. c ..s. e? t0 10 CABLE #2 e M/ FAULT LOCATION CABLIS 1 AND 3 FAULT LOCATION f AND AREA,OF C HIGE LEAKAGE ,y ~ 'f'8O YOR CABLE 4 ~ N 6b' '\\. = l ,s a .g. e 8 I \\ / gg ?. p a ~ g e s gt g 80 CONDUIT CONFIGURATION
H I -F:' O T TEST F:' A I L U R E ANALYSIS l NO - CONTAMINANTS NO AGE-RELATED-DEGRADATION CABLES WERE WITHIN SPEC OD DAMAGE FOUND. WAS HIGHLY LOCALIZED MOST LIKELY. CAUSE WAS IMPACT S MILS REMAINING INTACT INSULATION S
n_ ) T s\\ } L 'K
- ,,/
U A C C s'. I ~ P Y / T ,I. l. ( =. E ~. G s A r-t'l M e, A ._,7 D L / x'~%,'s. T L C A A W P M i l D I E C U D E R l1 lll 1
de cap for ha dling he ghts ./ g L l Acrylic tube Weights f R / Ruler ~ 2" dia. impact place Cable 1" Steel plate Sample e ~ FIGURE III
= =
VDAC.. .=
== 9 E e 4
] SON UNIT 2 INSIDE CONTAINMENT 10CFR 50.49 SILICONE RUBBER INSULATED CABLE
SUMMARY
ANOMALIFS MANUFACTURER CONDUCTORS POLARIZATION INSTAllFD TESTED
- BREAKDOWN INDFX AIW 282 32 6
1 ANACONDA 305 18 0 0 ROCKBESTOS 32R al 1 2 TOTAL 961 91 7 3 FOLLOWING TESTING OF THE ORIGINAL 16 AIW CONDUCTORS FOR THE VERTICAL SUPPORT ISSUE, ALL ADDITIONAL CONDUCTORS WERE SELECTED AS A WORST CASE SAMPLE BASED ON THE LONGEST CONDUCTOR LENGTH BY MANUFACTURER. l l l l -i DNE4 - 1035W )
J .y l " ADVANCED POWER CABLE TECHNOLOGY" J T. TANAKA AND A. GREENWOOD CRC PRESS 1983 VOLUME 1, PAGE 169 FIFCTRICAL TESTS DC HIGH VOLTAGE METHOD THIS METHOD JUDGES THE CABLE INSULATION PERFORMANCE FROM l THE TEMPORAL CHANGE IN CURRENT WHEN A DC VOLTAGE IS APPLIED BETWEEN CABLE CONDUCTOR AND SHEATH (TABLE 3.3.1). THE CURRENT GENERALLY COMPRISES THREE COMP 0NENTS:THE DISPLACEMENT CURRENT WHICH IS LARGE BUT DECREASES DRAMATICALLY A SHORT TIME AFTER THE DC VOLTAGE IS APPLIED, THE ABSORPTION CURRENT WHICH DECREASES' GRADUALLY OVER A COMPARATIVELY LONG TIME, AND THE LEAKAGE CURRENT WHICH DOES NOT CHANGE WITH TIME. THE ABSORPTION CURRENT IN THE SECOND STEP 0F CURRENT RESPONSE IS ACCEPTED AS AN inPICATOR OF A CERTAIN INSULATION PERFORMANCE IN POWER CABLES: IT IS EXPRESSED BY SEVERAL MEASURES SUCH AS THE POLARITY INDEX, THE INTERPHASE IMBALANCE COEFFICIENT, THE WEAK-SPOT RATIO, AND THE CURRENT KICK.- THE POLARITY INDEX IS THE RATIO 0F THE CURRENT 1 MIN AFTER VOLTAGE APPLICATION IN THE CURRENT AT A SPECIFIED LATER TIME (7 TO 10 MIN AFTER VOLTAGE APPLICATION). THERE IS AGREEMENT THAT THE POLARITY INDEX TENDS TO DECREASE AND THE LEAKAGE CURRENT INCREASE AS OIL-IMPREGNATED PAPER INSULATION DETERIORATES. 310Ui MEASURF. HOWFVFR. SFFMS TO BF INFFFFCTIVF FOR EXTRUDED CABIF INSUIATION. I DNE4 - 103SW b
SILICONE RUBBER INSULATED CONDUCTORS ANALYSIS OF TEST ANOMALIES BRFAKDOWN VOLTAGE MANUFACTURER AMERICAN INSULATED WIRE 7,500 V DC 8,000 V DC 8,000 V-DC 9,600 V DC 10,000 V Dc 10,800 V Dc 7,000 V DC ROCKBESTOS ) ) r DNE4 - 1018W
TVA PLAN FOR RESOLUTION ATTEMPT TO QUALIFY REDUCED WALL CABLE 'THIS WOULD; 1. PROVE REDUCED WALL IS OF NO
- CONCERN, ORg 2.
ALLOW FOR REDUCED TEST VOLTAGES POSSIBLE METHODS TO REDUCE CABLE WALL 1. EXTRUSION i 2. SHAVING 3. BUFFING 4. IMPACT TARGET THICKNESSES 8, 15, 20, 25, AND 30 MILS 10 YEARS TO TARGET QUALIFICATION FACILITATE SEQUOYAH RESTART i
D ~ xI E C UL E DL L EA P RW M A S D R E E R P I P S O E C D
R ~ O F K N E D O T E I E S T Z U A L E U U S Q N I I -',1 N H M C U E M T I N \\ T I N M E R G R N U I C R U Y S D A D E E M
n l' L E . TEST PROGRAM AT WYLE LABS IFFE 323 1974. IFFE 383-1974. 4 1 l J TVAL PREPARED SAMPLES ] 1 NO THERMAL-AGING REQUIRED o L l l 10 YEARS NORMAL RADIATION PLUS l ACCIDENTLRADIATION PLUS MARGIN 1 (LESS THAN 1.0 E08 RADS) L l '100 DAY LOCA CHEM-SPRAY-l i " ACCEPTANCE CRITERIA - HOLD RATED VOLTAGE AND CURRENT THROUGH LOCA L IR TAKEN WET AND DRY PRE-AND POST-LOCA 1 POST ~LOCA HI-POT TESTS WET AND DRY WET DC BREAKDOWN TESTS ) i .l DNE4 - 1018W l'__-.---_--____ u
W s Qualificcci:n Picn N3. 17460-59 REVISION A 6 Pago N.9. I 9.t$d 3 BOSS 3,Wd O 30 tr, v.i *. a f .-x -o T. E .Vb g [ ..M d' . - _.. l . dl .g ~., l g 3 , i.. _.l ,g..- p. ..._ g- ..I.. ~y... -C, g gg 3 ~.- - _ .~ p ..g.. r_ - t.... l ..m l .g ~. .. y y4..._. y g .eg -.1 i. ,l __f ._ g g ( ., P'* .....,f.. .5 ..,g - +... /. p...... _ _ _.. f. 2 .t . 4. l _.... _ -u,- ._,. /.... ;. - _4 .l .a ....J +,. at . _.s. -. o .W 3. nC..
- n...
.. ~. - - - / _... ..g .= ~../.. T'.
- P c
. f _.. ..... j. 4L. ._g .4,.. ..a.... g ...W.Z . i..._. _ 1.. I
- i,,.
=N
- a. -
. _ X >=. ._......./. m ..U j ...e. ._%,J. _L -_....g .e y,_. .w.. ...O. a O ..l j__ -t m, ......_.z...... ~ ... ]... w -..,... ~. - _.7.o. 4 ; ; .....*~~.} ~_ - ~ ~. ~ ...g ..g R -...-..-.f = an --.o l_ V0 n .. j u6,....- ... g_i g g_. y a g r e l _.. M .+ i..._.._ ...-~*~1....
- ~.
~ n -{.......- y, e .. a., .l n a .. g ..4 ...l ~. .I m ,.d ensus-i o .g -na ~ ~ ~ ~ ~ m 5 3 __. S # 8 8 8 R a p a I a u s n n n .-1, 3M.LWGdH3.L
WET / DRY "HI-POT TEST VALUES 'S MINUTES 500 VDC i 1 , MINUTE 1 0 0 0' VDC 1 MINUTE 2000' VDC 5 MINUTES' { i ~3000 VDC 1 1 MINUTE '4000 VDC 5 MINUTES 4700 VDC' 4 i ADDITIONAL STEP AT 240V/ MIL DC BASED ON THE MINIMUM MEASURED THICKNESS. REMAINED AT THAT LEVEL 5 MINUTES l s \\ ) O
TEST SPECIMEN DESCRIPTION ORIGINALLY ESTIMATED MINIMUM MEASURED REDUCED-INSULATION INSULATION THICKNESS MANUFACTURER SPECIMEN THICKNESS (MILS) (i 2 MILS) AIW 1 8 i 3 15 4 4 20 4 5 25 8 6 30 9 ANACONDA 1 8 2 2 15 4 3 20 4 4 25 4 { ROCKBESTOS 1 8 4 3 15 4 4 20 6 5 25 9 7 N/A* N/A* 8 N/A' N/A*
- IMPACTED SPECIMENS DNE4 - 1018W
') s WYLE TEST PROGRAM POST LOCA HIGH POTENTIAL BREAKDOWN-1 4 TFST RESULTS I MINIMUM MEASURED MINIMUM j BREAKDOWN INSULATION' DIELECTRIC-VOLTAGE (VDC) THICKNESS STRENGTH i MANUFACTURER SPECIMEN a 500 VDc/SEC (MILS)* (VDc/ MIL) AIW l' 18,000 4 4,500 3 42,000 6 . 7,000 j 4 60,000 6 >10,000 f j 5 60,000 10 > 6,000 l 6 60,000 11 > 5,455 I ANACONDA 1 12,000 4 3,000 2 28,000 6 4,667 3-40,000 6 6,667 4 36,000 6 6,000 ROCKBESTOS 1 12,000 6 2,000 I 3 25,000 6 4,167 4 40,000 8 5,000 5 58,000 11 >5,273-7 55,000 N/A N/A 8 5,500 N/A N/A l t MEASURED THICKNESSES HAVE CONSERVATIVELY BEEN INCREA BY 2 MILS TO ACCOUNT FOR THE ACCURACY OF THE MEASURING j i DEVICE. .DNE4 - 1018W
NEW CABLE HIGH POTENTIAL -ERFAKDOWN TEST RESULTS NOMINAL INSULATION DIELECTRIC BREAKDOWN GROUND THICKNESS STRENGTH MANUFACTURER VOLTAGE (KVDc) PLANE (MILS) (VDc/ MIL) AIW 50 F0IL 45 1111 53 F0IL-45 1178,. l l 59 F0ll 45 1311 64 FOIL 45 1422 l f l 68 WATER .45 1511 ) 70 WATER 45 1556-l .f 87 WATER 45 1933 s ANACONDA 30 WATER 45 667 35 WATER 45 778 37 WATER 45 882 ROCKBESTOS 95 WATER 45-2111 i 100 WATER 45 2222 >100 WATER 45 >2222 1 DNE4-1031W l l- )
j ) SIGNIFICANCE OF THE DIELECTRIC STRENGTH TEST 1 (ASTM D149 APPENDIX X1) (ASTM D3755 APPENDIX X1) POSTULATED MECHANISMS OF DIELECTRIC BREAKDOWN ] BREAKDOWN CAUSED BY ELECTRICAL DISCHARGES THERMAL BREAKDOWN 1 i INTRINSIC BREAKDOWN l 1 TVA TEST RESULTS WERE COMPARED WITH INTRINSIC BREAKDOWN VALUES l 0F THE-INSULATION FAILING TO CONSIDER OTHER MECHANISMS l l DIELECTRIC STRENGTH IS GREATLY DEPENDENT UPON THE SPECIMEN THICKNESS. l VARIES INVERSELY AS A FRACTIONAL POWER OF THE SPECIMEN THICKNESS -- TVA TEST PROGRAM ASSUMED A LINEAR RELATIONSHIP, ESTABLISHING TEST VOLTAGE a 240 VDC/ MIL THIS PLACED A GREATER STRESS ON THICKER INSULATIONS SUCH AS SILICONE RUBBER AT 45 MILS NOMINAL BASED ON INSULATION THICKNESSES WHICH HAVE BEEN DEMONSTRATED T0 i BE EXTREMELY CONSERVATIVE FOR BOTH ELECTRICAL AND MECHANICAL PROPERTIES DIELECTRIC STRENGTH DECREASES WITH INCREASING TEMPERATURES TVA TEST TEMPERATURE INSIDE CONTAINMENT EXCEEDED THAT OF " BASELINE" TESTS WHICH WERE AT LABORATORY ROOM AND WATER TEMPERATURES AND MAY HAVE RESULTED IN' LOWER. WITHSTAND TEST VOLTAGES BY COMPARISON BREAKDOWN VOLTAGE WILL TEND TO INCREASE WITH INCREASING RATE OF VOLTAGE APPLICATION THERMAL BREAKDOWN MECHANISM IS TIME DEPENDENT DISCHARGE MECHANISM IS USUALLY TIME DEPENDENT TVA TEST RESULTS WERE COMPARED WITH DIELECTRIC BREAKDOWN VALUES PERFORMED AT 500 VDC/SECOND 1 IN-SITU FIELD TESTS PERFORMED AT APPR0XIMATELY 120 VDC/SECOND SHOULD EXPECT LOWER DIELECTRIC STRENGTH ON SLOW RISE WITHSTAND TEST THAN ON SHORT TIME (RAPID RISE) BREAKDOWN TEST 1 DNE4 - 1035W
SIGNIFICANCE OF THE DIELECTRIC STRENGTH TEST (ASTM D149 APPENDIX X1) (ASTM D3755 APPFNDIX X1) DIELECTRIC BREAKDOWN VOLTAGE INCREASES WHEN THE TEST SPEC IMMERSED IN A LIQUID DIELECTRIC, MINIMIZING THE EFFECTS OF SURFACE l DISCHARCES-RESULTS IN ONE MEDIUM CANNOT BE COMPARED WITH THOSE IN MEDIUM LACK OF A SURROUNDING LIQUID DIELECTRIC MAY HAVE RESULTED -SURFACE DISCHARGES AND REDUCED WITHSTAND VOLTAGES i EVALUATION "THE DIELECTRIC BREAKDOWN VOLTAGE TEST REPRESENTS A CONVENI PRELIMINARY TEST..." ... BUT THE FINAL CONSIDERATION MUST ALWAYS BE THAT OF THE l PERFORMANCE 0F THE MATERIAL IN ACTUAL SERVICE." 1 I WYLE. TESTS DEMONSTRATE THAT THE DIELECTRIC STRENGTH 0Fi RUBBER DOES NOT DECREASE WITHIN THE SPECIFIED AGING PAR REDUCING THE NEED FOR HIGH POTENTIAL TESTS AS A MEASURE FUTURE ADEQUACY TVA HAS EXPERIENCED NO SILICONE RUBBER INSULATED CABLE IN SERVICE l l 4 DNE4 - 1035W
) COMPARISON OF TEST VOLTAGES BASED ON-VARIOUS WITHSTAND STRENGTH REQUIREMENTS VS ACTUAL BREAKDOWNS EXPFRIFNCED IN TVA FIFLD TESTING-(VDC) MANUFACTURER (MINIMUM 00AtIFIFD THICKNESS-MILS) SUSTAINED WITHSTAND REQUIREMENT AIW (4) ANACONDA (4) ROCKBESTOS (6) 1 ENVIRONMENTAL QUALIFICATION; IEEE 383-1974 240VDC/ MIL (5 MINUTE WITHSTAND) 960 960 1440 NEW' CABLE AT MANUFACTURING: ICEA S-19-81 300 VDC/ MIL (5 MINUTE WITHSTAND) 1200 1200 1800 NEW CABLE IN DESIGN STAGE: TYPICAL PRODUCT DATA SHEET 600 VDC/ MIL (5 MINUTE WITHSTAND) 2400 2400 3600 MINIMUM BREAKDOWN EXPERIENCED IN TVA FIELDLTESTING i (SLOW RISE BREAKDOWN) 7500 NO FAILURES 7000 I l l 'DNE4-1031W
POST LOCA OBSERVATIONS 4-OBSERVATIONS NO. AGE RELATED' CRACKING 0CCURRED NO LOCA INDUCED CRACKING OR SPLITTING OCCURRED-NO CRACKING OR SPLITTING OCCURRED DURING THE MANDREL REBEND NO SIGNIFICANT DIMENSIONAL CHANGES WERE OBSERVED FOLLOWING LOCA EXPOSURE. couct us10Ns SILICONE RUBBER IS:A HIGHLY STABLE COMPOUND FOR THE AGING, ACCIDENT AND MECHANICAL-STRESSES IMPOSED I 4 9
LJ i CURRFNT INDUSTRY ACTIVITIES o INSULATED CONDUCTORS COMMITTEE-(IEEE POWER ENGINEERING SOCIETY). TASK FORCE 14-4, " EVALUATING INSTALLED STATION CABLES."- NUCLEAR-POWER ENGINEERING COMMITTEE-NPEC (IEEE POWER
- ENGINEERING SOCIETY) 1 AD HOC WORKING GROUP ON IN-SITU CABLE TESTING.
ELECTRIC POWER RESEARCH INSTITUTE - EPRI EPRI/ INDUSTRY CABLE MONITORING PROGRAM. l l i 4 l \\ l l DNE4 - 1018W l )
a ^r-w ~ CONCLUEION ~ NO FURTHER WORK WARRANTED CONCLUSIVE EVIDENCE SUBSTANTIATES THAT 4 CABLES.WILL PERFORM THEIR INTENDED-SAFETY. FUNCTION ~ Jl 1. MOST EXTENSIVE CABLE TESTING OF ANY NUCLEAR FACILITY- ,}
- 14. INSITU TESTS HIGHEST VOLTAGE MOST CONSERVATIVE ACCEPTANCE CRITERIA 1
LARGEST SAMPLE EXCEEDS INDUSTRY' FIELD TEST. STANDARDS AND REGULATORY GUIDANCE ASTM.D149 AND D37S5 - . DEMONSTRATES MORE CONSERVATISM' B. _AND EVALUATIONSEXTENSIVE LABORATO DIELECTRIC, ENVIRONMENTAL, CHEMICAL, MECHANICAL 2. DEMONSTRATED BY APPLICATION OF i MINIMUM EQ THICKNESS TO ORIGINAL i DIELECTRIC STRENGTH REQUIREMENTS THAT CABLES ARE AS. GOOD AS' ORIGINAL i MANUFACTURING AND QUALIFICATION REQUIREMENTS O O 6
l 3. DEMONSTRATED BY TEST 100 PERCENT OF SAMPLE WILL HOLD I FAR IN EXCESS OF RATED SERVICE VOLTAGE 100 PERCENT OF SAMPLE EXCEEDS j MINIMUM EQ THICKNESS j 4. PROGRAM MODELLED TO EXCEED ALL LICENSING PRECEDENCE GIVING ADDED ASSURANCE PROBLEM NOT SIGNIFICANT AS TVA HAS QUALIFIED AS LOW AS 4 MILS OUT 5. 4 OF 45 MIL NOMINAL WALL EXCEEDED ORIGINAL NRC REQUEST. j 6. REMOVED MANY WORSE CASE AND i ALL / LABORATORY TESTED. ANOMALIES _EVALUATEpSATISFACTORILY. 7. EXTENT 0 TESTING ASSURES THIS IS SAFEST A.TERNATIVE. CONTRARY TO NRC NOVEMBER 13,-1987 LETTER, NO FAILURES OCCURRED BELOW .8. THIS SHOULD REMOVE NRC'S 7Kv. " HEIGHTENED LEVEL OF UNCERTAINTY." o e 4 4 W 0
- . g. q,p.,,
I 2 4
- SIGNIFICANCE- 0F 'THE DIELECTRIC STRENGTH TEST
-(ASTM'D149 APPENDIX X1) (ASTM D3755 APPENDIX X1Y POSTULATED MECHANISMS'0F DIELECTRIC BREAKDOWN BREAKDOWN CAUSED BY-ELECTRICAL DISCHARGES THERMAL BREAKDOWN INTRINSIC BREAKDOWN TVA TESTERESULTS WERE. COMPARED WITH INTRINSIC BREAKDOWN-VALUES OF THE INSULATION FAILING TO CONSIDER OTHER MECHANISMS; DIELECTRIC STRENGTH'IS GREATLY DEPENDENT UPON THE SPECIMEN THICKNESS VARIES INVERSELY AS A FRACTIONAL POWER OF THE SPECIMEN THICKNESS TVA TEST PROGRAM ASSUMED A LINEAR RELATIONSHIP, ESTABLISHING. TEST VOLTAGE-a 240 VDc/ MIL THIS PLACED A GREATER STRESS ON THICKER INSULATIONS SUCH AS SILICONE 1 RUBBER-AT 45' MILS NOMINAL BASED ON-INSULATION THICKNESSES.WHICH'HAVE BEEN DEMONSTRATED 1TO-BE EXTREMELY: CONSERVATIVE FOR BOTH ELECTRICAL:AND" MECHANICAL! PROPERTIES DIELECTRIC STRENGTH DECREASES WITH INCREASING TEMPERATURES TVA TEST TEMPERATURE INSIDE CONTAINMENT EXCEEDED THAT OF-i " BASELINE" TESTS WHICH WERE AT LABORATORY ROOM AND WATER TEMPERATURES AND MAY'HAVE RESULTED IN LOWER WTTHSTAND TEST VOLTAGES BY COMPARISON BREAKDOWN VOLTAGE WILL TEND TO INCREASE WITH' INCREASING RATE OF VOLTAGE 1 APPLICATION u THERMAL BREAKDOWN MECHANISM IS TIME DEPENDENT DISCHARGE MECHANISM IS USUALLY TIME DEPENDENT TVA TEST RESULTS WERE COMPARED WITH DIELECTRIC BREAKDOWN VALUES' PERFORMED AT S00 VDC/SECOND IN-SITU FIELD TESTS PERFORMED AT APPR0XIMATELY 120 VDC/SECOND SHOULD EXPECT LOWER DIELECTRIC STRENGTH ON SLOW RISE WITHSTAND , TEST THAN ON SHORT TIME (RAPID RISE) BREAKDOWN TEST DNE4 - 1035W
"* SIGNIFICANCE OF THE DIELECTRIC -{ STRENGTH TEST 1 (ASTM D149 APPENDIX X1) (ASTM D3755'APPFNDIX X1) l DIELECTRIC BREAKDOWN VOLTAGE INCREASES WHEN THE TEST SPECIM IMMERSED IN A LIQUID DIELECTRIC, MINIMIZING THE EFFECTS OF SURFACE DISCHARGES RESULTS IN ONE MEDIUM CANNOT BE COMPARED WITH THOSE IN A l MEDIUM j I LACK OF A SURROUNDING LIQUID DIELECTRIC MAY HAVE RESULTED I SURFACE DISCHARGES AND REDUCED WITHSTAND VOLTAGES ] EVALUATION "THE DIELECTRIC BREAKDOWN V0LTAGE TEST REPRESENTS A CONVEN PRELIMINARY TEST..." ... BUT THE FINAL CONSIDERATION MUST ALWAYS BE THAT OF THE PERFORMANCE 0F THE MATERIAL IN ACTUAL SERVICE." l l WYLE TESTS DEMONSTRATE THAT THE DIELECTRIC STRENGTH O l RUBBER DOES NOT DECREASE WITHIN THE SPECIFIED AGING PA1 REDUCING THE NEED FOR HIGH POTENTIAL TESTS AS A MEASURE FUTURE ADEQUACY TVA HAS EXPERIENCED NO SILICONE RUBBER INSULATED CABL IN SERVICE l i I 1 ~ DNE4 - 1035W l
P u stes.es -.a- -. g r; T n,
- ==.====....c,,
S Standard Test Method for DIELECTRIC BREAKDOWN VOLTAGE AND DIELECTRIC STRENGTH OF SOLID ELECTRICAL INSULATING F MATERIALS UNDER DIRECT. VOLTAGE STRESS' N meade,4. -d.aencenwean = omt.nw 6.n .n,4,,m w, kawiin, we.e use wm,.o.4 enpass amenwa ur. is inn saw at maus. ear ww w ams enama 4 aww w m.tJL 1 A gesp.WWige selshwe 4.$ seda seet de ebenerget alhamp.AmT ifW last if%ttenut.8 anspgWn D:4M Symfothm for Fiwent Conswtion
- 1. 'icope laboratawy Osens (w Elecawal Insulaimm' 1.1 This test method casers the determinanon D,tas7 spmenaik kw Mineral Insulaims of dielecttw twws&down volcaer and dickxt.icOilt'salin fkurnal Apqtaratus' strength of solid ciectrwai insulatms mawrtak 2 2 I"4'Nwd T*8th*8td X'd'648'd* l*'#Ne' under direct.woitage stress.
S8"ddal? l.2 Since sonw materials mavire servial treat. AN51 Cost.l Technnaua for'Onivtrw Tots. ment, reference shoukt also tw made ki ASTM IFI:listandard No 4. specifications or to the test method directi) ap. plicaNe no the matenal to be tested. 5x Tess y,g g g Method D 149 for stie determmauon tidniorwil eh Assw l'nsda.un inhoec-Refer to meength ofelectneal insulatins matenals as osm, Detiniskms D 1711. snevetal power (miuencies. 12 dh*"* '"v"e'h-Refer to Dermunw.s IJ 77si.utunduidnwrmndri ha:oriAwienw. U IIII tevrals.esamuseau.amlar tmusw.11rn varuhmt A) 'I"d*" * *0'**INII"""*' U l'II' m ska s me purtwar ten addern allet sk suksrImh Aws meresotal m*h we av. Is is sk nmrevuebd. Je r o< the tous ed shoutunderJ on estaNu Jr upres '8.1 The nrwunen. helit in a properly sloagnol h prmare us&fr unalhads4 twintan med divernure.shrulydasvAsht. red nsrtdweerrlimste JPP GD'* 's an incresung dirWt stiltage unut I sw. Specifee presaution stancments are snen in internal twakakmn awurs. The test soltay a Sartion 7. apptwd as a unikwm rate o(inaense. The darwt sediapr is <*itanned from a high st*ap suppi) of
- 1. Refervered Derweness adotualenserent apaat) and regulatum reanim.
2.1.I.V13t StundanA; att nople free. wah (acilities few messunns and D 149 Test Mettmid Gw Dnterw BaraLiksen "*8"I'nS the snapui stway. Yidtaer and Onistrw Strensch 44 Edul Elentwal Insulauns Mawnals at Comnwr. esal Pivearr 1 mournors'
- Th== a=*='..aarr
- svand=== at asm (*
D 176 Methista 64 Tosing Edni Filling and Treasing Coenpounds 18 erd liv Electsul m."m."ne "..d sw.s.' names," ope st en a"nsw=4 t r r.n asy weJs.n. me eweinnswi Insulatam' '"***'*"*'P*""'8*'8***1**8"'*= D a77 Tesa Meitual Ew Die 6attw Breakdoen TM[.,. siv 3,, v.e mn: Yoltaar of Insulaun6.aquads Umns Disk
- ===' w e.orw w vmso.oe 1
Elenrodes* U",,",",,d,,8j'f,,,',,'*,",,'",;,',81,,,,,,,,,, '. D 1711 Drfinnkms of Terms Retaung to Elec. tem asume.or. m vaa.wr iestsL trwallneulatese* o = - ..I d W Cm 1079 g.g3 e,
I M w e oms S. Signiflemate and Use L2 l*n/ ruse Gnund, that will enable the test V 3.8 This test niethod is latended for une as a voltage to be inemanad at a linear rate. Preference control and acceptamos test for dinct.vohage ap, should be given to a m", f motor.dttven pescations, it may be used also in the partial wohage control owr a manual control The rate. S evaluation d material for specific end uses and o(-nse o(test voltage shaK not very more than a as a means for detacting changes in matenal duc 20 'E from the speci6ed rate at any point 6.3 iWrmerer, to measure the vohage directly to specific detenorstmg causes. S.2 Espenence indicates that the breakdown applied to the electrode system. The responer of value obtamed with direct voltage usuaHy will tw the voltmeter shall be such that its time lag shall approximately 2 to 4 times the rms value of the not introduce an error gnaler than 1 % of fuu 60 Hr alternaung.vuhase breakdown. scale at any rate.of.nse used. The owran accu. F 3.3 For a nonhomogeneous test specimen. the racy of the vehmeter and the vohage.measunns disinhuuon of snitage stress within the specimen device used shall be such that the measurement erfor will not esceed 22 *E of full scale and he in is determined by impedance (largefy capacitiwi accordance with ANSI C618.l. with alternating wohase. With an incnasing di. 6.4 E4=vneter rett voltage. the voltage distnbution may be sull 6.4.1 For those cases when the insulating ma. largely capacitive, but depends partly on the rate of voltage increase. AAct sacady applicauon of terial is in the form of flat sheet or 12pe. or is of the nature of a semisolid (for esampie. smase direct vouage the voltage divimon across the tesa specimen is determined by resistance. The choece potting matenal etc. the electrodes may he se. lected from those listed in Table 2 a(Test Method o(direct or ahernating vohage depends upon the D 149. The electnute contas pressure shall he purpose for which the hreakdown test is to he used. and to some essent. on the intended appii, adcquate to obtain simul cicctrical contact. 6.4.2 Where curtlent electrode contact is cation c(the matenal. 3.4 A more compicte discusseon of the signar. oinsadcred important. paint or vapmud metal ekvtrodes are to he used. Such ciectrudes may icance of deelectne breakdown tests is given in Appendia XI c(this method and o(Test Method also he used when wcomen geometry prevents the use or nt, solid metal cics1oden. The resulta id D 149. Those appendas sections of Test Methnd D 149 that refer to alternating voltage are not obtaincd with minted or sprayed elcoudes may mn he comprahie wnh thine uhtamal umns
- applicahie to the direct. voltage method.
other types o(cisvinujes. 0* 6.5 Trv Chune/w-l'ir tests under other 6.1 As.ur (hrrev.l'nhvec Mme Nupphes. or than ambient conditions, the stuximen met he daciectric test sets of vanous voltage ratings, that platd in a suitahie envinmmental chamber of can operate with one u(the two output terminals ' adequase Swe. For tests at elevatof temperature. grounded, are commonly available:cummer. an oven that mes the ruluirementn o( 5mvire.. cially. Such appantum customanly sneludes the canon D 24% may tw cimvenient. The test necessary yohage.contnd, voltage. measuring. chamber must he nauspped with makty devices ~ and circust.enterrupung equipment. A proviseun, (Svuon 71. for retasning the treakdown voltage reading aAer 6.h finnens/ Nward-The power supply 4all breakdown is devraide. he equipped with a smunding switch that is 6.l.1 For a direct voltage derived from a ree. gravity operated and dcsegacd to cisne in less tireed and fihered pe=er (squency source. nppie than 0.5 s. The smunding switch Wall connect on the output vahagr generaHy should he less the high-vuhage output terminal of the parwer than a 'E. The criterian a enet 6(the timecunstant supply and ground terminal through a lisw resent. of the cirevet is at hast 0.4 s. The time constant anee when the mout supply pn=er is remmed or is product of the fiher capacitance plus the spee. the test chamher dimw is oprned. imen capacitance is macrofarada, and the speci. men insulation renstance lin mephms) corte.
- 7. Nelety Preesaniaan sponding to the parallel combinauen of the won.
7.1 WARNING-lethal sohaps may he meter croset sesistance and the specimen reses. present denna this test. It is essennal that the test
- ance, apparatus and all assocsaned equipment that may 2
.M Do g PRC8LEM HA,RO COPY __m_____ _ _ _
I l M "m "./ $ o rras be electneally connected to it may he progetty the matenal under test. its operatum mas N a V deugned and insalled for safe ciperatase. All puestise indacatim dhreakam n, 2.2 Fadure o(a sirevst breaker tooswrate mas metal parts that neight contact persons during not he a pseatise entenon of the aheme of thes test shed he solidly grounded. twr Ldoen. A breaker mas fad W anp bivause et S piwd to the test specimen, knh the spreimen 7.2 When a directeoitage test has been ap. is set for too great a current tv hmuse d mal. fun 61hm. On the other hand. if the inpping or. and power supply can remain charged aAer the cuit is set for uni km a arrent. arrents due to test soltage source has been de.cnersited. Thas M may present a haratd to test penonnel then1 leakage or punial discharye lominal may cause soltage toting may he more hasardousthan test. it w mp Nfore breaL&m n sintage is rea hed. N.3 Oherse the spromen dunny the wit w ing oth ahernaams soltase..here the charge on aserrtain that tripping d the breaker or currem. F the cumen is rapidly dasupatal in the kiw. emmvarme omdans of the test transforma aner sensing orwal is not aussi tw 4tahoser. % hen dashoser is a pn*4cm. it udt he navuan m the kN es de.enerysted. 7..t The seu spumen and high.sidtape output pnnade for nuwe ercepupe distansv around the d the emer supply must he emind in a ekvinnk s m avfrase senimen theeknem. or 86:. immerse the specimen in a in4ueJ dwkvine IN1N grounast musallie serven. Aarss to the test sn. sinuse must he dependent upon pruw grounding tam 1.4L of the pmn supply and test smurnen through a W.4 Otwenathm 4 A1ual rupure or asom. low ressstarnv as referred W in h.h. penstkm is omstise es miensv d sisurnen brnk. 7.4 A manual pruunding stil must tv umi emn. In W pastmm. h6 meter, thew phssnal so mmpactsty dischaspe the tot smumen armi esnienedbrukamn nsas fug N apturent. lf po.er supply aner the tot and prut to handling bewLam n is in sauotum it es osmmosi prA1we them. The gnainding samt ihmeki N iett in om. to n. peat the tot im the same vmmen. StoL-tat with the tot somerwn and high-sidtage &mn in omdrmal when reappheatum 4 tW transformer termmals for as long as feasable, s oitap iesults in a substantiate kmer brnka= n 7.3 WARNING-.Orone in,a physiolosseally sidtape. hasardous gas as cervsted on6tmrations. Lects E I'"8
- Ti"'**
at acwptable industrial espusure hate been es-4.1 1 or a doenptmies inf tot issunteris sif nia. tahdished by the Amenese Conferemt of Gov. tenals and their pnyuratnm. referrisy shall he ernment and laduarial Hypreist? Orone has a m.mle to the ufM methimis appewabie to ahe datinctive oder ihet is inihaay da6vrnebir at low concentratacus, has tempnnwy loss of the sense makanah to be testol. 4.2 t he smumens shall be trpresentatne d of smeel esa occv. k is hkely w he prearnt ediesever wohages esas that are sulTaient to cause the matenal to N tesksi. Sulfkiesit material shall partial ur runphree discharges in aw or other be asadahie to permit making the tesa in the atmosphem enntaimias osygre. When the odor prereratam d not sperimens (nwn solad mast. of sunne is -, "; preurns ur shen swone nais. take are that the surta6vs in mata61 *sth the ekunsks are putattet and as plane and prnerating conditions renumme, the mn6vntra. inte uf tuune in the alsnosphere shuuld he snesa snknith as the matenal permits sand mang c..
- 4 avadahle monotonns 4.4 17####.%fut.llaretrali t.%vre um/ /Wro devwen Appropnate means, samt as installatiost l.6 thuer / emer 17:n WThe test sotumen shall he d suffeesent area to present flashoter under of ethaessa vevita, shall he taken to masntasa tvuar concentistices in workin6 areas within the nindithms d tent.
4..l 17mi &dul.lfaraveulw The breakdown r.plahle arvedt d theek sided matenals is grneralls so high that 8 I'i'" d N the sonimen must he immened'in entulating duni to presem llaJdner and to mmamite puttial N.! Ihdmtse brokdown is penerally muwn. penskli hs an inklYane in curtTfit in the Ra ofruit dagQ. fmT $rG1hift I,1. (hhCT technhlues that may aisvale a senung element sue's as a ohich mag he usal to pervent dashoser are: oecust hersker. a fune, or currenMenung cistuit. If senutevwy n(ihe riement a metl annianated . sa=== rearmese a commmiu and imamasi Hs. with the charmienmact of the test entapment and area.r41 a Iest.cten ,ou c.in. 3 -1 Pit 08LEM HARD COPY
1 q M E $ 03735 9.4.1 The machining of a rocas in the test incation requemments in oeder to obtain repro. V specimen for an electrode. ducilde tueuits. 9.4.2 The um o(sh'ouds on the test specimen.
- 33. 5""ading Mediese 9.4J The application of a seeling apparatus S
~ rials in the mediurn in which they are to be used. under pressure to the upper and lower faces of 13.1 In erneral. it is preferable to test mate-the test specimen. whether asr. other gas, or an insulating liquid.
- 10. Thichaens Refer to the ASTM method appicable to the M
10.1 The thickncu used in computing the di. partcular matenal to tw tested to determine electne strength shall he the ascrate thskness of whether a medium is speci6ed. the specimen meawred as specified in the test 13.2 Where conditions of use are not well F method for the maten d insulsed. If not speci. defined. test matenals in air unlets an escessive fied, the thwknew mmrement shall be made amount of matenal is required to prewns flash. at nuam temperature i 25 2 !*C. over or excessave burmns o(the surface, la that 10.2 If the materr, is laminar or known to case. it is common precuce to test the matenal sary in daelectne strength with onentation. such under oil. Do not une liquid dielectnes as a as caused by grainancss. the specimen should be surrounding medium for porous matenals that cut 50 that its thskness is in the direction of the are intended to be used in air. Other gas, or electne field under use conditions. vacuum. Compenson d test results shall he lim. IU'S When than masenals. such as laminates. ited to those results made en the same surround. are to he tested in the detection of their width or ing medium, length, special proceduses may he needed to 13.3 When tests are to he made under ud, avoid flashower. some of which are desenhed in provide an sul hush u(adequate size. Use a giumi 9.4. Provisions for such tests may also be in-grade of dean transformer oil or simdar tul, in ciuded in the methods for specific matenais. accordance with Specification D.Wi? unks other sul is sectided. 14.larneedere , 11.1 Unless usherwne specifu:d. five tests shal 14.1 Inmaw the udtage at a undorm rate ' he made. innn eeru to brcakdown. Fiv a given maienal.
- 12. Ceedillemies refer to the apphcable spmfnatum or test 12.1 The dicicctne strensth u(rnust insulating methial for the rate. Llw a rate of.% n V/s if matenals vancs with temperature and humidity, applicahic. and 6 not otherwiw sptufnd. Cal.
Such matenals should he conditioned in a suita. culate the rate from measurements d time tr. Idy contrulled chamher. For information con. quired to rasw the udtage betwsvn two sekvicd cermns the conditioning treatmens. refer to the values. For a given matenal. refer to W apple particular methuJ for a seven matenaL Keep the cahie matenal somdention or test metheid few test specimens in the chamher long enough to the rate. Consader breakdown to have invurred reach a umform temperature and humidity he-unty when the vunditums d5atmen N hate h6vn fore tests are saaned.14 may he desirahie to met. Rmwd the test udtage at hveakdown. descrmine the dielectne hehavior of a matenal 14.2 Refer to Summ 2 for critena of break. over a ran e d temperature and humidity to down. which is a hkciy to he subsected in use. If con-14.3 Rmird W sidtare at breakdown. s ditsomns is performed under condatenas where
- 35. N'P8'1 condensation wdl uma. st may he desirahie to wipe off the surfaar ut the test specimen carefully l$.1 Unless otherwiw evirnd. the refert imenediatei) before tesung, as this will sencrally shallindude the 6dkowins' 13.1.1 Averzer thwkrw dihe *vinwn.
tend to minimue flashower 12.2 Since some masenals require a long time 13.1.2 Breakdown sidtage at each purwture. to astaan equshhnum at normal conditums ad l$.I.3 Average. masimum and mammum temperature and humnhty specify conditioning breakdown soitage for each speumen. when specimens are to he subpected to tests for 13.1.4 Aserage dicicctne strength o(W sptv. evaluation o(quahay control and purchase spec. irnens. 4 1082 E-2 PAOSLaiM HARD COPY 9 k
{ e d ~ h ~ ~ > vt v eji I5*1ai t I 2i:. g l Ii!a ]l i i mll! m }g!l 3 Il s lei 4hj i s2 i 2, u. i: cII !I i?5 :_ c;in ii! 5 ijj!l1 mi l a i!_ I1 p} 'h! ipa ilrl 3vlpaj i!j}I,j ie ti } #n jI d r i i m u> N Nlui i
- g j e tgP i
ldl l li i il l l 1i4i 't in3j n!{ i i } i - 1i3 af Il H 1, Ii j -) j i{ i; llij n $ g i,I i i c i(a, g t it iilll! i ii l , i4iii! a j i N i hi ul t lra 5 i a l!jsldsl 2 l !h i 4 pf i I3l f $ga i hh bdil j i i;l iI!jhgi )[
- .1; m.l l = !di H :!fi ] !f li l d3 t
1 a1i ']=i 5 n h]<ha%j n N i i lM4i}f l l5 = ! d}} $ [; i y i. b j! m$ ll,p}l\\l jit a$ i iis j1l t mI i*i i n!. iI W!"p!h1 1 h))}4gtl a! 8[ 1J 1l t i! i inj i,.,}2i4 ml 2j l i pj l i ! nf =: i u! aj i '1l t a 1';1 t a{
- a{
I ~, 23$:i 5 li
- = = +.
+ r L B ~~. e.- . ~.s _.mg. m. h.e me..r d hm e, y - V mamme usamms ammme er exhaus es te Isa seen. ieal aAme d enehage a che summer af names. memeammL Eamps e auss ehme er shus afsammuner se exhumt XLO T___ ___ The immemumme af the tes.umga a hung suussagent is a summmmuy se amamme summmmme and as asummetag usene amtmans the er tems es sutumus hemmesy aAmes by emadmui ama. exhauer swunge. ankmush ist man esammes amms % gemanhum6 waramumme af amamses - _ _ may haue a mapip. Der edsms. de esmusuL es estunne swough me8 dausmar mee insammes tempsuusases has the eswee to weste 18J Esubesim, the e iner degenes usan the meernal under tous. When Xlj.8 A Amadesmestal summuummes 4 the sumuda. se a haswa thes a maavue ont he voeussed te (maname time as alunnemi appermaus a thus e outuammd shr as asher them normal room temperosuse. n is omnetsel - oshage insamed en as an annum. Themisse thee as a that the duHertne stroph.4empersense retaessushes for pens ased for a tem to evolmee thr omsformoner af the masenal he driermined owr the range of esponed partarster smassumis as high schmer esemb The skatermer F essessing temorrasuset hsuchdene sehsgr tems sussuusmas a amersumum pudsm. Xl.4.4 Ame -Tem svissais wie er inAusemad, is inary ese to densamme ekseur a mansnel musus im> sosper essent, by the rear e( vainage appismaams. Dess to ther W hus is fuhn east af a emmelser a irnese degree thaa ahernessag vehage buumhdswa. A4 evehumsmo am tus sagannes sugesub Muu. sht emmes. the duer wehase changsk sim aanntisesem ef etunnemi has af a mansnia as enmanned e asummes e must cwseus se and en the samfame af the tem & ages AmNuess tem e amadassa en em tems. puunduty weh rhesgr oath Isans, le same assa, and nameuel delet= suged to thr samagnumma af the etmetrar Ruhl and slut huesse ef stannual chssges em the eerdssa may cassme sum a( mannel asemed en a, penalshutungr samsment as amssualms surfam Amasser, auschemani asuuk asseuses amadsmsm. and 1 Xl.4.3 Is'ser Fens s -la armerus the demiennt met audter mansnut6 5mseed, en sense ihmer use dres. sewegnb es iaAusased by me user isnm a' the aemised neemme smausseek homs, smartummel assuk punmi en. ouhaus. Watana its immmen samsuned in thss eschad the stargr taummel med as pnsdenk _ ear. mAusene af uese (ens a met soumament, weeste smay sehms ihr luuhdemo senser (er futus as Xl.4A aressnesher Meu6sme -The mervemasises euler a agemsy inessed. Sauer af etmar edsrs can smedmum can aNast er has summier seas esaarmai she-he muergusused as hhunmary esma. and a hrener sus-eksegrk t?ur (asid wedensmey. thmuhy pumasy seAusrere smear af shr samensi aus suumh has tem amel summedre. ing the tus runnsdes. Ramsts en amar ampuse commes he assma mese mimeys he thus es the ' af the sumissed wish themr as a shSmeus mushassa. nsaarnal se anumi survism. XI.4.7 Arderse fissuader-The oudmuse hasameny XIJJ The dartsant fuumanbewe tres susy he umud smAusases the dadurent usease en the eseems thus as a maestimi smasmans er asumia, namsud eras. m a asussue ahnread he. er as the murdue of. the mawnsi nimum af amirrywg saher twelemas test as venslehov. e mudar tem admets the estenne ime and marisse remhsr. nr en indirear enensrusses suureuse tant a namounal ersesy. Hruse, ses empertamme end dessed to a large apaqL la threr uses of the tne e a late nieser taher uf esases name aims assuse s( the annunal heing tsumuL the laun6dowe sanser tems e amensstame ender them ' Heureur, twee semanals thes staart betes er as anses. the alusiuer veder. TIvensveam sowee er Tromar and Adammumb am6ae mesnamsma segmeems airnahnse.#ame amwar muere owedm.-see ese est asuu._ f a seus eW l'eset se sese essadeders%pme%sfr messed sent 1. se ser omhdpr se ser em4 passes smart med ser asel af musesmuusser ad esse sumet asr assuade shser woe argmamatsher Time wandsela esduw se spesemus a mer esser er ser spesmeneshir essemned. . suedmuse Ar se*estsif asser ser evers emJ sf me arosesd amtse seems=ned se usedeum F.or. n se assend eatse av smose ad sese assuhsd se na addsmead usmuseds sud abandd Ar sud$ reed as ATT3d lieushmusesse F.usp _ _ _ sid secesaresmed sommineensee er e as= esse ed ser ogmaneantsssreas af rearsumme. eene seus mer assed # som 4,v aber emur. u esermeansee=deAssasusese== W suuhr pour earwe amass es ser 417M femmenseqssus.9meshpudh #986 Asse.9. testads$mes f f ivfed g e e es 6 E-4 PAOSLEM M4AS CQPV[ ..,, 1084
H-49-01 1 b b g ASTM D149-81 27 April 19 81 SUPERSEDES I ASTM D149-75 13 January 1976 ACCEMANCE NOTICE p si E The above nongovernment document was adopted on 27 April 1981 and is approved. '1 f or use by Doo. The indicated industry group has furnished the clearances i ,? I required by existing regulations. Copies of the document are stocked by Doo Single Stock Point, Naval Publications and Forms Center, Philadelphia, PA N 19120 for issue to Doo activities only. Contractors and industry groups must l obtain copies f rora ASTM 19' 6 Race Street, Philadelphia, PA 19103. 1 l', ' TITLE of DOCUMEWrs Dielectric Breakdown Voltage and Dielectric Strargth of Solid Electrical Insulatirq Materiala at Conumercial Power t Freque ncie s {. l TATE or SPEC!r!C-ISSUE ADOMZ3: 30 January 1981 RELEASING INDUSTRY CICUP: American Society for Testing and Materials I, custodians: Military coordinating Activity i A rmy -- MA Af"Y ~ "A Air Force -- 11 Project No. 9330-0952 Review activities: A rmy - AV, ER, ME Navy - AS 6 Air Force - 99 OLA - CS User activities: Navy - SH, 05, YD E-ll ~ FULL TEXT FOLLOWS 0039
7 H -4 7-o f f Designation: 0149 - 81 ~' AMERICAN sCCIETY FOR TESTING AND MATERIAL.S 1918 Race St..Philadelphte.Pe.19103 h Reprinted from the Annual Book of ASTM Stendards. Copyright ASTM If not listed in the current combined index.,will appear in the next edition. Standard Test Method for DIELECTRIC BREAKDOWN VOLTAGE AND DIELECTRIC STRENGTH OF SOLID ELECTRICAL INSULATING MATERIALS AT COMMERCIAL POWER FREQUENCIES' This standard is lasued under the naed designation D 149; the number immedisiety rottowing the desi nation indicsies the 5 year oronginal adoption or,in the case or revtsion. she year or last revision. A number in parenihenes indscsies the yest or tasi respprovat. TMs method has been approvedfor use by agenetes ofthe Department of Deferue se ref eet Meshed 4031 of federel Test bleshed l Standard 406 andfor hanng in she DeD Indes of Spec lfleattoss and $sandards E
- 1. Scope 1.8 The method appears in the fo!!owin5 sections:
1.1 This method covers,the procedure for determination of the dielectric strength of solid subjwt sniion insulatin5 materials at commercial power fre-Apparsius 6 PP cable Documenia 2 ti A quencies under specified conditions. 8 f 1.2 Unless otherwise specified the tests shall $,I,'l,."' 'd be made at 60 Hz. However, this method may conditioning il Procedure 12 be used at any frequency from 25 to 800 Hz. i At frequencies above 800 Hz dielectric heating E,"p,"d** *"d ^***'7 0 i may be a pret,lem. sar,iy erecauiions 7 1.3 This method is intended to be used in SamP ng 8 li conjunction with the AS".'M standard or other (*,P'ny,,,, } docuc.cnt that refers to this method. References summary or utihad 3 to this de,cument should specify the particular Terminology 5* Ten specans l options to be used (see 4.4). 1.4 It may bc used at various temperatures.
- 2. Appilcable Documents and in any suitable gaseous or hquid surround-D 374 Tests for Thickness of Solid Electrical 1.
s tnethod is not intended for measur. Insulatien* ing the dielectric strength of materials that are Plast,cs and Elcetncal i D 615 Condit,onin5 i, fluid under the conditions of test. Insulating Materials for Testin5' I 1.6 This method is not intended for use in D 877 Test for Dielectric Breakdown Voltage determining intrinsic dielectric strength, direct. voltage dielectric strength, or thermal failure of Ins,ulatin5 Liquids Usin5 Disk Elce. trodes under electrical stress (see Method D 3151). a 1.7 The method is most commonly used to D 1711 Definitions Q,Tums Relating to Electrical Insulation determine the dielectric breakdown voltage through the thickness of a test specimen (punc. ture). It may also be used to determine dielec. ' This meihad is under the junsdietion or Astu com. mitice D 9 en Elecineal tuulating Maienalm and is ihe dires: tric breakdown volta 5e alon5 the interface be, ,,,pon,b.i.i> or subcommiiin Dov.i2 on Elecincat Tesu. tween a solid specimen and a gaseous or h, quid corrent ed,iion approved san. 30. i9si podiahed us,ch 'i*" D !*,,'j,,tly Pubtahed as D Idv.22 T. t.us prewmu 19stOnsja surroundin5 medium (llashover). With the ad-dition ofinstructions modifying Section i1. this j,, ,,,,f,3 3, 3,,,s,s,, p,n 3,. method may be used for proof testing. ' Ann..s so.4 of Astu ss.ndo,ds, ran 40. l 0040 E-12
M N I W .~ g). o 14g = D 2413 Preparing and Electrical Testing of conditions, or other manufacturing or environ-Insulating Paper and Board Imp.regnated mental situations. This method is useful for . l with a Liquid Dielectric 8 process control, acceptance or research testin5- ) 8 Sl D 2436 Specification for Forced Convection 4.3 Results obtained by this method can sel. 2 Laboratory Ovens for Electrical Ovens' dem be used directly to determine the dielectric D 3151 Test for Thermal hlure Under behavior of a material in an actual application. Electric Stress of Solid ElectricalInsulat-In most cases it is necessary that these results M I D 3487 Specification for MineralInsulating be evaluated by comparison with results ob-8 ing Materials tained from other functional tests or from tests on other materials, or both. in order to estimate 8 Oil Used in Electrical Apparatus F I
2.2 OtherStandards
their significance for a particular material. International Electrotechnical Commission:* 4.4 Documents specifying the use of this s. Pub. 243 Recommended Methods of Test for method should also specify: ^ II E Electrical Strength of Solid Insulating Ma-4.4.1 Method of voltage application. g. 3 terials at Power Fre,quencies 4.4.2 Voltage rate-of rise,if slow rate-of rise American Natlanal Standards Institute method is specified. 4.4.3 Specimen selection, preparation and Standard:' 7 C68.1 Techniques for DielectricTe.s.IEEE conditioning.
- lt 4.4.4 Surrounding medium and temperature Standard No. 4 durin5 test. '
- 3. Summary of Method 4.4.5 Electrodes, and
,q l 3.1 Alternating voltage, at a commercial 4.4.6 Any desired deviations from the rec. q ~ 'l power frequency (60 Hz. unless otherwise spec-ommended procedures as given. I g 4.5 If any of the requirements listed in 4.4 ified)is applied to a test specimen. The voltage are missing from the specifying de:ument. then is increased from zero or from a level wc!! l below the breakdown voltage,in one of three the recommendations for the several variab!cs j j shall be followed.. prescribed methods of voltage application.until 4.6 Appendia XI contains a more complete q } dielectric failure of the test specimen occurs. discussion of the si nificance of dielectric 5 3.2 Most commonly, the test volt:5e is ap-plied using' simple test electrodes on opposite strength tests. ? faces of specimens. The specimens may be
- 5. Terminology molded or cast, or cut from flat sheet or plate.
3.1 Refer to Definitions D 1711 for defini-Other electrode and specimen configurations [ may be used to accommodate the geometry of tions of the fo!!owing terms: 5.1.1 die!cetric breakdown voltage, i the sample material, or to simulate a specific i 5.l.2 dielectric failure. f application for which the material is being 5.1.3 dielectric strength, and evaluated. 5.1.4 nashover. L
- 4. Significance and Use
- 6. Apparatus
'5 4.1 The dielectric strength of an c!cetrical I 6.1 Voltage Source-Obtain the test voltage insulatin5 materialis a property ofinterest for from a step up transformer supplied from a any application where ar; c!ectrical field will be variable sinusoidal low voltage source. The present. In many cases the dictectric strength of a material will be the determining factor in the transformer, its voltage source. and the associ-I ~ sted controls shall have the followin5 capabil-design of the apparatus in which it is to be ities: used. 4.2 Tests made as specified herein may be 6.1.1 The ratio of crest to root mean. square used to provide part of the information needed (rms) test volt:5e shall be equal to d 2 5 % for determining suitability of a material for a (1.34 to 1.48), with the test specimen in the given application; and also, for detecting changes or devialions from normal Character
- e Avaitable from Amerwan Nawn.1 Sundards tumuu.
istics resultittg fro,m processin5 variabics, sgtng too sto.4.sy. New York. N.Y.10018. S* 0041 E-13 s
t M "J o - @t D 149 1 circuit, at all volta 5es 5teater than 50 % of the discharge (corona) currents or, when the sens. breakdown voltage. ing element is located in the primary, to the 6.1.2 The capacity of the source sh:11 be step-up transformer magnetizin5 current. S I dielectric breakdown occurs. For most mate. sufficient to maintain the test voltage until 6.2 Voli g e Measurement-A voltmeter must be provided for measuring the rms test rials, using electrodes similar to those shown in voltage. A peak readingvoltmetermay/2 to be used, i in which case divide the reading by Table I, an output current capacity of 40 mA i is usually satisfactory. For more complex clec. rms values. The overall error of the voltage. trode structures, or for testing high loss mate. measuring circuit shall not exceed 5 % of the rials, higher current capacity may be needed. measured value. In addition, the response time .F The power rating for most tests will vary from of the voltmeter shall be such that its time lag 0.5 kVA for testing low capacitance specimens will not be greater than i % of full scale at any l st voltages up to 10 kV, to 5 kVA for voltages rate-of rise used. uo to 100 kV. 6.2.1 Measure the voltage using a voltmeter ^ 6.1.3 The controls on the variable low volt-or potential transformer connected to the spec-age source shall be capable of varying the imen electrodes, or to a separate voltmeter supply voltage and the resultant test voltage winding on the test transformer, that is unar. smoothly, uniformly, and without overshoots rected by the step up transformer loading. or transients, in accordance with 12.1, Under Nort-Those test sets making use of test trans-no circumstance shall the peak of any voltage former primary voltage measuremem as an indication transient exceed IA8 times the indicated rms. of the voltage across the specimen electrodes la in test voltage. Motor driven controls are prefer. violation of this issue of the method unless calibrated i ab!c for making short time (sec 12.1.1) e*. at th,e specific breakdown voltage and current. Call. rate ot nse (see 12.1.3) tests. bration wdl be dependent upon transformer second. ary current and voltage at the time of breakdown. 6.1.4 Equip the voltage source with a circuit. t i breaking device that will operate within three 6.2.2 It is desirable for the reading of the j cycles. The device shall discennect the voltage maximum applied test voltage to be retained sourec equipment from the power service and' on the voltmeter after hreakdown so that the i i protect it from overload as a result of specimen breakdown volt:5e can be accurately read and breakdown causing an overload of the testing recorded. apparatus. If prolonged current follows break. 6.3 Electrodes-For a given specimen con. down it will result in unnecessary burning of figuration, the dielectric breakdown voltage the test specimens, pittin5 of the electrodes, and may vary considerably, depending upon the contarnination of any liquid surrounding me. geometry and placement of the test electrodes. dium. For this reason it is important that the elec. 6.1.5 The circuit breaking device should trodes to be used be described when specifyin5 have an adjustab!c current sensing element in this method, and that they be described in the the step up transformer secondary, to allow for report. adjustment consistent with the specimen char. 6.3.1 One of the electrodes listed in Table i acteristics and arranged to sense specimen cur. should be specified by the document referrin5 rentiSet the sensing element to respond to a this method. If no electrodes have been speci-current that is indicative of specimen break. fled select an applicabla ene from Table 1, or down. use other electrodes mutually acceptable to the 6.1.6 The specimen current.sensin5 element parties concerned when the standard electrodes ,~ may be in the primary of the step up trans-cannot b= used due to the nature or configu. former. Calibrate the current.senstng dial in ration of the material being tested. See refer-terms of specimen current. ences to Appendix X2 for examples of some 6.1.7 Exercise care in setting the response of special c!cetrodes. In any event the electrodes the current control. If the control is set too must be described in the report. high, the circuit will not respond when break. 6.3.2 The electrodes should be in contact down occurs; if set too low, it may respond to with the test specimen over the entire area of leakage currents, capacitive currents, or partial the electrode, except for such obvious arets as h5 3 0042 E-14 y n
"um\\ ^**J' q'h D 149 a. the rounded edges of Types I,2,3. and 6 of the energy released at breakdown may be st ficient to shatter the container. Metal bar Table 1. 6.3.3 Keep the electrode surfaces c!can and must be grounded. it is recommended smooth, and free from projecting irregularities mineral oil meeting the requirements of Spc S fication D 3487. Type 1 or 11, be used. It shoi. resulting from previous tests. If asperities have have a dielectric breakdown voltage as de I developed, they must be removed. mined by Method D 877 of at least 26 U. 6.3.4 Whenever the electrodes are dissimilar In size or shape, the one at which the lowest Other dielectric Guids may be used as su: concentration of stress exists, usually the larger rounding mediums if specified. These inc in size and with the largest radius, should be at but are not limited to, silicone fluids and o liquids intended for use in transformers, cirer F grEid potential. 6.3.5 In some special cases liquid metal elec-breakers, capacitors, or cables. 6.4.1.1 Breakdown values obtained usi: i trodes, foil electrodes, metal shot, water, or conductive costin8 electrodes are used. It must liquids having different electrical propeni 2 be recognized that these may give results dif-may not be comparabic. Sec XI.4.7. If tests r fering widely from those obtained with other to be made at other than room temperatur. the bath must be provided with a means f. j types of electrodes. 6.3.6 Because of the effect of the electrodes heatin5 or coolin5 the liquid, and with a mear to ensure uniform temperature. Small bae .on the test results, it is frequently possible to can in some cases be placed in an oven (s obtain additional information as to the dielec- !~ tric properties of a material (or a group of 6.4.2) in order to provide temperature contre materials) by running tests with more than onc if forced circulation of the fluid is provide. care must be taken to prevent bubbles frc: type of electrodes. This techniqr *..f panic-being whipped into the Guid.The temperatu: ular value for research testing. shall be maintained within 5'C of the spu 6.4 Surrounding Nedium-The document calling for this method should specify the sur-ified test temperature at the electrodes, unter rounding medium and the test temperature, if otherwise specified. In many cases , tests are not to be made in air at ambient that specimens to be tested in insulating oil a temperature and humidity. !t is normally pref-to be previously impregnated with the oil a: etable to test materials in the medium in which not removed from the oil before testing (n Method D 2413). For such materials, the bn-they are to be used. When conditions of use are must be of such design that it will not b. not well defined, materials should be tested in air, unless the breakdown volt:5e is so high as. necessary to capose the specimens to require cacessively large specimena, or to testing. 6.4.2 If tests in air are to be made at oth-cause heavy surface discharles and burning than ambient temperature or humidity, an ove prior to failure. Flashover must be avoided. or controlled humidity chamber must be and the effects of panial discharges prior to vided for the tests. Ovens meeting the require-r failure minimized. even on short. time tests.Thements of Specification D 2436 and provide. mater'.at of the seals or shrouds around the ciectrodes may innuence the breakdown val-with means for introducing the test voltage wh ues. For specimens haviig a high Wakdown be suitable for use when only temperature is : voltage it is frequently preferable c. even nec-be contro!!cd. 6.4.3 Tests in gasses other than air will p essary to make rests in insulating el. Break-etally require the use of chambers that can down values obtained in insulating od may not be comparab!c with those obtained in air. In evacuated and filled with the test gas, usus. many cases, particularly for research or quall-under some controlled pressure. The design e fication purposes, it is desirable to test at the such chambers will be determined by the natu: capected temperature of operation. of the test program to be undertaken. 6.4.1 When tests are made in insulating oil. 6.5 Ten Chomber-The test chamber or at: an oil bath of adequate size shall be provided. in which the tests are to be made shall be The use of glass containers is not recommended sufficient size to hold the test equipment. ar for tests at voltages above about 10 kV, because shall be provided with interlocks to prew f.s. 4 0043 g F-1 j r _I.. r :3- ;rw=- -~c .m y r-.mg .a
i ,M W @) D 149 V accidental contact with any electrica!!y ener-ratory or other test area is begun. gized parts. A number of different, physical 8.3 Fcr the purposes of most tests it is desir. arrangements of voltage source, measuring able to take samples from areas that are not S. equipment baths or ovens, and electrodes are immediately ad,lacent to obvious defects or dis-possible, but it is essential that (1) all ates or continuities in the material. The outer few lay-E doors providing access to spaces in which there ers of roll material, the top sheets of a package M I .are electrically energized parts be interlocked of sheets, or material immediately next to an to shut off the volta 5e source when opened;(1) edge of a sheet or roll should be avoided, unless clearances are sufficiently large that the tic!d in the presence or proximity of defects or discon-F I charges (corona) do not occur except between the area of the electrodes and specimen are not tinuities is ofinterest in the investigation of the distorted and that flashovers and partial dis-material. 8.4 The sampic should be large enough to a the test electrodes; and (J) insertion and re-permit making as many individual tests as may 13 -. placement of specimens between tests be as be required for the particular material (see simple and convenient as possible. Visual ob-12.3). servation of the electrodes and test specimen during the test is frequently desirable.
- 9. Test Specimen 9.1 Preparation and Handling:
- 7. Safety Precautions 9.l.1 Prepare specimens from samples col-7.1 Lethal voltages are present durin5 every lected in accordance with Section 8.
6 dielectric breakdown test. It is essential that the 9.1.2 When flat faced cicctrodes are to be test apparatus and all associated equipment used, the surfaces of the specimens which will that may be electrically connected to it 6 oc in contact with the electrodes shall be designed, installed, and operated so that it is smooth parallel planes. insofar as possible with-not possible for any person to make contact out actual surface machining. with energized conductors. 9.1.3 The specimens shall be of sufncient 7.2 Sclidly ground all metal parts that any size to prevent flashover under the conditions person mi ht come into contact with durin5 the of test. For thin materiais it may be convenient 1 6 g test. to use specimens large enough to permit mak. 7.3 Thoroughly instruct all operators in the in5 more than one test on a single piece. h, proper way to conduct the tests safely. 9.1.4 For thicker materials (usually more m 7.4 When makin5 tests at high voltage with than 2 mm thick) the breakdown strength may ] large area electrodes, particularly in com-be high enou5 that flashover orintense surface h [.. pressed gas or in oil, the energy released at partial discharges (corons) may occur prior to breakdown may be sufncient to result in fire. breakdown. Techniques that may be used to explosion, or rupture of the test chamber. De-prevent flashover. or to reduce partial discharge sign of test equipment, test chambers, and test (corona) include: specimens should be such as to minimize the 9.1.4.1 Immerse the specimen in insulating f possibility of such occurrences, and to eliminate oil during the test. See Xl.4.7 for the surround-l the possibility of personal injury, ing medium factors influencin5 breakdown. This may be necessary for specimens that have
- 8. Sampling not been dried and imere5nated with oil, as f
8.1 The detailed sampling procedure for the well as for those which - been prepared in material bein5 tested should be defined in the accordance with Method u.413. for example. t I specincation for that material (See 6.4.) 8.2 Sampling procedures for quality control 9.1.4.2 Machine a recess or drill a flat bot-purposes should provide for gathering of sur-tem hole in one or both surfa es of the speci. ficient samples to estimate both the average men to reduce the test thicknest !f dissimilar quality and the variability of the lot bein5 electrodes are used (such as Type 6 of Table 1) examined; and for proper protection of the and only one surface is to be machined, the t samples from the time they are taken until the larger of the two electrodes should be in contact preparation of the test specimens in the labo-with the machined surface. Care must be taken i .^- 0044 E F2
W @) D 149 in machining specimens not to contaminate or ' procedures in Method D 618. mechanically damage them. 11.3 For many materials the moisture 6 9.lA.3 Apply seals or shrouds around the tent has more effect on dielectric strength t' S electrodes, in contact with the specimen to re-does temperature. Conditioning times for : duce the tendency to flashover. materials should be sufficiently long to per 9.1.5 Materials that are not in flat sheet form the specimens to reach moisture equilibrium M shall be tested using specimens (and electrodes) well as temperature equilibrium. appropriate to the material and the geometry 11.4 If the conditioning atmposphere is < of the sample. It is essential that for these that condensation occurs on the surface cu materials both the specimen and the electrodes specimens. it may be desirable to wipe i be dertned in the specification for the material. surfe*.s of the specimens immediately bc 9.1.6. Whatever the form of the material. if testing.This will usually reduce the probab tests of other than surface to surface puncture of surface flashover. strength are to be made, define the specimens
- 12. Procedure and the electrodes in the specification for the material.
12.1 Methods of Voltage Application: 9.2 In nearly all cases the actual thickness of 12.1.1 Method A. Shoer. Time Test-Ar. f the test specimen is important. Unless otherwise voltage uniformly to the. test electrodes fr. specified. measure the thick. ness after the test zero at one of the rates shown in Fi. I A ur.- 5 in the immediate vicinity of the area of break-breakdown occurs. Use the short. time tes: down. Measurements shall be rnade at room less othwise specified. ternperature (25 5'C), using the appropriate 12.1.1.1 When establishing 5 a rate initiall;. procedure of Method D 374. order for it to be included in a new specine tion, select a rate that, for a givra wt of spo
- 10. Calibration mens, will give an average time to breakdr 10.1 In making calibration measurements, of between 10 and 20 s. It may be necessar take care that the values.of voltage at the run one or two preliminary tests in orde electrodes can be determined within the accu-determine the most suitable rate.cf rise.
racy given in 6.2. with the test specimens in the many materials a rate of 500 V/s is used. I circuit. 12.1.1.2 If the document referenrin8 (! 10.2 Use an independently calibrated volt-method specified a rate of rise it shall be u meter attached to the output o,f the test voltage consistently in spite of occasional average th I I source to verify the accuracy of the measuring to breakdown falling outside the range of it device. Electrostatic voltmeters, voltage divid-20 s. In this case, the times to failures sha!: I crs, or potential transformers having compara-made a part of the report. l ble accuracy may be used for calibration mes-12.1.1.3 in running a series of tests comrr ing different material, the same rate of r surement. 10.3 At voltages above about 12 kV' rms shall be used with preference given to a ra i (16.9 kV peak) a sphere gap may be used to that allows the average time to'be between : calibrate the readings of the voltage measuring and 20 s. If the time to breakdown cannot t-I device. Follow procedures as specified in ANSI adhered to, the time sha!! be made a part of i; C68.1 in such calibration. report. l2.1.2 Methad B. Step.by Step Test-A
- 11. Conditioning volt 85e to the test electrodes at the prett 11.1 The dielectric strength of most solid starting voltage and in steps and duratier
} shown in Fi. IB until breakdown occurs. insulating materials is influenced by tempers-l ture and moisture content. Materials so affected 12.1.2.1 From the list in Fi. IB select c l should be brought to equilibrium with an at-initial volta 5e to be the one closest to 50 9 mosphere of control!!cd temperature and rela-the caperimentally determined or capa:.. tive humidity before testing. For such mate-breakdown voltale under the short time te.i rints, the conditioning 5 should be included in 12.1.2.2 If an initial voltage other than er the specification referencing this method. of the preferred values listed in Fig.111 11.2 Onless otherwise specified, follow the selected, it is recommended that the volt.y 6 0045 F-3
) Ml 4 O g @ D149 pi steps be 10% of the preferred initial volt:5e group or specimens breaks down in less than y[ immediately below the selected v'alue. 120 s, reduce either the initial volta 5e or the d 12.1.2.3 Apply the initial voltage by incress-rate-of. rise, or both. 'j in5 the voltage from zero as rapidly as can be 12.1.3.4 If more than one specimen of a N accomplished without introducing a transient group of specimens breaks down at less than @f volta 5e or ovenhoot beyond that permitted in 1.5 times the iditial voltage, reduce the initial ~ 6.1.3. Similar requirements shall apply to the value. If breakdown repeatedly occurs at a O procedure used to increase the volt:5e between value greater than 2.5 times the initial value $l successive steps. Aner the initial step, the tim r (and at a time of over 120 s) increase the initial p required to raise the voltage to the succeedin5 volta 58-step shall be counted as pan of tne time at the 12.2 Criteria of Breakdown-Dielectric fail-f succeeding step. ure or dielectric breakdown (as defined in Def. 12.1.2.4 If breakdown occurs while the volt-initions D 1711) consists of an increase in con-l age is being increased to the next step. it shall ductance. limiting the electric field that can be be recorded as having occurred at the neat sustained.This phenomenon is most commonly lower step and the actual breakdown volta 5e evidenced during the test by abrupt rupture also reported. throu5h the thickness of the specimen, which 12.1.2.5 It is desirable that breakdown occur can be seen and heard, and which results in a in four to ten steps, but in not less than 120 s. visible puncture and decomposition of the spec-If failure neurs at the third step or less, or in imen in the breakdown area. This form of less than 120 s whichever is greater, on more breakdown is generally irreversible. Repeated, than one specimen in a group, the tests should applications of voltage !cvel (sometimes un. be repeated with a lower initial voltage. If measurably low). usually with additional dam-failure does not occur before the tweinh step age at the breakdown area. Such repeated ap-or greater than 720 s, increase the initial volt-plications of voltage may be used to S ve posi-i i a5e. tive evidence of breakdown and to make the 121.2.6 Record the initial voltage the volt-breakdown path more visibic. age steps, the breakdown voltage, and the 12.2.1 A rapid rise in leakage' current may j length of time that the breakdown volt:5e was result in trippin5 of the voltage source without held. If failure occurred while the voltage was visible decomposition of the specimen. This bein5 ncreased to the staning voltage the fail-type of failure, usually associated with slow. i ure time shall be zero. rise tests at elevated temperatures, may in some 12.1.2.7 Other time len8ths for the voltage cases be reversible. that is, recovery of the I steps may be specined, dependin5 upon the dielectric strength may occur if the specimen is purpose of the test. Commonly used len$ths are allowed to cool to its ori inal test temperature 5 f 20 s and 300 :(5 min). For research purposes, before reapplyin5 voltage. The volta 5e source it may be of value to conduct tests usin5 more must trip rapidly at relatively low current for than one time interval on a given material. this type of failure to occur. 12.1.3 Method C. Slow Rate.of. Rise Test-12.2.2 Trippin5 of the volta 5e source may Apply voltage to the test electrodes, from the occur due to nashover, to panial discharge 6 stant.e5 voltage and at the rate shown in Fig. current, to reactive current in a high espaci. IC until breakdown occurs. tance specimen. or to malfunctioning of the 12.1.3.1 Select the initial voltage from shon-breaker. Such interruptions of the test do not time tests made as specined in 12.1.1. The constitute breakdown- (except for Dashover initial volt:5e shall be res:hed as specified in tests) and should not be considered as a satis. 12.1.2.3. factory test. 12.1.3.2 Use the rate.or. volt:5e rise from the 12.2.3 If the breaker is set for too high a initial value specified in the document calling current, or if the breaker malfunctions. exces. i for this method. Ordinarily the rate is selected sive burni35 of the specimen will occur. to approximate the average rate for a step.by. 12.3 Number of Tests-Make Gvc break. step test. downs unless otherwise speciGed for the panic. 12.1.3.3 If more than one specimen of a ular material. l l F-4 I 0046 / I
^~~/' l ) a,, h D149 14.1.3.7 Surrounding medium,
- 13. Calculations 14.1.3.8 Test temperature.
13.1 Calculate for each test the dielectric 14.1.3.9 Description of electrodes. strength in kV/mm or V/in11 at breakdown. 14.1.3.10 Method of voltage application, and S and for step.by. step tests. the gradient at the 14.1.3.11 Date of test. highest volta 8e step at which breakdown did not occur.
- 15. Precision and Accuracy M
13.2 Calculate the average dielectric strength and the standard deviation. or other 15.1 Sing /c. operator prectslan-Dependir measure of variability. upon the variability of the material beir tested, the specimen thickness, method of vo! F
- 14. Report age appl cation, and the catent to which tran 14.1 The report shallinclude the fo!!owin5:
sient voltage surges are contro!!cd or sur 14.1.1 Identification of the test sample. pressed, the cocmcient of variation (standar 14.1.2 for Each Spreimen: deviation divided by the mean) may vary from 14.1.2.1 Measured thickness. a low 1 of 2% to as high as 20% or more When makin5 uplicate tests on five specimer.: d 14.1.2.2 Maximum voltage withstood (for from the same sample, the coemeient of varia. step.by. step tests). 14.1.2.3 Dielectric breakdown voltage. tion will usually be less than 7 %. 14.1.2.4 Dielectric strength (for step.by. step 15.2 Multilaboratory precision-The preci-sion of tests made in different laboratories (er tests), 14.1.2.5 Diclectric breakdown stength, and of tests made usin5 different equipment in the ( 14.1.2.6 Location of failure (center of c!cc. same laboratory) may be quite yariabic. Ifiden. l trode, edge, or outside). tical types of equipment are used. with speci. l i 14.1.3 Ear Each Samp/r. men preparation, electrodes, and testing pro. [ i 14.1.3.1 Average dielectric withstand cedures closely controlled, the single.operato: precision may be closely approached. Wher
- strength, k
14.1.3 ? Ave:gt dielectric breakdown direct comparison of results from two or mort laboratories is to be made. it is recommende, j
- stren5th, L
14.1.3.3 Indication of variability, preferably that the precision between the laboratories in-1 the standard deviation and coemeient of vari. volved be evaluated. 15.3 Accuracy-This method does not deter. I acion. f 14.1.3.4 Description of test specimens, mine the intrinsic dielectric strength. The ten 14.1.3.5 Conditioning 8 and specimen prepa-values are dependent upon specimen geometry. electrodes. and other variable factors,in addi.
- ration, 14.1.3.6 Ambient atmosphere temperature tion to the properties of the sample, so that it and relative humidity.
is not possible to make a statement of accuracy m i i A 0047 7g i
e w @ 014e u TABl.E I Typical Electrodes foe Dielectrie Strength Testias of Varises Types of lasalatlas Metectals d Elec. trode Description of Electrodea # lasulating Materials 8 S. Type I Opposing cylinders $1 mm (2 in.)in diameter. 23 mm flat sheets of paper, filma, fabrics, rubber, molded plas. (1 in.) thick with edges rounded to 6.4 mm (0.25 in.) tics. laminates, boards, glass, mica. and cersmic radiua M, 2 Opposin5 cylinders 25 mm (l in.)In diameter. 25 mm same as for Type 1. particularly for 3 asa, mica, plutic, 1 (1 in.) thick with edges rounded to 3.2 mm (0.123 in.) and cernanie radius J Opposing cylindrical rods 6.4 mm (6 3 in.)in diameter same as for Type 1.particularly for varnish plutic and F with edgea rounded to o.8 mm (U.0J.J in.) tadius' other thin fdm and tapes: where small specimens necessitate the use of smaller electtmies, et where testing cf a smau area is desired 4 Flat plates 6.4 mm (0.25 in.) wide and 108 mm (4.23 same as for Type 1. particularly for rubber tapes and E in 1 loas with edses square and ends rounded to 3.2 other narrow widths of thin matenais g mm (0.123 in.) radius 5 Hemispherical electrodes 12.7 mm 10.) in.)in diameter # filling and treating compounds, gets and semisolid com. pounds and gresses, embedding. potting and encap-sulating materiala 6 Opposing cylinders; the lower one 75 mm (3 In.) in same as for Types I and 2 diameter.15 mm (0.60 in.) thick: the upper one 25 mm (1 in.) in diameter. 25 mm thick; wuh edges of I both rounded to 3 mm (0.12 in.) radius # 'These electrodes are those most commonly spuitied oc referenced in ASTM standards. With the esception of Type $ electrodea. r.o attempt has been made to suggest electrode systems for other than flat surface material Other electrodes may be uwd as specified in ASTM standards or as agreed upon between seller and purchawr where none of these ettetrodes in the table is suitable for proper ersivation of the material being tested. f 8 Elutrodes are normauy made from either brsas or stamless steel Reference should be made to the standard governing [ the material to be tested to determine which if either. materialis preferable. I, The electrodes surfaces should be polished and free from irregularities resulting from previous testing. s " Refer to the appropriate standard for the land force applied by the upper electrode anembly. Unless otherwiw specified the upper electrodes shau be 30 + 1 g. f' ' Refer to the appr.,na.e standaru aw the proper gap witinp 1 'The Type 6 ele.trodes are thow given in IEC Pubhcation 243 for testing of flat sheet matenals. They are less entical as to concenancity of the electrodes than are the Types I and 2 electrodes. l vj. _ _ _ _ _ _ _ _ I f> AY i At l l t t o td 9 i Rates ( (Y/s)
- 20 4 k
100 200 SW 1000 2000 5000 FIG. l A Voirage Prottle of Short.Tirne Test W-9 004d F-6
hM $ D 149 iV Vg... 1sl y: ~ lRMI i !!!! ,I I I . kg O I 2 s 5 tart Step Yolta e g1 )$.' V k f is ( ) ~' k 0.25 3 or less 0.25 It, - r.). (r, - til - tr. - s,) - 60 2 5 s 0.50 over $ to 10 0.50 I over to to 25 1 20] 2 over 25 to 30 2 300 l 'I'"" *'* I I I 5 over $0 to 100 5 120 s tw 5 720 s 10 over 100 10 } (twwill usustly esseed 720 s if 300 s time 20 1 step is used.) 50 100 l d Y. - 0.5 (Vw for Short Tirne Tenth unless constraints cannot be anet. FIC,15 Voltste Pronle e( Sterbr. Step Tess 9 h i D. 1 y AV [ f V,^ at t 0 bd g ( Rates (Y/s) 2 20 4 Constraints I t., > 120 s b 2 5 10 Vu = > l.5 Y. 12.5 ? 20 25 50 100 FIC. IC Yohste Pronle of Sow Ra:W Alse Tess l to 0049 p_7
o y @l D 149 APPENDIXES XI. SIGNIFICANCE OF THE DIELECTRIC STRENGTH TEST M XI.1 Introduction area. this area effect being more pronounced with A l. l.1 A brief review of thic:: postulated mech. thin test specimens. Test results are also affected by anisens of l>reakdown, namely: (1) the discharge ne the electrode geometry. Results may be afTected also corona mechanism. (1) the thert ! mechanism, and by the matenal from which the electrodes are con. (J) the intrinsic mechanism, as well as a discussion of structed. since the thermal and discharge mechanism t the pnncipal factors affectin tests on practical di,, may be innuenced by the thermal conductivity and lcetrics, are given here to aid interpreting the data. the work function, respectively. of the electrode ma. "Ibe breakdown mechanisms usuall operate in com. terial. Generall speakmg. th effect of the electrode materialis difn uit to establish because of the scatter bination rather than singly. The rol owmg discussion applies only to solid and semisolid matenals. CI{8Pjnmental data. strength of sblid commerical electrical insulating ma. XI.2 Postulated Mechanisms of Dielectric Break, down terials is greatly pendent upon the specimen thick.
- P*
" * * " * "" ' ' ' '*' **" 0 ' " A * *
- I*
Xl.2.1 Breakdown Caused Electrical Dis. solid materials, the dielectnc strength varies inver:4.ly charfes-in many tests on commercial materials,u a fractional power of the specimen thickness, and brea down is caused b ciectrical dischstges, which there is a substantial amount of evidence that for produce high local fic s. With solid materials the relatively homogeneous solida, the dielectric strength discharges usually occur in the surroundmg medium, varica a remitsately as the reciprocal of the square thus incressmg the test ares and roducing failure at root of t thickness. In the case of solids that can be or beyond the electrode edge. D harges may occur melted and poured to solidify between Axed elec. in any mternal voids or bubbles that are present o' trodes. the efTect of etectrode sepr ation is tess clear! develop. These may cause local erosion or chem-defined. Since the electrode seprcation can be Gxe m 7 decomposition. These proccues may continue at will in such caus, it is custr. mary to perform sea .intil a complete failure path is formed between the dislectric strength tests'cn liquids and usually on electrodes. fus ble solids, with electrodes having a standardized XI.2.2 77:ermal Breakdown-Cumulative heating Gard spacing. Since the dielectric strength is so de. 2 develops in local paths wnhin many materials when ndent upon thickness it is meanin less to report [ they,are subjected to high elcetne fic!d intensities, ielectric strength data for a material thout stating h causmg dielectnc and ionic conduction losses,which the thickness of the test s cimens uud. generate heat more rapidly than can be dinipated. XI.4.3 Te erature-he temperature of the test Breakdown may then occur becauw of thermal msta* spes men and is surroundin medium innuence the bdity of the material. dielectric stren th although r most materials small XI.2.3 /ntrinsic Breakdown-!f electric discharges variations of a bient temperature may have a ne. or thermalinstability do not cause failure, breakdown g;&ibic efTect. In general, the dielectric strength w ll will still occur when the fic!d intensit becomes sur. ficient to accelerate electrons throus the, material.. decreau with incressmg temperatures but the [ This critical fic!d intensity is called the intnnsic die
- test. When it is known that a matenal will be required i
!cettic stren th. It cannot be deterrnined by this t,est to function at other than normal room temperature, method, alt ough the mechanism stulf may be in* t is essential that the dielectric strength temperature volved. relationship for the material be determined over the ran e of e sted operating temperatures. XIJ Nature of ElectricalInsulating Materlata l.4.4 e-Test results will be mnuenced by XI.3.1 Solid commerical electricalinsulating ma* the rate of voltag application. In geners!. the break. terials are general nonhomogeneous and may con
- down voltage wii tend'to increase with increasing tain dielectric de ets cf vanous kinds. Diclectric rate of voltage application. This is to be espected breakdown onen occurs in an area of the test speci*
because the thermal breakdown mechamsm is time. j men other than that where the fic!d intensity is dependent and the discharge mechanism is usually greatest and sometimes in an ares remote from the g,ne dependent, although in some cases the latter material directly between the cicctrodes. Weak spota mechantsm may cause rapid failure by producing within the volume under stress sometimes determme critically h h local field intensities. the test results. x t,4,3 ,,e form-In general, the dielectric strength is influenced by the wave form of the a pplied i XI.4 I Auence of Test and Spee!mes Conditlosa voltage. Within the limits specified in this method [ XI.4.1 Electrodes-In general, the breakdown the induence of wave form ts not si nificant. 5 volts ge will (end to decreau with increasing electrode XI.44 Treg* ene)-The dielectric strength is not f- ]1 il ( 0050 p_g
L-L- i_ h D149 medium. by frequency varistions XI.4.8 /letartre //umidity-The realtive humidiiv significantly influencedwithin the range uf commerical power frequencies innuences the dielectric strength to the extent tha's 7 provided for in this method / However, inferencesmoisture absorbed by, or on the surface of. the ma. concerning dielectric strength behavior at other than terial under test affects the dielectric loss and surface E_ commerical power frequencies (50 to 60 Hz) must A
- {p not be made from results obtained by this method.
conductivity. Hence,its importance will depend to a XI.4.7 Surrounding Medium-. Solid irtsulating large extent upon the nature of the matenal being tested. However. even materials that absorb tittle or materials having a high breakdown voltage are usu-no moisture may be affected because of greatly in. .I ally tested by immersing the test specimens in a liquid creased chemical efTects of discharge in the presence dielectric such as tranformer oil, silicone oil, or chlo-of moisture. Eacept in cases where the effect of h
- i. fluorocarbons, in order to rninimize the efTects of caposure on dielectric strength i E
su: are da'uarges prior to baakdown. It has been shown by S. Whitchead* that in order to avoid dis-it is customary to control or limit the relative humid. ~ [- E charges in the surrounding meutum prior to r, cachingity efTects by riandard conditioning procedures. J-the breakdown voltage of the solid test specimen,in XI.5 Evaluation g-alternating voltage tests it is necessary that XI.S.! A fundamental requirement of the insula. g E.c'. /O.' + 1 > E.e*, JO.' + 1 tion in electrical apparatus is that it withstand the voltage imposed on at in service. Therefore there is a If the liquid immersion medium is a low losa material. great need for a test to evaluate the performance or e-2 the criterion simplifies to particular materists at high voltsge stress. The die-7/-- lectric breakdown volta ge test represents a convenient = e E.c'. > E.e'. /O.3 + 1 k preliminary test to determine whether a material '1 and if the liquid immersion medium is a semicon, ments further consideration. but it falls short of a ducting material the criterion becomes complete evaluation in two important respects. First. i fI'-- E.o. > 2wfsa.E. the condition of a material as installed in apparatus is much different from its condition in this test. j g particu'atly with re$ard to the configurat,on of the i g i-where: H 3-E = clectric strength, electne field and the area of matenal exposed to it. 80f0""'n"'w'.hamcal stresa, arnbient medium p ; f = frequency, s cistio ith other m i there are deteriorating,stenals. Second, n 0-e and s' = permittivity. mfluences, heat, mechanical D = dissipation factor, and = conductivity (S/m). stresa, corona and its products, contaminants, etc.. a , a, - Subscripts: which may reduce the breakdown volta $c far below (% 2 m refers to immersion medium. its value as enginally installed. Some of these effects U'- r refers to relative, can be mcorporated in laboratory tests, and a better f the material wi!! result, but the final
- f O refers to free space, estimate fo = 8.554 x 10-'8 F/m) and c nsideration must always be,that of the performance i
h,y_- s refers to solid dielectric. of the materialin actual service, b-XI.4.7.1 Whitchead points out that it is therefore XI.5.2 The dielectric breakdown test may be used desirabic to increase 1. and c., or o., if surface as a material inspection or qua,lity control test., as a dixharges are to be avoided. He also mentions lp. means of infernns other conditions such as variabil. b-261) that the use of moist semiconducting oil can I'I* ' '" indicate deteriorating processes such,as preciabic reduction in edge discharges. thermal aging. In these uses of the test it is the relative effect an abreakdcwn path between the electrodes is value of the breakdown voltage that is important E Unless the solely within the solid, results in one medium cannot rather than the absolute value. be compared with those in a difTerent medium, it should also be noted that if the solid is porous or ~ P-capable of being permeated by the immersion me.
- Whitehead. 5.. DMIrrme ArruAdv.a n/ Sons. Carord p
.dium, the breakdown strength of the solid is directly b affected by the electrical properties of immersion Univermy Press.195 t. e v-f m X2. STANDARDS REFERRING TO METHOD D 149 F cases the manner in which the reference is made to dit-- X2.1 Introduction this method is not in conformance with the require. X2.1.1 It is not espected that the documents listed ments of 4.4. Do not use another document such as in this appendia comprise a complete list of ASTM those given in this appendia, as a model for providing f-p-standards refernns to Method D 149. This listing is reference to this method, unless there is conformity included herein for the purpose of providing refer, with 4.4. 1 ence to a broad spectrum of documents concerned with dielectric strength at power frequencies. In some s- { 12 0051 Q F-9 as
N .y d @ 014e j .m X2.2 ASTM Standards Rohrrlig la Method D 149 i t _. ASTM ASTM ASTM i Daig. Subject Dais.
- Subject Daig.
Subject g nation nation nation + fahrte. Merr. Espr. Tape. Film. /fre D 1430 Polychlorotrifuoroethylene D 2230 Halogensted Organic Sol. oble Compostres end Coated Tobries: IPCTFE) P1astic vents D 1636 Allyl Molding Compounda D 3214 Coating Powders D 69 Friction Tap D 1674 Polymer (sable Embedding D 119 Rubber Tape g 7,,j, D 201 Insulating Paper Compounda D u su a D 295 Va,rnisW Conoe Fabric uge,, cf.,;,,,g p.,,,f,j,f o g .gn D 116 Vit.. lied Ceramic Materials D 530 Hard Rubber D373 Blach Blas Cut Varnished D 332 Need Mies D 1048 Rubber lasulating 8tankets I o I, loth and Tap, D 619 Vul anised Fiber D 748 Natural Block Mies D 1049 Rubber Insulator Hoods D 902 Rese-Coated Glasa Fabrics D 1039 Class. Bonded Mica D 1050 Rubber Insulating Line D 1677 Untreated Mica Paper Hose ano Tapes D 1000 Pressure.Sansitive Adhe. D 2442 Alumina Ceramics D 1051 Rubber insulating Sleeves D 3391 Rubber Tape sive Coated Tapes gf,,,tas. Tubes. Shorts, and Aeds-D 1373 Orone. Resistant Rubber Alling Com ungsf D 229 Algid Sheet and Plate Ma. Tape D 1389 Thin Solid Electric Insulat, terials D 176 Solid Filling and Truting
- ng g,g,,;,3 0 348 Laminated Tubes Compounda D 1458 Silicone Rubber Coated D 349 1.aminated Round Roda Adhes/ rest f
D 330 Finible Treated Sleeving D 1304 Adhesives Relative to Use S Class Fabric and Tapes D I439 Silicone Varnished Glass D 709 Laminated Thermosetting as Elestricallas
- eon Materials war,,,J c,61, Cloth and Tape D 876 Nonngid Vinyl Chloride S,
D 1830 Coated Fabite la Thermosettingand Jacketed W.sulated D 470 D 1930 Kraft Dielectric Tissu, Polymer Tuhing ire and D 1202 Cellulose Autaie Sheet and D 2305 Polymeric Film D 1676 Film 7 nsulated Magnet f t D 1675 TF fluorocarbon Tubing D I pre na P r D 3391 Rubber Ta D 17f0 TFE.Fluorocathon Rod D 2307 Film. Insulated Msgn't s D 2671 liest.5hnnkable Tubing Win 3-t Holding and Embedding Compounds: D 3394 Insulating Board D 2633 Thermoplastic lasulated I D 700 Phe n.. lie Molding Com. and Jacketed Wire and V8'"'8A'8' ##l""'#' 8"# I'8""gst pounda I D 704 Melamine Formaldehyde D 115 Varnishes D 3032 If kup Wire lasulation Molding Compounds DIM Silicone lasulaung Var. D 3353 Fibrous insulated MaInet D 703 Utes. Formaldehyde MotJ. nuhes i z
- i" ing Compounds D 1932 Thermal Endurances of O'"""U D 729 Vinylidene Chloride Mold.
Fleanble Electneal inau. ing Compounda lating V$rnuhes D 2304 Thermal Ev stustann I The Ameroran Serietrfor Testmg and.tlaterrels taire n.o pesotoon resperting the rahdoor of uny patent erghts auerted in rennerrien enk anr esem mentioneJ m thss usandard Users of thss standard we e apressfr adrued that determenattom of the vahdarr of one such pateni rights. and the ri,sk ofInfrmgement of such rights. are entsreir these onn respanhditv. Thss stanJard us subject to revisme et ans time br the verymushir technovel commetter and mun be erruemrJ everrfire vrars and of not revueJ. ruher reapproveJ or m uhdren n. Your comments are oneneJ ruherfor resusan af thu trenderJ orfor addatoonal stanJarJs and should be addressed to ASTM ligadquarters Yow rumments m-JIrrettre terrfulrunssJeraroon ut a mertmg of the responsible treharral commstrer, a hoek pu mar attend. Ifpufrelthat row comments have not erreserJ afasr hearmg yuu should make yew veras known to the ASTM Commatre un Standarda, tule Rare St.. Yhdadelphsa. Pa. luttl.t. mhsch moll se Ardule a I. furohre hearrag rezardmg pw comments, Tadmg satufortwa therr.pu mar appal to the ASTM RuarJ nf threrters. 't 1 -t ~ 1 DROBLEM HARD COPY 13 I 0052. .s F-10 x,,,.. .r )
- }}