ML18114A818

Heaf Information Sharing Workshop Transcript Day 2

ML18114A818
Person / Time
Issue date:	04/19/2018
From:	Kenneth Hamburger NRC/RES/DRA/FRB
To:
Shared Package
ML18114A814	List: ML18114A815 ML18114A817 ML18114A818
References
NRC-3640
Download: ML18114A818 (294)

Text

1 NUCLEAR REGULATORY COMMISSION 1+ + + + +2 NRC HEAF PHASE II INFORMATION SHARING WORKSHOP 3+ + + + +4 THURSDAY 5 APRIL 19, 2018 6+ + + + +7 8 The NRC HEAF Phase II Information Sharing 9 Workshop met in the 02A14 Classroom of Three White 10 Flint, 11601 Landsdown Street, North Bethesda, 11 Maryland, at 8:34 a.m., Michael Cheok, Deputy 12 Director, NRR, presiding.

13 14 STAFF PRESENT 15 MICHAEL CHEOK, Director, Division of Risk 16 Analysis, Office of Nuclear Reactor 17 Regulation 18 THOMAS AIRD, General Engineer, Division of Risk 19 Analysis 20 THOMAS BOYCE, Branch Chief, Regulatory Guidance 21 and Generic Issues Branch 22 ROBERT DALEY, Branch Chief, Region III 23 STANLEY GARDOCKI, Program Manager, Regulatory 24 Guidance and Generic Issues Branch 25 2 NICHOLAS MELLY, Fire Protection Engineer, 1 Office of Nuclear Regulatory Research 2 KENN MILLER, Office of Nuclear Regulatory 3 Research 4 MARK HENRY SALLEY, Branch Chief, Fire and 5 External Hazards Branch 6 DAVID STROUP, Project Manager, Office of 7 Nuclear Regulatory Research 8 GABRIEL TAYLOR, Senior Fire Protection 9 Engineer, Office of Nuclear Regulatory 10 Research 11 MICHAEL WEBER, Director, Office of Nuclear 12 Regulatory Research 13 14 ALSO PRESENT 15 JENS ALKEMPER, FM Global 16 SCOTT BAREHAM, NIST 17 JANA BERGMAN, Curtiss-Wright 18 ROBERT CAVEDO, Exelon 19 FRANK CIELO, KEMA Laboratories 20 MARK EARLEY, NFPA 21 KENNETH FLEISCHER, EPRI 22 DANIEL FUNK, Jenson Hughes 23 FRANCISCO JOGLAR, Jenson Hughes 24 CASEY LEJA, Exelon 25 3 ASHLEY LINDEMAN, EPRI 1 DAVID LOCHBAUM 2 SHANNON LOVVORN, TVA 3 MATTHEW MERRIMAN, Appendix R Solutions 4 ALICE MUNA 5 FRANCESCO PELLIZZARI, EPM 6 ROD PLETZ, AEP 7 SUJIT PURUSHOTHAMAN, FM Global 8 ANTHONY PUTORTI, NIST 9 ROBERT RHODES, Duke Energy 10 BRENDA SIMRIL, TVA 11 THOMAS SHUDAK, NPPD 12 STEPHEN TURNER, Independent Consultant 13 BAS VERHOEVEN, KEMA Laboratories 14 BETH WETZEL, TVA 15 16 17 18 19 20 21 22 23 24 25 4 AGENDA 1 Page 2 NRC Request for Test Parameter Ranges &

3 Equipment Selection...............7 4 5 Review of Phase II Draft Test Plan &

6 Comments Received...............227 7 8 Public Comment Period.............288 9 10 Adjournment..................293 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 5 P-R-O-C-E-E-D-I-N-G-S 1 (8:34 a.m.)

2MR. SALLEY: Okay, good morning. We'll 3get started here with Day 2. I take it, I trust 4 everyone rested well last night, watch my Penguins 5beat up on Nick's Flyers. Played well. That was a 6 very enjoyable hockey game.

7So, let's get started today. We had a 8 busy day yesterday, we covered a lot of information.

9 Again, I want to thank Mark Earley for the NFPAs 10presentation. A lot of that reference, there is a 11 lot of work that Gabe is going to be doing, looking 12 at stuff, like you, too, Nick.

13 And then Ashley in the presentation, Nick 14read the White Paper for everyone. I think that's 15 very valuable, thank you.

16 And Bas, you'll give presentations to a 17 fire protection engineer like me was very eye-opening 18about the electrical side. Much better handle on 19 testing than on the body protection side but it was 20 a great presentation. I learned a lot from it.

21 So, yesterday, like I said, it was a lot 22 of information that we were exchanging. Day 2 on 23 this workshop, we really want a lot of your 24 participation and we want to have a lot of good 25 6 active discussion today.

1 And again, it's forming how this project 2moves forward. We've got a good mix of people in 3here with utilities, the test labs. We got the NIST 4folks, FM, NFPA, some of the regulators. So I really 5 want a good discussion.

6 Last evening when we were looking at this 7 after the meeting, Nick, Gabe and I were kind of 8going over it. And we looked at the agenda, and what 9 we thought for today, to make it a little productive 10 and maybe a little quicker, was to flip the agenda 11 around.12So if you look at your agenda, what we 13 would like to do this morning is for the NRC request 14 for test parameters, ranges and equipment selection, 15 we'd like to start with that.

16 And I think with a lot of that discussion 17 we can actually answer some of the questions and 18things that we got on the test parameter. So, again, 19that's where we'd like to start, for the people on 20 the webinar, so that they know where we're at.

21So with that, I'm going to turn it over 22 to Gabe, Nick and Kenn Miller and we'll get the 23 discussion going. So, Gabe, you got it?

24MR. TAYLOR: So, just a couple of 25 7 administrative things before I get started is, again, 1we have the court reporter transcribing the meeting 2 so we would like you to use the microphone so that 3both he can capture as well as the sure phone. With 4 the speaker right there we can pick up the audiences 5 communications a lot better for those on the webinar.

6 So please use the phone, introduce yourself.

7That's the microphone. And on this one 8 you got to get pretty close, so you don't have to 9kiss it but almost. So, I'll just leave this back 10 here in the back corner.

11It's off. You have to hold those two 12 buttons down, it will do a countdown, three, two, one 13and then it will go on. So it's off right now, 14 you're not being recorded.

15PARTICIPANT: Not hot mic you're saying?

16 MR. TAYLOR: No hot mic, just this one.

17 (Laughter) 18MR. TAYLOR: Okay. So, as Mark said, we 19 want to get into the test parameters.

And really 20 that's one thing we got feedback from Phase I, is a 21 lot of people thought that we should be doing 22 different things than what we did.

23Part of that was, part of the real 24 reasons of what we did was because of what 25 8international group wanted to. Given the importance 1 of the aluminum impact, we need to have a little more 2 control on what we're doing here on Phase II.

3 So we felt it was important to understand 4the parameters that we can change when we test. We 5 get feedback from the audience on what those ranges 6for parameters should be, as well as the types of 7 equipment that we should be testing.

8 So the test plan, as Nick mentioned 9 yesterday, went out for public comment last year.

10And we have those comments. And we've resolved most 11 of those comments, however, there are a few that 12we're looking for feedback. So he'll cover that this 13 afternoon.

14 My point is that the test plan is still 15flexible, so it's not final. There is a lot of time 16 for changes before we actually go to the testing.

17 So hopefully we get valuable feedback and 18 make the changes and do the testing that we need to 19 address the problem.

20MR. MELLY: And again, we are doing this 21one with the OECD. And I believe we have ten-member 22 countries who are going to be joining on.

23 So next week I will be also sharing a lot 24of the insights from this meeting with the 25 9 international members to try and come to a resolution 1 of the parameters, what we're going to be testing at, 2 voltage current, duration, things like that.

3MR. TAYLOR: And the NRC is the biggest 4 player in this program so we have the biggest weight 5 of directing the program where it needs to go.

6 Obviously, if the country donates equipment or money, 7 they're going to have some say.

8 But if we can influence them with a good 9 background or a technical explanation of why it 10should be covered a certain way, I feel that that 11 should make it easy on Nick when he goes and tries to 12 negotiate the test plan and agreement on the 13 international folks.

14Next slide. So, what we really want to 15 get out of this presentation, I kind of already went 16 over it, but feedback, lots of discussions, all 17right. I'm not up here presenting like yesterday.

18 We went probably over 200 slides yesterday giving you 19 a lot of information.

20 Today we want to focus the discussion but 21 really hear from the audience and get feedback to do 22 what we need to do, as I mentioned earlier.

23 Understand the range of operating 24conditions. You know, all plants are very, aren't 25 10 the same. There's a lot of variations out there.

1 So understanding the different parameters 2and configurations and the importance. So we'll look 3 for feedback on that aspect.

4 And then also, getting into the equipment 5that we'll be testing. As I mentioned, there is 6 likely going to be some equipment donation from the 7 international participants, however, it's not going 8 to be as extensive as it was in Phase I.

9 Right now we're looking at two countries 10 that are going to donate equipment because they could 11not donate monetary funds to the program. So, all 12 the other equipment we'll have to go out and procure.

13So we're kind of working off a blank 14slate. We're looking for feedback from the audience 15 on what equipment is typical, what aspects of the 16 equipment are important for procurement.

17Next slide. A few weeks ago we put out 18this document. And the real intent and purpose of 19 the needs and objective of the document was to a very 20 simplistic high level, identify what the hazard is.

21You know, in general, common terms. As 22well as the second part of the document, identify 23 parameters that we'd be interested in trying to 24 understand the range for current, voltages, 25 11 durations, that sort of thing.

1 So, really, the whole purpose of getting 2that document out there was to get the audience to 3 look at it, look at their plants, help understand the 4 range of equipment variations and bring that 5 information here so we can have a thorough discussion 6 on understanding those variations.

7 Okay, so if we look at the hazard, and I 8 took this from a presentation that's referenced 9 there, but it provides a general description of the 10 arc events and the different types of effluence that 11 comes out of it.

12 So, the things that we're interested in 13is obviously the thermal effects. So that's going to 14 radiate energy out and potentially damage equipment 15 from the thermal aspects.

16 Also, this says copper vapor but 17obviously aluminum vapor is the same issue. The 18 conductor vapor that can go out much farther than the 19 thermal aspects, as far as the damaging concern.

20 So the vapor that gets expelled from 21 these events is going to be important to understand, 22 both how far it can go out and the nature of that 23vapor. Can it cause damage or shorting to other 24 electrical equipment that could cause functional 25 12 failures and then impact the ability to shut down the 1 plant or control the plant.

2 Also, another thermal aspect is the 3molten metal that gets expelled from the event. So 4 you melt the aluminum, you melt the copper, it gets 5thrown out, how far does it go. That's basically 6 another ignition source besides just the thermal 7 energy from the arc.

8 As Nick showed you yesterday from the 9 Turkey Point and the other events, pressure can also 10have an effect on the plant. Breaking down barriers, 11 opening doors.

12 So, that defeats certain fire protective 13barriers. So understanding that is also an important 14aspect. And I think there's been a lot of work done 15 in that area, but also something that we're 16 interested in.

17 And similar to the molten metal but also 18somewhat different, the shrapnel. So non-melted 19 material that gets ejected from the event and could 20 cause physical damage to equipment in the surrounding 21 area.22 So, the goals of the program, we talked 23yesterday when I was going through the different 24 matters, really we have a bounding method that's in 25 13 6850 and we wanted to see if we can refine that.

1 Whether it's for aluminum or copper or anything else, 2 we want to see if we can make it more realistic, more 3 representative for the plants.

4 And then also, if we modify that base 5 model, can we then take that model and modify it for 6 aluminum, the effecting impacts of aluminum.

7 So that's really what we want to get out 8of this overall program. And the objectives that we 9 need to complete to reach that goal are kind of 10 listed here.

11 Some of these we won't be talking about 12today, but a lot of them we will. So the first 13 bullet is really the important thing, understanding 14 the test conditions, or making realistic test 15 conditions based on plant configurations.

16 So understanding the electrical 17distribution. And we have Kenn Miller up here who is 18 the electrical to help us along with that discussion.

19 And also understanding operating 20experience. A great operating experience in, because 21it provides valuable information. Specifically, when 22 we talk about the duration aspects of some of these 23 events.24 We're going to go through some of the 25 14parameters that we can change. And the way that 1 we're going to go through it, is I'm basically going 2 to take one parameter, provide some background 3information and have discussion. And we'll go 4 through that, through the different parameters.

5 And at the end, because the parameters 6 kind of play on one another, we had a general 7discussion of all the parameters we support. But 8 again, it's a free for all, if you have questions 9 just ask them as they come up.

10 On the collection and measurement, and 11 also the developing application and measurement 12devices, we're not going to get into too much 13 information on there.

14 If there are questions related to that, 15please bring them up. But we've been working on 16 developing measurement devices, heat flux gauges and 17 whatnot for quite some time.

18 There was Sandia report that got put out 19in 2014. I know the Japanese have been working on 20 measurement devices and Tony Putorti and Scott from 21 NIST have been working on that.

22So I think we have what we need from 23 that, but if you want to have some discussion, 24 obviously we can it's just not going to be a focus of 25 15 the lead presentation.

1And then the last part, some more later 2 on, AMIs and the data and using that to improve the 3 models and revising the models, which are outside of 4 the actual testing program but it's an important 5 piece to consider how we're going to use the data.

6Next slide. So, here's some of the test 7 parameters that I've identified for discussion today.

8 So the important ones are duration, voltage and 9 current.10 And as we talked about yesterday, voltage 11 is not necessarily the system voltage that we're 12 going to be concerned about but the arcing voltage.

13 And we look at that, get into other things like bus 14 bar spacing and closure configuration, the impact, 15 the actual arc voltage.

16 Grounding configurations and X over R 17 ratios are also going to be important to understand 18 for the types of available fault current that you can 19 have in the system, as well as your DC time constant 20for the ACs asymmetric portion of the current 21 profile.22 And one thing I want to talk a little bit 23 about also is the arc initiation phase, phase angle.

24 So when you initiate the arc, again, on a asymmetric 25 16 aspect of the current profile, it may have some 1 impacts on the pressure from these events to run the 2 testing.3 And then the last one I want to talk 4about is arc location. For testing realistic 5 equipment, where should we initiate that arc.

6 So, should it be back in the bus bar 7 section, should it be up by the breaker slabs or 8somewhere in between. And that's something where we 9 need some, a lot of the discussion because when we 10 get into the measurement aspect, we want to make sure 11 that we can capture the energy coming out of the 12equipment. And we're going to be limited by the 13 number of instruments that we can put out.

14So we can't like have a slab power all 15 over, we have to kind of position them where we think 16the arc is going to come out of the cabinet. So 17 knowing the equipment and how the magnetics effect 18 arc, travel and movement and then being able to 19 position equipment where we can measure those, those 20energies, is going to be important. So we'll spend 21 some time on that.

22Next slide. Okay, so thermal energy.

23 That's an important piece for the thermal aspects of 24 the zone of influence.

25 17 If you look at the IEEE type calculations 1 and the NFPA tables, the calculations really rely on 2 both arc voltage, arc current, duration as well as 3 the heat transfer coefficient as a simplified 4 assumption, or a simplified model to characterize the 5energy. So, we'll start off talking about those 6 variable first.

7Next slide. And also, what I've done 8here is taken our test matrix. And we can come back 9 to this slide as needed.

10But this is currently how we have it 11broken, broken down. This is just showing the 12electrical enclosures with the cabinet. Experiments 13 that we have tested.

14 We have two spares at the bottom which we 15haven't decided what to do. Those could be, possible 16 used for pressure or they could be used for 17 additional replicates or something else.

18 And we haven't decided what to do with 19the spare tests right now.

But basically, if you 20 look to going toward the right material, we split it 21 up evenly between copper and aluminum.

22Voltages levels. We have half of them 23 being low voltage, 480, and the other half being 24 medium voltage. Either 4160 or 6.9.

25 18 The 4160 are likely going to be the 1equipment that's donated from Korea. They've already 2 identified what equipment, and it's right between 3here and the breakers. So unless they can be tested, 4 unless the enclosure is rated for a higher voltage, 5 it's likely that equipment will be tested at that 6 range. And then the rest is at 6.9.

7Current. This is going to be a little 8different from what you see in the previous test 9 plan, and I'll get into currents in some later 10slides, why we've selected these, but we're looking 11 for feedback.

12 15 and 25 for low voltage and 25 and 35 13kA, that's thousands of amps, for medium voltage.

14 And then the durations range from anywhere from two, 15four, eight seconds for the enclosures. One, three 16 and five are actually for the bus bars.

17Gaps. You can see that column is empty, 18 so we'll have some discussion there because, from 19 what I found, there is no standard that says this is 20 what the gap should be for low voltage or medium 21voltage, it's basically up to the manufacturer to 22 design their equipment to meet the performance 23 requirements and various standards.

24 So, if there is any information to 25 19 support a specific gap, for the enclosures, we're 1 more interested in learning that.

2 And then a lot of the energy, I think we 3 want to put the energy in there before we go test 4 them and understand what it is, but there is numerous 5ways to calculate that. And it's dependent on, 6highly dependent on what your arc voltage is. So I 7 didn't fill that one in right now.

8 MR. MELLY: Question.

9 MR. TAYLOR: Yes.

10MR. VERHOEVEN: Good morning, Bas 11Verhoeven from KEMA. I was just looking at this 12table and normally at low voltage panels the short 13 circuit current are much, much higher.

14 An earlier rate of 18, 19 kiloamps, 15especially in the industrial area. While in the 16 medium voltage, the currents are a little bit lower 17 in the value that is here.

18 So how come you don't have higher short 19 circuit currents at low voltage amps?

20MR. TAYLOR: So, I can get into that, but 21 if we can go forward a few slides, the current will 22 be done.23Okay. So, back up a little bit. All 24right, go, okay, go forward. All right, so on 5.4, 25 20 we don't have all the plants information that we need 1to identify what the arcing, all the fault currents 2are. You know, that's going to be specifically 3 classified as a need to, system needs and whatnot.

4 But what we do have is some information 5 on transformer powers, impedance to the transformer 6and the system voltage. So you can do an impendent 7 bus type calculation, which is what I have done here 8 for some of the plant.

9 Do you see the sample is pretty small?

10Only 18 for low voltage. Basically, what I did is, 11 for plants that I could find that information, I took 12one or two buses and ran that calculation. And then 13 this is just a composite.

14 So, with a low sample number, I wouldn't 15say this is highly representative but it's the best 16 that I can do to come up with kind of understanding 17 where the fault currents are.

18 So what, on the low voltage side what 19 this is showing is just a histogram of, the blue 20 would be your voltage fault current and then the 21 orange would be your arcing fault current, based on 22 the calculation IEEE 1584.

23 So based on the orange arcing fault 24 current, we picked basically a mean value of the 25 2114.5. So that's where we got 15, and then 25 was 1 somewhat higher.

2Obviously, the way that I did it, I 3 didn't include any motor contributions or anything 4else. And that's implicit, but it's what I had to 5 work with.

6 So, you are seeing some multiple currents 7that are much higher.

But typically, when you go 8 from medium voltage to low voltage, at least from the 9plants that I saw limited by the power of those 10 transformers being, maybe on the high end, 2 megavolt 11 amps, ratings of those transformers.

12 I don't know what the experience is here 13but typically that's current on the higher end of 14 those transformers.

15MR. MELLY: And that's exactly the type 16 of information that we were looking to collect is, 17 with plant experience, what are your fault current 18 levels for these low voltage pieces of equipment, 19 medium voltage pieces of equipment.

20MR. VERHOEVEN: Yes. Also, to give a 21comment. When we test the buses that we get from 22KEMA, and we test low voltage circuit breakers, 23 normally the current value is, for normal circuit 24breaker testing, are in the values of 60 kiloamps, 25 2218, sometimes even 100 to 30 kA for industrial 1 circuit breakers.

2And I can't imagine that this is an 3 industrial implication.

4MR. FUNK: I guess my input would be, 5this is probably not unrealistic. And I do agree 6 most 480 volt systems, or even 600 volt systems, the 7 standard rating would typically be 42 kA, 65 and 85.

8 That would be a tide test.

9 But nuclear plants are a somewhat unique 10 from a broad perspective of all industry in that 11they're medium voltage to low voltage step down 12transformers tend to be small. Typically ranging 13 from 750 kVA to 1.5 MVA. It's pretty normal.

14 So, I think that maybe explains the 15 difference that you're looking at it.

16 PARTICIPANT: Okay.

17MR. FUNK: Which means the gear typically 18will be subjected to fault currents. Hopefully, 19 considerably below their rating.

20 Maybe Shannon or some other utility folks 21 can weigh in on that.

22 MR. TAYLOR: Yes, I can speak to --

23MR. FUNK: There will of course be 24 exceptions, but I think Gabe's approach is not 25 23 unrealistic.

1 PARTICIPANT: Yes.

2MR. FUNK: Yes, let's go over here real 3 quick. Shannon Lovvorn from TVA.

4MR. LOVVORN: Yes, Shannon Lovvorn, TVA.

5 I would agree that the median looks about right to 6me. Our low voltage fault currents range around the 7 16,000 amps or maybe some 18 or some lower.

8 The switchgear is typically feed off of 9like a 1000 kVA transformer for us. So that's, I 10 think for us, I think you're in the right range.

11MR. MILLER: I guess the other factor 12 that can affect that too is, I'm thinking about the 13 plant I was at, I think our 480 volt system was 14 solidly grounded as opposed to the 4 kV and 13 kV.

15 So that can tend to drive the fault currents up on 16the lower voltages. And I can't remember what our 17 numbers were but --

18MR. LOVVORN: Yes, we're on, our 480 19 volts are delta ungrounded.

20 MR. MILLER: Ungrounded, okay.

21MR. FLEISCHER: Ken Fleischer from EPRI.

22 So, since I anticipated a workshop, I brought several 23 nuclear utility diagrams and I've already spot check, 24 I've seen 750 kVA up to about 1.5 MPA. We can --

25 24 MR. MILLER: For transformer size?

1MR. FLEISCHER: For transformers for a 4 2 kVA to a 480.

3 MR. MILLER: Yes.

4MR. FLEISCHER: I can continue to look at 5 them and we can maybe adjust those number ranges but 6 that seems to be right spot on with what Dan Funk was 7 just saying.

8 But, I did bring drawings so that we can, 9 if we need to fact check something that may be 10 available on a one line, just let me know and I can 11 do that here.

12MR. TAYLOR: Okay, thank you. While 13 we're here, let's just go to the next slide to show 14 the medium voltage.

15 And I did it for 4160 because if you look 16at the times, I mean, that's about 85 percent, 17 somewhere around there, have 40, or excuse me, 4160 18volt gear. And maybe about only a third of the 19 plants have 6.9 kV here.

20 So 4160 is much more common in plants for 21medium voltage on distribution. So I put it up that 22 way just to see if there's any difference between the 23 two.24 Again, the sample size is not that great 25 25but similar to what I did for the low voltage. And 1 just a histogram showing voltage fault and arcing 2 fault.3 One thing that is different here is that 4 you saw a much bigger shift from voltage to arcing 5fault in the low voltage. And that's primarily 6 contributed to the IEEE method.

7 There is a calculation for low voltage 8and a calculation for medium voltage. And that 9 approach is based on a statistical analysis, or a 10regression analysis, of the data that they had to 11 develop those models.

12 So, because this is a histogram, you're 13 still getting a reduction from your voltage to your 14 arcing fault, but it's not enough, at least for this 15case, to make it drop bins. And my bins are 10,000 16 amp width for the bins.

17 So you can see I didn't put the mean and 18median for the voltage and the arcing fault. And you 19 see the arcing fault currents does reduce but not by 20 as much as you saw in the low voltage.

21 The low voltage is almost half of the, 22 the arcing fault current was almost one half of the 23 voltage fault current.

24So, again, not that many samples. I took 25 26one, maybe two buses, two buses. And I took two if 1 I saw a difference in the power rating of the 2 different transformers, whether it was a startup or 3 a rad or whatever the transformer was.

4Typically there wasn't too much 5difference. So in those cases I only take one 6 sample. But that's for the 4160.

7Next slide. This is for 6.9 kV. So 8 again, slightly higher but not by really that much.

9Right around 30 kA on the mean. And there is some 10change here, just it jumped in when you went from 11 voltage to arcing.

12 So, when we looked at what we wanted to 13 test for this, 25 and 35 is what we picked. We 14 wanted 25 to be, we wanted one point on the medium 15 voltage to be in line with one point on the low 16 voltage side.

17 So 25 is the common voltage between the 18 two, or common term, sorry, 25 kA current between the 19two. And then we picked one higher, which is 35 in 20 this case.

21 Again, this doesn't take into account 22motor contributions but, again, it is a non-23 conservative assumption but then, again, it doesn't 24take into your opinions on your source. Which is the 25 27 opposite.1 So, another factor of the real system 2response, somewhat undetermined. It really depends 3 on what motors you have running, what your system in 4 HEAF test for your power source.

5Where is the handheld? Sorry about that.

6 But with the information that I had 7 available, this is at least one way we can get a 8 basis or at least a discussion here and get feedback 9 from that.

10So I know it looks like Kenn is looking 11through some diagrams, if you guys want to look 12 around while I kind of go back and talk about that, 13 at least you know now where we're coming from on the 14 current problem.

15 So if there's not any questions, I'm 16going to go back to where I left off. We'll hit this 17 again.18 (Off microphone comment) 19 MR. TAYLOR: Nick, can you go back?

20 MR. MELLY: Back to --

21MR. TAYLOR: Wherever I left off. I 22 guess the HEAF plan --

23 MR. MELLY: Here, yes.

24MR. TAYLOR: Yes. So again, 1525 low 25 28voltage, 2535 for medium voltage. And then the 1 durations, we need to spend some time talking about 2this. Some of the limits are based on what available 3energies from KEMA and other ones are based on OpE 4 for these types of events.

5 Obviously, if the breakers work that's a 6 sign your talking cycles and not seconds. And if 7 they worked on designed you wouldn't have the HEAF.

8And there's also been a lot of research 9 done from NFPA and IEEE and others, on the shorter 10end. So it's another reason why we're focused on the 11 longer durations. We have slides on that as well.

12 One thing before we go, EPRI's report on 13 review, the testing in the OpE, as well as the other 14 report on the mean, it did identify that there is not 15 that many events in low voltage, and right now we 16 have it split 50/50 between medium and low voltage.

17 So I think one thing that's up for 18discussion is, should we reduce the focus on low 19 voltage and still run some test on low voltage but 20expand the number of tests that we do at medium 21voltage since it seems, from the information that 22 we've been reviewing that there is more medium 23 voltage type events occurring in the plant that have 24 a safety significance, or a potential safety 25 29 significance.

1 So, I want to just think about it right 2 now, but I want to come back to that because, as I 3said, this is fluid, we can make changes now. But 4 once we get an agreement signed with the 5 international group, we're then limited to what we 6 can change.

7 Because the test plan actually goes into 8the agreement and it gets signed. And to make 9 changes to that, from then you have to get all the 10 countries agreement.

11MR. MELLY: Seventy-five percent of the 12 project group.

13MR. TAYLOR: Okay. You got to get 75 14 percent of the project group to agree to that change.

15 So, right now the majority of the testing 16is in that project. There are some, actually, Nick, 17can you go to the website? Pull up the website with 18 your test plan.

19MR. MELLY: Which website do we have it 20 on?21MR. TAYLOR: Just go to, it's already 22 pulled up unless you closed it.

23 MR. MELLY: Here we go.

24MR. TAYLOR: So, the test matrix that I 25 30 showed, it just takes Nick's test plan and puts it in 1this form. But you can see the blue and the orange, 2 the blue cases are what's going to be sponsored with 3 the international group, and the orange cases are the 4 ones that the NRC are going to have to fill in on 5 their own. Just from a financial standpoint.

6 We couldn't get enough interests for 7 contributions for an international group to cover 8everything. And because, as Nick mentioned 9 yesterday, aluminum is more of a U.S. interest than 10 most international countries, they elected not to 11 focus on aluminum as much.

12 So, I guess, just for your awareness, 13that's where this came from. And we do have a little 14more flexibility on some of the tests. You can go 15 back.16 MR. DALEY: Hey, Gabe?

17 MR. TAYLOR: Yes.

18MR. DALEY: I just got to ask a question, 19 maybe it's premature, but where are you seeing the 20 aluminum? Where are we seeing that?

21 Are we seeing it in both 480 and 4160 or 22are we seeing it primarily in one application as 23 opposed to another and are we seeing it in newer 24 plants via older plants?

25 31 And the reason I'm asking is, because the 1 separation issues that older plants are much 2 different than newer plants, so I'm really kind of 3curious what you want to look at, from both an 4 aluminum and a copper.

5MR. MELLY: So I can take a shot at 6answering that. Based on the NEI survey, we can't 7 tell whether it is in older plants or newer plants 8because the plants were anonymous. Just numbered as 9 they were given to us.

10 MR. DALEY: Okay.

11MR. MELLY: However, in terms of the 12 pieces of equipment that we've seen in them, it's all 13over the board. I was expecting to see it more in 14 the low voltage equipment based on the OpE that we've 15 seen, however, it was in many pieces of medium 16 voltage equipment.

17 And in many plants, it's used primarily 18 as the enclosure material for the bus ducts 19 throughout the entire plant.

20So, from the NEI server that we did 21receive, we saw more aluminum than we expected.

22 However, only 52 plants were involved in that survey.

23MR. MILLER: Did they break it down like 24aluminum used as bus work versus just enclosures or 25 32--1MR. MELLY: Yes, they did. For, in terms 2 of the bus ducts that they used, they gave us whether 3 it was the conductor material within the bus duct or 4 the housing.

5 In terms of the electrical cabinets, if 6 aluminum was involved in any piece with the closure, 7 conduct or for the cradle units themselves, that was 8 also included.

9 But in terms of vintage of plant, we 10 don't have that information. No.

11MR. DALEY: Yes. I'm just trying to 12 think, I'm really getting ahead of ourselves, but I'm 13 trying to think, where is this going to go, right, 14 and the effects on different plants when you got 15 certain plants that have separation for all the 4160, 16 really good separation.

17 Some plants have minimal, you will have 18 40, you might have 4160 switchgears separated by 19 rooms by you might have some of the other division 20kind of roaming through that. And then there is some 21 of them where you got 4160 that are in the same room 22 and you're separated by distance, right?

23 MR. MELLY: Yes.

24MR. DALEY: So you're really dealing with 25 33three different factions there. So, in the 4160, I 1 mean, to me it's just a clearly, from a fault 2 perspective in how the fault is going to affect 3 things, it seems to be the, it seems to be more 4 important.

5 And the other thing, now this conductive 6 property, the conductive vapor in that, that would 7probably just be talked a little bit more about 8 because that could affect everything.

9MR. MELLY: And, Bob, I think we may be 10 giving you a call when we start the pilot plant 11 discussion for the selection of plants when we get to 12 that stage, because these are the exact things that 13 may come up in that discussion of how we select which 14 plants we're going to try to evaluate the risk impact 15 of this, where is this going.

16MR. DALEY: Covering a broad spectrum of 17 types.18 MR. MELLY: Yes.

19MR. TAYLOR: While I have it up here, 20 also on the bus duct testing, the majority of it is 21aluminum. So you can see the NRC is picking up a lot 22of that cost. So feedback on that has been very 23 important for us to make sure that we get the right 24 bus ducts configuration and whatnot.

25 34 And so we have, Slide 8 will into this, 1 but right now, for your awareness is, we're breaking 2 it down by bus material and by enclosure material.

3 So copper bus fuel enclosure, copper bus aluminum 4 enclosure, aluminum bus steel enclosure and then 5 aluminum bus and aluminum enclosure.

6 So, I think we feel that those are up in 7 the plans, but obviously if there's one of those 8cases isn't it, we would want to test it. So 9 feedback when we get to those slides, on that aspect 10 and the importance of the testing matrix.

11Can you go to Slide 10? So, what I just 12showed turned into a table form. We had a couple 13spares that we could do. You know, possibly do a 14 longer duration. Again, it's flexible.

15 But everything at 4160, no changes in the 16current. So that could be something that we could 17change, we could use multiple currents. And it's 18 just a variation of bus and duct material.

19 Again, gap spacing, I don't have any of 20that information. So I'm not sure there are 21 standards out there that says it should be five, 17 22 years apart, whatever it is. So feedback on that.

23Design specs. Again, it's solely 24 representative of --

25 35MR. MELLY: And the one thing that we 1 also need to discuss when we're talking about the bus 2 ducts, is whether the conductive material is 3insulated or not insulated. That can have effect on 4 where they place the arc and how we place the arc.

5 Because we are dealing with a limited 6 segment of bus ducts that we can put in the test 7 enclosure itself, we need to be able to determine how 8to hold the arc in place. If it has the insulated 9 material on the conductors themselves, we've seen in 10 testing that has been done for the Cooper event that 11 we can hold the arc in place.

12 However, if there is no conductive, 13 there's no insulation material, we may be just 14 creating a rail gun and having all of the energy go 15 out directly one end of the arc travel down towards 16 the end of the bus.

17 So that's why the test plan currently had 18a hard break in the conductive material to try and 19hold the arc in place. However, we'd like to know 20 whether your plants have the insulated material on 21 the outside, inside the plant and kind of get a feel 22 for how that is across the fleet.

23MR. TAYLOR: And no bus bars have been 24 procured yet so we're open on what the configuration 25 36 is. Question from Mark Salley?

1MR. SALLEY: Yes. This is a good place 2 as any to jump in here, but going back to this 3original testing we did with the internationals and 4 the GENESIS on how this all came together, we knew 5 one of the biggest expenses was going to be procure 6 nuclear grade equipment to basically destroy it in 7 the testing.

8 And when we had the original agreement 9 with the countries, the concept that we went in with 10 was, hey, find a plant in your country where they're 11 putting in some new switchgear, they're upgrading, 12updating something, get that old equipment. It's 13 going to be scrapped anyhow, so let us use it for the 14 testing and that will be a whole cost saver.

15So that's why you saw a lot of that 16 equipment in there, for example, Korea were the first 17 ones that were on it, and those original cabinets we 18 did were from Korea.

19 Again, Gabe and I through Zion Solutions 20 were able to procure the duct that was tested, the 21 bus duct from Zion.

22 But when I, I guess the plea or the 23 question I'm throwing out here is, we have some 24 plants that will be going into decommiss ioning, and 25 37 as you guys are doing upgrades to the plant, we'd be 1 open to say equipment donations that you could or if 2you'd want to put into this program. So, again, keep 3 that in, we can work through EPRI, and equipment that 4 you would maybe be doing an upgrade, would be 5 scrapped out, could be used or for plant safety 6 commission, we'd welcome that equipment.

7 MR. TAYLOR: Thanks, Mark.

8MR. SHUDAK: Tom Shudak with NPPD. So 9 the comment I keep hearing coming up is, what we have 10 as our plan configuration, will it be one that has 11 insulated aluminum buses.

12 So, we really don't know what's out 13there. We have an anonymous survey out there. You 14 keep bringing up the point, we need to know that 15 information, not all plants are represented here in 16 this room.

17 Is there another effort to gather that 18 information from industry to make sure we get the 19 testing right, as far as that configuration?

20MR. TAYLOR: I will let Mark Salley 21 answer that question.

22MR. SALLEY: You know, as far as going 23 through that information, we've kind of backed off 24there. And we would look to NEI, as we did with the 25 38original one. Or to EPRI or EPRI and NEI together 1 working in that.

2 So, again, that's where we would look for 3 that to come from.

4MR. TAYLOR: And not that it's that 5 important, but process wise, what process these could 6be used to get that information. And really, I don't 7 think where we're suited by, I'm not sure that we get 8the support to go out and issue a 50, for a type 9 letter or a temporary TI for the inspectors to go out 10 and look for it.

11 So we're really trying to leverage EPRI 12 or NEI to help us understand what's out there.

13MR. SALLEY: And again, that's the 14 purpose of today's meeting and having your attendance 15 here, is to learn what's out there and what we should 16 be testing. So, again, that's a big part of today.

17 We've done similar things with cables.

18 If you remember years ago, looking at cables, what 19was the typical thermoset cable out there. We did 20 the biggest bang for the industry, our data, and it 21was raw asbestos fire wall III. I think that was in 22 like 80 percent of the plants.

23 So, again, those were the things we 24learned so that our research is applicable. So, 25 39 again, that's a key point of you all being here at 1 this meeting here today is to help us guide how we do 2 this testing.

3MR. TAYLOR: And I guess the other option 4 would have been a bulletin but, you know, those are, 5take at least a year. And then trying to keep this 6 within the GI program being timely response is also 7 something that we're trying to work with.

8 Question in the back with TVA.

9MS. WETZEL: So when do you anticipate 10 the agreement to be signed and the test to be locked 11 in where we can't provide input?

12MR. MELLY: The agreement right now is 13currently held up with lawyers. There is some 14 viability issues in signing with that many countries 15 and the NRC internal ability to accept testing, 16 liability and indemnification clauses within an 17 agreement.

18 We were hoping to have the agreement 19 fully signed in hand towards the end of summer.

20August time period. That is when the test plan will 21 be included on it.

22 Small changes can be made but they will 23need to be discussed with the entire group. And 24 small changes I consider to be, if we decide to put 25 40 insulation on the bus bar materials.

1 After, if we receive more information, we 2 would have a meeting with the international members.

3 And it's a very collaborative working group that we 4 have established for this testing program.

5 We usually do not run into any 6 controversial issues where we can't breach consensus 7agreement. So I don't see it as a huge stopping 8 point to make changes as we move forward.

9 Like we say, I mean, testing, we have the 10 spares in there because we're bound to have other 11 cases where we just see something we do not expect.

12 So, I don't think it's going to be a hard stop.

13 That said, we still are under an 14agreement. But the end of August is what we're 15 shooting for to have the agreements signed by the 16 member countries.

17MR. TAYLOR: Okay, thanks, Nick. Can we 18 go to the next slide then.

19 Okay, so for the first parameter, one of 20 the most important, in my mind, well, arc holders is 21 very important too, but one of the also more 22 difficult ones to predict from the HEAF standpoint, 23 is the arc duration.

24 So, one thing that we wanted to 25 41 understand was, your clearing times for your 1protection systems. And if you look at the needs and 2 objectives document that went out, we tried to break 3 that down to understand what was more current for a 4 typical case and then when you're needing a breaker 5or your protective equipment fails to clear on the 6 first clearance protector, device point.

7 Such that the fault would travel upstream 8 and then you have a second protective device to clear 9and what the clearing times for that are. For both 10voltage fault as well as arcing type conditions. So 11 that's one aspect of the duration that we're 12 interested in.

13 The other aspect that might have more 14 weight is actually the operating experience that 15we've seen. The IN has information on the aluminum 16events. And I think there has been presentations 17 that have discussed the duration of other events in 18 the duration.

19 So, because those are really, those types 20 of events are really what the frequency is being 21 contributed to for the Bin 16 type events, we're 22 looking into a lot more on that.

23 But we wanted to understand if the 24 systems worked as designed, what are those starting 25 42 times for voltage and arcing fault currents.

1So next slide. Just to reinstate what 2 we're planning on right now, and this is flexible, 3 low voltage four and eight seconds, medium voltage 4two and four seconds. We can go longer on the low 5voltage because of the system, the energy that we 6 have at KEMA.

7 And then the bus bar is one, three, four, 8potentially up to five seconds at medium voltage.

9 But most of the majority is medium voltage bus bar.

10Testing right now is one to three seconds in 11 duration.12 So, some information --

13 MR. MILLER: We have a question.

14MR. LEJA: Thank you. So, one thing that 15 comes to my mind is, if you go back a slide, for the 16 480 volts, or low voltage switchgear, we're saying 17four and eight seconds. And I can't think of where 18 the eight seconds come from.

19The only incident that comes to my mind 20is Fort Calhoun. And that situation you had a bad 21 design, so misaligned status.

22 And also, you had zone select interface 23jumpers that were not being disabled. So you lost 24 all protection in the incident.

25 43 Other than that, I've never seen a 1 coordination diagram where you would actually have 2 eight seconds before the second level protection 3would pick it up. So maybe two, three seconds, maybe 4even four seconds at some plants. I haven't seen it.

5 But eight seconds seems very excessive.

6And I don't think anybody really supports it. So for 7 the realism scenario that we try and represent, I 8 really don't think it's representative.

9 And in any statistical analysis, when you 10 have an event like Fort Calhoun where it's so off the 11 wall, you know, everything else is four to eight 12 seconds, maybe 12 seconds and you have an event 13 that's 42 seconds, you were to remove it because it's 14 anomaly, right?

15 So I'm thinking, if we're trying to be 16 real and trying to really represent what's out there, 17 I don't think there's anything there that supports 18eight seconds. And that's just my opinion but I 19don't know how anybody else feels about it. Thank 20 you.21MR. FUNK: Yes, along that similar that 22Gabe gave, we've talked about this, and Nick and I 23 too, that for a properly designing system, we really 24 wouldn't anticipate exceeding the two second range 25 44 even with your backup protection under most 1conditions. But again it is, as he just mentioned, 2 statistically based because there is no way to know 3 exactly what the fault current is.

4 And so, coordination folks work under the 5concept of a range of available currents. You get 6your voltage for your theoretical max all the way 7 down to an overload condition and we're just not good 8 enough to predict exactly what's going to happen for 9 every scenario.

10 So, what strikes me with the testing 11 that's been done so far is it's performed on a basis 12 that your primary and secondary protection will 13 either fail the function properly or were improperly 14 designed to begin with, based on normal 15 considerations.

16 Is there any value in performing some 17 tests that, number one, is through primary protection 18 work property or the primary protection failed and 19 the secondary protection was relied upon?

20 In other words, you would limit the 21duration approximately two seconds. I mean, you 22 could work a range on that but that kind of concept.

23 Or is there, the problem with that is, 24 well, you don't get necessarily great results as far 25 45 as maximum information gained on equipment suffering 1catastrophic failure, I agree with that. And with 2 limited equipment, you want to maximize that.

3 I think that my concern, to sum it up is, 4 every test we've run and projection for future 5 testing is we assume loss of both primary and 6secondary protection, which has happened. I'm not 7 going to say it hasn't, but it's just the unusual 8 case versus the normal case.

9 MR. TAYLOR: Kenn had a question --

10MR. MELLY: Before we get there I wanted 11to touch on the eight seconds question. We do have 12 international experience where we do have one low 13 voltage event that occurred in Germany that lasted 14 the 8.5 seconds.

15 We do understand that these are anomaly 16 events but the frequency is fairly low on these 17events, which brings you into the model. We're 18 looking for more informa tion on that event in 19 particular.

20 And we do know that the low voltage 21 events, we've been discussing it is, they are rare 22because you have to defeat so many levels of 23protection to get one of these long events. So we 24 see that in the frequency in that when we had done 25 46 the split for the low voltage versus medium voltage 1 equipment, you see more of these higher arcing faults 2 in the medium voltage because you don't have to fail 3 these many levels of protection.

4 One of the reasons for testing at the 5 eight second range for the low voltage equipment, is 6because it does give us the larger energy profile.

7 We've seen in the operating history for Robinson, for 8 instance, we have had the ten second to 11 second 9 range of duration for that medium voltage case.

10 And we cannot replicate that at the KEMA 11facility. We're limited by the duration that we can 12 go for medium voltage 3.5, four seconds, kind of 13 stretching our generator there.

14 So there were several reasons why we went 15 to eight seconds, but that's why we're here today.

16 We want to discuss the selective values and take note 17 of, we do understand that eight seconds may be fairly 18 excessive for the low voltage cases, however, we have 19 seen these off the wall events in operating history.

20MR. DALEY: Do you want me to, I'm going 21 to put on my regulatory hat here to respond to, Dan.

22 I think you're right, Dan, I think we have to, yes, 23 you have to be realistic with it.

24 I just look at this whole encompassing 25 47thing and I see where the major effects on it. You 1 know, we go back to one of the first slide 2 presentations we had and it talked about fire 3 protection and it talked about single failure, right?

4 Okay, but you're talking, when you talk 5about those two things you're talking two different 6events, right? Fire protection you don't assume that 7 something is just going to fail.

8 So, I think the idea that doing a 9 realistic, some type of realistic arcing failure for 10 fire events seems to make a lot of sense, because in 11 the end, I think that's where it will be from a 12 regulatory perspective.

13 Now, the single failure, you assume a 14single failure so having something, having a fault, 15 this type of fault with a breaker, or at the breaker, 16 is kind of, it's kind of an all-encompassing event so 17 then you get into the situation, okay, I can go this 18long with it and how does that effect other 19 divisions, right, if there --

20 And like I said, I'll go back to what I 21 was saying earlier, you're dealing with a lot of 22 older plants in that regard because, later vintage 23plants. I'm a BWR guy so I think BWR-5, BWR-6 just 24 ain't, you're not going to get there.

25 48So, that's kind of my thought. So, I 1 agree with you, Dan, I think you should have a lot 2 more realistic ones but you should probably put some 3 bigger ones in there too.

4 And for the conductive stuff, for when we 5 actually get to the point where we do the conductive 6 cloud, the cloud of death --

7 (Laughter) 8MR. DALEY: -- then we probably need some 9 longer durations in there.

10MR. SALLEY: Yes, Dan, and if I could 11 just put one thing up, we have the fire forum last 12 week where I gave a presentation on this and a couple 13 guys grabbed me afterwards and said, hey, could you 14 just bring one thing up in the meeting, and I said 15 sure.16 And one of the things they told me, or 17 they thought was a vulnerability to look at, is when 18 they're switching over to standby generators, that 19 that would be a time when we could see these longer 20durations. They just asked that I bring that up, so.

21 Again, I don't know if it is or not, but 22 that was something I thought about.

23MR. FLEISCHER: Ken Fleischer with EPRI.

24 So, there is a technical explanation, theoretically, 25 49 as to how you can get to eight seconds and beyond.

1 It's governed in more of our White 2 Papers, and none of the OE, there is not sufficient 3 data in the OE to confirm any of these events 4 occurred.5 But, with the available fault current in 6 a three-phased voltage fault with protection being an 7 instantaneous element and an overload element, you 8 should trip instantaneously.

9Now, if you get to the point with an 10arcing fault and the current is below the 11 instantaneous setting, now the instantaneous setting 12cannot see that so a couple things can happen. Now 13 you have to start writing up the time curve until you 14 hit your inverse, overload element.

15 Which could take anywhere from one second 16 up to 15 minutes, depending on where you arrive that 17curve. Similarly, even if that breaker failed, your 18 next upstream breaker, that current, again, cannot be 19 seen by the short-term instantaneous, or short-time 20 delay or short-time delay element, again, waiting for 21 a thermal element to pick it up.

22 And so that's why I can see these occur.

23 However, the OE just doesn't have the sufficient data 24 to confirm it so that's why these were that long.

25 50 But I do still support that we should 1 really scrub our OE, understand the best we can and 2 develop realistic tests.

3MR. LEJA: So, just a response to that.

4 It is true that it could take from one to 15 seconds 5 depending on where you are in the thermal region.

6 But, again, for sake of realism, I looked 7 at coordination plots with the fault currents that 8they're simulating and I couldn't get there. I 9couldn't get to four seconds. I'm saying, yes, you 10 could get from one to 15 seconds.

11 So yes, you could get there in one to 15 12 seconds, however, the fault currents that we're 13simulating is just not realistic. So, again, it's 14 the realism scenario, you know.

15 MR. FLEISCHER: Yes, and I agree.

16MR. LEJA: And another thing about Fort 17 Calhoun is, if we take that into consideration and 18 say, you know, this piece of equipment was mis-19 designed and this install and we're testing it 20 because it could happen, then what's stopping us from 21 taking the protection in the turbine and having the 22 turbine explode and say, hey, this could happen as 23 well. So I don't want to go that far.

24 I think we will get a lot more out of it 25 51if we had it at two and four seconds. The reason I 1 am saying two seconds is because that's kind of what 2 the gear is rated for so I think we'll get agreement 3 and people will be very interested in those results.

4And also, also because if you look at 5 relay coordination, I was actually looking at that 6 last night as well, could we get there, could, 7 because normally when we do relay coordination you 8 just want to, you just want to have things 9 coordinated, you don't really worry about this arc 10 fault arc kind of stuff.

11 But if I was worried about it and 12 somebody gave me a project and told me to redo the 13 relay so that we do get down to some reasonable 14 amount of separation.

15 So let's say we want to get down to those 16two seconds, we could get there. If you're talking 17 about higher than that, I mean, get down to like 18 three seconds I mean, I don't know if it's going to 19 happen, so, I think we could get down to two seconds 20 and that's why I think it would make more sense to do 21 it at two and four seconds versus four and eight.

22 Because I'm having issues articulating 23 myself but maybe I'll get better. Thank you.

24MR. TAYLOR: Any other comments to our 25 52 conversation, but do we have a response or --

1MR. MELLY: It was a suggestion we go to 2 two to four because the currents that would be 3 associated with having a fault that sticks in for 4 that longer duration, 15 seconds, are not the fault 5 currents that we are proposing for our test program, 6 they're the ones that are riding more like the Fort 7 Calhoun and the 5 kA range.

8MR. TAYLOR: Okay. So basically from 9 your coordination curves you looked at, for these 10 durations, this is what the fault current would have 11to be to get that duration. Is that kind of 12 summarizing what I missed?

13MR. MELLY: At the fault current that you 14 are receiving you wouldn't --

15MR. LEJA: The duration would be much 16 shorter and much less.

17MR. TAYLOR: It would be much less of 18course. Which would be much less than four seconds.

19Which is not realistic compared to what, I don't 20 know.21MR. FUNK: Yes, I would agree with that 22 but I do, the equipment, I think there is some force 23 to suggest that testing between one and two seconds 24 also has value, which would assume loss of your 25 53 primary trip function which, in a lot of cases would 1 be differential trips which are measured in a few 2 cycles and they're reliant on maybe a less sensitive 3or a current relay back for something along those 4 lines.5 So I think that adds to this realism that 6I know that I have been looking for. One to two 7 second range I think stays within the rating of the 8equipment and it provides maybe a realistic data 9 point.10 These other cases we're talking about are 11the created events. So, not to belabor the point, 12 but I think there is value in seeing what we would 13expect to happen. What's more closer to the median 14 than the extremes. Thanks.

15MR. MELLY: I think the, well, the reason 16 that the one to two seconds were even lower, the .5 17 seconds that's recommended in the guide or the metal 18 clad switchgear.

19And the reason that's not showing up in 20 the current test program is because those events are 21 not what I would consider to be part of this higher 22 in the arc fault damage zone bin.

23 MR. FUNK: Okay.

24MR. MELLY: These are the ones that are 25 54 going to go into Bin 15, where you just have the 1 internal cabinet damage creating an internal fire 2 rather than breaching the cabinet and have externals 3 under the influence of damage.

4 MR. FUNK: They won't be a HEAF.

5MR. MELLY: They won't be considered a 6HEAF and that's not going, that's what I do not 7 intend to be in the HEAF frequency bin.

8 So that's more in line with what happened 9 at Turkey Point where you have the 32 second, or 32 10 cycle event, .5 second event.

11 Also, the Cooper event was right on the 12threshold. That was a one second event. And these 13 are the events that they're questionable, they're not 14 HEAF events because you do have damage but you don't 15 have the type of damage that we're looking to get 16 towards modeling.

17 If I had endless amounts of resources, I 18 would love to put in the one to two second range so 19 we have those events as well.

20 However, we're talking about the 21 destruction of a piece of equipment, extreme 22refurbishment, a full day in the test lab and the 23 full apparatus of test equipment and measurement.

24 Which is why we were focusing more to get what we see 25 55 in line with the Bin 16 bin.

1MR. LEJA: And I agree with that. I 2 agree .1 second, .5 second testing would be a waste 3of your money. IEEE C37.20 says, .1 seconds to one 4 second.5 MR. MELLY: Yes.

6MR. LEJA: But I still find value in two 7 seconds.8 MR. MELLY: Right.

9MR. LEJA: Because I haven't really seen 10much tests at two seconds, because that's what we 11expect gear to be rated for. So that's why I see 12 value in that.

13MR. FUNK: I understand, Nick. And, 14 again, I understand the tradeoff of "a waste of test" 15 because you don't expect to see the substantial 16 results.17Be aware though that that's out of line 18with the frequency calculations right now. That's 19 the misalignment I see.

20Because they're all included. Even those 21 events that would now be your, a Category 1 or even 22 a Cat 2 that you've defined in this meeting.

23 So we would need, or my recommendation 24 then is, okay, kind of a new baseline we're working 25 56 from here so let's go re-scrub the events and get the 1 frequencies that are in alignment the way we're 2 defining this testing.

3MR. MELLY: That's exactly what we plan 4 on doing with the collaborative work with EPRI on the 5 HEAF program.

6MR. TAYLOR: Yes. That's one of the 7 reasons why we're working tightly with NFPA and 8 everyone here to try and come up with the definitions 9 because, you know, I think five, six years ago there 10 was talks about, well, let's define a MEF and let's 11 define a LEF. Or a low and a medium and then high.

12 So trying to take a definition that, with 13the categories that we've proposed right now, you 14 know, Cat 1, well, that wouldn't be included in the 15 HEAF. It may not even be included in 15.

16 Cat 2, maybe it's 15, maybe it's 17somewhere else. And a Cat 3, really what we're 18 interested in here is the high energy arc fault type 19 bin.20 So, trying to work with, to formalize 21 those and then, as the comment was, tie it into 22 frequency, make sure everything is consistent in the 23 PRA model.

24 Okay, so what I hear, two to four seconds 25 57seems, was one proposal out there. Is there any 1other feedback on the duration from the group? Okay.

2MR. PLETZ: Hi, Rod Pletz from American 3 Electric Power.

4 I don't have a good electrical 5 background, my background is classical fire 6protection. And what I'm used to seeing is during a 7 typical fire curve you'll get to a peak HEAF release 8 rate and that will be sustained at some point and 9 then it will drop off.

10 When we're looking at the duration of 11these HEAFs, is there a point where we're going to 12 see that energy level peak out and then the duration 13 of this event we're just going to see more product 14thrown against the wall, basically, and should we 15 limit our duration to when we're going to hit that 16 peak energy release rate?

17MR. TAYLOR: Okay, so I'll go ahead and 18try and take a stab at that response. Yes, I think 19 for most of the testing that we've done, and what we 20 see, it's pretty constant, the energy release rate.

21The one caveat being the aluminum. If 22 you look at the test, was 23 the finish equipment?

23 MR. MELLY: Yes.

24MR. TAYLOR: If you look at Test 23 and 25 58 you look at the timeline, you had the initiation of 1the arc up to about, right around one second. And 2the energy seemed to be pretty constant through 3 there.4 And then right around one second you had 5 a large increase of what's going to be expelled and 6 the energy that's coming out of that event.

7 We haven't got down to what caused that, 8 was it because you got up to a certain temperature, 9 did something change internally in the cabinet that 10caused that additional energy release. But that is 11 one case, that I'm at least aware of, where the 12 energy wasn't constant and you had a change during 13 the event.

14 I don't know if, Nick, had anything else 15 on that.16MR. MELLY: In terms of the energy 17release electrically, if you want to talk about the 18 transformer breakdown where right now at KEMA we have 19 the energy to look at the current constant throughout 20the event. Whereas in the plant you have the degree 21--22MR. TAYLOR: If you're cutting from a 23 generator?

24 MR. MELLY: Yes.

25 59MR. TAYLOR: Okay. So I guess the point 1that Nick brings up, if you look on your medium 2 voltage bores and you're being feed from a generator 3 during unit aux, typically if you don't clear that 4 fault, as the EPRI document, you can back feed from 5 the generator into the fault and the generator will 6 wind down.

7 Versus how we test the KEMA where we 8basically have a constant current for the entire 9duration that we need. So you don't have the decay 10 of the current.

11 I don't know if KEMA has anything else on 12 their setup. Or Tony.

13 MR. MILLER: Yes, question.

14MR. PUTORTI: Yes, the other change that 15 we saw during the experiments, as far as the ZOI and 16 the threat to surrounding equipment was, if you went 17 long enough to breach the exterior of the switchgear 18then there was a change in what we would see as the 19threat to the surrounding equipment. So that would 20 be one point in time where there was a change.

21MR. VERHOEVEN: Bas Verhoeven from KEMA 22Laboratories. With regards to the current that we 23 make during the test, we normally arrange our systems 24 in such a way that we have a super expectation on the 25 60short circuit generator. That means it will not, it 1 will maintain the steady state current.

2 In normal power networks, they say the 3 feeding power is sufficient to have a constant short 4 circuit current. That's why we can make it.

5 If you would like to see decaying current 6to reasons, just ask because we can do it. We can 7 set the machines in such a way with the citation that 8 the current can decay.

9 And also, the voltage level, let's say 10 that is for the ignition of the arc after has 11occurred. And also, the current value. That's all 12part of means we can handle. So don't be too limited 13 in your design of what you would like to see tested, 14we can make quite a bit. In terms of current values, 15 the supporting voltage and the expectation occurs.

16 And also the borings where we switch on, 17in design way with regards to the DC offset. So 18 don't bother too much with our capabilities, we can 19adapt according to your needs. So let that be your 20 primary goal.

21MR. FLEISCHER: Ken Fleischer from EPRI.

22 Then I would like to emphasize that that's something 23 we consider, and I can see Gabe putting it on the 24 board.25 61 Because the difference between an 1 infinite bus feeding a fault and a decaying voltage 2 from a generator where its exciter switchgear breaker 3has been open is a tremendous difference. Yes, the 4arc is sustained, yes, damage is occurring, but I 5 anticipate that the actual energy, totally energy at 6 the end of the extinguishment of the arc, will be 7 significantly different and the damage should be 8 different.

9 So, that's a capability that we should 10 consider.11MR. TAYLOR: Okay, I think, as far as the 12 NRC world, we're open to making those changes but 13 again, to make a realistic from representative.

14 However, the question is, okay, how 15 should we set it up, how should we have the equipment 16 be, setup to model the plants.

17MR. MELLY: And that will go in for 18 medium voltage rate.

19 MR. TAYLOR: For that case, yes.

20MR. FLEISCHER: Yes, it would be for 21 medium voltage, you wouldn't see it at the 480.

22To emphasize, think about this. Ashley 23 has talked to me a little bit about this for some 24 other aspects of the testing, but remember, the gear 25 62 that you probably are going to get is going to have 1 multiple cubicles.

2 And as I've seen in the other tests that 3I have reviewed, the entire switchgear is not 4 entirely damaged, there are other cubicles that can 5be used. So one of the options would be, is to do a 6 decay test and set that configuration up for a decay 7 test and then setup an undamaged cubicle for the more 8 infinite bus test.

9 So that might be how you can arrange your 10test configurations on one piece of switchgear with 11 multiple cabinets, as long as they're not destroyed 12 for one of the tests.

13MR. MELLY: Yes. And in the first series 14 of tests that we did we did try to utilize as many 15 cubicles as possible during the test phase.

16 So for instance, we had cabinets donated 17from Japan which had multiple cubicle units. And we 18 essentially worked backwards through that cubicle 19 doing, I believe four tests per cubicle trying to 20 maximize the testing within the same compartment, 21 varying the parameters.

22 So we are looking to potentially do that.

23 It really comes down to what equipment we buy, how 24much it needs to be refurbished and how quickly we 25 63 can accomplish that.

1MR. TAYLOR: Yes, and the other thing 2 that the German regulator is interested in is 3prorogations to adjacent cabinets. So if we do get 4 damage, we're limited to use of that second cabinet.

5 I think it's a valuable comment of 6comparison. So that's something that we should 7 consider.8MR. MELLY: That also brings into 9 question, so, one of the goals for this test program 10 is to really limit variable so that we can have a 11 consistent data collection process.

12So by changing the cubicles, we need to 13 change the arran gement of the instrumentation. And 14 that brings into play trying to predict where the 15affluent is going to release from the cabinet. Which 16 in testing is difficult.

17MR. LOVVORN: Shannon Lovvorn, TVA. I 18 guess somewhat of I hear in the feedback is, we're 19 talking about a lot of different permutations here 20 of, some of us are wanting to take a realistic case 21 and then I think the NRC is somewhat interested in 22the failure case where the device does work as 23 expected.24 Even in Kenn's definitions of a HEAF 25 64 yesterday was suggesting that there is an adequate 1design or some sort of failure that contributes to 2 the event.

3 And so, I guess I'm okay with 4 understanding that phenomenon and how much damage 5 that can be, but we also need to be looking at, there 6is a lower probability of that and putting a 7 probability number on to that kind of failure, if 8 we're going to take that larger damage.

9 So, I guess, that's why the duration 10 discussion seemed pretty difficult because we all 11kind of had a different mindset of what's realistic 12and what we're trying to accomplish. I think there 13 is different permutations of what you can consider.

14 One is that your system acts as design.

15 I mean, I shouldn't have a HEAF really but if there 16 is some sort of failure then hopefully there is some 17 sort of probability that we can put on to that.

18 Because, the presentations last week in 19 the fire forum, we all know that fire risk is 20dominating the risk picture. And part of that is 21because of over conservatism in the, what I'll call 22the immaturity of these models. So, we just need to 23 be thinking about that.

24MR. MELLY: So I'm going to, let me grab 25 65 that one real quick is that right now we currently, 1this ties in with the definitions, the frequency in 2 all these pieces.

3 As the current Bin 16 frequency, or where 4 we want it to go, I believe that every one of these 5 events here, where you see these longer duration 6 HEAFs, the ones where we would consider these 7 damaging events with the zone of influence, every 8 single one of these did have a protection level 9 breakdown allowing it to occur.

10 So that reliability of the component, the 11 failure of the component, will inherently be captured 12 by the frequency of the events that we use for the 13input to the model. And that's currently the way 14 that we're trying to handle it rather than going the 15 approach of putting the reliability number somewhere 16 within the equation itself.

17MR. TAYLOR: I think that's a good 18 comment and breakers and tracking devices are highly 19reliable. And so we were thinking along the same 20 lines but then to get that conditional probability of 21 becoming a HEAF becomes typical.

22 So I think what Nick said, it seems 23 reasonable that if we can have a good definition of 24 what these events are we can develop a frequency from 25 66 the fire events to characterize the frequency and 1 then the data from this just characterize the hazard 2 that we have kind of continuity.

3MR. MELLY: And from the OpE I think I 4 only, I can only off the top of my head remember one 5single event in an ISO phase where there wasn't 6 levels of breakdown where there still was huge damage 7states. But primarily, every single case that we've 8 seen in the HEAF bin did have failures of protective 9 levels.10MR. MILLER: And we did want to have an 11 understanding of the extent of damage when we do have 12 a HEAF like that.

13 MR. MELLY: Yes.

14MR. MILLER: And we know there's been 15 failures and it's a lower probability, right.

16 MR. MELLY: Yes.

17MR. MILLER: So part of this is 18 understanding that sort of influence of that high 19 energy event rate.

20 MR. MELLY: Yes.

21MR. VERHOEVEN: Bas Verhoeven from KEMA.

22 The topic that has not been mentioned so far, at 23 least not to my knowledge is, is that if you have an 24 arc in a cabinet, and let's say the cabinet remains 25 67 closed and it stays within that box, at some point of 1time it will burst somewhat out. The doors will open 2or it burns through the cycles. That means that the 3 arc and the flash will go into different segments.

4 Is there thoughts being discussed of 5 having the cubicle also inside the space, the volume, 6 that is mimicking the, let's say the control room of 7the switchgear or is that just an open space where 8 basically the fumes can go out because it will be a 9 secondary buildup of pressure if that's a confined 10 area?11MR. MILLER: Like with the one event 12 where the fire door, the fire barrier door was 13 breached, I guess that's what you're referring to?

14 MR. VERHOEVEN: Yes.

15MR. MELLY: So we haven't discussed that 16 right now because that is one of the goals is to 17 understand the pressure.

18 So, the current test plan has us 19 measuring the pressure of just the internal enclosure 20itself. And we have been discussing ways to try and 21 extrapolate what the room pressure increase would be 22from that pressure reading. We don't have a good 23 handle on whether that's going to be possible or not.

24 During the discussions we have had with 25 68 KEMA, as well as in the PIRT, there was the 1possibility of creating a mock room around the 2 electrical enclosure and measuring the pressure rise 3 of that room so that we can extent it to plant 4 specific room designs so you know your, what the 5 pressure rise in that room would be so it can 6 challenge fire doors.

7 Another challenge there is fire doors 8 typically aren't pressure rated, so making that 9connection is another effort. It's something that we 10 are considering putting in limited number of tests so 11we can get a room pressure rise increase. Again, 12 that's going to be up for discussion within the next 13 weeks.14MR. TAYLOR: Yes, I think it comes down 15 to the balance point of what's worth the cost on the 16 effort on the NRC side versus how big of an impact it 17 makes on the safety side.

18 MR. MELLY: Right.

19MR. TAYLOR: So we haven't come up with 20a decision. We have been thinking about, I know 21 Steve Turner has some information he shared and is 22 available, and the literature on different ways that 23they tried to model pressures.

So there might be 24 some analytical approaches to model it but I'm not 25 69 sure of how much are those, approaches are.

1 Do you want to add anything?

2MR. TURNER: They are hard to do. At the 3minimum, it's very cheap to do. At least stick one 4 microphone out at 15 feet and one microphone out at 520 feet and you get data, the acoustics are all 6 terribly off, but you get some idea if you can see 7 that in some results from a February 2017 test that 8 we did. We got that data.

9 MR. MELLY: Yes.

10MR. TURNER: So, just those microphones 11 are cheap.

12MR. MELLY: And we're also similarly 13 starting that from scratch on the pressure 14 measurements because during our first attempt, in the 15 first test series when we measured the pressure, we 16 had a lot of interference, electrically, to our 17 measurement devices, so we weren't really relying or 18we can't rely on the pressure measurements from our 19 first series to try and extrapolate.

20 And that's why in the OECD report you see 21 that we've reported the pressure on the same graphic 22that we did for the electrical energy. And you see 23 a direct comparison, or spikes in the pressure 24readouts, both on, the power turns on and when it 25 70 turns off.

1 Which we would not expect when we turn it 2off. That huge jump and spike that we were getting 3 in the pressure readings if it was just based on the 4 pressure itself.

5So we were getting interference. And a 6 lot of that actually has been reduced from testing 7 that we've seen from Steve and the Japanese.

8 KEMA, in their facility, had moved from 9 a fiber optic measurement device which has given much 10cleaner results. And I think that it's true pressure 11 reading that they received in their later testing.

12 So, that's one thing to keep in mind if 13 you have reviewed the pressure that we have collected 14 from the first series.

15MR. TAYLOR: So, from what I hear, I 16 think getting back on topic, we're talking about 17duration right now. I heard a proposal to change low 18voltage to the four seconds. I didn't, does that 19seem reasonable for the group? So I think that's one 20 change we should make.

21 The other thing is, I didn't get much 22traction on the proposal to try to do the lower 23duration event, to have a second or whatnot. Is 24 there any more discussion we wanted to have on that 25 71 aspect?1 MR. FLEISCHER: For the low gear?

2MR. TAYLOR: Yes. For the low voltage.

3 Or even medium voltage.

4MR. FLEISCHER: I have some input on 5that. In a way, this is still on topic but bear with 6 me.7One of the comments that I was, had 8 offered up, and I think it's in the EPRI comment is 9 that, probably the first thing we should do before we 10 take off with all this testing, right, is to 11 demonstrate that the gear that we're going to use, we 12 understand its capability and its ratings and that we 13go ahead and take on the compartments which should 14 not damage anything else and you shouldn't worry 15 about, propagation of the arc outside its cubicle is 16 to do a full standard C37.27.7 test.

17 The right size arc wire, the duration of 18 only half a second protective device being used to 19 limit the fault and all that and do a standard test 20 in conformance with that IEEE standard, and then just 21 say this gear has been certified suitable for use for 22this exploratory type testing. And so therefore 23 that's my traction for the .56 seconds.

24 Yes, I do believe we should do it, but it 25 72 would be part of the certification that the 1 switchgear is healthy to continue with the testing.

2MR. MELLY: I'll take that one. We may 3 be able to do that for the low voltage cases.

4 For the medium voltage cases, we will not 5be able to do that because of the current contract 6 that we have with KEMA, that will be essentially 7 doubling every medium voltage test that we have.

8 Which is expensive.

9 So based on our current restrictions and 10 budget that we have moving forward, I don't know if 11 we have the money to do a full another 16 or 32 12 tests.13MR. FLEISCHER: We're going to have that 14many switchgear components? I was think of 15 certifying one as one design.

16MR. MELLY: Yes. We may be able to do it 17 if we buy one standard design for the other tests.

18 For the donated pieces of equipment, I do not know if 19 we can do one for them.

20MR. TAYLOR: Yes, so donated equipment, 21 we don't know exactly what we're getting from 22 Germany, but from Korea we're getting 5 kV Magna-23 Blast with the protocol track and breakers and gear.

24MR. MELLY: Also, I don't think, for the, 25 73 are you discussing the metal clad switchgear standard 1with the, where we're not going to be ejecting the 2 hot gas out the front with the cotton waste stabs and 3 things like that?

4MR. FLEISCHER: I'm trying to digest the 5question. I'm just thinking of the IEEE C37.20.7 6test for metal class switchgear. Yes, it would be 7 the medium rolls.

8MR. MELLY: That's the guide on arc 9 resistance?

10MR. FLEISCHER: That's the guide on the 11 arc resistance certification.

12 MR. MILLER: Yes, with the cotton.

13MR. FLEISCHER: Yes, those are cotton 14 targets, I do remember that.

15 MR. MILLER: That's right.

16MR. MELLY: And if it failed that test, 17 what would we consider?

18 MR. FLEISCHER: Well, that would be bad 19 now wouldn't it.

20MR. MILLER: What's the purpose of that 21though? We're basically confirming that the 22 equipment was built to the original, per that --

23 MR. FLEISCHER: Right.

24 MR. MILLER: -- out there.

25 74MR. FLEISCHER: And so, as we learned 1 from Bas yesterday, only 25 percent of those, if you 2don't test them, you don't know. It might not work.

3What's the purpose of the test? The 4 purpose of the test is to demonstrate to us, and our 5 stakeholders and the utilities, this test going 6 forward is using gear that was sufficiently designed 7 and rated, and verified rated, to be able to, for the 8 test going forward.

9MR. MILLER: But wouldn't that apply to 10 ever piece of gear that we are, going to do the test?

11 Do we need to do that for all?

12MR. FLEISCHER: No, I don't know if we 13 need to do it for all of them, but I do understand 14now the challenge. If we have different pieces of 15 gear and it's a mixed bag, I understand the 16 complexity and the cost impact.

17 I guess I was a little more idealistic 18 that we would procure, until today I didn't know, but 19 when I was thinking these comments through, I figured 20 we were going to design and procure maybe a GE Magna-21Blast switchgear breakers for all of the medium 22voltage tests. And you would just take one of those 23 switchgear and verify that it was arc resistant for 24 the IEEE standard.

25 75MR. MELLY: But I'm also more worried 1 that if I did that, let's say I reached out and I 2 said, this is the most common piece of equipment that 3is in 20 plants and I'm going to go do the test. And 4 then I run the certification at KEMA and I come back, 5this is not, it didn't pass the test. Is there 6 implication for those 25?

7MR. FLEISCHER: Well, that's a good push 8 back.9 (Laughter) 10MR. FLEISCHER: The thing is though, is 11 that then we'd have to look to see if it's a 12 production issue versus a design.

13MR. MELLY: It could open a can of worms 14 though if we --

15 MR. FLEISCHER: It would.

16 (Off microphone comment) 17MR. MILLER: Can I speak? But that's not 18 the purpose of this testing either, we're not trying 19 to certify the equipment that's built per the IEEE 20 C37 standard, meets the standard, we're looking at 21 stuff beyond that.

22 MR. FLEISCHER: So is --

23MR. MILLER: So I'm still not seeing the 24purpose of what, of doing what you're saying.

25 76 Because that's not the purpose of the testing.

1 MR. MELLY: It's the baseline.

2MR. FLEISCHER: Kenn, correct, it's not 3the purpose of the test. What it does is it verifies 4 that the equipment we have chosen to do the test 5 would have survived a standard arc fault at .5 6 seconds or less. That's all.

7 Just that the piece of equipment we've 8chosen is robust enough so that the results we get 9 know that under ideal conditions of all protective 10 device sparking, it was okay, now we're taking this 11 beyond its rating --

12 MR. MILLER: So at baseline --

13 MR. FLEISCHER: It's a baseline.

14 MR. MILLER: -- this test.

15MR. MELLY: Do the plants have that 16 certification right now certification right now on 17 pieces of equipment that are installed?

18MR. FLEISCHER: I don't know. I don't 19 think they all have it.

20 MR. MILLER: Well --

21MR. FLEISCHER: It's a guide, right?

22 IEEEs are a guide, so this is voluntary.

23MR. MILLER: Kenn, I mean, when plants by 24 equipment per a spec, they get a certain indication 25 77that it meets the requirements of the spec. well, it 1 is.2 I mean, that's the whole purpose of 3buying something per our spec. You get a certificate 4 of performance.

5MR. CIELO: Yes, you might get a 6 certificate of conformance or you might get a COC.

7You're not going to get what Bas was describing 8 yesterday.

9 Which is a certification from an 10 independent test lab that it met 100 percent of the 11requirements of whatever the standard is. Which is 12 a full type test, which is an independent test.

13 Which means that the manufacturer submits 14 that piece of equipment, in our case, to our test 15lab. Our test engineer and our test team verify the 16 device to the drawings.

17 They actually do a physical verification 18 against the drawings and then they run it 100 percent 19 to the requirements of the standard without any input 20 from the manufacturer. It's to the standard.

21 We pick the circuits, we pick, you know, 22 in accordance with the standard, all the voltages, 23 the amperages, all those test parameters, every piece 24 of equipment that we use in our test labs is 25 78 verifiable backed NIST, in the U.S. anyway, because 1 of our ISO 170.25 certification.

2 So we can vouch for every piece of 3 equipment that we use to measure all of the, the 4results basically. And so when it passes 100 percent 5 in accordance with the standard, then a lab like KEMA 6Laboratories would issue a certificate. But that's 7 not necessarily what you're getting when you get a 8 COC.9MR. FLEISCHER: On the interest of time 10I was going to try and get it back to Gabe here. So 11 my thoughts are, why don't we just put it as a 12 parking lot, a breakout session if you will, as a 13 consideration.

14 And if it's viable or there is a way to 15make it work, that would be great. But it's 16 essentially, it's a baseline.

17 The other thing to your question, I now 18know what you're asking. When you do a purchase spec 19 for a piece of switchgear, you're going to put in the 20 design standards.

21 MR. MILLER: Yes.

22 MR. FLEISCHER: Right.

23MR. MILLER: This is all the standards 24 that they're testing against.

25 79MR. FLEISCHER: Right. But you may not 1--2MR. MILLER: They're putting, GE is going 3 to say this breaker will meet those requirements.

4MR. FLEISCHER: Right. But you may not 5put this testing standard in. In other words, you're 6 only going to get what you put in your specification.

7 So I think the answer to your question 8 is, probably a lot of the early nuke designs probably 9weren't aware of this testing and probably didn't 10have this done. So their certification search is to 11 a design standard as opposed to a testing guide.

12MR. MILLER: Yes.

So, to me, I would 13 attribute this to an assumption on our part with the 14beginning of the testing. Otherwise we're talking 15 about a lot more testing and scope increase of what 16 we're doing.

17 MR. FLEISCHER: Yes, okay.

18MR. DALEY: I got this microphone about 19 20 minutes ago so this will probably sound like a non 20 sequitur at this time.

21I just got to keep on going back. I see 22 this, we got to know what the end of this is and 23 what, keep the end in mind of what we're looking at 24 and look at it in terms of, I'm just thinking right 25 80now. I'm thinking your design basis event and then 1 your external events.

2 So you got your design basis single 3 failure type thing which has specific events, 4 specific portions of this could potentially affect 5that. And then you got your 805 portion which is a 6 totally different animal.

7You're dealing with probabilities. I 8mean, some of the same things in deterministic, and 9 the Appendix R plants are going to be similar but 10 you're dealing with probabilities too so you're 11 looking at that.

12 And then you got the deterministic 13 plants, the Appendix R plants, which would be some of 14 the things that you're looking at in 805 as a subset 15 of that.16 And then the last thing, and I haven't 17 given a lot of thought to this is, is you got the 18 seismic type event where you got, where you would 19 almost half to be dealing with a non-1E bus that's in 20proximity to your Class 1E buses. Which is not that, 21 that's not that uncommon particularly with older 22 plants.23 So, anyway, I just think you got to have 24 the end in mind.

25 81MR. CHEOK: So, this is Mike Cheok. A 1quick comment. And you know I am always interested 2 in everything I just heard.

3 So the beauty of risk analysis is you do 4couple all the consequences with the frequency. So 5 when you do PRAs you have large, medium and small 6 LOCAs. Loss of coolant accidents.

7 And each one of these things come with a 8frequency. And for the large LOCA you have a much 9 smaller frequency.

10 Enough even for a loss of offsite power 11 so you have different durations, how long you lose 12your offsite power. And obviously, the longer you 13 lose offsite power the smaller the frequency.

14 And you know, the other extreme is 15seismic PRAs. You know, you do the different hazard 16 curves, you have the smaller earthquakes all the way 17 to the large earthquakes and obviously the frequency 18 curves would be, you have a lot smaller frequencies 19 for the large earthquakes.

20 So I think the discussion here is a very 21 good discussion and the results, I think one thing we 22 have to be very careful about is when we're 23discussing tests and results and when we have the 24 results that we want to display, let's say for a four 25 82 second duration or even a eight second duration, we 1always have to couple that with a frequency. We 2 cannot just go out and tell people the zone of 3influence is X without saying this was a very 4 infrequent event that is coupled with several 5 different failures and probably with a frequency of 6 ten to the minus eight or whatever.

7 So I think we all have to keep the 8 consequences on the damage states where the 9 frequencies in our test and in our communications of 10 results and things like that.

11MR. SALLEY: It's a good point, and I'll 12 give it back to Gabe, but you say, Bob, with the 13 earthquakes. Look how we do earthquakes, right.

14 We have our design basis earthquake, 15which is what we expect to see in the plant, and we 16 do that and then we have our safe shutdown 17earthquake. If you want to draw a parallel to this 18 with the short durations, what we expect to see with 19the equipment and when it does work, that safe 20 shutdown earthquake is what the HEAF is.

21 So, again, going back to the definitions 22 I think is going to be key. Here you go, Gabe.

23MR. TAYLOR: I think this is a great 24 discussion, exactly what we want.

25 83 MR. MELLY: Yes.

1 MR. TAYLOR: So, next slide.

2MR. TURNER: I got one more comment on 3 the --4 MR. TAYLOR: Steve Turner.

5MR. TURNER: As far as talking about half 6 second duration and one second duration, I think one 7of the things we're missing in this presentation is 8 we look at our Phase I test, there is a huge data 9 bank of half second tests out there that's in the 10 IEEE and FDRA flash tests.

11 There have been hundreds and hundreds of 12tests at half a second and one second. The weakness 13it has is it barely touched on them. But I would 14 recommend that you guys go back and grab the IEEE 15 papers and write that up in some way to convince 16these people that at least it will cover half a 17second and one second is kind of covered and we're 18 just going to focus on this new bin for the really 19 bad stuff.

20 And as far as going back and doing the 21 test and saying, look, it passed, my view on those 22 tests have always been, and you saw it in his video 23 yesterday, the first one of those I ever saw that 24 passed I said, geez, if that's a success I'd hate to 25 84see a failure. Because it still looks pretty 1 dramatic.2 Those flames can still come out and stuff 3can still get lost. And so, it's just not a very 4 appealing thing to say, look, this passed the test.

5 And the other thing is, the thought may 6be that if you do that easy test first you can do a 7test in it again, you probably can't. It's going to 8 damage it enough, you won't hit it again.

9MR. MELLY: And again, one of the videos 10 we saw yesterday was a pass of the test because it 11 didn't come out the front facing where the personnel 12 would be, which is how that test is designed.

13 However, for our zone of influence, I'm still worried 14 about the back and the top and where the cables are.

15 So, running that standard test, if it's 16 to a cabinet design to have a detachment plate or 17 something along those lines, I'm still going to be 18worried about my cables, my other equipment and 19 things like that.

20 While I do still care about personnel, 21 it's generally not in the PRA calculation.

22MR. TAYLOR: Okay, the next slide. And 23 I think I've gone through this enough but the longer 24 versions ties into our energy and the thermal aspects 25 85 of that.1 Obviously, that type of modeling is going 2 to be different from expelling the molten materials, 3the shrapnel, the pressure, things we also need to 4 consider.5 But I think that's what a large scale 6 testing, how good we can capture that through 7 document it via video as well, on thing I didn't 8 mention yesterday is, in the first series of tests 9 we're going to bring up Sandia's, the plan is to 10 bring out Sandia's video equipment and they can 11 actually track particles and run them into their 12 computer and they can get you miles per hour, where 13it's going, what it's bouncing off of. So we want to 14give that a shot and see if that works. That might 15 help us on a non-thermal aspect.

16 Again, to characterize better the ZOI.

17MR. MELLY: I think we've covered this 18 part.19 (Laughter) 20MR. TAYLOR: Yes, we've heard a lot of 21 discussion, maybe I should have showed this on the 22 second slide.

23 (Laughter) 24MR. TAYLOR: But if there is anything 25 86 else, we want to talk about duration right now, 1please, please do bring it up. If not, we will go on 2 to the next topic. Unless we need a break.

3MR. SALLEY: Yes, why don't we take a 4 break, Gabe, 15 minutes. Okay.

5MR. TAYLOR: All right. So, let me just 6check the webinar real quick before we break. And it 7 shows no questions coming in.

8 MR. SALLEY: Any more out here?

9MR. TAYLOR: It doesn't look like we have 10 any questions on the webinar, so let's go ahead and 11 take a break for 15 minutes and we'll come back at --

12 MR. SALLEY: 10:30.

13MR. TAYLOR: -- 10:30. So almost a 20 14 minute break. 10:30 we'll be back.

15 (Whereupon, the above-entitled matter 16 went off the record at 10:11 a.m. and resumed at 17 10:32 a.m.)

18MR. SALLEY: All right. How do we turn 19this on? Okay, so we will pick it back up and, 20 again, thank you, this is kind of what we envisioned, 21 organized chaos.

22 I mean the discussion is great, you know, 23we want this input. Also we're reaching out to our 24 folks on the webinar, you know, you all are part of 25 87 this, too, so we got young Tom there on it, so please 1 feel free to type any of your questions in and bring 2it in, so we want the people on the webinar to be a 3 part of this process, too.

4 So with that, Gabe, let's go until 12 5 o'clock.6 MR. TAYLOR: All right.

7MR. SALLEY: At 12 o'clock we'll take 8 lunch because the discussion is going to be a good 9 break point and we'll pick it back up.

10MR. TAYLOR: Okay. So the next parameter 11is voltage. Obviously, we are looking at low voltage 12 as 480, the most common in U.S. plants although there 13 is some 600s out there.

14 And then medium voltage, 6.9 is a lot 15more common in European plants, so, you know, there 16is a big push from the group to go at 6.9. Also, 17 some of the later U.S. plants were putting in 6.9 18 safety buses, but there is a lot of variation.

19 6.9, as I mentioned earlier, about half 20 of the plants in the U.S. have it, right around a 21 half. 4160 is much more common in U.S. plants.

22 So if the plant is medium voltage, 6.9, 23 they test it at that, however, if donated equipment 24 doesn't go there we'll go to 4160, but, again, it's 25 88 another topic for discussion.

1 If we want to do something 4160 or some 2 other 7.2, 12, 13kV, that is up for discussion, but 3 that would be something that Nick would have to take 4 back and negotiate with the international, so just a 5 little background on where we are with that.

6 I have some slides that might help the 7discussion along, so we can go to the next slide. So 8 we talked about this a little bit yesterday, but what 9 I did is I went back to the Phase I that the NRC ran, 10 those test results, and the orange is the system 11 voltage, so it was either 480 that we ran for the low 12 voltage or, it's a little hard to see, but we ran 13 6.9, 7.2, one 4160, and then two 10kV tests in Phase 14 I.15 So the orange is showing the system 16 voltage and the blue is showing the arcing voltage, 17 so the voltage collapse during the arc, and you can 18 see, although we have some variation between 4160 and 19 the 6 or 7, 6.9 or 7kV, and even 10, the arc voltage 20 doesn't change too much, too dramatically.

21 And some of the reasons behind that could 22 be the bus bar spacing has an impact on what the arc 23voltage is. So just to make that clear, when you are 24 looking at energies it's not the system voltage that 25 89 you're concerned with it's the arc voltage that goes 1 into the energy calculation.

2Next slide. So just another way to look 3at this. What I am comparing is in the blue is our 4 arc voltage and the orange is basically the distance 5 between our buses in the gear.

6So based on that trying to see if it 7 tracked, and in general it tracks, but, again, there 8 is some variation that doesn't align with, you know, 9 a correlation on just distance between the buses.

10Next slide. So there are some methods 11 out there to predict what your arc voltage is based 12on your spacing. So this method here, without 13 getting into too much detail, comes from the CIGRE 14 document.15 And basically based on your spacing it 16predicts a voltage per unit distance, or voltage in 17this case, kilovolts per centimeter. So the gray is 18 basically that model and the orange is our medium 19 voltage test and the blue is our low voltage test.

20 So if you look at the medium, you know, 21 the medium, the general tendencies that the model 22 tracks and however there is some spread for, more so 23 for low voltage than medium voltage, and on the 24 medium voltage side it tends to over predict the arc 25 90 voltage compared to what we are seeing in the 1 testing.2Next slide. This is showing the same 3 information but now instead of distance, what I 4 showed in two slides back, I am showing what the 5 CIGRE document would calculate as your arc voltage, 6 comparing it to blue, which is what we actually 7 measure the arc voltage.

8 And so it is an important parameter, but 9 just because you have a certain systems voltage 10 doesn't mean you are going to get a certain arc 11 voltage.12 I think distance does have impact, but 13there is some scatter there. So being able to 14 predict that is a complication that we are going to 15 run into.16Next slide. Okay, so that's really what 17I had on voltage for background discussion. So as 18 far as our testing, you know, I want to hear, get 19 some feedback on, you know, is 480 reasonable, and 20then the approach for 6.9kV for the medium voltage 21 gear.22 Are there other aspects of the testing, 23 the system lineup, the power supply that we are using 24 at KEMA that comes in and impact the arc voltage.

25 91 So I want to go ahead and open it up to 1 the room if there is any feedback on the voltage for 2 the test.3MR. FUNK: Gabe, yesterday you and I 4 briefly discussed, you know, fully being able to 5characterize the supply system. Can we get the 6 characteristics for the KEMA generators so we can 7 actually do a model with those, an impedance model, 8 and that way we can predict voltages, the degradation 9 in voltages with the new software that's out there?

10 It's pretty easy to do.

11So if that is available it would be of 12 interest I think for some of us to get access to 13 that, maybe just build it into the test plan.

14MR. CIELO: Yes, this is Frank. When you 15 say "characteristics," I mean it's not just a 16 generator, it's everything in between, too, right?

17 I mean because we can --

18 MR. FUNK: Yes, absolutely.

19MR. CIELO: And that would be for the 20entire lab? I mean I am not sure what you are 21 asking.22MR. FUNK: Not really. You know, in the 23electrical world when we model this we model the 24 generation or the source and that's usually done with 25 92 the transient and the sub-transient reactants, cable 1 length, and standard distances between the phases.

2It's not very difficult to do. So you 3 black box it and you should be able to characterize 4 it for us. You need the transformation.

5MR. VERHOEVEN: Bas Verhoeven from KEMA.

6 It will be very difficult because our generators are 7 not normal generators.

8MR. FUNK: That's why I am asking, I 9 understand that.

10MR. VERHOEVEN: And like I said earlier 11 that we have a way with our excitation to counteract.

12 But I don't really see the point of adding that into 13 a model because most of the times the network is a 14very strong source with a certain impedance, of 15 course, and that is basically the reference that you 16 can calculate and that we make sure that our 17 generators will mimic exactly that part, that's why 18 we have that excitation.

19 The only part is when you were referring 20 to that you have let's say being fed by the generator 21itself, that then drops down the frequency. That's 22something different. But for average short circuit 23 calculations the system voltage are basically the 24 impedance, in that case, the maximum short circuit 25 93 value, and that's what we can adapt for strong 1 networks, low networks, with very high DC offset or 2 small DC offset.

3MR. FUNK: Yes, I partially agree with 4you but maybe not completely. When we do a model you 5 have a starting point, be it an infinite bus or a 6generator, once again the model of the generator, 7those standard parameters you would use, and I 8 understand that your generator is different, so doing 9 so in a conventional manner is maybe not as 10 realistic.

11 But as a starting point the black box 12 needs to be the voltage or typically an MVA available 13 from the system and an X to R ratio which when 14 married up within the in-house part of the 15 configuration cables, et cetera, which we should be 16 able to characterize, that's what gets you the, as 17 Gabe had on his screen, the voltage, the X to R 18 ratios, and the current.

19 And with those three parameters then you 20can move forward with an electrical analysis, so 21characterizing that shouldn't be completely 22 impossible.

23 I mean if you are going to offset it with 24 your equipment then you know what you are offsetting.

25 94 So, again, to do a short circuit model, to look at 1 all the different quantities that are of interest, 2 being able to characterize the generation source I 3 feel is important.

4 It's not absolutely critical, but it 5 would sure help when we went to pair up past events 6 with test results as we try to define what categories 7 things are in and using a predicted model, the 8 software out there, eCAP, EasyPower, there's a 9 million of them, they all work the same way and they 10 are all sort circuit calculations conducted in 11 accordance with IEEE. 12 So I guess that's where I leave this that 13 we want to be able to replicate IEEE modeling to 14correlate the test results to what we have seen in 15 our past events and without that information from the 16 test then we cannot do that.

17MR. VERHOEVEN: Yes, well maybe we have 18 to take this offline a little bit, this discussion.

19MR. TURNER: Let me try one quick one.

20We measured the generator voltage and then we 21 measured the test object voltage, so we get a pretty 22 good idea of what the I squared R losses are from the 23 generator to the test object.

24 That's something, we get those white 25 95 forms to go along with things. We generally just 1 report the test object voltage, but I think maybe 2 with those two things you ought to be able to come up 3 with some sort of -- it would be the black box but 4 just based on the voltage you're dropping the I 5 squared R losses.

6 We lose about 50 percent sometimes of the 7 power just in I squared R losses from the generator 8 out to the test object.

9MALE PARTICIPANT: Would that give it to 10 you?11 MR. FUNK: No, probably not.

12 MALE PARTICIPANT: No?

13MR. FUNK: The I squared R losses -- I 14understand the I squared R losses. What we are 15looking for is to characterize the impedance of the 16system because that's what drives for the medium 17 voltage and high voltage AC systems, the entire 18 analysis and also the functioning to some degree of 19 the protective relays.

20 I agree, let's not get bogged down in 21 some detailed discussion here, but it's something 22 that I think we would want to pursue a little further 23 offline. Thanks.

24MR. VERHOEVEN: Like I said we can do 25 96 quite a bit with our generators in terms of settings.

1 Basically it's the situation in your system, how it 2 is, what is the impedance there, short circuit 3 impedance, and the transients and sub-transients that 4 we see, offsets and so forth.

5 MR. FUNK: Right.

6MR. VERHOEVEN: So basically that's the 7 value that you should use in your calculations of the 8 short circuit in your system and those values we use 9 to set our machines so that we mimic your situation.

10 So it's basically the other way around I 11would say. It's your system designs or dictates what 12is happening and in case it results in a short 13 circuit code and duration depending on the voltage, 14 and let's say for the test itself we set our machines 15 in a such a way that it will represent that situation 16 because we have impedances to that we can play around 17 with the DC offsets.

18MR. FUNK: I understand. You're saying 19 let's reverse engineer the black box and you can do 20that. I think that's reasonable. I think that's 21reasonable. Please understand I am trying to foresee 22where ultimately the utilities are going to be. They 23 are going to have to try to make their case based on 24 your test results which aren't precisely what their 25 97 system has.

1 So the closer we can draw that lane, the 2 stronger argument, technical basis, they'll have for 3applying the results however this may work out. I 4 hope that makes sense.

5MR. TAYLOR: Okay. So on voltage 6.9kV 6 is there any discussion or proposed changes from that 7 approach for medium voltage?

8 (No response) 9 MR. TAYLOR: Anything from the webinar?

10 (No response) 11MR. TAYLOR: Okay, what about 480, we're 12comfortable with testing at voltage? I see heads nod 13 yes, so -- and one thumbs up, okay.

14MR. MELLY: And this is one difference 15that we had. In the version that went out in June 16the medium voltage was designated as 4160 with the 17 international interest more on 6.9.

18We have moved it to that and we felt 19comfortable doing it because when we looked at the 20first series of tests between our 4160 and 6.9 the 21 actual arc voltage did not differ very much between 22 those two ranges.

23MR. TAYLOR: Okay. So I'm not hearing 24anything, so we'll get to the next topic, current.

25 98You've already -- Back up. All right, so this is our 1test ranges at the rate pressure 15/25 for low 2voltage, 25/35,000 amps for medium voltage. Go to 3 the next slide.

4MR. MELLY: Can we, just so -- We did 5show already we have the end values of how many 6 plants we have collected this information from.

7 As we have said previously we don't have 8 a very good a venue to go out and actually request 9 additional information, so this is an area where we'd 10be looking for more information to kind of enhance 11 our histograms and the analysis that we have here.

12 It would have to be done through EPRI, 13 NEI, a voluntary, or any plant in the room, or other 14 information to enhance our dataset to make sure these 15 values are appropriate.

16MR. TAYLOR: Yes, and on that, you know, 17 I have one question for the utilities, I am not that 18 familiar, although I have been at plants where I have 19seen the arc flash labels on the equipment saying 20 this is the category of PP and this is the energy 21 from this equipment.

22 Has the utilities gone out and done arc 23 flash studies for their equipment that you are aware 24 of?25 99 MR. LOVVORN: Yes.

1MR. TAYLOR: Shannon is saying yes.

2Anybody else? Okay, so there are some plants that 3have gone out there. That might be a reasonable area 4 to look to see if you could open those analysis up 5 and see where the currents are going out, because to 6 do that you are going to come up with a voltage fault 7 and you're also going to follow the IEEE, assuming 8you follow the IEEE guidance to do those analyses 9 you'd have the arcing fault current.

10 You know, one thing that differ there 11 though is the duration they are assuming the 12protection works. So that might not be as 13applicable, but, you know, that might be a useful 14 source of information to help confirm or deny where 15 we are sitting there.

16 And if there is even databases available 17 for your equipment maybe that is something that may 18 work with EPRI or NEI or whoever to at least compare 19 with what we were working off of to make sure we are 20 in the ballpark.

21 You know, we want to be, again, a 22 realistic representative and the more information the 23better we can be at achieving that. So that's a 24 request.25 100 I don't know if there is any formal way 1 we can follow up on it, Mark, but, you know, again, 2 please feel free to contact myself or Nick if you do 3 go back and look at that and say, oh, my gosh, we're 4 10,000 amps too high or too low or whatever.

5MR. MELLY: And we can send out a meeting 6 summary after this to the participants with almost 7 the request action item below list and things like 8 that. Tony has a question.

9MR. PUTORTI: Since you are making that 10 request and following on the previous discussion does 11 it make sense to ask for a sample circuit that might 12 be appropriate for low voltage and medium voltage and 13then provide that to keep the match up for the 14 experiments?

15MR. TAYLOR: So as far as the 16 characteristics for the sample circuit --

17 (Simultaneous speaking) 18MR. PUTORTI: Something that is 19 representative of the plants.

20MR. TAYLOR: Okay. I mean we could make 21that request. I think it would be worthwhile, you 22 know, knowing there is a lot of variability.

23 But I think what where we are looking at 24 and based on some of the EPRI stuff is the medium 25 101voltage boards coming right off of the step down 1 transformer from the generator and then offset load 2 extender right below your load extender, your regular 3 load extender.

4 The other thing, too, that we haven't 5 really got into, Mark brought it up, is your easels 6 for the safety-related boards, you know, if there are 7 any voting concerns there that tension could cause a 8 longer duration type event.

9 Again, that's on duration, but, you know, 10 that would be the other area where there might be 11 some interest to get the circuit characteristics 12 there to support this testing.

13 So I wouldn't -- That's I good point, I 14wouldn't make that offer. When we get those 15 characteristics from the plants then we can make sure 16 that the test generator is set up to mimic those.

17MR. SALLEY: Gabe, we have a question on 18 the webinar by the way.

19MR. TAYLOR: Okay. Let's let Kenneth 20 Fleischer from EPRI --

21MR. FLEISCHER: Ken Fleischer from EPRI.

22 This just hit me so it's more of a brainstorm comment 23on standby EDG. It's a worthwhile comment, but we 24 talked about realism here and I don't think any of 25 102 the OE we saw there was a diesel that was on the 1 system and it is so infrequently on the system that 2 I am wondering if it should just screen out from a 3 probabilistic standpoint.

4It's a thought. Keep it on the board for 5 now, but that's just a thought I had on the standby 6 diesel.7MR. TAYLOR: Okay, thanks. Tom, on the 8 webinar?9MR. AIRD: This question is from David 10Lochbaum. The question is with NRC seeking voluntary 11 info from industry how will the NRC ensure any info 12 provided is representative of overall industry and 13 not a non-representative best case?

14MR. TAYLOR: So I think to respond to 15 that we'd have to work with some of the experts at 16the NRC to run it by the electrical group. I am not, 17 I wouldn't consider myself to be one of those 18 experts.19 But we're working with either in research 20 with Kenn's group or even out in the regions with the 21 inspectors that have the background and have the 22 experience to confirm the information that we are 23 receiving would be probably the most reasonable 24 approach to addressing that concern.

25 103 I don't know if Mark or Nick has anything 1 else that --

2 (Simultaneous speaking) 3MR. MILLER: Now I would assume if we're 4 going like through NEI and EPRI they would be doing 5that for us, too, right? I mean they would 6 filtering, or checking data for us, also.

7MR. SALLEY: Yes, I think you covered it.

8 And, Shannon, I was just thinking of Kent Bryan at 9 TVA when he would solve these types of problems the 10 word he would also use was garden variety and I think 11 that's kind of what we are looking for is that, you 12 know, middle of the road garden variety thing.

13 Again, I'll fall back to request this 14Firewall III cable. We knew from all the questions 15 we asked when we looked at that when we were testing 16 that cable that was getting pretty much 80 percent of 17 the plans for the thermoset.

18 So, again, we'd be looking for that same 19 type of input and, again, the steel line, would be 20 that garden variety circuit that we have been looking 21 for.22MR. MELLY: Yes, and to reiterate what 23 Mark said is that we will be getting that garden 24 variety but that's also why we are testing at the 25 104 current ranges that we have selected so that there is 1the potential when the model does come out it will 2 have this, the dynamic zone of influence where you 3 can use what's applicable to your plant-specific 4 lineup with the research that we are conducting so 5 that you can create a plant-specific modeling state.

6 We can't cover it all in our test 7 program, that's why we would like to be right in the 8 middle of the road with a range so that it can be 9 applicable generically.

10MR. TAYLOR: Okay, so good feedback on 11that. Thanks for the question. We can go to the 12next couple of slides. I have already gone over how 13 I develop these and what we are basing it off of, 14 but, again, any additional information to help 15 support that by all means please provide that to me 16 and Nick.17 And, you know, we'd probably have to have 18a discussion whether we can have a follow-on go-to 19 meeting or a webinar just to when we receive that 20 information and further discussions moving forward to 21kind of remain the collaborative environment and to 22 make sure that we are understand the information that 23 we are getting and that we are using it appropriately 24 to modify the test plan moving forward.

25 105 So that's just a thought that came into 1my mind. I don't know if the group thinks that this 2 reasonable to have a call in the future, but I think 3 it would help us out.

4Next slide. Okay, so let's go back one.

5 Any additional discussion on fault currents, where we 6 are currently at in the information we have at hand, 7 so it's 15K and 25K for low voltage, 25 and 35 for 8 medium. Anything to add from the audience here?

9MR. MELLY: Is there anything on the 10 webinar?11 MR. TAYLOR: Okay.

12 MR. MELLY: Okay.

13MR. TAYLOR: So that's what we'll go with 14 for the time being until we have any information to 15go otherwise. So the next question is on the system 16 connections and to the generator they can configure 17 and either be Y or delta connected.

18 Most of the testing that we did in Phase 19I was delta. However, there can be wye connections 20 made at KEMA, so as far as the plants and the uses 21 here, you know, which one do you think, or what 22 combination of the two do you think would be 23 appropriate, and that's something that we haven't 24 specified yet in our test plan.

25 106 Obviously, there is a lot of transformers 1 that are delta, wye, primary, secondary, vice versa, 2so looking at where the faults are is there any 3 opinion on one over the other, one configuration over 4 the other?

5MR. MELLY: And just as a note from the 6 testing in Phase I, the ground it seemed to use was 7 primarily related to the test, the limited test we 8 did in wye connection was because we were testing at 9 a high voltage for a long duration and the generator 10 and the setpoints we were looking for were more 11 easily obtained using the wye configuration.

12MR. LOVVORN: If I can just say that, you 13 know, for those that are frustrated that the Y would 14 be representative of the medium voltage connection 15 and the delta would be representative of a low 16 voltage connection.

17So if I was going to pick for my plant 18 I'd say wye on the medium and delta on the low, but 19 I don't know if that's representative of the 20 industry.21MR. MILLER: Another plant I was at it 22 was wye for both.

23MR. FLEISCHER: Kenn Fleischer from EPRI.

24 I can go back through my drawings, but if I recall at 25 107medium voltage there is both, there is a lot of 1 combinations of delta and wye.

2 So to get down into low voltage I see a 3more commonly grounded wye. There are a few systems, 4 I know of one plant that is delta, and there is I 5 think maybe a few others, but I would think that 6 unless we can assure ourselves that there is no 7 difference in the arc performance or damage between 8 a wye and a delta I think we should test both 9 configurations, particularly at the medium voltage 10 level.11 The low voltage level, unless I find a 12 greater preponderance of delta systems in the U.S.

13 nuclear power plants maybe we can just limit it to a 14 grounded wye, and when I say "grounded" that may also 15 have some impedance.

16MR. MELLY: And one thing to note is that 17when we did have the, hold the PIRT this was one of 18 the lower importance rankings to the eventual damage 19 states.20We didn't see it as being a very high 21 factor in any of the tests that we ran as the 22 eventual damage state, so it is a lower ranking for 23 the PIRT and I think on what I am hearing is I think 24 we agree that it's going to be a lower importance in 25 108 here.1MR. MILLER: It's more about the amount 2 of all current.

3MR. TAYLOR: Yes. So just to reiterate 4 I think maybe that might be a secondary or even 5tertiary parameter and to make sure that the KEMA 6 equipment can be set up to achieve our durations and 7 our currents and then see, you know, whether we can 8 support wye or delta for those.

9 And if we can't get a combination of 10 both, we should attempt to do it to provide the 11 comparison, but, again, it's not a, from what I am 12 hearing it's not a primary important parameter.

13MR. CIELO: Yes, this is Frank from KEMA.

14 Yes, once you decide on what you want to do then it's 15 a matter of, then we'll take that and we'll see how 16 we need to configure the lab to be able to do that 17 and we'll tell you if we can or cannot achieve it and 18 then you'll be able to judge from there.

19 MR. MELLY: Yes.

20MR. TAYLOR: Okay. So we can go to the 21next slide then I think. So moving on for the wye 22 connections and grounding, what I saw was just the 23 reactive or ungrounded, you know, what's the 24 preference, and maybe I'll start out with KEMA, you 25 109 know, what's your capabilities for grounding the 1 system, the power generation system?

2MR. CIELO: I would want to have some of 3 our facility people be able to answer that question, 4 but, again, I think also I'll just reiterate what I 5 just said.

6 I think the way we work is you will come 7 to us and say this is what we are trying to 8accomplish, this is what we need to do. We'll work 9 with you with given the limitations of our equipment 10and the lab and so forth and we'll tell you what we 11 can and cannot do.

12 So there is so many different 13 possibilities here, so many different configurations 14 that we could use, and, you know, at this point we 15 don't see any limit.

16 The biggest thing that we are worried 17 about is the thermal effect on our generators because 18 we're not going to put that, you know, mocking up the 19 test cell is one thing --

20 MR. MELLY: Yes.

21MR. CIELO: -- doing anything to our 22generators is a completely different story. That's 23 a completely different story.

24 MR. MELLY: Yes.

25 110MR. TAYLOR: Okay. So from the audience 1I think we have a comment in the back. That would be 2 Robert Rhodes from Duke.

3MR. RHODES: Yes, this is Robert Rhodes 4from Duke, or Bob. Harris has, and I'm a Harris 5 person so I am plant person, not a testing person or 6a theoretical person, we've got wye and it's high 7 resistance grounding which limits your fault current.

8 We had the high energy fault on the ISO 9 phase and the output was just exactly what was 10expected by or explained by KEMA. It went out, we 11 had faults on the ISO phase, it blew the main 12transformer, it fed back because it's a high 13 resistance ground with a grounding transformer, did 14 the same thing on that grounding transformer and blew 15 that out.16 The bus stayed intact except for the 17 piece that got ripped down through the neutral 18 grounding transformer because the vehicle or the 19 enclosure got destroyed.

20 So that resistance -- And I believe there 21 is a lot of resist of grounding schemes in the 22 industry that would tend to limit your ground fault 23 and change your whole characteristic on a solidly 24 grounded system and your results are going to be all 25 111 different.

1 And I would warn the testing people 2 because when they did this on our ISO phase it fed 3 back into the generator and we blew out our hydrogen 4 seal, so I would be very concerned about your 5 equipment because I don't want to damage your 6 equipment.

7 It didn't hurt us very long but we had to 8 replace our hydrogen seals, but there you are.

9MR. LEJA: While the microphone is here.

10 I can't speak for the entire fleet, but for the sites 11 I am comfortable with or know well, usually on a 12 medium voltage you always have a grounding 13 transformer or a resistive, a ground resistor, on the 14 low voltage it would usually just be solid ground.

15 But I have seen that throughout, not just 16 Exelon.17MR. DALEY: Yes, I was just going to say 18 that I think Duke, the Duke plants are a little 19different. I agree, they do have a high resistor and 20 I think that will affect how these events carry on.

21 But I don't know that that's very -- I 22 mean for the plants that I have seen it is not clear 23 to me that that's the trend in the industry.

24 I think you see some type of resistive 25 112 grounding or you see some solid grounding out there, 1 I think that's more prevalent than a high resistance 2 ground like you guys have, but that's just from the 3 plants that I have seen.

4MR. FLEISCHER: Ken Fleischer from EPRI.

5 Yes, there's going to be all kinds of grounding 6 schemes, but also HEAFs can also be initiated phase 7 to phase which then takes that out of the picture.

8 So I'll see if I can do a little bit of 9 work and maybe kind of get a representative of what 10 I am seeing through the various U.S. nuclear power 11 plants trying to bin that up for info, but, again, I 12 think that if we look at a HEAF from phase to phase 13 the grounding schemes will take us, that will not be 14 too much of a big factor.

15MR. MELLY: And we are initiating phase 16 to phase in every case.

17 MR. FLEISCHER: Right.

18MR. RHODES: One more comment. This is, 19again, Bob Rhodes from Duke. One other thing I have 20noticed through the years is when you see the high 21 energy arc faults the arc faults tend to find or seek 22 out the weak points in the system and the weak points 23 in the system may be the bus bars but more likely 24 they are transformers, bolted connections, things 25 113 like that that go and the test seems to be ignoring 1 all of that.

2 So I would at least look at that impact 3 and see how you see it, because if you look at the 4 Robinson fault it was the cable over the bus that was 5 completely destroyed and the conduit that was steel, 6 you just had a gap in the conduit, it's not there 7 anymore.8 So the weak point, wherever that point 9 is, not necessarily a bus bar being the weak point.

10MR. TAYLOR: Anyone have anything on the 11 transformers?

12MR. MELLY: Transformers were Mark's 13 planned phase --

14 (Simultaneous speaking) 15MR. TAYLOR: Yes, I think right now we've 16kind of excluded that from the scope of this work.

17 I think internationals, Germany, in particular, is 18 interested in that.

19MR. MELLY: Yes, there has been several 20 discussions of testing transformers at the higher 21voltages. It's a little bit outside of our scope 22 right now, we're just trying to limit it to what was 23 currently, what we saw as more dominant risk from the 24 OpE that we have looked at, it being the bus, or the 25 114 electrical enclosures themselves as the initiating 1 points.2 MR. TAYLOR: Okay, so go ahead and --

3MR. AIRD: Gabe, there is another 4 question from the webinar.

5 MR. TAYLOR: Sure.

6MR. AIRD: This one is from Preston 7Cooper. He says from the original arc flash 1584 8 equations the grounding is basically broken into 9 effectively and non-effectively grounded systems.

10 Are we wanting a more detailed view to 11 the effects of grounding?

12MR. TAYLOR: I'm not sure if we have come 13to a decision on that. I kind of threw it out there 14 to add the, you know, realism to the event, but, 15 again, I think it comes down to realism from your 16 plants and what type of fault currents you can get 17 and duration.

18 And from what I am hearing from KEMA they 19 can basically configure their system to meet those 20 parameters and for, you know, ground, phase to phase, 21 as EPRI brought up, you know, the grounding is 22 probably not as big of an impact.

23 So from that discussion, you know, I am 24 kind of concluding that we're probably not going to 25 115 evaluate the grounding versus non-grounding in the 1 test parameter, but on, you know, the representative 2 side, but that's really what the discussion was 3 focused on.

4MR. MELLY: And another additional factor 5 to keep in mind as you look at the test matrixes that 6 we have developed and how we are focusing a lot of 7 the discussion here today is that you see us trying 8 to limit our test variables and have direct 9comparisons between tests so we can draw one-to-one 10 comparisons.

11 With such a limited number of tests as 12 well as the difficulty in collecting data, especially 13 the measurement of the heat flux, the slugs, ensuring 14 that we are within the line of affluent, or the plume 15 of the arc itself, we are trying to ensure that we 16 can have comparable results which is why we are 17 trying to do replicate testing, limit the number of 18 variables.

19 So if we can through discussion or 20 consensus deem that one parameter is less likely to 21 have an impact it's of benefit to exclude it from the 22 test scope.

23MR. TAYLOR: All right. Any other 24 feedback on that one? Ken?

25 116MR. FLEISCHER: Yes, actually I'm getting 1this idea from you, Nick. I think the brainstorming 2 whiteboard we got going on here is good.

3 I think at some point what we may want to 4 do is bin these into what the priorities are, which 5 are the most important, what's your absolutely must-6 do's for the test to get meaningful results, and 7 which are the secondary and tertiary level items that 8if we can do we will and if we can't we don't. I 9 think we should probably bin this at some point.

10MR. TAYLOR: That's a good suggestion.

11I think I agree. Let's get through this and then see 12 where that takes us and if we have to come back over 13 lunch you can come back to this to do the binning, 14 that's fine, but we still need an hour or two to get 15 through Nick's test plan and comment resolution.

16MR. MELLY: We may be able to handle that 17 during the test plan discussion, because right now we 18 have a large slot to discuss the comments that were 19 received on the test plan and we're kind of knocking 20 them down one-by-one as we're going around this 21 discussion right now.

22 So I see us -- Right now we're moving at 23 a good pace.

24MR. TAYLOR: Yes, I think this is the 25 117 most important part of this whole workshop so I would 1 like to spend a lot of time working on this 2 discussion.

3MR. MELLY: Yes. And the binning I think 4 we can do during lunch.

5MR. LEJA: So this is kind of going on a 6 tangent I guess, and I'm not PRA person, but I can 7 understand the fact of frequency for like the 8 frequency level PRAs.

9 So for these medium voltage systems you 10 do your tasks how you do your task phase to phase, 11 but when you develop your frequencies I think that 12 should be one of the things you guys kind of consider 13 as this resister systems if it's phase-to-ground 14fault it's going to limit the current and typically 15 that's how most faults start, it's phased to ground 16 and then it flashes over.

17 Not to say it's not going to flash over 18if you have that resister ground, but it's probably 19something that you should keep in mind when you 20 develop your frequencies for PRA.

21 So it's less likely that it will be a --

22 it'll be phase to phase fault if you have that 23grounding resister or grounding transformer. I am 24amenable to hear which is typical. I believe it's 25 118 difficult.

1MR. MELLY: I agree and I think that's 2 something that we need to keep in mind when we do the 3re-review of the OpE for the binning and the 4 frequency as to how much information can we get on 5 the events that have occurred, type of grounding of 6 the system, the duration, the current.

7 For a lot of these the information that 8 we have is coming from the LERs which typically focus 9 on the plant response rather than the electrical 10 characteristics of the event itself.

11 I know that now that these events are 12happening if we can go out and collect as much 13 information, for instance the Cooper event, Turkey 14 Point, the Robinson, we have a wealth of information 15 on what actually occurred, specific information on 16 the duration, the current, the voltage, the 17 configuration, if we can go back and collect that 18 through the OpE of the previous, of the past events 19it would really help in nailing down the frequency 20 and nailing down the OpE and how it correlates to 21 what we're trying to do.

22MR. TAYLOR: Yes, and I agree with that 23discussion. Just to clarify though, you're basically 24 saying that if you have a high resist of grounded 25 119 system the arc would stay phase to ground and it 1 wouldn't flash over, was that the --

2MALE PARTICIPANT: I think it was less of 3 a chance.4 MR. TAYLOR: Less of a chance.

5MALE PARTICIPANT: But it probably will 6 flash over.

7MR. MILLER: This is Kenn Miller. Less 8 of a chance of the initial line to ground fault.

9 (Simultaneous speaking) 10MR. TAYLOR: Becoming a phase to phase 11 fault is that --

12 (Simultaneous speaking) 13MR. TAYLOR: Okay. Okay, I just wanted 14to be clear on that. But, yes, I think that is 15something that we can look into. Next slide. Nick?

16 MR. MELLY: Oh, sorry.

17MR. TAYLOR: You know, one thing that I 18 think is important, and I have mentioned this 19 multiple times already, you know, today is where do 20 we put that arcing wire to start the arc.

21 It has an impact on, you know, where it 22 will transition if it will move, or, also more 23 importantly, we need to collect data external to the 24 equipment to characterize what type of thermal source 25 120 is coming out from the event as well as our 1 photography and videography to capture where any of 2the molten material is going and the vapor, the 3 conductive cloud, cloud of death, whatever you want 4 to call it.

5 So if you look at the guide, the IEEE 6 guide, it tells you basically put it as far away from 7 your incoming power source as possible, so that's one 8 way.9I think a lot of the reasoning behind 10 that, and I don't have it, so, you know, the 11 background behind is that basically the magnetics 12 that was going to push the arc away from the source.

13 So we have various locations where we put 14 in Phase I, and this is actually a diagram out of the 15Phase I test report. In the back section of the 16 cabinets we could always put it somewhere on the bus 17 bar, so we could put it on the horizontal or the 18 vertical bus bars.

19Next slide. And this is was in a low 20 voltage piece of the gear, or we can put it on where 21 the stabs are on the front of the gear to initiate 22 the arc.23So those are kind of the locations. I 24 think there has been some OpE for where you have bus 25 121 cables coming in to the back of the gear, there has 1 been some failures there, and, obviously, you see 2 failures at the stabs where the gear goes into.

3 So, you know, those are two places we 4 want to focus in on, but also the bus bars in the 5 back there has been some OpE on failures there, 6 right?7MR. MELLY: Yes. Yes, and there is some 8challenges associated with both. So if we went with 9 the approach putting it into the arc location of the 10 front of the cabinet, which we did in a few cases in 11 the first test series essentially the breaker 12 misalignment, the stab misalignment going in and 13 things like that, we can initiate the arc in the 14 front of the compartment except the instrumentation 15 becomes so mewhat of a challenge in that orientation 16 because the door will swing open, sometimes throw the 17 door and our equipment stands, which would be located 18 directly in front it, will take the brunt of the 19 impact of potentially the door swinging open and 20 knocking them out, whereas in the back of the cabinet 21 or the sides of the cabinet we have less of a risk of 22 impinging on our instrumentation stands.

23 So I know Tony has been giving this a lot 24 of thought as to how to collect repeatable datasets 25 122 and this was a challenge for if we do the front.

1MR. PUTORTI: In Phase I if you take a 2 look at the experiments we used on the equipment that 3 was donated by Japan we were actually able to have 4 the same HEAF result four times.

5 We placed the arc wire on both of the 6 connections in the rear of a switch gear which was on 7 the secondaries of the breakers where the voltage 8 connection was between a bus bar and the cables that 9 would be exiting to the load and we were able to, the 10 HEAF exited right out the back in all four of those 11 that we did.

12 So that was an example of a scenario that 13 we favored because it would allow us to know exactly 14 and hopefully be able to repeat again where the HEAF 15products would exit the enclosure and we would be 16 able to put our instrumentation there to characterize 17 it well.18MR. MELLY: Yes. In the front 19 compartments there was a lot of, it was a little 20 chaotic, either the door blew open and the energy 21 came out the front or we would breach internally 22 compartment-to-compartment or the exterior wall and 23 it was somewhat of a guessing game as to the location 24 of the arc agenda.

25 123MR. FUNK: One quick comment, and I'll 1 put Mr. Earley on the spot on this a little bit, but 2I believe in the most recent issuance of IEEE 1584 3 they found a definite distinction between the energy 4 that was emitted form an arc created on a vertical 5 section versus a horizontal section and they have 6 changed some of their guidance based on that.

7 So that, again, I apologize I'm not as 8 familiar with that as I should be and maybe I'll, 9 again, put Mr. Earley on the spot to comment on that, 10 whether that should be considered for our testing.

11MR. EARLEY: Yes, that is true that the 12results from the vertical and horizontal test did 13 yield some different results.

14 MR. FUNK: Thank you.

15MR. TAYLOR: So any other feedback on 16 that or is -- I mean in the vertical orientation was 17 it more severe than the horizontal or -- I don't even 18 know.19MR. EARLEY: I believe that the vertical 20 was more severe.

21MR. TAYLOR: So that would be, you know, 22something we would need to consider. Is there any 23 other feedback on where we should initiate the arc?

24 You know, one additional comment was that 25 124 when we did do the finish piece of equipment we put 1 the --2 MR. MELLY: This wasn't there.

3 MR. TAYLOR: No, this has been --

4 MR. MELLY: No, that's our --

5MR. TAYLOR: You had the incoming power 6 supply from KEMA coming here, we put the arc down 7 here.8 MR. MELLY: Right in the center.

9 MR. TAYLOR: Oh, in the center?

10 MR. MELLY: Yes.

11MR. TAYLOR: Somewhere in here on the 12gear, so for that cubicle let's say, and all three 13phases tied together initiated the event and it 14 quickly promulgated up into the back part of the 15 gear.16MR. MELLY: So it promulgated from the 17 front of the switch gear towards where you had more 18aluminum material and the horizontal bus bars and 19 then you see the entire, from the video you saw the 20 ejector coming out the side following the magnetic 21 path.22 But, again, that was -- We initiated the 23 arc in the front cubicle and it promulgated up to the 24 top.25 125MR. FLEISCHER: Ken Fleischer from EPRI.

1 Yes, I like this more detail and thoughtful 2 discussion where you put the arc.

3 I can't remember which report, or test 4report I read where some of the arc initiation was 5 just not in accordance with any operating experience 6 we have, removing the 24 inches of arc chute to try 7 and get an arc to work was just in my mind not 8 realistic.

9 So what I recommend we do is not place an 10 arc somewhere where it is just not conceivable that 11 we can even initiate an arc there, if that makes 12 sense.13MR. MILLER: Keep the original 14 configuration of the equipment.

15MR. FLEISCHER: Right, right. Yes, I 16 understand we have to cut a little bit into an 17 existing gap that's in the insulation between maybe 18 a bolted connection to get a good wrap of the wire, 19 but not go into a part that's already got insulation.

20 It's kind of like I think the discussion 21 about the no n-safe bus, is it insulated bus or is 22not. You're not going to initiate an arc where the 23 insulation is covering the buses.

24MR. MELLY: Right. For our series of 25 126testing what we plan to do for the Phase II, we did 1 not alter the cabinets in any way besides removing 2 the initial insulation at the arc connection point.

3The only thing that we had that was a 4 little strange in the initiation of one of our tests 5 was in Tests 4, 5, 6 and 7 the Korean cabinet donated 6 equipment we had a lot of difficulty in maintaining 7 the arc itself, or initiating the arc where we had 8 extinguishment immediately a few milliseconds after 9 the test.10 So Test 7 we have coined the Frankenstein 11 test because you'll see bolts in there just as trying 12to have enough material to vaporize initially to 13 maintain the arc.

14 That was the first series of tests that 15we ran and it was a very trial and error. It did not 16 quite work and we have learned a lot more as we have 17 moved forward.

18MR. FLEISCHER: Actually, I'm glad you 19reminded me of that. And I am okay with 20 experimentation, so maybe one thing we can do is if 21 we decide to a test that is what I call experimental 22 or exploratory and it's not expected in real life 23let's make sure we just very clearly state that in 24 the final test report so that in the future the 25 127 utilities can say, hey, this was not representative, 1 this was an exploratory test.

2 MR. TAYLOR: I definitely agree.

3 MR. MELLY: Yes, we agree.

4MR. FLEISCHER: The other thing we can do 5 is, and if you say you're going to get to this later 6 we will, but I also, I'm having a hard time getting 7 my head around the size of the arcing wire using for 8 medium voltage.

9MR. TAYLOR: I was just talking to Ken 10 about that.

11 MR. MELLY: Yes.

12MR. TAYLOR: So if you look at guide, the 13 IEEE C-37.20.7, it has two different wire sizes in 14 there. I believe it's six and 24 --

15 MR. MELLY: Ten and --

16MALE PARTICIPANT: Twenty-four is the 17 medium voltage and ten for the low voltage.

18 MR. MELLY: Ten K strand.

19MR. TAYLOR: Okay, ten for the low 20 voltage and 24 for medium.

21MR. MELLY: And the ten is K strand wire.

22MR. TAYLOR: And the ten is K strand wire 23 meaning that it has a lot more small strands to make 24 up that ten gauge conductor.

25 128 So, you know, moving forward we want to 1 understand, all right, the arcing wire is only there 2to create the conductivity in the plasma and really 3 all we need is enough material to get the arc to 4 initiate and then maintain it to allow the event to 5 occur for the duration.

6 So Kenn Miller actually reached out to 7 IEEE to understand what's the basis, what's the 8 reason behind the different sizes, and we haven't 9 gotten a -- They forward it on to their committee.

10 Maybe some of the individuals here on 11 that can try to get some basis, but if there is 12anybody in the room that has some information, you 13 know, we're interested to learn more on that.

14MALE PARTICIPANT: Do you have any 15 background on that?

16MR. FLEISCHER: No. I can go back to the 17 standards and see if I recognize anybody, but what I 18 offer up is that I am not saying take exception but 19 it would say, hey, we're going to initiate an arc for 20 IEEE C37.20.7 and then we use a different wire size, 21 I would just say if we're going to use a different 22 wire size and a different wire type let's just 23 outright say that it's going to be a deviation from 24 the standard or not.

25 129Yes, I'll take a look, Kenn. I will go 1 look into the standard, who's on the standard board, 2 and see if I recognize anybody from my past non-3nuclear IEEE life. There might be someone on there 4 I know.5MR. MELLY: Yes. And when we did look 6 into it we noticed that there was a specific callout 7 as to why they used the ten gauge K stranded wire to 8initiate the arc in low voltage because they had a 9 greater success of maintaining the arc for the 10 duration, the desired duration, and there was no 11 explanation as the reduction in the wire size for the 12 medium voltage.

13 I assume it's just because it's much 14 easier at holding the arc in, but because we did see 15 that explanation as for the holding for the desired 16duration our initial thought was we went with that 17 for all the tests because that was one of our main 18 parameters of interest.

19MR. TAYLOR: And that's in the 2017 20edition of that standard. They put that in in an 21 annex.22 MR. MELLY: Yes.

23MR. TAYLOR: So I think m oving forward, 24 and Nick and I have talked about this a little bit, 25 130 is that we plan on using what the standard, or guide, 1 I guess, says to use for either medium or low voltage 2 but we still want to hear back from that standard 3 committee to see what the basis or logic is.

4MR. FLEISCHER: Yes, and I'll try and 5 look see if I can find something.

6 MR. MELLY: Yes.

7 MR. TAYLOR: We appreciate that.

8MR. FLEISCHER: I'll get with you by 9 either way.

10MR. MILLER: Yes, I've sent an email to 11 the chair working for them but I haven't heard back.

12MR. TAYLOR: Okay, so go ahead and the 13next slide, Nick. Another picture I have taken out 14of the press report, again, trying to get the same 15 thing, where do put the arcing wire.

16 This is just a medium voltage piece of 17 the gear, so you have the bus cables coming in the 18back making connections to the bus bars within the 19 enclosure in the middle section and then you'll have, 20 this is a medium voltage, you'll have two sets of 21 breakers in this case up front.

22 So we could potentially put the arcing 23 wire right at the bus connection, the bus cable 24connections to the gear, we could make shorting in 25 131 the bus section of the gear in the middle region, or 1 we could make the arc connection in where the stabs 2 are, where the breakers are located.

3So those are possibilities. We are 4 looking for feedback and --

5MR. MELLY: And this is the test that 6Tony was talking about. We had very good success of 7getting similar arc conditions when we initiated it 8 in the rear of the compartment than we did when we 9 tested it in the front cubicle.

10MR. TAYLOR: And I know I am probably 11 stealing a little bit of Nick's thunder here, but 12 moving forward we do plan on including breakers in 13 the enclosures.

14 They may not be functioning and we might 15 not have them racked in but they are at least there 16 for their mass and thermal characteristics --

17 (Simultaneous speaking) 18MR. MELLY: Yes, that was one of the 19 resounding comments we received from internationally, 20 NEI, EPRI, is that the picture that was depicted in 21 the test plan did not have an actual breaker unit in 22 it.23 We have changed it so it does and we do 24 plan on putting non-functional breakers in so we have 25 132 the thermal mass, the fire load potential, as well as 1 the pressure influence.

2MR. TAYLOR: And breakers are the most 3 expensive piece, so the non-functional piece is 4important. If we can get something that is the right 5 size to put in there we will.

6 Obviously if we can get a one-to-one 7 match that's what we'll use, but we just have to 8 watch on the budget side because if we're going to 9 have to spend, I don't know, X number of dollars on 10 the gear and then we got to spend ten or 20 times 11that just to get breakers, you know, it's going to 12 limit the amount of test that we can do.

13MR. MELLY: Which is another reason why 14 we planned on not testing at the breaker unit itself 15 so we don't have to buy repetitive pieces of breaker 16 equipment.

17MR. PUTORTI: Tony Putorti from NIST.

18 And in Phase I these enclosures and cabinets did have 19 breakers.20MR. TAYLOR: Okay, next slide. So this 21 is just the same piece of gear, just the schematic of 22 it, so, again, in this case that's where they put the 23 arcing wire.

24 Is there any feedback from the group of 25 133 proposed or suggested locations that initiate the 1 arc?2 (No response) 3MR. TAYLOR: Okay. On the webinar, Tom, 4 do we have anything?

5 (No response) 6MR. TAYLOR: All right. Again, if you go 7 back home and something comes in your mind please 8 email us on that, otherwise I think you know those 9 are the three locations that we were interested in.

10 OpE shows that events are occurring there 11 and we'll probably pick, we probably won't just do 12one, but, you know, maybe two and possibly three, 13but, again, we want to weigh, you know, how 14 repetitive we can get the, I'll call it effluent, the 15energy coming out for our measurement gauges. Next 16 slide, Nick.

17MR. MELLY: One more point there is that 18 some of the parallel work that is being done by Japan 19 as well as their testing agency CRIEPI, which is the 20 parallel of EPRI in the U.S., they are doing a lot of 21 testing as well to look at the energy levels to limit 22 the amount of energy and they are doing predominantly 23 most of their testing in the breaker cubicle itself 24 or very close to the breaker cubicle because they are 25 134more interested in the threshold amount of energy 1that is initiating an ensuing fire which they are 2 using as one of their damage states.

3 So by initiating at the breaker unit 4 itself you have a higher chance of starting the fire 5 in that breaker unit which is why they are doing the 6 testing there.

7 So we had been working with the 8international community and we are going to try and 9 leverage some of their testing experience through the 10 use of IAs and other avenues so there is data that we 11 might be able to obtain from that arc location given 12 their test series.

13MR. TAYLOR: All right, next thing is bus 14 bar spacing and I show you that it is important for 15 arc voltage, or it's one of the parameters that 16 influences arc voltage.

17 The standards don't specify, at least I 18 am not aware of any, a nd typically spacing that I 19 have found is IEEE has some tables in 1584 that gives 20 you either a range or a single point value for 21 different types of equipment for bus bars.

22 So from that, you know, six inches, four 23 inches, or one inch about for the different types of 24classes of gear. And then I also found a White Paper 25 135from one of the, one company that had some 1information as well. It was similar to what was in 2 1584 but it was, you know, typical bus bar spacing.

3 So I think it's really going to depend on 4 what type of gear that we choose will depict on what 5 the spacing is, but if there is any opinions in the 6 room of, you know, this is the type of spacing we 7 have in the plans or we bought this gear because of 8 this rating we'd be interested in knowing that now or 9 moving forward such that we can procure our gear that 10 was representative.

11MR. MELLY: The other important aspect of 12 the spacing itself is that the way that we have set 13 up the test plan that we'd want to compare the copper 14results versus the aluminum result, the spacing can 15be different based on the sizing that you need, 16 copper versus aluminum, we don't want to do a this 17 exact same size copper versus the exact same size 18 aluminum because they are going to be differing in 19 the field.

20 We want to potentially buy equipment that 21 is spec'd out for either material and there will be 22 slight differences in the spacing, depending on what 23 material you are using.

24MR. LEJA: So I understand, you're going 25 136 to be limited to the spacing based on the kind of 1gear you're able to get, but I would probably just 2 note it in the report and maybe reference that table 3and just say based on, you know, this is standard, 4 and here's what we tested, just so people have a 5 comparison and could maybe have done som ething that 6 might have skewed the results a little bit.

7MR. TAYLOR: Okay. I think that's very 8reasonable. Okay. Not hearing anything, nothing 9from the webinar, we'll go to the next slide then.

10All right. Bus insulation. We can test bus bars 11with insulation or without. I think there's been 12 some recent event where they had insulated buses in 13their bus duct. But not only bus duct, even in the 14 gear, do you guys experience insulation on the buses 15 or is it typically non-insulated or a bare bus or bus 16bars? So that's the kind of feedback that we're 17 looking for because, again, we have prepared 18 equipment, we could go either way or we can get a 19 cross-section of insulated or non-insulated.

20MR. FLEISCHER: One plan I'm intimately 21 familiar with, the non-seg bus, it is insulated.

22It's that Norrell type insulation. But the fault 23 that they had was another one of these maintenance 24errors. Apparently, they had done an inspection, and 25 137 where the bus goes horizontal and then goes to a 90-1 degree with bolted connections, they forgot to put 2the boot back on. So that was exposed, and that was 3where the arc occurred. So, therefore, my one 4 experience with an arc on that would be an exposed 5 area of an insulated bus.

6MR. TAYLOR: Okay. I think if we just 7 add on, I think if we did test insulated buses, we 8 could either test it where a connection is where 9there's an insulation and there's a boot. You might 10have to get the arc to stabilize. But also I think 11there was recent testing done where they removed 12 insulation within the bus bar and caused a three 13 phase there, so that's another approach we could 14take. And with the insulation removed, I think it 15helps stabilize the arc so it doesn't, you know, 16shoot down the bus bar. And that's one of the 17 reasons why Nick put in the test plan for the 18 uninsulated cases we have a cap so that the arc 19didn't just shoot down to the end. I think there are 20 some cases where the arcs are stabilized and haven't 21 moved that far.

22 But, again, from a testing perspective 23 and to setup the instruments, we kind of have to make 24 some modifications to make sure we can get the arc 25 138 where we want it.

1MR. MELLY: Right. Because if you 2 remember the video that we showed of the bus duct 3 that we did test, you saw this just plume just 4shooting out looking like a rocket engine. And if 5we're trying to refine the current model for bus 6 ducts, it's this cone in a downward direction and we 7 see in the modeling, as well as the OpE, is from 8 these events where the bus ducts hung overhead.

9 That's why the gap break was there because if we 10 didn't put that we were worried it was just going to, 11 the arc was going to migrate and move all the damage 12 downwards.

13We have learned a lot from the Cooper 14testing that was done. And they were able to 15 stabilize the arc and center the cubicle or center 16 the enclosure and get much more damage consistent 17 with what you would see in the FAQ method, so we are 18 trying to move forward in that respect.

19 MR. DALEY: Ours were variable.

20MR. TAYLOR: That one comment was 21 utilities would have bare conductors.

22MR. MELLY: Does it matter whether it's 23an external or internal? Like for the Cooper tests, 24 the Cooper bus duct was outside versus a lot of the 25 139 bus ducts that we're worried about postulating for 60 1 to 50 are internal.

2MR. TAYLOR: Okay. Not hearing any 3feedback, it might be reasonable, I'll put this on 4 the table, to test some bus bar, bus bar 5 configurations with insulation and another segment 6without and then have a comparison. So does that 7seem reasonable to the group? I'm seeing some yes, 8some yes nods in the yes direction. All right. So 9 that might be something we want to change then, Nick.

10 MR. MELLY: Okay.

11MR. TAYLOR: While I'm writing this, what 12about, for test bus ducts, what about gear? Is there 13 configurations out there in the plants that have a 14 lot of insulations on the bus section of the gear, or 15 are they mostly uninsulated?

16MR. MELLY: If you go back to this 17 picture of the Japanese equipment, this was insulated 18bus bar material internal to the cabin. When we did 19 initiate the arc, we removed, roughly, an inch of 20 insulation around and connected the arc and wire at 21 that location. Tony.

22MR. PUTORTI: So there were a couple of 23experiments on these enclosures where we did remove 24the insulation when it was in, let's say, a middle 25 140 compartment.

1 MR. MELLY: Yes.

2MR. PUTORTI: When we did the four on the 3 rear, the bolted connection was uninsulated, so we 4 put the arc wire on the uninsulated portion.

5MR. TAYLOR: Okay. I'm just looking at 6 time and I think we got enough time to get through 7this. Was there any other feedback on insulation 8characteristics? All right. Not hearing any, if you 9 do find that the gear is insulated or commonly 10insulated, please let us know. I think we have a 11 path forward for the bus ducts.

12Okay. The next couple of slides are 13looking into the pressure influences. So where the 14 thermal is more of the energy, from the pressure it's 15more the arc power.

Some of the literature we've 16 looked into emphasize the DC time constant has a big 17 role in how the pressure effects from the event occur 18or propagate. Asymmetric current to a point has some 19 impact, but also looking at the volume of your 20 enclosures and your openings, given a door opened or 21 just the ventilation, also have an impact on pressure 22because you need some type of containment to get 23 pressure obviously.

24 So, you know, I think, Dan, you had a lot 25 141 of feedback on the asymmetric nature of these time 1 currents or these current profiles and looking at the 2DC time constant and the asymmetric nature of these 3profiles. Is there any influence or interest in 4 exploring those aspects of these events from a 5 pressure standpoint?

6MR. FUNK: I guess we could. You have 7 characterized it well so we at least know exactly 8what we've got. To be honest, the testing is pretty 9 benign in this regard to the offsets and the time 10 constants are really low, and you don't include any 11feedback from motor contributions, etcetera, 12etcetera. So I guess that's my feeling on that 13 particular part of this.

14MR. TAYLOR: Okay. Anybody else have a 15-- 16MR. FLEISCHER: Just help me understand, 17 DC time constant, are you saying DC offset which is 18 part of the asymmetrical current? Okay.

19MR. TAYLOR: Yes, I think it's the same.

20 No, I think I'm referring to the CIGRE report, 602, 21 and they identified that the time constants, how fast 22 you decay.

23MR. FLEISCHER: Right. So it's a DC 24offset. It's the standard asymmetric half cycle to 25 142 the first --

1 MR. TAYLOR: The decay --

2MR. FLEISCHER: Yes, okay, all right.

3 I'm used to it referred to as DC offset, but CIGRE 4may call it different, but it's the same thing.

5 That's all I needed to know. Thanks.

6MR. TAYLOR: Is there anything on the 7webinar or in the room on that? It seems like it's 8not a high priority. All right. Just quickly on the 9 equipment, a few things I want to go over on the next 10slide with Germany is planning. They're still 11working with their utilities. Korea already has 12lineup of what they want to provide us. All their 13 equipment will be procured or donated, so if there's 14 any plants out there that are upgrading or modifying 15 their plant and then it gets representative 16 equipment, you know, feel free to contact Nick or I.

17 We can see if we can help arrange for that donation 18 of equipment and get it into our storage facility up 19 in Pennsylvania before that.

20MR. FLEISCHER: Kenneth Fleischer from 21EPRI again. If it's going to be really impactful to 22 do that based on the testing we talked about, we want 23 to make sure that all the donated equipment, we get 24 the vendor manuals, as Kenn Miller and I talked 25 143about, to get the vendor manuals, the factory test 1 reports, everything that's related to that equipment 2 so that we have something to at least review to make 3 sure that what we got is what we got.

4MR. MELLY: Yes, we're going to reach out 5 and get as much information as possible about the 6donated equipment. There have been challenges in the 7 past, either language barrier or vintage of procuring 8that information from country to country. We're 9 going to, I think with this round of equipment that's 10 going to be donated, it's going to be much easier to 11 obtain that.

12MR. MILLER: This is Kenn Miller. I 13 know, too, when we were getting stuff from the Zion 14 plants, you know, that plant was already down and 15 being demolished and we couldn't get much 16 information.

17MR. TAYLOR: Yes, a lot of the 18 documentation has been removed for, you know, non-19safety or non-required to maintain. I want to put 20 that out there because that can save us a lot of 21 expenses and, if you save us expenses on that aspect, 22 that means we can potentially do more tests or more 23analysis or more in other areas. So I want to put 24 that out there, but I think, you know, Kenn's comment 25 144 on being, you know, getting equipment that's very 1 similar so we have some repetition is also something 2we have to weigh. And, you know, if we go out and 3 procure equipment, I think that's what we tried to do 4is try to pick out we want this manufacture, this 5 model number, we want X number of those types of 6equipment. We'll probably be going to basically 7 salvages equipment or refurbished, as that's what 8 we've done in the past.

9MR. MELLY: There are potentials to have 10 challenges with proprietary information, as well, in 11 publishing those vendor manuals and things like that.

12MR. FLEISCHER: Yes, I guess my point is 13 it would be a shame on us not to ask at least.

14MR. TAYLOR: All right. Next slide. So 15 real quick, without getting into too much detail, 16Korea is on the left. I guess they're 5 kV rated 17breakers. We're not sure of the gears rating, but 18we're assuming it's also rated to 5 kV. As for the 19vendor manuals, we got a generic manual from them 20 that's not specific to this set of what they have 21 procured here, so I think Nick has asked them a 22 couple of times to get more detailed information on 23 that. But we know we have -- is it two or four?

24 MR. MELLY: Four.

25 145MR. TAYLOR: Four sections of switch 1gear, so four cabinets and four breakers. These are 2 the vertically racked-in medium-voltage breakers from 3 Korea.4 And then on Germany I'm not, we met with 5them a few weeks ago. They just switched plans with 6 what utility was donating, so they didn't have all 7the information available yet. But right now I think 8 they're donating four low-voltage switch gear 9 sections and four medium voltage.

10MR. MELLY: Yes, with the knowledge that 11this plant is shut down. The difficulty in getting 12 equipment out is that the entire plant is dark, and 13they said there's rats everywhere. So they're trying 14 to get a company to get out the equipment now.

15MR. TAYLOR: Nick, do you know if these 16 are air breakers, docking breakers --

17 MR. MELLY: I do not know.

18MR. TAYLOR: Okay. So that's the only 19information we have to collect, but in these two 20 countries they could not donate money, so this was 21the alternative to include their participation. And 22 both have worked with us in past projects, so it was 23 valuable to continue that collaboration.

24All right. So this is something to think 25 146 about, but, if we go procure equipment, what is 1 representative, what's realistic, and what should we 2go out and -- and this isn't a complete list. This 3is just things that I've run into in the past. You 4know, you have the Gs, Allis Chalmers, the 5 Westinghouse IT ADDs, other manufacturers obviously.

6 But, you know, if I wanted, today, to go out and 7 procure 30 pieces of gear or 30 sections of gear to 8 do these tests or however many we need, where should 9I focus in on? What's representative of the plants?

10 So do we have comments on that right now, 11 or we can wait until after lunch to give you guys 12time to think about it. I think it's very important 13 for us, when we go for a procurement, to get the 14 right equipment.

15MR. MILLER: This is Kenn Miller. I'm 16 familiar with the GE Magna-Blast and Westinghouse VS 17 switch gear.

18MR. TAYLOR: All right. And then on low 19 voltage, my list was, there's going to be a lot more.

20 But the common ones I've run into is, again, the 21Westinghouse, GE's. There's going to be other ones.

22 Next slide.

23MR. FLEISCHER: The previous one, the GE 24 Magna Blast and Westinghouse were the most common for 25 147 the fleet that I represented.

1 MR. TAYLOR: Magna or Megne?

2MR. FLEISCHER: Well, actually, GE Magna 3 Blast.4 MR. TAYLOR: Magna?

5 MR. FLEISCHER: Yes.

6MR. TAYLOR: All right. Next slide then, 7please. All right. And then the other thing that's 8 important, you know, we talked about it a little bit 9 today is you have the arc, you have your source term 10 from the arc, but then to damage things outside the 11cabinet you need to get through the cabinet. And 12 looking into enclosure thicknesses for the burn-13through times, you know, what's typical? And I 14 think, from the information we had yesterday, it 15 seems like they're changing to make the minimum 16 requirements to meet certain standards.

17 The only thing I could really find out 18 there on material thickn ess was from the IEEE 19 standards, and it's the C37.20.1 and also C37.20.2, 20 so 20.1 is low voltage and 20.2 is medium voltage, 21and they have the same information on the minimum 22 steel thickness of the enclosures being 1.9 23 millimeters thick for the enclosure and then 24 partitions between the primary circuits would be 3 25 148millimeters. So that's the minimum that's specified 1in that standard. I don't know what's out there in 2 the plants.

3 So feedback on that would be, I think, 4 important because one of the reasons why is the 5 finish test that we did, we didn't have a complete 6lineup of gear and all we had was a section. And 7 where the arc occurs, the partition seemed to be very 8thin. I don't know if you have a measurement of 9that. It wasn't representative of what we see in a 10 lot of the other equipment.

11MR. MELLY: We did not have a measurement 12 of the thickness of the partition at that point, and 13 it was a thinner material, as it was going to be 14buttoned up to another section. Additionally, 15 something that gave us another indication was that we 16 did use a thicker steel material to button up where 17 the incoming power supply would have gone, and we've 18also vaporized that. But it was a question mark in 19 our minds as to what was the importance of the 20 various thicknesses for the external cabinet walls, 21 given that we saw an immediate breach and larger 22damage state from this equipment. So it's something 23 that we're going to be focusing in on, and we want to 24 really get feedback on what's out there because we 25 149 know it could be across the board.

1MR. TAYLOR: We're getting close to lunch 2here so not much discussion. So another thing, you 3 know, maybe after lunch we can come back and hit on 4it. But we'd like to at least, you know, be 5 representative in this circumstance because it does 6 have an impact on the energy that can get out of the 7 enclosure.

8 And ventilation, again, we might be bound 9 by what the manufacturer used for the equipment that 10 we procure, but it does have an impact on pressure 11and we want to be representative. So any feedback on 12 ventilation, we'd be interested to learn more on that 13 if there are a lot of variations or if we should 14 just, you know, with the equipment we select, use 15what's provided. Obviously, I don't think we should 16be out there modifying the enclosures to meet a 17 certain ventilation but we want to have some feedback 18 on that.19 And the last thing before we break is on 20bus duct. We've just spent a lot of time on the 21 enclosures, and I think that's where a lot of our 22testing is, but, from the informal survey that went 23 out, it seems like there's a lot of bus ducts with 24aluminum on them. So I already talked about the 25 150configurations that we planned on doing, the bus 1 versus enclosure configurations, so we're looking for 2 feedback on that, as well as what is your bus bar 3 configurations, square bus bars like we tested in the 4 previous program, testimony six, rectangular, 5circular, or what's your bus bar configurations? I 6 don't think we're getting into ISO phase buses, so 7 we'd be looking more on the medium-voltage bus bars 8 and not the ISO phase going down to your main 9transformer. And then the rating or the size of 10 those buses.

11So I think now we're almost close to 12breaking here. Those are the things maybe we go pig 13 out during lunch, come back and have this discussion, 14 unless anybody wants to bring something up right now.

15 So we'll have the gear we should be procuring, 16 enclosure thicknesses, bus duct configurations, and 17 come back after lunch and have a discussion on that.

18 And then we'll get into Nick's test plan response.

19MR. SALLEY: Okay. So before we break 20for lunch, are we losing anybody this afternoon?

21Bob, Shannon, Greg. Well, thank you guys for 22attending. I really appreciate it. We'll soldier on 23without you after lunch, but we appreciate you being 24 here. 25 151 Before we go, though, Frank and I were 1talking, there was a number of questions about the 2facility generator. Dan, you had a lot of questions 3about it. And one of the suggestions and actually an 4 offer that Frank and Bas are making here, if we 5 wanted to do a day trip or an afternoon to KEMA to 6 see the facility, if you think that would help, we 7could potentially set something up. Ashley, I work 8 through EPRI with you and Victoria, with NEI and 9Mark, if any of your folks think it would be 10worthwhile seeing it, if you could briefly get back 11to me. Frank would be gracious enough to set a field 12 trip up and we could go and you could see the 13facility, see what they're doing. If you think that 14 would be of value, how about getting back to me and 15 letting me know from there?

16MR. CIELO: We'd answer any questions 17 that you have, show you anything, you know, have our 18 facility people there and our test people available.

19 Yes, I mean, it's not that we don't want to provide 20the information. We can provide the information, but 21 we want you to be able to put it in a context of the 22 lab and how the lab actually works and how we 23actually function. So we would, you know, we would 24 absolutely like to, if you guys -- NRC is going to 25 152spend all this money, taxpayer money, and you're 1 going to invest all this time and this effort and 2 it's that important, I mean, I think you'd really 3 want, I would think you'd want to see the lab and 4 you'd want to see what the lab is capable of doing 5 and how it functions and to meet some of the people 6that are actually doing the testing. So that's an 7 open invitation, and I'll leave it up to Mark to let 8 us know when that is and we'll work it out.

9MR. SALLEY: All right. Mark, for the 10 NFPA folks, if it's of value on that, and we'll talk 11to Victoria for NEI and see if there's value. We do 12this trip a lot. It's a day trip. It takes what?

13Three and a half hours from here? It's right up by 14 our Region I office outside of King of Prussia and 15 can do it in an afternoon. So we'll put that out.

16So with that, it's almost 12:00. Let's 17 break for lunch and be back here by 1:15 downstairs 18 and let's pick it back up at 1:30 in the afternoon.

19 (Whereupon, the above-entitled matter 20 went off the record at 11:55 a.m. and resumed at 1:28 21 p.m.)22MR. MELLY: Okay. Starting up, we wanted 23 to discuss, as we're going to get started again --

24MR. SALLEY:

Before we get started, the 25 153 discussion we were having was pretty good, so we're 1 not going to try to stick too much to the agenda now.

2I think it's more important to get the information 3exchange. We want to get that, Gabe, we got a couple 4 of tables up here that, hopefully, when we rank some 5 of these and bin some of these, we'll be able to take 6 that away and we'll include that with this meeting in 7 some of the information as we move forward.

8Also, in the same idea of that, as we 9 look at it, we had a thought over lunch that, you 10 know, this is similar to what the week-long PIRT 11 exercise did, and we have that result and we got that 12report published. So when we're done with that, Nick 13 is going to pull out the PIRT and, just for 14 curiosity's sake, this is what the internationals 15 have been working on, we'd like to see how this group 16 ranks the same things as, you know, what's important, 17 what's medium importance, low importance, and just 18 see if we're getting alignment between what we've 19 done a year or so ago with the PIRT versus what we're 20doing today. So, again, we'd like to do that this 21 afternoon.

22 So with that, Gabe, I'll turn it over to 23 you and pick it back up.

24MR. TAYLOR: From this morning's 25 154 discussion, I think it's important for us to go and 1 kind of rank some of the parameters so we know what 2 to focus in on or what's most important for the arc, 3arcing events, and also get into this because I 4 understand, if you haven't had a chance to look at 5the PIRT report on this, to understand what they 6thought was important. That might help us with our 7 ranking here.

8 Before we do that, Kenn Fleischer brought 9up some information. He looked at 28 plants over 10 lunch to get information for their transformers from 11the unit aux. And from what he found, it's not a 12 complete of all the plants, but it's a sample that 13 kind of gives you a representative, hopefully a 14 representative snapshot that a delta to Y, there's 15about 20 plants out of the 28 that has that 16configuration. Six of the plants had delta-delta, 17 two of the plants Y-Y, and he didn't see any Y-deltas 18 there, but that doesn't mean they don't exist, it 19 just might not have been included in the sample.

20 So, again, just for the background and 21 the follow on from the discussion this morning, so 22what that might tell us is that we maybe we need to 23 spend more time looking at the Y configuration since 24there seems to be more of a population in that 25 155configuration versus the delta. So when we look at 1 the split fraction of events, if we can modify the 2 test plant to be representative of that, that might 3help us out. So I appreciate that information and 4 we'll take that back as we go and change the test 5 plant.6Okay. So with that, I think, Nick, can 7 you go over quickly the PIRT results and findings 8 from that effort?

9MR. MELLY: Right. And what you're 10seeing here is the summary table. It's in Chapter 3 11 of the PIRT, essentially trying to look at all the 12 phenomenon on the three scenarios that were ranked 13 during the study as to what were the most important 14characteristics. They have put together level 1 15 phenomena as the highest importance, level 2 and 16 level 3, as you go down, decreasing in importance.

17 And the report goes into specifically 18what all of these mean. I can quickly go over kind 19 of what's included in the terminology there.

20 The target characterization essentially 21 was what is the fragility of the targets that is 22 being exposed to the HEAF effects, such as is it 23 thermoset, is it thermoplastic, what's the cable, 24what is the damage state, what's seen as a very high 25 156 parameter to determine what the zone of influence of 1 damage would be?

2 Moving down, we get to the arc 3 characterization that included things such as the 4current voltage of the arc itself, the arc 5 mitigation, this would be your protective systems and 6 devices as to what that can limit the duration of the 7 event, where am I within the lineup of my specific 8 plant, and how important is that to determining what 9 the zone of influence will be?

10 Also, cabinet lineup effects, this was 11looking at cabinet damage for damage or, sorry, 12 whether I'm going to damage an adjacent cabinet from 13the zone of influence. And you can see it's missing 14 from the first two scenarios and only included in a 15 third because the first two scenarios did not have a 16lineup as part of the scenario. If you were to look 17 in each scenario's detail, we had a lone cabinet to 18 a cable tray to a cable tray or a lone cabinet to a 19 both trains of equipment cabinet lineup and it varied 20 from case to case.

21Moving down from that, we looked at the 22 internal ensuing fire, such as what combustible 23 materials would be in the cabinet as to how large of 24 a fire we're going to have in that cabinet; external, 25 157what's outside of the cabinet, cable tray again, 1material came into play there; pressure effects, am 2I going to damage anything in the room from the 3 pressure effects, and that was largely room 4 dependent; the electrical configuration, a lot of 5 factors came into that and they're explained in the 6 report so I'm not going to dive deep into that one 7 because it gets a little in the weeds.

8Then when we get a little bit lower in 9 importance, you see suppression effects, room 10configuration, and fire detection. Those are 11 primarily listed as low because we didn't see them as 12 an important factor for determining what the zone of 13influence would be. So for example, fire detection, 14 these events happen so quickly that the detection is 15 not going to play a role in determining if I have 16 damage because, clearly, these events create so much 17smoke and they're happening so quickly that the 18 detection is not going to limit the damage that we 19 see from the initial blast.

20Room configuration, as well. While we 21 did see that as important, we tried to be in the 22 mindset that the zone of influence can be used in any 23 room configuration if you figure out, it's going to 24 be scenario specific, which is why it's down low 25 158 again.1 Suppression effects came into play 2 because there hasn't been much work on is my 3 sprinkler system going to be reliable given that I've 4 had one of these events, am I going to damage the 5 capability of my system with an event this scale?

6 So this was the perspective of the 7 members as to how important things were. It goes 8 into great detail in the appendices on each 9 participant's ranking, and it goes into the 10scenarios. So to scenario one, you see Train A, 11 Train B, cable tray, and this is kind of how we 12approach the problem. And this is really in parallel 13 of what we were discussing earlier of binning the 14 importance of the characteristics, and I feel like 15 this was the stab to do just that based on the result 16 of phase one.

17 The report also goes into detail of these 18 higher importance events in the summary conclusion.

19 But it's definitely something that we would like 20 feedback on if it hasn't been something folks have 21looked at. It's a pretty all-encompassing report and 22 a lot of work went into it.

23 Does anyone, it's difficult to have the 24 discussion without understanding exactly what goes 25 159 into each of these categories, but does anyone see 1 something that doesn't align with what they're 2 thinking would be in terms of importance for 3 determining a zone of influence?

4MR. SALLEY: Before you go there, Nick, 5 one other thing before you go into that is, in the 6PIRT exercise, this is to help us drive research. So 7 the questions that are being asked when they're going 8through these scenarios is do we know how to do 9something? If we know how to do something, we've got 10a lot of experimental data. Then that shakes out 11because we can do it. A question will come up with 12 phenomena, hey, this is the phenomena, now can we run 13 the experiment, is it something that's reasonable to 14do that we can get information? So where does that 15lie in our rankings? This is where the 16 identification and ranking table piece comes in, and 17 so it goes back to this work that we're talking here 18with the experiments. What experiments can we do to 19 get data to fill holes so that we can better do the 20risk analysis on? So that's kind of the underlying 21 premise that the group was going in.

22 Now, if you wanted something that was 23 just pie in the sky $10 million to do it, you'd 24 obviously washed it off the table because it was not 25 160achievable. So the goal was to get to achievable 1 research to help better the problem, and that's the 2 underlying premise that PIRT starts out.

3MR. MELLY: Yes, and you can see that the 4 questions also ask here that Mark was getting to is 5 that's the state of the knowledge of the parameter 6itself. So is there -- model adequacy was is there 7models that can handle this currently? Data 8 availability is does the data exist to apply into a 9model? And the ability to collect the new data, so 10 how easy it is to collect data on this particular 11parameter. It would be easy to collect the data, it 12 would just be extremely expensive to collect the 13 data, so that was also taken into account as the 14 ability.15 But this might actually be a good place to show the 16 breakdown that was included in each of these 17 phenomena.

18 So this was probably the better place to 19go. For the electrical configuration, we're talking 20power supply, electrical tension coordination, 21 internal cabinet configuration with things like the 22 breaker configuration, cabinet compartmentalization, 23 cabinet combustible loading, bus bar configuration.

24 This gets to the horizontal/vertical discussion that 25 161we were having as affecting the energy release.

1External cabinet. Again, ventilation, structural 2design, cabinet penetration. Arc characterization, 3 arc migration, breaching the enclosure, electrical 4 characterization, thermal effects of the arc, 5 magnetic effects of the arc, electromagnetic 6interference. This one, as you can see here, low, 7 medium, and these are the rankings of the individual 8 members and they're compiled in that summary table 9 you saw above.

10A lot of them also have notes as to why 11 some folks differed from others in their rankings.

12Not all of them, though. Fire protection, presence, 13characterization of the system. Fire suppression 14presence, the fire suppression effects. Pressure 15 effects, projectile, missile damage, pressure wave, 16 internal ensuing fire, fire ignition, 17 characterization of the fire source, fire 18development. This gets into the T=0 that's currently 19frequently used. Smoke generation, altered 20ventilation effects, the ensuing fire. Very similar, 21 smoke generation and things like that.

22 Then we get into the room configuration, 23 that was room integrity, the room arrangeme nt, room 24 ventilation. Then we get back up to the summary.

25 162 But when we do our discussion here, it's 1going to be interesting. And we can go back to this 2 PIRT if we think that the current in the voltage, as 3 well as the protective systems, are the highest 4 important parameters, we can compare with what the 5 PIRT has ranked those on the individual parameters 6 throughout this report.

7 And, again, this was a week-long 8 phenomena ranking elicitation exercise with, I 9 believe, seven countries who participated.

10MR. PLETZ: The conductive cloud, is that 11 included under the pressure effects?

12MR. MELLY: That will be included, I 13 believe it is in the arc migration. We discussed 14 that in the arc ejecta.

15MR. SALLEY: Nick, why don't you walk 16 through the parameters to see, generally, how the 17 people ranked them, if they're in similar alignment 18 with what . . .

19MR. MELLY: It's a little difficult to do 20 that with the three scenarios differing.

21MR. SHUDAK: Can I get a clarification 22from this morning? The medium voltage testing, are 23 you still planning on doing it at the 4 kV and 6.9 24 kFV, or did you say you're only going to be doing it 25 163 at the 6.9?

1MR. TAYLOR: So the question was, for me, 2 voltage testing, are we going to be testing it 4.160 3 or 6.9 kV?

4 MR. SHUDAK: Or both.

5MR. TAYLOR: Or both. Right now, the 6 majority of the testing will be 6.9, except for that 7equipment that was donated. From the specifications 8 that they sent us, it looked like it's only rated for 9 5 kV, so we'll test that at 4.160 if we receive it.

10 MR. SHUDAK: Okay.

11MR. TAYLOR: But right now, the plan 12wasn't to test anything else at 4.160. And if you 13 look at the arc voltage, the presentation I provided, 14 it seemed like the system voltage didn't have much of 15an effect on the arc voltage. That's one of the 16reasons why I put that together. So I don't see, we 17don't have any real preference, but, right now, the 18 plan was to go 6.9.

19MR. SHUDAK: If you're looking at other 20 donated equipment, if it comes from, you might get 21 other donated equipment that would be 4 kV equipment.

22 I'm assuming you'll test it at that?

23MR. TAYLOR: Yes, correct. So the 24 question was if we get other equipment donated that's 25 164 not rated for the higher voltages, what are we 1testing that? We're not going to test it beyond what 2 it's rated for, so if we get a 4 kV or 5 kV piece of 3 gear, we're not going to test it at 6.9 or do 4 anything higher. We'll test it at 4.160.

5MR. MELLY: And to answer the previous 6 question, the gas and the cloud conduct with 7 particulate is included in the arc characterization 8 portion of what you saw in the above table, and that 9 was one of the high parameters ranked for importance 10of determining the zone of influence. So it's one of 11 the relatively high ones.

12MR. FLEISCHER: Kenneth Fleischer from 13EPRI. The one thing I don't see in there that I 14 noticed in the background on page eight of the phase 15 two test plan is it said that the PIRT panel covered 16 three distinct HEAFs, the first was an HEAF occurring 17 in an electrical enclosure with a cable tray passing 18 over the enclosure, but I don't see anything where 19 the test plant gets into considering the enclosure as 20 being a target in the zone of influence.

21MR. MELLY: The enclosure. Do you want 22 me to speak to --

23MR. FLEISCHER: I'm sorry. It's the 24 cable tray.

25 165MR. MELLY: The cable tray. When we get 1 into my presentation as to how the test, the phase 2 two will be performed, we'll get a little deeper into 3that. But I can speak to it briefly here is that we 4 hadn't looked into putting a physical cable tray 5above the cabinet that we're going to be testing 6 because we were going to try and focus on the ability 7to put instrumentation into measure the effect. By 8 putting a physical cable tray above the cabinet, we 9 would be limiting our ability to collect information 10at varying distances away from the cabinet. We were 11trying not to shield our instrumentation with the 12physical cables above the cabinet. So we're 13 attempting to get all the information we needed which 14 could then be applied to determining target fragility 15 and determining whether that cable tray would be 16 damaged or ignited in that case.

17 We've received several comments on that 18 on whether we would have a physical cable tray above, 19 as well as what materials would be put into that 20 cable tray, whether it be thermoset or thermoplastic 21cables. However, the current plan was to focus on 22 collecting the heat flux data at varying distances so 23that it can then be applied to the target 24fragilities. Gabe spoke a little bit to that, but 25 166 that would be a secondary piece to make the link-up 1between the data collected in phase two and the 2probable damage states. So that's what we had 3 currently planned for the phase two of testing.

4MR. TAYLOR: And just to add a little to 5 that is that, you know, let's say we go out and want 6to put in, hypothetically, a test tray. Obviously, 7 we're going to have limitations with measurement and 8 devices, as Nick mentioned, and also what cable do we 9test? You know, if we test one cable, does it 10 represent X, Y, and Z, and also, you know, you're 11 looking at does the HEAF damage the cable from a 12 functionality stand point and does it ignite a cable 13from a fire propagation standpoint. And both the 14 material and configurations are going to have a lot 15of influence on those two aspects, so it's another 16 parameter that's just going to add a lot of 17 variability and we can't reproduce enough tests to 18 capture all those variations.

19 So we felt that hanging good measurements 20 with the instruments would serve us better than 21trying to get a cable tray in. But, you know, if 22 there's ways around that or if there's ways to bound 23 things, you know, we're still open ears.

24MR. TURNER: It's so cheap just to stick 25 167 on a couple of cable samples, like I used to on 1tests. One of my hindsight things was I should have 2 been sticking little cable samples around all over 3the place and didn't. You know, you just do the 4 resistance before and after if you want to, but it's 5cheap and it's quick. So tests are so expensive and 6 so rare, in hindsight, I would have done a lot more 7 of that if I had a chance, so you ought to think 8 about doing that.

9 The other thing is, even though you're 10 measuring these fluxes, the exposures are so short, 11 I don't know how you can relate it to something, like 12 cable ignition where we not only have to have fairly 13 long durations of exposure, like 30 seconds, you 14would say here's the flux it takes at 30 seconds. We 15 got this flux over two seconds, I don't think you're 16 going to be able to figure out if the cable without 17 the actual cable.

18So it's cheap. I put cable samples up 19there. If you look at the IE we're working on now 20 where I did a lot of cable sampling, I would just 21 change that from, rather than solid metal sample 22 holders, use expanded metals so it gets air behind it 23 and it's got more of a chance to ignite it when it 24 gets hot enough.

25 168 MR. MELLY: Yes, that's a good point.

1MR. PUTORTI: That's currently in the 2 test plan is to put samples of cable along -- Tony 3Putorti, NIST. The current test plan has small 4samples of cable. We just have to make sure that we 5 don't pack things so tight that we disrupt, you know, 6 the various levels of measurements that we're trying 7 to do and different distances.

8MR. TAYLOR: Okay. So I guess the change 9then is the cable tray. All right. That is one 10 change that we're making.

11MR. MELLY: And this is what Tony was 12referring to. In the current plan, we do have cable 13 coupons on the mock test stands that we've discussed, 14as opposed to the full tray. We didn't receive much 15 feedback on the PIRT when it was published and made 16publicly-available. We're open to any review or any 17comments on differing opinions as to the importance 18 of certain characteristics, and it is a pretty in-19depth report. It's very difficult to skim over the 20 entire thing within a few minutes of today's effort 21 without wasting too much time but --

22MR. TAYLOR: Yes, and the feedback 23doesn't have to be different. If you think there's 24 a conclusion in here that you guys agree with, I 25 169 mean, that's just as valuable feedback.

1Okay. So not hearing anything on the 2webinars, all right. So then, Nick, can you pull 3 back the presentation from this morning and go up to 4 an earlier spot where it had all the variables?

5 Actually, let's stay here because we kind of, it was 6kind of getting towards lunch and we were kind of 7 winding down.

8 So after lunch, has anybody thought of 9other equipment models that are common that you see 10out in plants? Right now, we have identified the GE 11 Magna Blast and the Westinghouse DB-50s that I 12 believe are common from earlier discussion.

13MR. MILLER: Another low voltage one is 14ITE. It's pretty old. This would be, like, 60s or 15 early 70s.

16MR. MELLY: I believe that was the one 17 that was at Fort Calhoun, as well, the ITE series.

18MR. TAYLOR: Okay. Well, again, if you 19 go back to your plants and you have that information, 20 please get it to Nick or I.

21Next slide. Keep going. Thicknesses.

22You know, these are minimum in the standard. I don't 23 know when the standard had these in it versus when it 24didn't and when the standard came out. So given the 25 170 vintage of the plants, I'm not sure that this was 1 applicable at those times, but this is kind of what 2we're working on. I think what we do for equipment, 3we make sure that we're meeting these minimums. But, 4 again, the realistic side, does anybody have any 5 thoughts or feedback on what the actual equipment 6 enclosure characteristics are?

7MR. MELLY: That's a tough one off the 8 top of your head.

9 MR. TAYLOR: Without looking.

10MR. MELLY: But this is one where if 11 anyone digs into this from spec sheets or things like 12 that as to what the actual equipment thicknesses are, 13 it would be a good comparison to what we plan on 14 purchasing.

15MR. TAYLOR: Enclosure ventilation. I 16 think for this morning it's going to depend on what 17we procure. I'm not hearing other feedback, so I 18think that's kind of where we, I believe, want to 19 comment.

20MR. FLEISCHER: Ken Fleischer from EPRI.

21I was thinking earlier about what you said. You were 22 making a distinction whether it's indoor or outdoor, 23right? I'm familiar with outdoor units that actually 24 have heaters and vents, so they actually do circulate 25 171 the heat rise through them, whereas I'm also familiar 1with indoor. They're more sealed. I don't know, are 2 we, I don't think, are we considering trying to 3 represent any outdoor bus testing?

4MR. MELLY: We were not going to, we had 5 not focused on that as a major parameter to look at, 6whether it's indoor design versus outdoor design. We 7are trying to stay middle of the road again. From 8 our phase one of testing, when we did have to create 9 a mock cabinet, we were donated breakers and we built 10a cabinet around it. We created typical venting at 11the top of the cabinet, which we've seen throughout 12several plants. It wasn't louvered design, but it 13was almost checkerboard top vent. So we're really 14 just trying to be representative of what's out there, 15 and we've seen a lot of cases where you just have the 16 louvered design on the back face of the cabinet.

17MR. FLEISCHER: Okay. My thought would 18be is maybe we don't put a lot of importance on the 19 outdoor, just similar like the large transformers, I 20 think, that were discarded because they're outside 21 pretty much, you know, the air is like a heat sink, 22 it's just not really to just go right into atmosphere 23and pretty much be attenuated. You could treat the 24 outdoors non-safe bus the same and really consider 25 172 what the representative types are indoors.

1MR. MELLY: So would you say that they're 2 representative type for indoor switch gear and medium 3 voltage would be sealed, not vented?

4MR. FLEISCHER: Yes, the ones I've seen 5are pretty much sealed, non-vented. This is non-safe 6bus, right, we're talking? Not enclosures. We're 7 talking -- did I get on the wrong topic here?

8MR. MELLY: I was talking about the 9enclosure, but if we're talking about ducts, yes, I 10 would say no vent.

11MR. FLEISCHER: All right. My mistake.

12 But I have seen, we do have outdoor turbine decks 13 where we do have some of that equipment outdoors, but 14 it's non-safe.

15MR. RHODES: This is Bob Rhodes, Duke 16Energy. The non-safe bus ducts that I'm familiar 17with at Harris Robinson, they all have a breather 18 about the size of a quarter on the bottom of the 19duct. It's sealed at the top, but it's not airtight, 20 which is one of the reasons that we assume that the 21smoke would get into there and cause an issue. Of 22 course, ours all have the coating that was mentioned 23earlier, so they're all insulated. Norrell, yes.

24MR. MELLY: The one we pulled from Zion 25 173 also had the boots, as well as the insulation on it.

1MR. MILLER: Could we go back to the 2 panel thickness one again? So are we asking that 3 because we're going to be fabricating --

4MR. MELLY: No, we do not plan on 5 fabricating the cabinet --

6MR. MILLER: We're just going to be 7 testing the equipment we've got which has already got 8 a defined enclosure, right?

9 MR. MELLY: Yes.

10MR. TAYLOR: So I think it's getting at 11 the phase one testing where we have the international 12 equipment that got donated and the partitions weren't 13these thicknesses. So, you know, we wanted to make 14 sure that, moving forward, we test equipment --

15MR. MELLY: In addition, we were hoping 16 to get some information as to when we do procure the 17 equipment to test that it's representative of what is 18 currently out there.

19MR. TAYLOR: I think, along your lines we 20 get equipment from whatever manufacturer and it meets 21 these, then we should be okay.

22MR. MILLER: Or I guess if it doesn't, we 23would just make note of that? Because, I mean, if 24they were bought to all the standards before these 25 174 took place or they changed --

1MR. MELLY: Yes, that's a good point.

2 And I guess it's definitely what we want to consider 3 when we're going through a procurement.

4 And then the last one on bus ducts, was 5 there any additional thought on the configurations, 6 how we had it broken down in the test plan and mixed 7tree of tests, aluminum bus, aluminum duct, and so 8forth, on different configurations? Does anybody in 9 the room say, hey, we don't have aluminum bus steel 10ducts or whatever their configuration is? I know we 11 have aluminum and copper, aluminum bus bars/aluminum 12 enclosure, copper bus bars/aluminum enclosure, I'm 13 assuming we have copper/steel.

So nobody is 14 objecting to any of these four configurations?

15MR. EARLEY: And while you're on this 16piece, I'll mention that I checked out one of the 17papers on this. The horizontal configuration is 18worse than the vertical configuration. And the 19 horizontal configuration, the magnetic effect propels 20 the plasma cloud outward.

21MR. TAYLOR: Thank you. So it is more on 22 the plasma cloud being more --

23MR. EARLEY: Well, the incident energy is 24 higher. 25 175MR. TAYLOR: Incident energy is higher 1also for the horizontal configuration. Okay. And 2 then, you know, ratings of the buses, I think that's 3 something that I do have access to in the information 4 we have, but if there's any feedback from the group 5 here on we know our buses rated to 1600 amps or 6whatever it is, we'd like to test something that's 7close to what's out there. So any feedback on that?

8All right. So I think I can work with our electrical 9group and see where the population is. Nick, do you 10 have anything else on the test parameter?

11MR. MELLY: No, not at this time. But 12 one thing in the test parameter that's not currently 13well defined is the physical arrangement of how the 14bus ducts will be tested. There's some pictures in 15there. So that's one area that we are going to be 16enhancing as we go forward with the test plan. It's 17going to be the second phase of testing, so we're 18 most likely a year away from even beginning any bus 19 duct tests.

20There was a very robust design and test 21 that was done by Cooper on their bus ducts, and it 22was very interesting and the results seemed to work 23very well with what we're trying to do. So we may be 24piggybacking off of tests that have already been 25 176 done.1 So this is actually, what was in the test 2plan, we received several comments as to why there 3was the physical break in the bus duct that was 4 depicted, and that was an attempt to try and keep the 5 arc in this location because we were worried about 6 just continuing down the magnetic path and having the 7 arc migrate to the very end of the bus duct.

8 After seeing some of their testing, at 9least with insulated bus ducts, we don't know if 10 that's going to be necessary to put that physical 11break in there if there's the insulation. If there's 12 no insulation, we still don't know whether that's 13going to be a problem. But, again, you can see that 14we would have the instrument tests then located at 15 various distances around the bus duct itself.

16 In another picture, we had a cabinet 17 determination point, and we do not know if we are 18going to need that arrangement set up. So this 19 picture here was we had it terminated again to a 20cabinet. We don't think that we will need that 21 arrangement.

22 We'll go over this in a little bit more 23 detail when I'm doing my presentation on the phase 24two test plan. But any thoughts on this would be --

25 177MR. TAYLOR: Yes, I think from this 1 morning's discussion, we talked about maybe splitting 2 it up half insulated/half not or some fraction of 3 that. And for the insulated --

4MR. MELLY: For the insulated cases, I 5 think, you know, we can do what was previously done 6 by Cooper and peel the insulation off from a section 7 where we initiate the arc and it should stabilize in 8that location. But for cases where we don't have the 9insulation, does anybody object to what we had here 10 where we actually physically separated, had a 11separation along the bus and we initiate the arc 12 obviously on the power side of the bus, and, again, 13 it's to help maintain the arc in one location so we 14 can get good measure, decent measurement in that 15 location, because without that the magnetic forces 16should push the arc out and away from where we 17initiated that. And, you know, not having it 18 confined and not having a more continuous 19 configuration is basically just blow everything right 20out the end and we don't have open bus ducts in the 21 plant. So that was kind of our approach.

22MR. SHUDAK: This is Tom Shudak from 23Nebraska Public Power. Related to the Cooper testing 24 that was done, the picture you showed there has the 25 178switch gear at the end. We had a lot of discussion 1 on what we tested ours, and if you put caps on the 2 end of them as you guys did on the Zion where you 3capped that end of it, we didn't feel that was 4realistic. And so we spent a lot of time and 5 modeling of it to see if we needed to have caps on 6 it.7 So I guess if you're not going to put the 8 cap on there, I guess we'd like to know, I'd like to 9 know if you're going to put something on the end of 10those. There's lots of supports that go along the 11 way that would slow down any pressure wave, and these 12things just keep going on and on, right? They're 13very lengthy. They don't end. So I guess if you're 14 going to turn that one at the end of it, that's going 15 to impact, I think, your test results.

16MR. MELLY: And that's also, that's one 17 good point you brought up is that, in that testing 18that was done, you did not have a cap on the end 19 because you did have those supports periodically 20beyond where the test was initiated. We have not 21procured these yet, so we don't know how they're 22 going to be physically arranged within the bus duct 23itself. If we do have those supports, it might not 24be necessary to put the cap on the end. However, if 25 179 it is just a continuous run without any support in 1 the way from procurement, we may need to physically 2 cap the end or have a termination into a cabinet.

3MR. TAYLOR: And it doesn't necessarily 4have to be a cabinet that we procure. We can build 5 a cabinet and put some ventilation on it and just 6 terminate it into that just to kind of represent the 7end effect of a long run. But I think it's a good 8 point on, you know, you have the support pieces, the 9 support members it goes down the length of the bus, 10 and, you know, the question I then was asked is do we 11include those? I think we would, depending on what 12 we procure and what the design spec is.

13MR. MELLY: And one other consideration 14 is that when we had put together this test plan, one 15 of the reasons why we had the cap in originally, as 16well as the reason why we had this physical 17 termination, is we were testing in one of KEMA's test 18 cells, it's test cell seven, which had their medium-19 voltage incoming power supply on the left-hand side 20 of the cell and the low-voltage on the right-hand 21 side of the cell, and we were very worried about 22 damaging the adjacent power supply from that cloud 23 that you saw in the test video.

24 We have moved all of our medium-voltage 25 180 tests, I believe, we moved them all to test cell nine 1where we don't have that concern. One of the test 2 cell walls is completely devoid of any kind of power 3 supply, so we'll just be able to not worry about 4damaging anything. And that cap did not work very 5 long, as you can see by the video.

6 But as you can see here, I'm getting a 7 little bit ahead of myself, but we have the varying 8 levels of instrumentation stands so that we don't 9 have just one static three-foot instrumentation 10point, so we can actually get different levels of 11instrumentation. And if we damage one, we'll have 12backups at different distances. For the Finland test 13 that we did run on test 23, we destroyed the 14 instrumentation tests they have, so we have very 15little information as to what the flux was at the 16 three-foot zone of influence, and we had no 17redundancy beyond that. So while we do have other 18 instrumentation at different points around the 19 cabinet, it gives us very little information as to 20 the immediate effect from that test.

21MR. AIRD: We have a question from the 22webinar from David Lochbaum. His question is do the 23 cameras provide information needed to assess the 24 severity of the HEAF event or is their role to help 25 181 market the badness of HEAF events?

1MR. TAYLOR: I can attempt to address 2that. So there's two types of cameras that we used 3in the past. There's the regular video cameras that 4take high speed photography video of the event. We 5 can use that partially to look at the shrapnel and 6any type of metal that leaves the compartment. And 7 then we have the thermal measurements, as well, so we 8 have some high-speed high-arc cameras that we've been 9using that NIST has. I think there's some 10 limitations with their initial camera that we used, 11 and we've been working with his to get a camera that 12has a larger, what they call dynamic range, and it 13 automatically changes the ranges, meaning it can 14 capture different temperature ranges of the event.

15 So with that information, we can better 16 characterize the temperature rise of different 17 components or surfaces of the enclosure and use 18information that way. I don't know if NIST or Nick 19 had anything to say.

20MR. MELLY: The high-speed camera that we 21 actually used, it was also possible to watch the arc 22 migration from point to point within the cabinet in 23certain cases. There was some early on testing where 24you could actually see the initial arc strike and 25 182 then watch the arc move to a different portion of the 1 cabinet, and you could see it re-strike and start the 2arc in a different location. So that video was very 3 helpful in watching what was occurring during these 4 events. 5MR. PUTORTI: Tony Putorti from NIST. So 6 one other thing that we plan to do with these extra 7 cameras that we've added is that not only do you get 8 a qualitative idea of the extent of the arc ejecta 9 but also most of these cameras are not going to be at 10 a right angle, so it's relatively easy to go and get 11 a quantitative view of where the arc ejecta is, what 12 the extent is, and where it is in relation to the 13 instruments that we have installed.

14MR. FLEISCHER: Ken Fleischer, EPRI. I 15 like how you set up the three-tier there so that you 16said before if you lose something you get the next 17level. Should the bias be significant, some of those 18coupons may fall off and onto the floor. They're 19 going to be individually marked, so you know when you 20 pick them off of the floor which rack they came from.

21 Good. 22MR. MELLY: The answer was yes to that 23 question.

24MR. FUNK: This is Dan Funk. A quick 25 183 question on --

1MR. TAYLOR: Let's go to Dan, and then 2 we'll come back to Stan.

3MR. FUNK: A quick question on the test 4setup. We're going to force the damage to be where 5you started obviously. That's the intent here. And 6 so you're going to prevent the natural phenomena that 7 I think we all agree takes place typically to a 8termination point. Is the intent of this design to 9 simulate a segmented bus duct to where you're failing 10 at a joint or is the intent for this to apply to both 11 segmented and non-segmented, and that's going to tie 12 back into our current practice, which for non-13 segmented bus ducts, because of the magnetic effect, 14we postulate the HEAF event actually at the main 15 breaker terminals and not anywhere along the path.

16 This, depending on how the results are applied, could 17 dramatically influence that.

18MR. MELLY: This was more taking in mind 19 the segmented effect, so we'd be doing it at the 20 joint or at the bolted connection in the bus duct.

21 For those that are occurring at the termination 22 points, we were trying to map that to the damage of 23 the enclosure testing, rather than the bus duct 24testing. So those plants which welded the bus ducts 25 184 rather than have the bolted connection, they would be 1 matched to the enclosure testing results, and we're 2 going to try and clarify that in the methodology.

3 For here, where we're actually trying to 4 focus where it occurs, it is to try and simulate that 5 bolted connection in that case.

6MR. JOGLAR: So before we run the test, 7 vision is to stay consistent with the FAQ on the zone 8 of influence for this thing?

9MR. MELLY: We are using that as the 10starting point. However, there's no way for us to 11 locate that 37 feet above the ground.

12MR. JOGLAR: No, no, but at least for the 13 non-segmented ones, I think you said just put it on 14 the breaker.

15MR. MELLY: Yes, we're staying consistent 16 with that because we have no OpE to make us think 17that anything has changed in that realm. Any other 18 follow-up questions from the --

19MR. GARDOCKI: Well, where are you going 20 to go into, you've got another presentation, 21 positions, the racks and where --

22MR. MELLY: So the question was how many 23racks are we going to place. The picture on the 24 screen right now depicts three above and three below.

25 185Tony, do you want to take that? Right now, the 1 current plan is to have three varying levels.

2MR. PUTORTI: Tony Putorti from NIST. So 3 are you referring to the bus duct here or to the 4 enclosures?

5MR. GARDOCKI: Test coupons. You're 6 going to have one position?

7MR. PUTORTI: For this particular setup?

8MR. MELLY: So for this particular setup 9 where you see the three racks, each rack will be 10depicted or placed with, this is what the proposed 11rack would include. So each rack would include these 12 instrumentation pieces, so multiple coupons and the 13 depicted instrumentation.

14MR. PUTORTI: And so we have slugs, the 15classic slugs for clothing around there. Click 16 thermometers are on there used for measuring heat 17 flux, as well as a new tee-cap sensor that we're 18 working on now which measures, we'll be able to 19 measure very high heat fluxes, as well as some 20 coupons to collect any types of materials that may 21 coat things in the vicinity, as well as the samples 22 of cable.23MR. MELLY: And, again, right now, what 24we're showing on the screen is for the enclosures.

25 186 Again, you see that the three test stands, again, 1 this is not done on scale so don't think that they're 2 all going to be that close together, but the 3intention is to have the varying distances of 4 instrumentation racks that are not cluttered enough 5 to actually impede the hot gases, so we have included 6 instrumentation at varying distances. That was one 7 of the questions we got that these were not 8 consistent with the mentioned spacing that was 9 consistent with the IEEE guide, and that was for a 10 reason in that we're going to be designing these with 11 minimal impedance for the hot gases coming out.

12 Any other questions before we move on?

13MR. TAYLOR: So, you know, we still have 14 Nick's presentation on the test plan and the counter-15resolution, and we're going to get to that. But one 16 thing that came out of this morning's discussion was 17 should we look at the parameters and the things that 18 we can change and kind of rank their priority?

19 So what I've done is I've put together a 20 list here on two sheets, everything from arc 21 location, grounding configuration, winding 22 configuration of your transformers, the supply, 23 whether we're going to model from their generator at 24 decaying current or a steady-state current as we've 25 187done in the past. Obviously, duration voltage and 1 current are going to be important.

2 And on the second sheet here, you know, 3 enclosure thickness, bus insulation, circuit 4 characteristics, and that was, you know, providing a 5 sample from the plant of the circuit characteristics.

6Bus bar, gap separation of the buses. Equipment 7separation. You know, older plants and newer plants 8have differences in designs. Is that important? DC 9 offset or phase initiation angle, ventilation, 10 aluminum alloy or bus material.

11There's more. We can put them up there, 12but what I'd like to do is kind of go through those 13 and rank them from a priority or importance 14 standpoint, how important is it to our test protocol?

15High, medium, low. So I think that's kind of the 16 suggestion earlier from Ken, so is it okay if we just 17 start going through those and kind of throwing out 18our plans? I'm going to use a different color for 19 the ranking.

20 So starting off on these three, I already 21think that they're going to be high importance. So 22 duration, does everybody agree it's of high 23importance? Anybody disagree? Okay. So we'll put 24high for duration. Voltage, high, medium, low? All 25 188right, okay. Any disagreement with high on voltage?

1All right. So the comment was the excitation voltage 2isn't going to change much. The arc voltage is what 3we're interested in. The batteries might be dead in 4 that one, Ken.

5Okay. So arc voltage, high. And then 6current, is that going to be high, medium, or low?

7High? Medium? Low? Abstentions? All right. So 8high on that one, too. I kind of thought that's 9 where those would fall.

10 Arc location, you know, I put horizontal 11 and vertical, and that was the feedback from NFPA 12 that showed, the recent research showed that 13 horizontal had a higher incident of energy than the 14vertical orientations. So that might influence where 15 we put the arcing wire in our experiments, maybe more 16 on the bus section of the enclosures.

17 So for arc location, how important is 18that to the overall HEAF event? Do I have any highs, 19high importance rankings? Medium? What about low?

20 I'm getting zero feedback.

21MR. FLEISCHER: I think it's still high.

22MR. MELLY: I have one question on this 23 as to, for the results that were during the NFPA 24 testing that showed horizontal, initiation showed a 25 189 higher energy in the plume, is that because the 1 instrumentation was located in a horizontal direction 2or was it not due to the instrumentation locations?

3 So that's what I'm trying to get at is that, if I 4 initiated the arc in a vertical orientation, were 5 there vertical instrumentation locations that would 6 have still been in the plume or was the energy, were 7 the results calculated at the front face of the 8 cabinet because personnel were the objective of the 9 testing? 10MR. TAYLOR: Well, of course, that is a 11 major objective of the testing is the personnel and 12how much it tends to impact them. And the plasma did 13move in that direction. Of course, the magnetic 14 effect pushed it in that direction, and it also 15indicated that the effect was worse when the fault 16 took place deep inside the cabinet.

17MR. MELLY: So that's, I think, a good 18 clarification because if we are going to be putting 19 our instrumentation both horizontally and vertically 20because we're worried about both the three foot and 21 the five foot, I think it's something to keep in mind 22 that it might not have the same effect because we're 23worried about the plume and the magnetic effects in 24 both the horizontal direction, as well as vertically.

25 190 So it may have a lower importance than what some 1people have typically seen in literature for our 2 specific case.

3MR. TAYLOR: Yes, it may not have the 4 effect on the overhead cable tray, whereas that was 5 non-arcal.

6MR. MELLY: Right. So if I can, I 7 actually have a test --

8MR. EARLEY: I think I know what test 9 he's talking about, and the sensors were in the front 10 so they were engulfed in a side plume that came out 11 horizontally.

12 MR. MELLY: And I can show the --

13MR. EARLEY: The vertical plume is just 14 looking at the side and it's not engulfed.

15MR. MELLY: And I'll show one of the 16tests right after the break. I can pull up where we 17 had a test that was initiated on a vertical section 18of the bus bar material within the cabinet. It was 19 our OECD NEA Test 11, and the arc was directed 20 vertically, in the vertical orientation, but we do 21 have test stands in the front of the cabinet, so 22you'll completely see that we're going to get lower 23 results on our horizontal facing where the personnel 24 would be versus where we're instrumenting at the 25 191vertical case. So that's just something to keep in 1 mind when we're putting this ranking on.

2Okay. So, again, for arc location, who 3 thinks that that is of high importance? Dan.

4MR. FUNK: I'm looking at the recommended 5 updates to 1584, and horizontal versus vertical, they 6say it's about twice the impact. So I think it's 7 probably important because if we're going to be 8 playing with zones of influence, it's something we 9need to know. Maybe it can all be handled by 10research of what 1584 did. And you're correct. If 11 they're relying on it for the tests, it's kind of 12what you would expect. But it's probably something 13 we should know a little bit more up-front than before 14 we spend all the money for the tests.

15MR. TAYLOR: So we have two votes for 16high. Anybody else? Three. All right. Raise hands 17for high. Now we have three, four, five, six.

18Mediums? And lows? One medium. Okay. So three 19 lows, a medium, the majority high.

20Grounding configuration. It seemed like, 21 because we were doing three phase, that the grounding 22configuration wasn't that important. So, you know, 23 we could go and we could solidly ground it or we can 24 higher-resistant ground it or we could leave it 25 192floating, right? And I think some of that has to do 1 with the machine team, what they can best protect 2from overheating losses. So they left, but that's 3 something we also have to consider.

4 But from the group that's still here, how 5 important do we think grounding configuration is for 6the test that we plan to do? Do we think it's a low 7 priority, a medium parameter priority, or high?

8Okay. So for high? Who think it's of high 9importance? We have one in the back, two. Medium?

10One, two, three, four, five. And low? Two. Two-11five-two. Oh, I missed Nick. Two-five-three. So 12 you're doing high, medium, low, right?

13 MR. MELLY: Yes.

14MR. TURNER: I'll just cover, the way we 15 do the test for Japan is we're mainly looking at 16energy and really high energy events. We don't care 17 what the grounding is, as long as we get the current 18we ask for and the duration we ask for. And the 19 reason we do that is KEMA is setting up with you to 20meet your best for 23 kiloamps. They look at the 21 variety of circuits they can do and what combinations 22 of things can give them a nice, steady 23 kiloamps 23with the duration you asked for. Sometimes, it's 24 steadier in Y and sometimes it's steadier in delta, 25 193 and so they'll just recommend a circuit to me and I 1 don't care whether it's Y or delta because we rarely 2 see anything --

3MR. MELLY: Yes, I'm with Steve. Change 4 mine to low, so change one of the mediums to low.

5MR. TAYLOR: I'll go back through. Dan, 6 you had a comment?

7MR. FUNK: Yes, I'm going to basically 8qualify my high. Some of what Steve said, I think 9 the grounding configuration, where it's important is 10actually in frequency. If you have an ungrounded or 11 a high-grounded system, you're likelihood of 12 protection failure is actually, in my mind, going way 13 down, so I think it's a frequency issue more than 14 maybe the testing which is what you're interested in.

15 Does that make sense?

16MR. TURNER: Yes, and the testing at KEMA 17 is, of course, complicated because the frequency is 18 going to go down as the rotor slows down to give you 19the power. So an example, a test where you're 20 running 40 megajoules, they might start out running 21 them at 60 hertz on the generator, but by the time 22you finish the test it's running 48. It slows down 23that much. When you're out there in the generator 24 hall, you just hear it grinding down.

25 194MR. FUNK: Yes, I'm sorry. When I say 1 frequency, I mean probability --

2MR. TAYLOR: Okay. So let's go ahead 3 around the room again for the grounding 4 configuration, importance of that parameter to the 5 actual testing, who think it's a highly important 6parameter? No hand raises. All right. What about 7medium? Medium importance? And low? One, two, 8 three, four, five, six, seven, so it looks like it's, 9for those who voted, it's low on that, Nick. We're 10 not putting your names up here, so don't feel 11 bashful. Please provide the feedback on this.

12All right. Grounding configuration, 13delta versus Y. I think the information that Ken 14 gave us is pretty useful for when we split up the 15 different configurations but then, again, you know, 16 comments we've heard, as long as we get the 17 characteristics of the fault that we want from the 18 laboratory, it may not be that important.

19 So for delta versus Y winding 20 configurations, what's the belief for its important?

21 Who thinks it's a highly important parameter for the 22testing? Medium? Low? So we have one medium, and 23 one, two, three, four, five, six, seven, eight, nine, 24 nine on the low.

25 195MR. FLEISCHER: I would like to ask a 1 question on that because I think maybe you guys have 2experience. So the face-to-face voltage between a Y 3 and a delta is different, but you're saying that the 4 testing you've seen is always the arc voltage and 5 have you seen the arc voltage similar, regardless of 6 whether it's a Y or a delta?

7 MR. MILLER: Yes.

8 MR. FLEISCHER: They're the same?

9 MR. MILLER: Yes.

10MR. FLEISCHER: All right. Then go ahead 11 and change mine to low.

12MR. TAYLOR: Okay. So that's another 13unanimous low, a unanimous decision on that one. All 14right. Another thing that we had for a variable is 15the supply. With KEMA super-excitation configuration 16 at their site, we can basically do a steady-state 17 current, which is what we've been doing in the past.

18However, for medium voltage with the unit conducted 19designs, the generators want to trip. It's going to 20 coast down and continue to feed the fault, so you'll 21 have a decaying current for as long as that fault is 22 locked in.

23 So the question here is how important is 24 it to test the actual decay that you see in the 25 196 plants versus the steady state that we've been doing 1 to the testing?

2MR. TURNER: How big is the decay before 3the generator can't hold it anymore? Ten percent, 4 thirty percent?

5MR. TAYLOR: The question is how big is 6 the decay before the generator can't maintain the 7 voltage.

8MR. FLEISCHER: So I don't want to steer 9anybody in the wrong direction. I don't know off the 10top of my head. I don't want to venture to guess.

11I'd rather get facts. I'm going to see if I can find 12 a digital fault record that would accurately record 13that type of information. I'm aware of one plant 14 where I may have access that was a generator, it was 15a six-second generator fit fault. No, Turkey Point 16suffered a flashover of the main power transformer.

17 Actually, the arcing fault was on the high-voltage 18terminals, so the main power transformer. But it 19 would show you how the generator responded to it 20 because remember we did a root cause.

21I'm going to see if that -- and I 22remember we pulled a DFR. I'm going to see if I can 23find that DFR. That would then tell me what happens, 24 but I would anticipate, again, let's walk through the 25 197scenario, right? You're basically, you've tripped, 1you got a HEAF. The plant has tripped. Part of that 2is a plant differential. You open the yard breaker, 3 so you have no voltage contribution from the switch 4yard. You're now just with the generator with maybe 5like a stuck breaker. So you're going to have an 6 immediate step change because the exciter fuel 7 breaker is going to half open, so really it's going 8to be the residual magnetism of the rotor. The rotor 9won't even be excited. So, therefore, you're going 10 to have an immediate step change down and then, from 11there, you will have a decay. And I think that's 12 what we saw on the digital fault recorder.

13 So I did make a note to go see if I can 14find that. Is it 70 percent? I don't know. I can't 15 remember.

16 But as far as going for a vote, I think 17right now, unless we run into a technical 18feasibility, I vote yes. That's a tie. Fifty 19 percent of the plants are on off-site power, fifty 20 percent are off the generator.

21MR. FUNK: I'd like to second Ken's 22perspective. It's not, you don't run into a 23 practical limitation because right now these unit-24connected designs or that coast down are, quite 25 198 frankly, what's driving some of the answers we're 1coming up with in our database checks. So 2 understanding that phenomena I think could be pretty 3 important if that holds true that the unit-connected 4 design is vulnerable.

5MR. TAYLOR: Any other comments before we 6 go to vote?

7MR. TURNER: I'm thinking we're mainly 8 trying to learn about energy and the ZOI, so I'd say 9 if the current only drops ten percent that doesn't 10change the energy very much and you don't learn 11anything. If it changes 80 percent, you may as well 12have run just a duration half as much. But I don't 13know if KEMA can do as much as 80 percent. They can 14 probably do probably a 30 percent drop-off.

15MR. FUNK: That's fine, but just 16 understanding the phenomena, I think, so we can go in 17 with our eyes open and not closed is what's 18 important.

19 Tests aren't ever perfect.

20MR. MELLY: One thing to note, we do plan 21 to make a test plan change to include this decay as 22part of the medium-voltage testing. It will require 23 plant information where we are looking to dig deeper 24 on this so we do know what to tell KEMA to do 25 199 because, currently, it's very easy to -- yes, we will 1 need a sample profile to give.

2MR. TAYLOR: Yes, and then we'll also 3 have to enter discussions with KEMA to see what 4they're capable of doing. And if they're not capable 5 of doing what's representative, we need to maybe see 6 what they can do, run some tests with that, and then 7 have comparison.

8Okay. So if there's no more discussion, 9we'll go to see where we are. So for the decay 10 versus that steady state, who thinks that's a highly 11important parameter? One vote, two vote, three, 12four, five. Five for high. Medium? Six for medium.

13And low? Two for low. I think that's it for this 14 sheet.15 You have a question from the webinar?

16 Okay. 17MR. MELLY: We have another low from Dave 18 Lochbaum on the phone.

19MR. TAYLOR: So one low from the 20 participant on the webinar.

21MR. MELLY: Also, for those on the 22 webinar who are listening in, feel free to also voice 23your opinion and we will include the counts. That's 24 very easy for Tom to relay that information to us.

25 200MR. TAYLOR: Yes. And just to add on to 1 Nick, if we missed you going through those, the 2 rankings, you can see them on the screen. Please 3 feed that to Tom and we can update.

4MR. MELLY: Yes, if we can update any of 5 these fields, if you would like to provide your input 6 on what's currently shown on the screen.

7MR. TAYLOR: Okay. Let's just work down 8my list. It's not in any particular order.

9Enclosure thickness. Who thinks that -- is there any 10discussion on enclosure thickness? So who thinks 11that's a highly important parameter? Medium 12importance? One in the back, two, three. And low 13importance? Eight. So three for medium and eight 14 for low.

15The next one is bus insulation. I think 16 what we talked about earlier is we tried to split the 17experiments up half insulated and half not. But, 18 again, do we think that's important for, an important 19variable for the HEAF phenomena? Do you want any 20 discussion on the bus insulation?

21MR. TURNER: Can you define it a little 22bit more? Are you talking about the really thick 23 ceramic we put on bus bars sometimes, or are you just 24 talking about some insulation that's thermal, like a 25 201 boot? 1MR. TAYLOR: I'm not an expert in the bus 2 insulation. I don't know if anybody --

3 MR. SHUDAK: Ours is a thermal boot.

4MR. TURNER: Yes. And so those, the arcs 5just tend to blow right through those. If you have 6 a heavier ceramic that surrounding, like a square bus 7bar, it doesn't tend to burn that back as fast. You 8won't burn as much of a bus bar. And if it's 9 aluminum, you won't need as much aluminum in there 10because you didn't run as much bus bar. But if it's 11 just a heat shrink wrap on there or a boot, it's 12hosed. It's going to go right through that stuff, 13 and it just makes the experiment messy.

14MR. TAYLOR: Are we aware of any that's 15 ceramic configurations out there?

16MR. TURNER: Well, that's what you tested 17 on the bus duct design, that real heavy ceramic that 18 was -- 19MR. FLEISCHER: Yes, probably Norrell, 20right? It's like a sleeve. It's actually, it's not 21 rubber, it's a sleeve you slide on.

22MR. TURNER: Yes, and it's pretty heavy.

23MR. MELLY: And where we initiated the 24 arc, we removed that wrapping.

25 202MR. TURNER: It's one of those, I think 1it's high. If it's just a boot, I don't think it 2 matters.

3MR. TAYLOR: Okay. So I think what I'm 4 hearing is it depends on type, but, you know, I 5 think, with that understanding, whether it's either 6 the shrunk on or the ceramic type, do we think that's 7 important for the HEAF phenomena?

8MR. FLEISCHER: I have one clarification.

9Ken Fleischer, EPRI. So I think what I'm hearing, 10 though, is maybe we separate that bus insulation for 11switch gear in bus insulation. We're not saying bus.

12It sounds like it makes a difference there. It 13 sounds like for one it doesn't make a difference.

14 You said, regardless, whether it's the rubber or the 15boot, it's going to blow by anyway. And I think a 16majority of the switch gear is either non-insulated 17 or if it is insulated it's that, whereas a non-seg 18will typically have that more robust insulation. I 19would recommend splitting that. Yes, so the first 20 one would be a bus insulation for non-seg bus, and 21 the second one would be bus insulation for load 22 centers and switch gear.

23MR. TAYLOR: All right. So on the first 24 one, bus insulation for enclosures, votes for high 25 203 importance? Medium? Low? Eight for low.

1 MR. MELLY: Do we have anything else?

2MR. TAYLOR: Nobody for medium or high.

3Bus insulation for non-segregated bus duct. High?

4High importance? One, two, three, four, five, six, 5seven, eight. Eight votes. Oh, nine votes for high.

6Medium, medium importance? One, one vote for medium.

7 Low? No votes for low.

8All right. Next is circuit 9 characteristics. And this, I think, from our 10 discussion was characteri stics from a plant to help 11 us inform and understand how to set up the generators 12that came up. I don't know if we want to have any 13discussion or clarification on that, Dan. I think 14 that was one of the comments you brought up.

15MR. FUNK: Well, I think you probably 16 did. 17 I think the discussions over the past two days, the 18 voltage current, time duration, detriment, we've 19probably beat that to death. And the idea of being 20 able to characterize the test circuit, you know, we 21got feedback from KEMA on that already. We've 22discussed that we may need to try to come up with a 23 representative system that we would expect them to at 24 least be able to replicate within reason.

25 204 MR. TAYLOR: Any other discussion?

1MR. TURNER: And I think they can 2 probably give you, once they set the circuit up, they 3 give you an effective reactance for that circuit, so 4 here's what it is and here are the components we used 5to get it. They're still confused on why you need 6 that information, but I said don't worry about it, 7 they do models and stuff.

8MR. TAYLOR: Okay. So from the ranking, 9who believes that to be highly important for the 10program? We got one high, two highs. Medium? Four, 11five, six, seven. Sorry, seven. Two highs, seven 12mediums, and low? Anybody believes that to be of low 13 importance? Okay.

14Bus gap separation? I think we've kind 15 of shown that it has an effect on your arc voltage, 16 and we already ranked arc voltage before as being a 17highly important parameter. So, you know, we could 18 group it there, but keeping it separate, who believes 19 that that's highly important to characterize and 20understand what we're testing in the program. So for 21high? Eleven, eleven highs. Medium? And low 22 importance? One low.

23MR. TURNER: And let me explain why I'm 24low. Usually, when these arcs set themselves up, 25 205they're trying to go phase to phase. What matters is 1 where they can find a way to connect the circuit so 2this phase can connect back to that phase. And it 3 could be a metal wall that's a certain distance away 4where it's just at an open space. It really doesn't 5 matter how far the bus bars are apart, even if it's 6a half an inch or an inch. It's the length it's 7 going to to find its way back to the phase it's 8 trying to come back to, so I don't think the spacing 9 has as much effect as the surroundings, but it's the 10 closest thing by which usually is what sets the arc 11 voltage.

12MR. TAYLOR: Okay. We're going to skip 13over this one. We'll come back to it. DC offset and 14phase angle. And phase angle when you initiate the 15arc, so it's this phase angle. You start the arc, 16 and it has some effect on your asymmetric nature of 17your current wave form. And I think just from what 18 I discussed earlier, it may have some effect on the 19pressure. But from the overall program, you know, 20 what do we expect the importance of the DC offset or 21 the asymmetric nature of the wave form on the HEAF 22event? Is there any discussion on that parameter?

23 So who believes it's a highly important 24parameter? Medium? Five mediums. Low? And three 25 206 lows. 1 I don't know why it took Kenny so long to 2do. We're doing this in an afternoon. I wonder if 3 Ken is still on the line.

4Next one is ventilation. So this is the 5 ventilation, and this one might be one we were going 6to separate enclosures and bus ducts, as well. But 7 ventilation for enclosures, who be lieves that to be 8 a -- well, first, is there any discussion on the 9 ventilation aspect?

10MR. FLEISCHER: This is more for 11clarification for me. So we're talking about a 12 ventilation, I think I heard earlier that it can make 13 a difference as to where the arc plume is directed.

14 15MR. TAYLOR: Some of the literature I've 16looked at, it can have an effect on the arc 17 characteristics: the diameter of the arc, the arc 18voltage to a point. I don't know if Steve has got 19 anything else on that piece.

20MR. FLEISCHER: All right. Since we're 21 worried about the zone of influence and whether it's 22 going to be directed at it or away from it, and it 23 sounds to me that, okay, so it sounds to me it's 24 really important.

25 207MR. TAYLOR: I think the other thing is, 1 you know, Nick brought up yesterday with the louvered 2 vents in some of those, and Mark touched on it, too, 3 some of the effects there, that might have some 4impact of where the arc projects energy out of the 5 enclosure and the breach occurs or a door opens.

6MR. MELLY: Well, what we've seen in 7 testing is that the ventilation points for a lot of 8 these cabinets are typically where the plume and the 9 plasma is directed just from that's where all the 10material is going. We have seen the louvers just 11completely vaporize, but that is the location where 12the louvers are that these zone of influence is 13typically directed. It's a weak point in the 14cabinet. It's where the easiest, the path of least 15 resistance.

16MR. TAYLOR: Okay. So any other 17 discussion on that ventilation?

18MR. RHODES: And this is Bob Rhodes, 19Harris. I've seen buses, I've got buses that have a 20 ventilation that you all haven't discussed which is 21 on the top, it's got a top hat arrangement, and so I 22 haven't heard any discussion about that.

23MR. MELLY: So we have had experience 24 with the top hat arrangement in some of the Japanese 25 208testing that was done. When they were trying to 1 recreate the event that occurred at Onagawa, they did 2 use a top hat design, and we, again, saw that that 3 was the direction that the plume and the plasma 4 wanted to head because of the ventilation.

5MR. TAYLOR: Okay. So for ventilation on 6 enclosures, not bus ducts but on enclosures, who 7 believes that to be a highly important parameter?

8Nine. Medium importance? One. Low importance?

9 Zero. 10Okay. What about the bus ducts now? I 11 think we've had a discussion of some of the 12variations. Who believes that to be highly 13important, ventilation on bus ducts? Medium 14importance? One, two, three. Low importance? Seven 15 on low. 16Okay. And on the last one that I have, 17well, second to last, aluminum alloy. We haven't 18really talked too much about this. In the test plan 19 that we had for the small scale, we had some 20 discussion in there on the different types of alloys 21that can be encountered. And I think, you know, we 22 picked one of the alloys in there, but I think 23 there's going to be some variation of what's used out 24 in the plants.

25 209 So just on that variable of different 1 alloys, aluminum, I don't think there's going to be 2 much variation, but do we believe that there's going 3 to have to be a big impact from the HEAF phenomenon 4that parameter? Any discussion on the alloy aspect?

5Okay. So who thinks that's a highly 6 important variable or parameter?

7MR. TURNER: Discuss it a little bit 8more. You mean an actual alloy, or do you mean the 9 type of aluminum, or do you mean --

10MR. TAYLOR: Type of aluminum. Type of 11aluminum in the bus bars, yes. Well, bus bars or bus 12duct. So who thinks that's of high importance on the 13HEAF phenomenon? I've got one. Medium? And low?

14MR. MELLY: And I'm going to explain my 15--16 MR. TAYLOR: Seven on low.

17MR. MELLY: Let me explain my high is 18 that if we did this testing for the first round and 19 someone asked me if aluminum versus copper was going 20 to be of high importance, I would have said not at 21all. And we thought that going into the first round 22of testing. So I think my state of knowledge as to 23 whether aluminum is going to be such a large impact 24is so low that I'm going to err on the side of 25 210 caution.

1MR. TAYLOR: So you're making the 2 discussion of we didn't think aluminum versus copper 3 would be much of a difference, but we saw a big 4difference. I think what the question here that 5 we're asking is does the alloys make a big 6 difference, and you're saying that, because we saw a 7big difference in the other things and also the 8 alloys, there might be differences between aluminum 9material. We're not comparing it to copper and 10 aluminum. So is that the --

11MR. MELLY: Yes. Because if we had a 12 state of knowledge ranking, as well, it would be low.

13MR. TAYLOR: So I think we had seven.

14One low -- sorry. One high and seven low, eight low.

15Who votes low on that one? Seven. So it was seven.

16 17 And then the last one doesn't really have 18 too much to do with experimental plants, but we did 19 bring it up when we were talking about the pilot 20 plants and that relates to the equipment separation 21 and understanding, you know, out in the plants, the 22 different separation of equipment that's important 23for either safe shutdown or redundant divisions. I 24wanted to bring up a discussion point. It may go 25 211 into some of the decisions of where we put the racks, 1 the instruments, you know, how we space those.

2 So I put up here is there any discussion 3 on aspects of equipment separation out in plants and 4 how we would then instrument the experiments as that 5being an important parameter. I think right now, 6 just a little more clarification, we're basically 7 looking at multiple distances, so we have maybe three 8 feet, six feet, nine feet, something along those 9 lines, from a horizontal distance out from the 10 enclosure is what we're planning on doing right now.

11 You know, is there some logic or some configurations 12 where we would look at something different?

13MR. AIRD: A question from the webinar 14real quickly back to the aluminum alloy. The 15question is from Preston Cooper from TVA. His 16 question or comment is that can we include the 17 testing of this alloy using a chemistry tape 18 analysis?

19MR. TAYLOR: I'm not sure I'm following 20 the comment.

21MR. AIRD: If we get better information 22 from Preston, we'll get back to you.

23MR. TAYLOR: Does anybody else in the 24 room have an idea of what their comment was --

25 212MR. PELLIZZARI: Well, I guess is the 1 intention to, I mean, are you going to intentionally 2 test different alloys with aluminum, or you're just 3 presupposing you're going to get different alloys 4 with aluminum in your various equipment?

5MR. TAYLOR: So we don't know what the 6population is out there. So we looked in a little 7 bit, and there's some variation among the alloys that 8 we could procure that are commonly used in electrical 9conductors. So we wanted to make sure that we were 10 being representative but also, if there is variation 11 among there, that, you know, we vary it accordingly.

12 And what we're asking here is do we think that that 13 parameter has a large effect on the HEAF phenomena, 14 HEAF energy?

15MR. MELLY: And as part of the Cooper 16 testing, it was also a question that came up as to 17 what the composition or the alloy of aluminum is used 18for the enclosure. This is very difficult 19 information to obtain from the specifications that 20 were done 30 - 40 some odd years ago as to 21 specifically what alloy of aluminum is used in the 22 manufacturing of the equipment, and there are 23 primarily two different predominant alloys that can 24 be chosen for use in the electrical distribution, as 25 213 well as the enclosure design.

1MR. TAYLOR: Did you get any 2clarification? Okay. We'll wait on that. Again, I 3 put this equipment separation up here because it was 4 on the board back there and I thought it might have 5some impact, at least for the instrument. So is 6 there any additional discussion on that?

7MR. SHUDAK: This is Tom Shudak from 8Nebraska Public Power. So just a sort of 9clarification on this question. Those rotation racks 10will have coupons on there that will be able to 11determine the impact on equipment. I think it's 12 important that we are going to have cables on there 13 to see if those can ignite.

14 The location of the equipment is also 15 important, too, and that they can penetrate that 16 equipment because that's one of the ways we're 17looking at it. It's not just a three foot or five 18foot. It's also to the point where it comes up 19 against something non-combustible that's going to 20stop the projectiles. So those facing that is, I 21think, critical. Right now, I think we're using 22 three foot and five foot, so I think we have to 23 follow on at least with the current, and the tighter 24those are the better. I'd hate to have to go from 25 214 three foot to nine foot if I modeled it in our 1 current plan.

2MR. TAYLOR: All right. So maybe I can 3rephrase this. Rather than equipment separation, 4 sensor separation or measurement device location.

5 That might be a better way to characterize it.

6MR. MILLER: Was that what the item was?

7 Like GDC 17 separation?

8MR. TAYLOR: So I think it was related to 9one of the comments that Bob brought up of over 10 equipment having both safety and non-safety equipment 11in the same room and the separation can vary. So 12 that's kind of where I got it from, but, you know, 13 from a test parameter, how do we relate that into 14what we're going to be capturing? I think it might 15 fit better on where we locate the instrument rack so 16 that we can then, you know, look at the impacts that 17 we're measuring from the instrument racks to the 18separation because separation is plant specific. I'm 19 not sure we can go out and capture everything for the 20 test years that we're going to do.

21MR. MILLER: This is Kenn Miller. When 22 I think of separation, too, I guess the conducted 23 plume aspect of this is of interest because of the 24 visibility to reach beyond the range, particularly 25 215 filling the room.

1MR. TAYLOR: So on that piece, we're 2 looking to use those aerial gels by carbon tape and 3 then do the XPS analysis to see what the conductivity 4is. We don't have a live measurement of 5 conductivity, like a temporal measurement.

6 Obviously, if anybody knows the way to do that, we're 7more than interested to learn about it. But that's 8 one way we're looking at measuring the conductivity.

9And, again, those coupons can be put on these 10 instrument racks and post-analyzed.

11 So if I can kind of rephrase the 12parameter to measurement, what did I call it?

13 Measurement separation or measurement locations.

14 Maybe in the comments, we can put target locations.

15And position, yes. So distance, as well as whether 16 it's above, to the side, in front for positions.

17 So anymore discussion on that parameter 18there? Okay. So who thinks that this is a highly 19important parameter for the experimental? Thirteen.

20So thirteen high. I don't think there -- that was 21pretty much everybody. Medium? Medium imp ortance? 22 And low?

23MR. MELLY: We also got the most votes 24 for this one.

25 216MR. TAYLOR: Okay. Make sure you save 1that file, Nick. I don't want to lose the progress 2 we've made so far. So I think that's all I had for 3the list. Were there any other parameters that we 4 thought were important, or not important but we felt 5should be considered and ranked? Okay. Not hearing 6any. Is there anything else from this morning's 7 discussion and early afternoon's discussion on test 8 parameters that we want to touch on?

9 MR. MILLER: Or definitions.

10MR. TAYLOR: Is there anything on the 11 webinar?

12MR. TURNER: Does anybody ever anticipate 13 using arc resistance switch gear in the future where 14 the flap opens and all the hot gas goes out a vent 15 and goes around a corner and gets dumped outside or 16something like that? Does anybody anticipate having 17that? Because we haven't done any testing on that 18 stuff, even though it really helps the worker who's 19standing right in front of the cabinet. We keep 20 telling the people that invented it that this is 21 really bad for us in fire protection because now you 22just put the flame over there. The response is yes, 23but we've got barriers in there. It captures most of 24 the heat, and those emissions only last for a couple 25 217of tenths of a second, so they're consumed. So if we 1 had one of these long arcs, here I am putting the 2 flame out 20 feet away from where the accident is and 3changing it. So I'm just curious if anybody ever 4plans on putting those things in there, which are 5 great for personal protection but horrible for . . .

6 MR. MELLY: Stan.

7MR. GARDOCKI: We are in discussion on 8our panels. We were discussing how the arc transits 9through the atmosphere, so one of the parameters I 10 was thinking, the atmosphere conditions, whether it's 11 humidity or dryness, temperature, if that's the case 12 of the arc transiting through an atmosphere but arc 13versus a thermal arc or something like that, is 14 atmospheric conditions going to influence how far 15 this heat goes with the aluminum that's in the cloud?

16 I would open that for discussion for anyone who 17 thinks that would be an impact on this aluminum heat.

18 MR. TURNER: A parameter of interest?

19MR. TAYLOR: Was that something that the 20 NFAA or IEEE will do for all their arc flash testing 21 that you're aware of?

22 MR. EARLEY: No, we did not.

23MR. MELLY: As part of our testing on 24 phase one, we did record all atmos pheric conditions 25 218 on test days: humidity, temperature --

1MR. MILLER: Those were outdoors, right?

2MR. MELLY: -- barometric pressure. We 3were outdoors. So we did have a range of conditions 4that we tested in. Rain the Japanese tested in, but 5a lot of snow. We were outside. We covered a fairly 6good range in Pennsylvania. We have not taken the 7 time to look into the effect, primarily because we 8 didn't have the electrical conductivity measure live 9 to make the comparison to humidity or other 10conditions. But that's something that I guess is 11 still open for discussion right now whether we think 12 that's an important event.

13For my opinion, I think it may be of 14 importance to the initiation and to the frequency, as 15well, and some of the events, the Monshawn 16 particularly saw the humidity, the atmosphere, and 17 the salt, potentially the outdoors, that could have 18contributed to the initiation of the event. I don't 19 know about the consequence of the event. Dan.

20MR. FUNK: Yes, I would second what Nick 21is saying. I've seen an amount of research done for 22 pollution and surface contamination but only as an 23initiator. Like, a lot of utilities have programs 24 where they go out and wash down switch gear and stuff 25 219 for just their very purpose for the measure, what 1 part of the country you're in or where you're at for 2 cumulative deposits of pollution and continue to 3 degrade the overall surface resistance, but I've 4 never heard it being a factor once you get an arc 5started. It's such a minor contributor to the 6 temperatures and forces.

7MR. TURNER: My perception of it is the 8 original atmosphere that starts in the cabinet is 9completely changed in five milliseconds. Everything 10 is completely difference because you push all of that 11out real quick. And then the plasma itself is behind 12the initial pressure front, and it's kind of doing 13its own thing in its self-made atmosphere. There's 14 not much difference in that self-made atmosphere, 15 depending on the original conditions.

16So I agree. It might have something to 17 do with the initiation of the fault that started it.

18 But once you get that arc going, it kind of, I don't 19think it's going to have much effect at all. And you 20 can see some of the effects on the atmospheres around 21it because there's a lot of tests with SF-6. So you 22 don't even have air insulated anymore, so you know 23 that SF-6 and air insulation see some differences in 24 terms of how the arc is located and the energy, but 25 220not as much as you'd think. So I don't know if you'd 1 want to handle SF-6 different than here.

2MR. TAYLOR: Okay. That's great 3feedback. Any other discussion on atmospheric 4conditions for the actual HEAF event? Okay. Let's 5take a quick vote then. So who believes that the 6 atmospheric condition parameter has a high impact on 7the HEAF phenomena? Medium? One for medium. Low?

8 Ten, ten for low.

9 Anything else that we've missed?

10 MR. MILLER: I'm sorry.

11MR. TURNER: Oxygen availability. Is 12 that an atmospheric condition?

13MR. GARDOCKI: You need to oxidize the 14 aluminum to get the --

15MR. TURNER: And we're not sure how that 16works in the testing we do. When I've made 17 calculations, I said let's just assume the oxygen is 18there. It might run out of oxygen, we just don't 19 know because I've run tests where you get a nice 20 plume going out the top, you could conceive that it 21 could actually draw air in, like your fireplace in 22your house when you've got plenty of oxygen because 23 arcs feed the oxygen going into the arc, not this 24way. They feed it from the ins where it's starting.

25 221 We plan to run some tests to try to 1 figure out how the oxygen level changes in these 2larger things. So that's a big effect that we're 3 testing for Japan.

4MR. TAYLOR: So any other discussion on 5 the oxygen availability ventilation aspects?

6MR. PUTORTI: Tony Putorti, NIST. But 7 that was something that I was thinking of when we 8originally were looking at ventilation. So, you 9 know, is that something that we've already covered in 10 ventilating the enclosure up top?

11MR. MELLY: I think they're linked but a 12 different aspect.

13MR. TAYLOR: It might be worthwhile to 14separate, as you've done here. So, you know, 15 ventilation is just more of a physical configuration 16 and the impact it has on the event versus, you know, 17 oxygen availability, especially on the aluminum 18piece. Do we have enough oxygen to oxidize the 19metal? So let's go ahead and -- I mean, it makes 20sense to me to keep them separate. Does the group 21 agree? I see some heads nod.

22Okay. So for oxygen availability for 23 aluminum, who believes that to be a highly important 24parameter? Eight. Medium importance? Two, three.

25 222 And low importance?

1We'll do the same thing for copper. So 2 high importance for copper, oxygen availability?

3Medium importance? Two. And low importance? Nine, 4 ten. Your hand is up, right, Nick?

5 MR. MELLY: Yes.

6MR. TAYLOR: Yes, ten. Anything else?

7Any other parameters? Anything from the webinar?

8Okay. So I think that kind of -- I appreciate all 9the feedback. It's really good information. I 10think, especially if we go and formalize the test 11 plan and finalize what we're going to do moving 12 forward, it's going to help provide feedback to us.

13 I think at this time it would be a good 14time to take a break. So it's 10 after right now.

15Let's come back at 3:30. And then Nick will get into 16 the test plan and the changes that he's made to it 17 that resolve the comments at the high level and then 18 anything we want to bring up with regard to the test 19plan and that aspect, we can do that. So we'll go on 20 break now until 3:30.

21MR. MELLY: Small-scale testing, you find 22 something or add, I mean, is that going to input 23 something to this, or you don't think you'll have 24 enough time after the small-scale testing to impact 25 223 this? 1MR. TAYLOR: So the question is we're 2 doing small-scale testing and how will that be used 3 to impact how we're going to change the test program 4for full scale. Okay. That characterizes the 5question. I don't really see a lot of the small 6 scale on the HEAF phenomena or the energy output 7 really supporting much on the large scale, but the 8 real purpose of the small scale is to characterize 9 the particles to help further develop the model that 10Maryland is working on. So there will be some 11 information there that can be used, but also it's 12 proof of concept so those aerial gels and the black 13 carbon tape, we can then use that for here and we can 14 put those at those test stands, as Nick was showing 15 you earlier, or we can even put those, because 16 they're passive devices, we can put them inside the 17 enclosure which is really where we want them because 18 that's where a lot of the additional energy from the 19aluminum is going to occur inside the enclosure. So 20 we'll have some feedback there.

21 But as far as, like, the HEAF phenomena 22 and the parameters that we're using for large-scale, 23I don't see the small-scale informing that much. I 24 don't know if Nick or anybody else have anything to 25 224 add. 1MR. MELLY: Yes, I agree. It's more for 2 the additional instrumentation which may be used for 3 the large-scale testing to focus in on that 4 conductive cloud of aluminum, as well as some of the 5 videography that we'll be doing to capture the 6 particulate.

7MR. TAYLOR: And that's a good point.

8 And, actually, the question that David Lochbaum 9 brought up, that's one thing we missed is that the 10Sandia equipment has the potential to be able to 11 track particles to get the velocity, which can be 12important for the model that we're working on. So 13we'll probably try that out. We are going to try 14that out on the full-scale side, so that's another 15 thing in the small-scale we'll be using.

16MR. GARDOCKI: How about the percentage 17of aluminum of the components? Are we going to be 18able to draw an analogy to how much aluminum grams 19 inside the cabinets, per se, and draw that out to say 20 the more aluminum the larger HEAF, or are we going to 21 do some kind of staging or percentages of aluminum 22 content? 23MR. MELLY: We can get into that a little 24 bit when I discuss the test plan as our plan for 25 225 evaluating that and trying to make some energy 1 calculations, yes, energy correlations based on how 2 much aluminum is missing from the bus bars after the 3 test program, how we're going to be collecting that 4information, and post analyzing it. And, hopefully, 5 there will be drawbacks to the small scale because, 6 as Dave said, one of the things that we're looking at 7 is the amount of aluminum that is removed in a small-8 scale testing to do our energy calculations there, 9 which is one of the reasons why we're downsizing the 10 bus bars at Sandia so that we can more accurately 11predict how much aluminum has been removed from the 12 bus bar.

13 MR. TAYLOR: Does that answer --

14MR. GARDOCKI: Yes, I mean, I don't know 15 if you want to incorporate in your test parameters to 16 get good representation of how much will be consumed 17 or if it happens too fast, consume all of it, or how 18much in a particular time period. I'm not sure 19whether the physics of these dynamics of how the 20 plasma develops and time spanning of the arc.

21MR. TAYLOR: Yes, I think that's 22 something that we'd probably measure on the back end, 23 so measuring before and afters.

24MR. GARDOCKI: If you don't have enough 25 226 aluminum in there, then you're not going to get the 1 spectra.2MR. MELLY: I think we have that covered 3 in both the factors that we have on the board, such 4 as the ventilation, oxygen availability with the 5 aluminum, as well as the duration of the event will 6 both influence how much aluminum is involved in the 7 process. 8MR. GARDOCKI: But your cabinets are 9 going to have a certain amount of aluminum or you're 10 going to add it into it.

11MR. MELLY: The cabinets will have a 12 certain amount of aluminum, and we will be measuring 13 the total weight of the bus bars before we run the 14 test and after we run the test so we'll know exactly 15how much aluminum was lost during the test. One of 16 the uncertainties, I will say, will be that we do 17 have this particulate and you can see the objective, 18which is solid aluminum, rather than oxidized or 19 vaporized aluminum, so there will be particulate when 20 we have these tests to see the multiple aluminum 21 pieces, as well as the slag thrown in all directions 22in the high-speed videos. That will be the 23uncertainty of the test. But in terms of the bus 24 bars, we will be doing putting in post-monitoring of 25 227the weight of the bus bars themselves. So that's one 1 aspect that we'll be looking into in greater detail 2 in the phase two test program, which does add a lot 3 of time to the post-processing and the analysis of 4each test that we plan on running. And I'll cover 5 that in a little bit more detail on my next 6 presentation.

7MR. MELLY: All right. I think we're 8 ready for a break, and we will come back in 15 9 minutes. 10 (Whereupon, the above-entitled matter 11 went off the record at 3:14 p.m. and resumed at 3:34 12 p.m.)13MR. MELLY: Okay, we're going to get 14 started here real quick on the -- this is the second 15to last presentation. It may end up being the last.

16 A lot of the comment review, we've kind 17 of knocked off one by one throughout the discussion 18 in the morning.

19 I'll quickly go through that last 20presentation, but I think that we may have covered 21 almost everything.

22 So, this presentation is focused on the 23 second draft test plan, what we've been talking about 24 mainly all this morning, a lot of the parameters and 25 228 things like that, but this will dig in to the actual 1 document that we released for public comment as well 2as the updated version that we released for this 3 meeting.4 So, some of the objectives that we have 5semi already covered is discussion of how the 6 phenomena identification and ranking table influence 7 how we are going to be doing the testing.

8 Some of the experimental variables that 9 we are going to be focusing in on for this second 10 phase of testing, how we're going to be doing the 11 measurement for the second phase of testing, what 12members we are currently working with inside the 13 OECD.14 We've alluded to the fact that we do have 15 10 members. We'll share those members right now.

16 We're going to discuss the test structure 17 which we've also shown a little bit, the experimental 18 approach and the time -- the current time line which 19 is very fluid, depending on how -- when we can get 20 arrangements, contracts in place, updated test plans 21 and things like that.

22 So, these were what the phenomena 23 identification and ranking table identified as high 24 important phenomena.

25 229 So, the cabinet to cabinet fire spread 1 and secondary arcs and cabinet line ups, this was 2seen as a high parameter of interest from several 3 countries, especially because we did see cabinet to 4 cabinet fire spread in the Onagawa vent as well as 5 additional secondary arc potentials in that lineup.

6 The thermal damage criteria and target 7 sensitivity, for the short high heat exposures which 8 we've discussed earlier, this gets to what we've been 9 discussing this week as the -- we've been terming it 10 target fragility as to how we make that link between 11 this high heat extreme exposure for a very short 12 period of time and whether that cable or secondary 13 piece of equipment, will it be damaged?

14 Also, the likelihood and severity of 15secondary fires. We've been discussing this as to, 16 if we have a short event, it's unlikely to actually 17 propagate a secondary fire or a fire within the 18 actual cabinet of origin, different methods of 19 getting about that.

20 One of the high parame ters was also the 21performance of HEAF shields. This is postulated 22method for mitigating the effects of damage from a 23 high energy arcing fault through the use of a barrier 24 between the cabinet of origin or the bus duct of 25 230 origin and your secondary targets.

1 The performance characteristics were 2 looked at as highly unknown and it was important to 3 gain more understanding as to how those would perform 4 in the actual HEAF environment.

5 Also, a high parameter was the likelihood 6 and severity of damage from the arc ejecta on 7electronic equipment. This is what we're referring 8 to for the aluminum cloud that's created as to what 9 effects that can have on potentially shorting out 10 secondary equipment.

11 Metal oxidation, that is in terms of the 12 effect of aluminum and copper and the added energy 13 from the oxidation of that aluminum material and the 14 electrical -- the arc electrical characterization is 15to the points of voltage, current and duration. And, 16 they're not in order of 1 to 10, these are just the 17 high parameters.

18 So, these are the focused variables and 19we have covered this in great detail. And, you can 20 see here that I've identified that this was going to 21 be a focus for a later discussion through Gabe's 22 questionnaire is that we're trying -- we were going 23 to try and pin down collaboratively the current and 24 the duration of testing which we have done, and 25 231 discuss the methodology that we created and put in 1 the test program.

2 So, the current that we discussed earlier 3 of using was a range from 15kA, 25kA for low voltage 4 and 25kA and 35kA for medium voltage.

5 The arc duration, I think we discussed 6 that and have decided that we should nominally be 7 testing at the 2 and 4 seconds for both the low 8voltage and medium voltage range and the bus ducts, 9 I think we maintain that we'll still be testing at 1 10 and 3 seconds.

11 There were two additional tests at 5 12seconds and the driver for those were some 13 international focus to have a longer duration event 14 that corresponded with some of the operational 15 experience, and especially on the generator-fed 16 events.17 Also, in the Phase 2, we've discussed 18 that some of the main priority in the U.S. is this 19 aluminum versus copper, so the electrical enclosure 20 conductor material as well as the different 21 configuration of bus ducts switcher I think that we 22 have covered adequately.

23Again, very little changes. You see a 24 lot of what we've done here is in the last exercise 25 232 we did that went to binning as well as I think that 1 we've covered every single one of these that's on the 2 board in terms of the importance to the zone of 3 influence calculation.

4 Equipment type, bus bar material, bus 5 duct material, voltage, current, frequency, power 6 configuration, equipment grounding, arc duration, arc 7 energy, bus bar insulation, bus bar spacing, bus bar 8 size.9 We did a pretty good job during our 10 discussion this morning.

11 So, the measurement parameters that we're 12 going to be focusing on, we had the presentation 13 yesterday from NFPA where they had to develop a lot 14of the instrumentation it was using because of the 15 challenge to actually collect the information.

16 We did this in parallel and it seems like 17 we actually ended up in a similar place as to how we 18 were collecting the information and what we were 19 collecting.

20 We're going to be -- in the test report, 21there's actually a great detail into how these were 22 developed.

23 Tony published a paper for the SMiRT 24 which covers in great detail how these heat flux 25 233 measurements are working and detailed.

1 We're going to be looking at the 2 temperature and the heat flux, the pressure we've 3 discussed at internal to the cabinet.

4For room pressure is an area that's not 5 currently in the test plan but it may be slated for 6 one of the spare tests if we have the budget.

7 We're also going to be documenting the 8furthest extent of damage. This was one of the tools 9 that we were using the video for as well as post-test 10 analysis.11 We're also going to be looking at the 12mass of material vaporized during the tests. We 13discussed that in the last presentation. We're going 14 to be doing pre- and post-test measurement of the 15 cabinets themselves for the potential to develop the 16 approximate energy release models from the classic 17 energy conversion models.

18 This also has ties to some of the small 19 scale testing and whether linkages can be made in the 20 small scale energy release versus the large scale on 21 mass loss.

22 We also discussed the cable sample 23materials. This is the cable hoop-ons that we 24discussed. We're not going to be doing full cable 25 234trays. And, we covered that one in pretty good 1 detail.2And, the byproduct testing. So, we're 3 also going to be looking to do the conductivity 4 measurements through, again, the proof of concept 5 small scale work that we're doing, capturing that 6 affluent that comes off and also coats material.

7 We're going to be doing spectroscopy on that.

8 One main difference between the RGR fault 9 Phase 1 and Phase 2 is we will not be measuring heat 10 release rates in the Phase 2 of testing.

11 There were several reasons for that, one 12 of them primarily is being that we only really saw 13 the benefit when we were collecting heat release rate 14information on that in during fire within the 15 cabinet.16 If it was actually under the hood, that 17 initial blast overwhelmed the hood with, I believe 18 our max capacity for that hood was around 1000 19 kilowatts, 3500, but it mainly was based on how much 20 smoke we collected in a short period of time and you 21 can see the entire facility was full of smoke, we're 22losing it. We had -- were exposed to the atmosphere 23 and wind conditions.

24 We used -- tried to use a lot of curtains 25 235-- wind curtains to capture as much as we could.

1 But, we didn't really have a good way to capture all 2 of that.3Additionally, it was extremely time 4 consuming to set that up and work around when we were 5 trying to put the instrumentation stands up and 6 things like that.

7 So, by removing that, we'll have a higher 8 focus on being able to collect the heat flux data at 9 varying distances away from the cabinet.

10 This goes into the measurement devices 11that we plan on using. For the temperature, we have 12 thermocouples, plate thermometers and the IR imaging 13 camera we referenced.

14 For the heat flux, we're going to be 15 using the plate thermometer for the time varying.

16 For the average we're going to have the 17 plate thermometer and the thermal capacitance slugs 18 which are new for this Phase 2 program.

19 Incident energy, we're going to be doing 20 the equivalent slug calorimeter so we have 21 comparisons to the NFPA testing as well and to the 22 IEEE.23 For the compartment internal pressure, 24 we're going to be doing the piezoelectric pressure 25 236 transducers.

1 We also, in terms of videography, we've 2 covered that fairly well.

3 We have the IR filter videography, the IR 4imaging. We have high speed cameras and high 5 definition cameras placed at various locations around 6 the test.7 And, again, we're going to be doing that 8 surface deposit analysis with links to the small 9 scale testing.

10 So, in terms of the international 11 members, we have Belgium, Canada, the Czech Republic, 12 France, both their IRSN and which is the regulatory 13 body and EDF which is their -- on the power 14 generation side.

15 We also have Germany, Korea, Japan, 16 again, we have CRIEPI which is their electrical power 17 research institute.

18 And, we have the NRA which is the 19 regulatory body.

20 We're also going to be including the 21 Netherlands and Spain.

22 So, from the previous test program, we 23 are gaining EDF, the Netherlands and Belgium and the 24Czech Republic as additional members from the first 25 237 round of testing.

1 The only member who is not participating 2 in the second round that was in the first is Finland.

3 And, this is primarily due to resource constraints, 4 personnel challenges and for the ability to 5 participate.

6So, we have covered this in fairly well 7 detail, but just to go over it again, is that the 8tests in blue will be part of the OECD test 9contribution and the ones in orange were those 10 specifically driven by the GI program.

11 And, our focus, the U.S. centric focus on 12 additional aluminum tests.

13Same goes with the bus duct testing. In 14 blue is OECD and in orange is the USNRC.

15 So, we discussed this in detail as well.

16 One of the main things we want to do is limit the 17test variables. And, this was so that we can have 18 the direct comparison between tests, a low voltage 4 19 second, medium voltage 4 second.

20 And, we can look at the differences in 21the zone of influence, the differences in the 22 calculation hoping to have an identical cabinet so 23 these can be single variable differences in hopes to 24 create that dynamic model based on the parameters 25 238 that we have deemed to be high importance and we can 1 adjust those to evaluate the zone of influence.

2 Again, we talked about the repeatable arc 3location. This becomes very important in the 4 discussion we had about the NFPA testing in that --

5 In the NFPA testing we discussed that the 6 horizontal direction showed a higher energy flux 7 based on the calculations.

8And, you can see on the tests on the 9 picture in the upper left-hand side that, if we had 10 a horizontal orientation and our instrument racks 11 were located in the horizontal direction, clearly 12 we're going to show that there is a higher influence 13 or a higher instant energy reported there.

14 However, if we have test stands 15 vertically above the cabinet, the vertical 16 orientation is going to most likely show the higher 17 flux from that testing.

18 So, right now, we want to be able to 19 ensure that our test stands are located in a 20 repeatable direction so we can get the affluent 21 coming out of the cabinet from one of these tests.

22If we -- if the arc location comes out 23 the top and we have all of our instrument racks on 24 the side, we're still going to be collecting data, 25 239 but is that data relevant to the energy release 1 during that actual test?

2 So, we want to make sure that we can have 3 repeatable tests on whatever cabinet we use, which is 4 why arc location becomes so important where we 5initiate the arc. If we miss a test, that's a full 6 test where that data is not really applicable or 7 relevant for comparison on the other ones.

8 Again, instrumentation will be the 9primary means of data collection. This is in terms 10 of we're not going to have that cable tray above, but 11 we're going to be using the instrumentation, the heat 12 flux, the temperature as a surrogate to make that 13 linkage between the actual materials.

14 And, again, I just wanted to reinforce 15 this, that no testing will be performed -- no testing 16 to be performed will be subject to any equipment to 17 get to conditions that exceed the equipment ratings.

18 And, what I mean by that is, if we have 19 a donated cabinet that's rated for 4160, we're not 20 going to subject that to a 6.9kV situation.

21 So, this was the general lab that you've 22been -- that we've shown a few times. Multiple 23 equipment rack locations to capture the direction of 24 where we believe the arc is going to go.

25 240 This is discussed in detail in the test 1 report in that we're going to have a -- essentially 2 a week -- the first week of testing will be a 3 shakedown week where we are going to go with the 4 proof of concept where we believe the instrument rack 5 should be located, where the arc location is going to 6 be.7 And we are going to try and confirm that 8 we can collect repeatable information from these 9 tests.10 We're doing that because we don't want to 11 go secure 40 lab days and go just 40 at one time.

12 Because, if we need to make adjustments to the test 13 plan and to how we're collecting the information, we 14 can adequately do that.

15 Test days are the most expensive part of 16 the project along with the equipment that we're going 17to be testing. So, we are going to approach it in a 18 careful manner.

19So, the time line for the testing, it's 20 not head workshop, but right in here, we're at the 21 April 18th and 19th, the workshop date.

22 Next week I have the meeting with the 23 OECD International Members where I'm going to relate 24this information to them.

And, essentia lly discuss 25 241 how we plan on changing the test plan and finalizing 1 the test plan.

2I will also receive their input. I know 3many of the members have been on the webinar. And, 4 I just want to make sure they have a fair chance to 5 have an equal level of discussion of any plan changes 6 or test changes that do come out of this meeting, 7 that they are fairly gone over with them.

8 We will be doing a comment resolution.

9 Essentially, we're still working on the comment 10resolution, we have to go over it a little bit. And, 11 we've received a lot more information during this 12 meeting.13 So, we hope to finalize all 91 comments 14that we have received by mid-May. And, we also plan 15 to revise the test plan based on feedback from this 16 meeting and the international meeting, again, in mid-17 May.18 We discussed earlier, the international 19 agreement is currently held up with the lawyers.

20 And, I left it a little bit vague as the summer is 21hopefully the data that we will have it signed, 22 towards the end of summer, August.

23 I think everyone in the room realizes the 24difficulties with getting past lawyers. So, I really 25 242-- that's a shot in the dark.

1 Again, one of the aspects of the project 2 is we do have equipment delivery and that becomes a 3 very complicated process from country to country as 4well. So, we're hoping to have everything shipped to 5 our storage facility which is down the road from the 6 Chalfonte facility in Pennsylvania in the fall, 7October time frame. And, we look to do the initial 8 test series in October, that is the target date 9 currently.

10 That first test series will most likely 11 be on equipment that we do purchase, so the 12 purchasing is in that range.

13 Depending on feedback from this meeting 14 as well as management how the agreement goes, that 15 October date is aggressive, but we hope to be able to 16 do testing then.

17 Again, the second series of tests, we 18 wanted to correspond that with an international OECD 19 meeting to have our international partners a chance 20to witness testing. Again, this is in Pennsylvania.

21 And, Mark has already opened this up to 22 try and set up a field trip up to the KEMA facility 23 so you can see the generator capabilities as well as 24 the KEMA facility.

25 243 There will be several rounds of testing.

1 Typically, we like to do a week or two weeks of 2testing in a row. That can usually accommodate one 3test per day because of the amount of effort that 4 goes into instrumenting the equipment, safety related 5 checks, high pot testing, prior to testing.

6 And so, we usually have one test a day 7 for that week.

8 So, there will be several rounds of 9 testing that goes on to complete the full series of 10tests which are currently slated at 36 tests. So, 11there will be opportunities in the future to 12potentially have visitors a the KEMA facility to 13 witness testing.

14 We'll coordinate all that under the EPRI 15 MOU or other avenues, I think.

16 And, that is our current time line of 17actions with the Phase 2. I think we covered most of 18 this actually in our morning discussion, but does 19 anyone have any questions of the overall picture of 20 how we plan on performing the next series of testing?

21MR. CHEOK: What do you mean by initial 22 test series?

23MR. MELLY: Initial test series is, 24 that's document and the test plan is our week of 25 244 scoping tests and to ensure that we're collecting the 1 correct information.

2Tony may want to speak to a little bit 3 more of that as to what we envision and why we are 4 calling them scoping tests.

5MR. PUTORTI: Because we're using some 6 new measurement techniques and also because we want 7to make sure that Nick alluded to earlier, that we 8can get the arc to exit the enclosure where the 9 instrumentation is.

10We wanted to just have one week of 11experiments. If you go and you book two weeks in 12 advance, that doesn't really give us -- if we have to 13 change something, it doesn't give us much time to 14 change it and test days are very expensive so that's 15 why we just want to do just one week to start with.

16 MR. FLEISCHER: Kenn Fleischer, EPRI.

17 The small scale testing, it's not 18integrated in this schedule? I thought it was going 19 to be a predecessor.

20MR. MELLY: Yes, the small scale testing 21is separate from the overall OECD arrangements. So, 22 that's an oversight on my part of not putting it in 23 here.24 That testing will fall on the time line, 25 245 I believe it's June 25th that we're planning on doing 1 the small scale testing at Sandia, that's the initial 2 date we will start.

3PARTICIPANT: They're not really tied 4 together, right?

5 MR. MELLY: No, they are not.

6 PARTICIPANT: They're independent?

7MR. TAYLOR: Yes, so, Nick's correct.

8June 25th, they start testing, that's the plan. And, 9we have the final report by the end of September.

10 And then, publication on our end will take some time.

11MR. MELLY: Any other questions on the 12 path forward?

13 (No response) 14MR. MELLY: Okay, I'll jump into the 15 comments and I have a feeling that almost every 16 single comment I'm about to show, we have covered.

17 Do we have any questions from the 18 webinar?19 PARTICIPANT: No.

20 MR. MELLY: Okay.

21 MR. MILLER: Well, that's good.

22 MR. MELLY: I'm hoping.

23So, again, we've seen this slide. We've 24got 91 comments in total. Again, that is through the 25 246 Federal Registry as well as the EPRI MOU, 27 comments 1 were received by EPRI.

2 I tried to breakdown the comments into 3 categories because we did receive a lot of redundant 4 comments as well as I wanted to find a way to cover 5 them here.

6 Some of the comments were on the 7 generator capabilities and the applicability for the 8HEAF testing. So, the decay of the generator itself 9 as we have received a few comments on as well as the 10 2250 MVA rating.

11 We received a lot of comments on the 12 protective relaying and the duration of testing, the 13 equipment ratings and how we're going to be selecting 14 the equipment.

15 The test conditions, the equipment set up 16 combustible load and cable trays.

17 Test parameters such as the voltage 18current grounding team and a lot of comparisons to 19 the IEEE guide for the metal enclosed switch gear 20 rating.21So, some of the main comments that we 22have -- that we haven't talked about, actually, is 23 that we received one comment that the stated 24generator capabilities at KEMA were 2250 MVA. That 25 247 is the maximum capability of KEMA's generator, but 1 that's not what we're going to be testing each piece 2 of equipment to.

3 So, there was a misunderstanding as to 4 how much power we're going to be using for each test.

5 I think that we've cleared that up through our 6discussion this morning. That is just their maximum 7 capability.

8 And, KEMA does essentially dial in the 9 current, the voltage and the duration that we're 10 looking for and we have discussed even the ability to 11have that decay put on to certain testing. So, we've 12 covered that pretty well.

13 We've also discussed the -- how they'd 14 use process of the SuperX excitation to compensate 15 for that decrease in voltage, and they have the 16ability to change that.

So, we're going to be 17 looking into that for future testing.

18 We also are still going to be putting out 19 a request for an example of what that decay looks 20 like.21 So, there were a lot of comments as well 22on the ratings of equipment. And, we've been looking 23 at the apparent power range of the values we selected 24 versus the industry sample averages that Gabe has not 25 248 shown.1 This was part of his evaluation to look 2at the capabilities. We were looking at the apparent 3 averages and the range that we're going to be testing 4 at.5 So, using the above calculation, the 6 square root voltage times current, we calculated the 7 range of applicability for the 4160 as well as the 8 6.9 versus the industry averages that we calculated 9 from looking at plant to plant variability.

10 PARTICIPANT: Sample averages?

11MR. MELLY: Sample averages, this was not 12across the fleet. This is tied to the amount of 13plants that we had access to this data. I believe it 14 was 18 and 23 plants, respectively.

15MR. TAYLOR: We had more information of 16these ratings, so it's -- I don't have the actual 17 numbers.18 MR. MELLY: Right.

19 Well, we can see from this quick look 20 that we were well -- we were inside the potential 21 sample industry averages for what we're going to be 22 testing at.

23 And, there were a lot of comments on the 24 fact that most equipment is equipment is rated to 500 25 249 MVA.1 I think that we've covered this one in 2fairly good detail. There was a lot of discussion of 3 the protective relaying and the potential to test 4 these are going to fault to .5 seconds and 1 second.

5 And, we discussed why we're not going to 6 be doing the testing in that short duration range.

7 And, the ability to separate the arc 8 faults in to bins.

9 We saw those short duration events go 10 into the Bin 15 frequency, not high energy arcing 11faults. This is a typical electrical cabinet fire 12 because you only have that cabinet damage.

13 And, the potential to split Bin 16 into 14 these arc blast categories as well as the full on 15 high energy arcing fault, larger extended damage bin.

16 Again, the frequency of the HEAF events 17 is a current area of work that we're going to be 18working on under the MOU. And, we plant to start 19 that official work, I believe it was stated this 20 summer on our plan.

21 MS. LINDEMAN: Yes, later on.

22MR. MELLY: So, later on this year, we're 23 going to start that work under the MOU.

24 Again, duration, I think we've beat this 25 250over the head quite a few times. So, we're going to 1 be basing that on operating experience as well as the 2plant specific designs. And, their circuit 3 protection scans.

4 That is one area that we want to focus in 5 on under the MOU as well.

6 And, a good place to look at the current 7 work that's been done on that is in one of those EPRI 8 white papers.

9 So, again, this is actually a picture 10from the Fort Calhoun event, the one that did last 11for 42 seconds. Just kind of wanted to show it as an 12 example of these low voltage events that do hold in 13 for an extended period of time.

14 Additionally, that event that I alluded 15to which is the 8 second German arc, and there is a 16 link to where there is more information about that 17 event.18 So, we do see that some of these can hold 19 in and there is an extensive amount of damage and 20 vaporization of the conductive material from one of 21 these events.

22 MR. MILL: Is that an aluminum bus bar?

23 MR. MELLY: That is aluminum bus bar.

24 So, in the Fort Calhoun event, the 25 251 cabinet itself had the aluminum bus bar and 1 conductive material. And, where the event actually 2 occurred was at a transition point where it went from 3 the aluminum bus bar to a copper stab for their 4breaker. And, they were using a silver oxide to --

5 as the transition point for the aluminum to copper.

6So, this is one that we discussed a 7little bit earlier. There were a lot of comments 8 received on the fact that the initial test plant had 9no breakers installed in them due to both pressure, 10 combustible load and things like that.

11 We have made a change and included the 12 breakers, all breakers will be included in the 13 cabinets that we're going to be testing.

14 Again, we talked about the cable trays 15and how we plan on making the jump from 16 instrumentation to what would be there if we tested 17 a cable tray and why we're not.

18 We also received a few comments that we 19 need to do a better job at the internal combustible 20 load, what is in the cabinet, how we load the 21 cabinet, what arrangement the cables are in, whether 22 we are introducing cables or we're using stock cable 23 material that come with the cabinet purchasing.

24 And, we will be documenting all of that, 25 252 including the size, orientation, mass, cable jacket 1material, cable instrumentation material. And, 2 hopefully, we'll be including pictures of the 3 internals of the cabinets with each test prior to the 4 test.5 The last category was the comparisons to 6 the IEEE guide.

7 We've covered that in a fair detail right 8 now, but the main point is that we're not trying to 9replicate the IEEE guide. We have different goals 10 for what we're trying -- the data that we're trying 11 to collect and different objectives.

12 Our objective is not personnel safety in 13 this case, so we're not going to be -- that's one of 14 the reasons where -- why we are not using the cloth 15 targets.16 What you could see from the test videos 17 that Bas showed yesterday is that those cloth pieces 18 of equipment or those cloth markers actually take up 19 a lot of room as well as can potentially inhibit the 20 flow of gases out of the cabinet.

21 So, we're not looking to do that because, 22 if we show that we burn a cloth target, it's not 23 giving us much information at the location they're 24specified in the IEEE guide which is very close to 25 253 the cabinet.

1 I expect that if we run one of these 2 tests for 2 seconds, those cloth targets will be 3 damaged in every single case.

4 Question?5 MR. FLEISCHER: Kenn Fleischer, EPRI.

6 Not so much a question, but this is good 7 info, you know, as we've gone through this workshop, 8 a lot in my mind as to the paradigm we came in with 9 has changed significantly.

10 I think what you're saying would really 11 serve the industry well from a historical perspective 12 is if this is all documented.

13MR. MELLY: So, we do have an official 14comment resolution format. And, I think our plan is 15to make that public. That's an option that we can 16 make it public as we move forward.

17To every comment received, we've 18 identified, if it were -- if it resulted in a change 19 to the test plan, where that change was, how will 20 that comment resolve or potentially an explanation as 21 to why the difference is there.

22MR. FLEISCHER: Yes, and that's the 23 important part.

24 MR. MELLY: Okay.

25 254 MR. FLEISCHER: Yes.

1MR. SALLEY: Nick, we can make that a 2 part of the CP.

3MR. MELLY: We can make that a part of 4 the CP, yes.

5 Again, one thing that we already did 6 discuss in detail was the wire sizes and why the 7 first round of tests was done at the 10 gauge class 8 case standard versus what's specified in the 9 standard.10 We are reaching out for more information 11on that. I think moving forward, unless we receive 12 some information from IEEE which is in conflict of 13 what we're trying to do, we plan on fol lowing the 14 guide and using the 10 gauge for our low voltage and 15 the 24 gauge for our medium voltage.

16 As we identified in the first series of 17 testing, we had no problem initiating and holding an 18 arc for the medium voltage tests where we did see an 19 issue with the low voltage tests even using the 10 20 gauge wire.

21 Another comment was on the arc location 22 and the arc initiation phase angle, how we're 23potentially differing from the IEEE standard. We've 24 discussed that as well as identified the reasons why 25 255 we're going to be potentially differing from that 1 location.2 Primarily so that we can collect the 3 information in a repeatable manner.

4 So, that's very important as to where we 5 put the arc.

6 Again, I pulled this from Gabe's 7discussion. We flip-flopped, so these were 8 everything we planned on talking about this morning 9 that we've already covered.

10 But, that is a good point that we did 11 receive all these comments and we have spent a 12 considerable amount of time going over each comment, 13 discussing with our partners, both NIST, KEMA, the 14 International members as to how to resolve these 15 comments, get clarification on what we need to 16 understand in the comment and how it could 17 potentially change the test plan.

18 So, it is a good recommendation that we 19publish it with this CP. I believe it'll give a 20 historical perspective on how the plan has changed.

21 We won't have Rev numbers, but it'll give an overall 22 picture.23MR. MILLER: It also documents the 24 interaction with the industries.

25 256 MR. MELLY: Yes.

1 Any comments? Questions? Stones?

2MR. GARDOCKI: What's your outcome of 3 this? Are you going to put out a NUREG or --

4 MR. MELLY: For the comments?

5 MR. GARDOCKI: For the tests?

6 MR. MELLY: For the tests?

7 So, there is a plan.

8MR. TAYLOR: You're not going to be able 9 to talk about it?

10MR. MELLY: I can get into it and you can 11 correct me if I --

12 MR. TAYLOR: Okay.

13 MR. MELLY: -- get it wrong.

14 So, the test is going to be covered in 15several different ways. Because this is a 16 complicated project where we are working with the 17 international community as well as we have different 18 phases where we are going to be -- some of the tests 19 are going to be OECD, NEA, some of the tests are 20 going to be NRC exclusive.

21The plan is to publish the OECD tests 22 which we are going to be doing internationally under 23 that collaboration the same way that we did with 24 Phase 1 where we have a Report of Test published in 25 257 the CSI -- CSNI documentation style much like the 1 Phase 1 test with some conclusions of the test 2 program and in that same format.

3 Additionally, this test will include two 4NUREGs. One NUREG including tester report of the 5 aluminum tests as well as the overall picture.

6 And, a NUREG on enhanced methodology.

7 That one's going to be the analysis of all the data 8 and the initial work towards refining the zone of 9 influence, the modeling approaches that we've 10 discussed and things like that.

11We do plan on that's a little bit down 12 the road after the test program and we still need to 13 discuss with EPRI -- it would be beneficial to do it 14 under the MOU so we can get involvement and do the 15 analysis much like the Helen Fire Program and the 16Rachel Fire Program were successful. And, that's a 17 good model to approach this.

18 Additionally, I believe we're going to 19 have a standalone NUREG for the small scale testing.

20 MR. TAYLOR: That's correct, yes.

21MR. MELLY: So, those are the four pieces 22 that are going to come out of this as well as the 23 NUREG/CP which documents this interaction, so five 24NUREGs, one CSNI. And two IAs on the international 25 258 work.1 MS. LINDEMAN: Two more IAs?

2 MR. MELLY: Two more IAs.

3 So, the Japanese have been performing 4 additional tests focused on parameters such as 5pressure. And, we're going to be working under an 6 MOU with the Japanese to publish two additional IAs 7to collect that information. So, we'll have the full 8 leverage of information.

9MS. LINDEMAN: And, I assume you'd have 10 a joint NUREG after your report for the briefing you 11 take for it?

12 MR. MELLY: Yes.

13MS. LINDEMAN: Are we talking about that?

14MR. MELLY: I should have made a slide on 15 this.16MR. AIRD: Nick, we have a question from 17 the webinar.

18 MR. MELLY: Okay.

19 MR. AIRD: It's from Mark Hewitt.

20 His question is, in the documentation of 21 the enclosure, would an assessment be made as to the 22 inclusion of any other aluminum pieces -- piece parts 23 other than the conductor or enclosure duct?

24 For example, it uses relays, control 25 259 cables and the frame structure?

1MR. MELLY: That is a good question and 2 we have started to think about that. And, the 3 questionnaire that went out to NEI was phrased in 4 such a way that if aluminum was part of the 5 conductive pathway that can experience the arc and 6become consumed during the arcing event, that was 7 what we were looking for in terms of, is aluminum in 8 the cabinet?

9 So, the real importance there was, can 10 the aluminum become involved in the event and can we 11 have the additional energy release from the oxidation 12 of that aluminum?

13 So, if we did have aluminum pieces, parts 14 that could be part of that pathway and consumed 15 during the event, we would be -- we are interested in 16 collecting information on that as well as having that 17 part of the test program.

18 That's going to become important for what 19 pieces of equipment we collect or procure and how --

20 where we initiate the arc within the cabinet.

21 So, yes, that is an important part and 22 it's still to be determined based on the procurement 23 of equipment.

24MR. GARDOCKI: Gabe, as far as small 25 260 scale testing, when do you think you're going to get 1 your report done?

2MR. TAYLOR: So, right now, like I said, 3 I write it down to the contractors at the end of 4 September.

5 So, we'll probably get a draft sometime 6late to early -- late July, early August. And then, 7 go through an internal review and then Sandia will 8 respond to those comments and finalize their report.

9 So, a final report from Sandia, I'm 10 expecting by the end of September and then it'll go 11 through the publication process which takes a few 12 months.13 MR. GARDOCKI: Months?

14MR. TAYLOR: Yes, that's been our recent 15 experience.

16MR. GARDOCKI: And, it won't be 17 publically available until next year?

18MR. TAYLOR: That's right, late this year 19 or early next year, depending on the politicians, the 20 publications branches.

21MR. MELLY: It took them six months to 22 publish the PIRT.

23MR. SALLEY: So, you know, did everyone 24 see Stan put on his project man?

25 261 (Laughter) 1 MR. SALLEY: If you didn't, okay.

2MR. TAYLOR: The other option which, you 3 know, this is more for the NRC management type 4 direction is, you know, we could always issue it as 5a Sandia report rather than a NUREG. That would get 6 it out a lot quicker, but it wouldn't have the, you 7 know, the NRC NUREG stamp on it.

8 I'm not quite right now, but that's 9 something we'd probably take back and discuss.

10MR. GARDOCKI: Well, just for the eyes to 11 see the report and the results at a quicker pace 12 before the --

13PARTICIPANT: Will this be released for 14 public comment?

15MR. TAYLOR: We weren't -- so, the 16question is when -- how fast we can get the 17 information out and also will we release it for 18 public comment.

19 For a lot of the testing that we've been 20 doing lately where we've issued it for public -- the 21 test plan for public comment, we haven't been going 22 through a draft version followed by a final version.

23 In a lot of cases, it takes over a year 24 between the draft version getting put out and getting 25 262 the final version.

1 So, we weren't planning on doing a draft 2for public comment on this one. And, Mike Cheok has 3 a comment.

4MR. CHEOK: So, I think, you know, we can 5-- as soon as we approve the reports and your report, 6 we can have a public webinar to announce the summary 7of the results and for results that might influence 8 the next set of tests we do through that publically, 9 we are now in a public meeting.

10MR. TAYLOR: And that's a very good 11 suggestion.

12MR. GARDOCKI: Yes, I was just concerned 13of getting the results of the small scale testing.

14 I'll be fully due to initial testing, in the time 15line of there. But, your initial testing, I've 16 forgotten the date.

17MR. MELLY: It's on the other 18 presentation.

19MR. TAYLOR: I think it's October, right, 20 the first phase?

21 MR. MELLY: Yes.

22MR. GARDOCKI: So, you think we can get 23 something publically out before that October?

24MR. TAYLOR: We can probably have a call 25 263 or a webinar to discuss the results, but the actual 1 get a document out in the public around, I'm not sure 2 we can meet that.

3MR. SALLEY: Gabe, with the schedule 4 here, I understand that Nick's schedule is very 5 aggressive.

6 MR. MELLY: They tend to be.

7MR. SALLEY: Yes, it's extremely 8aggressive. So, please don't hold us to this 9schedule. I mean, with some of it, the challenges 10 we've got with the --

11 You know, we've got to work with the 12lawyers. We've got to get that right and we've got 13to work with our international partners. And, those 14 things all take time.

15 So, we're not going to sacrifice quality 16for speed here. So, you know, we'll slow it down to 17 ensure that we get the quality.

18 So, again, this is very aggressive.

19MR. MELLY: For the first phase, I 20anticipated getting done three tests a day. And, 21 that did not pan out, and 5 tests a day for the low 22 voltage.23 We did not pan out meeting that goal.

24KEMA usually shuts down shop at around 3:00. They're 25 264 on a different schedule than most.

1 We got done one test a day.

2MR. TAYLOR: Yes, but I think, you know, 3 as soon as we can get agreement with the results, we 4 can communicate them in some form to the public and 5 move on and not have to be delayed by the publication 6 schedule.7MR. GARDOCKI: Especially with the 8important players. I know you're going to need the 9 results very quickly, the small scale test, right?

10 MR. PUTORTI: Not to do our part.

11 MR. GARDOCKI: No?

12MR. PUTORTI: You would need the -- I 13 think that you'd need the small scale results for 14 would be where would technologies you're going to 15 take from small scale measurement technologies and 16use them in the larger scale. That's what you would 17 need the results for.

18 But, as far as the results, I'm not sure 19 how much that's going to change the first round of 20 the full scale. I don't see that.

21MR. GARDOCKI: I was thinking more along 22 the lines of coupons and degrees of temperature, 23ranges to measure what the expectation of plasma 24 coming out of this would be.

25 265MR. MELLY: Right. And, we will have --

1 those are developed right now and we will have those 2 available to be putting in the first round of 3testing. We just won't have the results published by 4 the time the initial tests are taking place.

5MR. PUTORTI: All right, but if you're 6 referring to the techniques that Sandia's looking at 7 using for the small scale and bringing those into the 8 large scale, that will take time for them to field.

9 But, it won't change the rest of the large scale 10 plan.11MR. MELLY: For instance, those coupons 12that Gabe was referring to to collect the eject of 13 the aluminum, they are developed and available now so 14 we can still use them in large scale testing.

15 Whereas, the relevance and whether we 16 should be investing our time and energy is going to 17 come out of the results, that's where the main trust 18 and the main resource drag is for the small scale.

19 But, we can easily put them in the 20 cabinet right now, shelve them and wait to see if 21 they're worthwhile to be tested.

22MR. FUNK: This is not the first time we 23ran into this issue. So, what we did in the past, I 24 know, Gabe, you're familiar with this, once the NUREG 25 266 report has gone through tech review, and we're really 1confident there's not going to be any technical 2 changes, that's when we started using it.

3 Can a target date be established for tech 4review? Because I believe that would be within your 5 control where publication will not necessarily so.

6MR. TAYLOR: I think that's reasonable.

7 It's really going to depend on the quality of the 8 first draft that we get from Sandia and how much, you 9 know, feedback or changes the NRC reviewers see from 10 that document.

11 And, but, if there's not much change and 12 we don't see technical changes from the first draft, 13 they may well make quick changes to address those 14comments. And then, at that point, we could get 15agreement within the NRC to, you know, issue the 16 results or have the discussion of the results at that 17 point.18 So, I think that's the reasonable 19 approach.20MR. MELLY: Yes, and again, to keep in 21 mind that the small scale testing is going to be much 22 in the style of a report of tests where there will 23 not be any methodology changes proposed in that small 24 scale testing.

25 267MR. TAYLOR: Yes, they're not going to 1 try to, you know, do some type of modeling of the 2 results to extrapolate different variables and how it 3 affects particle sizes and whatever the other aspects 4 of the testing are.

5 It's just you test at these variables, 6this is the size distribution, this is the fraction 7 of oxidation, whatever they need to report.

8MR. MELLY: Any other questions, time 9 line, schedule, outcomes?

10MR. PUTORTI: Just one small detail, we 11 discussed this earlier in the day.

12But, when we decide or when we get the 13 enclosures that we're going to be testing, the 14 question out there about if they do arrive with bus 15 bars that are insulated, do we want to leave them in 16 that state or remove the insulation?

17MR. MELLY: I think that is going to 18 depend highly on some of the things we have discussed 19 here today.

20 For the internal or for the cabinets, the 21enclosures themselves, I think that we will most 22 likely be -- it depends on how we procure them, but 23 I see that typically they're not going to -- no 24insulation for an enclosure cabinet, we may be 25 268 removing.1 If they're for the bus bar or for the bus 2 ducts, I think the decision was to test half and half 3 with insulation and non-insulation.

4MR. TAYLOR: I wouldn't see us going out 5and making a change to equipment. I think we'd want 6 to procure it.

7MR. MELLY: Yes. Now, one question that 8 I think we may have missed in the discussion earlier, 9 we were talking about procuring these cabinets and 10 the initial thought was that if we're -- let's say, 11 we're running 20 tests, 10 at copper, 10 of them at 12 aluminum.13 If we can procure both copper and 14 aluminum from the same manufacturer, we will do that 15 and the bus bar spacing may be a little different.

16 However, if we can only procure aluminum 17 or a copper conductor, are we going to fabricate the 18 aluminum equivalent for the other test series or for 19 the other tests that we're going to run and what 20 factors would be important if we go down that route.

21 Because, our electrical contractor right 22 now does have the ability to wholesale and replace 23 all the copper conductors inside the cabinet with 24aluminum conductor material. I know that may alter 25 269 how the design is working based on the spacing of the 1 conductors internal to the cabinet.

2 Are there anything -- any concerns or 3 approaches that anyone has in the room that we should 4 keep in mind of if that's the route that we have to 5 go?6MR. SALLEY: You're driving into the 7 heart of Ken's discussion about the we won't buy the 8 equipment starting out.

9 MR. FLEISCHER: No.

10 MR. MELLY: Yes.

11MR. SALLEY: And, we can't take that.

12 And, that's was -- Ken, that's what I got from your 13 whole discussion on that.

14 And, we need to start out with something 15 that is certified.

16MR. FLEISCHER: Right. Well, that 17 discussion -- Ken Fleischer, EPRI.

18 That discussion was based on the premise 19 of my paradigm coming into this workshop was that we 20 weren't getting donations, I didn't know about 21 donations and we would procure the equipment.

22 And, therefore, we would have total 23 control over configuration.

24 In this particular case, I guess my 25 270 thought is, is why don't we just review what is our 1end game in mind? Our end game in mind is to see if 2 the zones of influence are affected by the additional 3 vaporization or oxidation of aluminum.

4 If our target is to set off or find out 5 what kind of switch gear is going to do that to set 6 off the equivalency between copper and an aluminum 7 switch gear configurations may not be as important.

8 But, I think we should strive to try and 9 keep some similarity with what we've got and minimize 10 modifications.

11 I think if we start modifying things, 12 then we have thrown in a new variable that we may not 13 have fully thought through and have unintended 14 consequences.

15 MR. MELLY: I agree.

16 And, on the flip side of that is that 17 because that is one of the main parameters that we're 18 looking at is the copper versus the aluminum, and the 19 delta of the damage states between them, trying to 20minimize the amount of variables that we have 21 different in that case is very important.

22 So, for instance, if I buy all aluminum 23 GE Magne-Blast and all copper Westinghouse, that's a 24 huge difference immediately as to the potential 25 271 damage state.

1 So, it -- whereas, I may be qualified 2 aluminum for the Westinghouse and qualified copper 3 for the GE, those are two immediately different 4 cabinets which may have distinct zone of in fluences 5 just on the cabinet design.

6 I wanted to try and limit the variables 7 and I would love to find a manufacturer that can spec 8 the copper conductors and the aluminum conductors in 9the same box. I just am not too familiar with 10 manufacturing processes or the catalogue availability 11 of this.12 MR. MILLER: This is Kenn Miller.

13I agree with you. I mean, if a 14 manufacturer can sell above waste of waste --

15 MR. MELLY: Right.

16MR. MILLER: -- that gets over this issue 17of making a new design yourself. It's not, quote, 18 certified.

19 MR. MELLY: Exactly.

20MR. MILLER: Whether that's true on that.

21 MR. FLEISCHER: Yes.

22MR. MILLER: You should certainly start 23 with that if possible.

24MR. FLEISCHER: Ken Fleischer from EPRI 25 272 again.1 Yes, yes, I'd buy that that's what would 2be preferred. I guess maybe the bigger question, 3 because I'm still not sure, are we looking to try and 4 do this testing program with 100 percent donated 5 equipment?

6 Or, are we looking at donated equipment 7 and then maybe with what we're going to get and we 8 find out what we have, we can contact, let's say, the 9 majority is going to be the GE Magne-Blast copper.

10 Can we contact GE to maybe replicate a unit in 11 aluminum?12 MR. MELLY: That would be the ideal.

13 So, to answer the first question, no, 14we're not looking to have the majority of the test 15program on donated equipment. We're going to be 16 using donated equipment for several of the countries 17who are not unable to provide monetary donations to 18 be all inclusive to our international members who are 19 interested in this work.

20 The rest of the equipment will be 21 purchased and procured through several of the 22 contracts that we have in place right now.

23MR. FLEISCHER: Okay. Well, then that 24sounds good. And, I think we can be smart about it 25 273 by looking to see what we have.

1 MR. MELLY: Okay.

2MR. SALLEY: Or, as we -- or, Ken, as we 3 discussed today, if we have some, you know, U.S.

4 participation, well, some of that equipment we would 5 be more than happy to work with you.

6 MR. MELLY: Yes.

7MR. SALLEY: So, again, if we're looking 8at some of the plants that are closed up or -- I 9 talked to Brenda Simril a little bit, TVA's cleaning 10out some warehouses, looking at some surplus. I know 11there's a lot of stuff out there, there's more out 12 there and it would be very open to working with the 13 industry on some of this.

14MR. MELLY: Yes, and if we do receive 15 more donated equipment, that is a good suggestion 16 that, whatever we do receive in donations, it 17 actually allows us to test more equipment because 18 that's large cost driver for the test program, but we 19 can reach out and see if there is an equivalent 20 design or specification for how the internals need to 21 be swapped out if we're going to be changing the 22 conducting material.

23 MR. MILLER: Kenn Miller, NRC.

24 So, that is captured, though, what we try 25 274 and when we try and go out and buy the additional, we 1 will look for that option of getting the same design 2 with both types of material.

3 MR. MELLY: Yes.

4 Sorry, I went off on a tangent there.

5MR. TAYLOR: Anything else then on the 6 test plan?

7MR. MELLY: All right, is there anything 8 else on the test plan, either webinar or in the room?

9 Anything we think we haven't covered or 10 something you had in mind prior to the meeting that 11 they were looking for a discussion on?

12MR. TAYLOR: I think there was one follow 13up question from EPRI. I think Ken, he went back and 14 looked at the fault monitoring equipment for the one 15 event he talked about earlier.

16 Did you want to add anything on that?

17 MR. FLEISCHER: Ken Fleischer, EPRI.

18 Yes, I need to go back and go through the 19data in detail. So, I can put someth ing together 20 concise, but I have very high resolution digital 21 fault recorder raw data that we can pull together.

22 It's for a 6 second -- it was not a --

23 well, I don't know, I mean, it was an arc but it was 24 an arc on the high side of the bushings of the main 25 275 power transformer.

1 And, I have the high data resolution just 2 to give you a quick summary of what it looks like, it 3 looks like in the first 26 cycles, it drops to the 4pre-position point for the voltage regulator. So, it 5 goes from about 22 kV down to 17kV for about 26 6cycles and it holds steady. It's a very short 7 duration.8 My belief is that that's where the 9 exciter switch gear breaker is still in service 10 actually exciting the rotor.

11 Then, after that, you see a quick drop to 12 about 50 -- so, that was about a 20 percent drop.

13 So, then you see another 30 percent drop 14down to about 50 percent. And, I believe that is 15 when the exciter field breaker now pops open so now 16 the field is no longer excited and it's residual 17 magnetism.

18 And, yes, it slowly decays from there 19just as I would expect. It's a, you know, an 20exponential decay. It's not as fast as I thought and 21it continues well beyond 6 seconds. But, the final 22 arc finally extinguishes at about 6 seconds.

23 What I can do is compile that together 24and maybe like a small mini white paper. I could 25 276 provide that back to the group and we can send it to 1KEMA to see if they can replicate that decrement 2 curve.3 Now, the thing is, is that it doesn't 4 have to be an exact replicate, but I can give them 5 the raw data, they can get as close as they want.

6 But, it's definitely clear, there's a lot less energy 7 coming out of that generator than if you were 8 infinite bus through a transformer.

9MR. MELLY: So, one of the things that 10would be a potential work around as well, if KEMA 11 cannot replicate that decrement curve, there is the 12 potential to calculate the energy release during that 13 event and align the duration to the equivalent total 14 amount of energy release.

15MR. SALLEY: Ken, yes, that's a great 16idea. And, if we can do that, you know, this CP that 17 we're talking about isn't going to happen overnight.

18 So, if we can work with EPRI and do that, we'd be 19 happy to include that and document that in here.

20 So, we're really open to that.

21MR. TAYLOR: Any other follow up from 22 this morning's discussion or this afternoon's 23discussion on the test plan? Are there any, 24 especially on the test plan, are there any comments 25 277 that Nick really didn't cover and you need a little 1 more information or background on the NRC's initial 2 disposition of the comment?

3 MR. MILLER: This is Kenn Miller, NRC.

4 I will do something probably with the 5 HEAF definition and adding back in some time data 6 based on the 2 and 3 seconds, based on the switch 7 gear and the breaker ratings.

8 And, maybe we'll put that out to --

9 that's the only thing I'm th inking of in terms of 10definition changes that we've talked about. But I'll 11 put that out there and we can look at it.

12 Unless anybody else had anything else on 13 definitions?

14 (No response) 15MR. TAYLOR: Just getting back to, if 16 there's nothing else, so, as far as the generic issue 17 program discussion, I think it's worthwhile on our 18 end to go back and formalize the milestones and 19 dependencies so that when we do get asked questions, 20 we have a more formalized and clearer response to 21 that and direction on of how and when these different 22 milestones are going to be completed.

23 On the PRA realism, I believe we need to 24 continue working with EPRI on the frequency and the 25 278classification as well as these definitions. I think 1 it's going to be very important especially from a PRA 2 model to make sure everything aligns.

3 And, I should probably also put up here 4the very important thing on the pilot plants. And 5 so, we'd like to identify some pilot plants for the 6 GI program, potentially go out to the pilot plants 7 and understand what the risk is for these HEAF 8 scenarios with aluminum.

9 So, identifying the right pilot plants.

10 Nick talked a lot about this about, you know, if they 11 already assumed top isolator, everything in the 12 room's lost, well, you increase ZOI, it's really not 13 going to have that big of a delta risk for those 14 scenarios.

15 But, there could be other modeling or 16 compartment configurations where an increased ZOI may 17 have an impact.

18 So, understanding what the risk change is 19in the plant for these events is important. So, I'd 20 look to EPRI and possibly NEI to help us out to 21 identify the pilot plants and support that 22 understanding.

23 So, I think that's very important in how 24we go about that. We'd have to work with the Stan 25 279and the GI program and include everybody that's in 1 that program working on it.

2 Also, we had some discussion of going 3back and understanding what tests IEEE and NFPA has 4 performed for their research and make sure that we're 5 not being repetitive and understand what came out of 6 that work. We have that action item.

7We talked about fault currents. So, for 8 the personal safety, I think there's a lot of 9 information that the plants have already done in that 10 analysis. And, this gets into the fault currents that 11 we're proposing for the test plan and the sample data 12 that I pulled up, you know, is that representative of 13 what we've seen in the personal safety assessment?

14 So, that's one for industry to go back 15 and look in and just verify what we're doing, if it's 16 in line with those studies.

17 Also, on sample circuit characteristics, 18 this is one that Dan and I identified if there's any 19 characteristics that are important for the tests 20 limiting X over R or anything else that we would want 21 how KEMA sets up their lab, we'd like that feedback.

22 We talked a little about the arc 23locations. Really need to get a clearer 24understanding of where we should put the arcs. So, 25 280 I think we have the three, at least for the 1 enclosures, the three locations where we potentially 2 put it.3 But, we haven't come up with a definite 4road we want to go there. So, any feedback on that 5 one would be important for us.

6 And, we agree, the equipment document 7 when we get, we need to understand what the, you 8 know, the manuals are for that equipment and document 9 it in the report, this is what we got, this is its 10rating, this is how we tested it, pretty 11 straightforward.

12 A more recent discussion, manufacturer 13 design verification for the aluminum versus copper.

14Once we identify what equipment we want to procure, 15 we have to reach out to the vendors and say, hey, we 16 will bind this in copper.

17 If you're going to have aluminum or maybe 18 they have aluminum in that product line, how can we 19 configure it or could you maybe point us in the right 20 direction to procure that equipment if we can't find 21 it.22 So, bringing the manufacturers in to make 23 sure that we can test real representative equipment, 24it's important. That's on the NRC side to perform 25 281 and go through procurement.

1 And, the last one I have here is I put 2EPRI on it, but the decrement curve. If we can get 3 some information on that, we can work with KEMA and 4 go from there for that.

5 So, I think that's the majority of it.

6 I don't think we have too much traction on the 7verification tests. So, right now, I don't see that 8as being forward on our end. If there's discussion 9 otherwise, we can bring that up now.

10 And, the other thing that we haven't 11really resolved is the grounding. So, you know, 12 should we just move to ungrounded, resistance 13grounded? I don't think we can come to a conclusion 14 here on that.

15 MR. MELLY: No, we have not.

16MR. TAYLOR: We did rank it the day 17 before, but I don't remember what that was.

18MR. MELLY: We ranked the importance 19 during -- we ranked it as low here with a few highs.

20 A high left the room.

21And, we had it ranked in the PIRT as 22 well, so we'll take that into consideration moving 23forward. And, I'll discuss it additionally with the 24 international members next week.

25 282MR. TAYLOR: So, I think that's kind of 1 the action items we have coming out of this meeting.

2 Is there anything else that I missed?

3 MR. MILLER: Kenn Miller, NRC.

4 One thing on the equipment documentation, 5I know you mentioned the equipment is being donated 6 by the Koreans and Germans.

7 Depending on who you're getting equipment 8-- who we're buying stuff from, if it's secondhand 9 refurbishment commonplace, they may have access to 10 documents that we would need for the other items that 11 are donated, assuming we get the information on what 12 exactly they are donating.

13 So, there may be ways to get 14 documentation through that same contract that we're 15 buying the equipment from to help us with this part 16 of the deal.

17 Unless, I mean, we may get stuff from the 18 people that are donating it, but if not, that may be 19 an option to get it.

20 MR. MELLY: That's a good point.

21 MR. TAYLOR: That is a good point.

22MR. MELLY: In terms of output from this 23meeting, what people can expect, we do plan on 24 issuing the NUREG/CP which will document all of the 25 283 presentations that have been given, some of the notes 1that we've taken during the meeting. That's a little 2 bit longer term.

3 Do we plan on putting out a meeting 4 summary?5MR. TAYLOR: So, this is a public 6 meeting, there will be a meeting summary, short two-7 page meeting summary of what took place. But, it 8 won't be as detailed as the CP that we'll put out in 9 the near future.

10 Questions in the back?

11MR. FUNK: Kenn, I think this question is 12 directed to you.

13 Yesterday, you mentioned that there is 14 going to be an interest in looking at this testing 15 and the results from a more traditional Class 1-E 16 general design criteria perspective with regard to 17 the generic issue.

18 What about your fire guys, for all 19 practical purposes and so you can see that's the 20 focus of everything here.

21 So, with regard to the pilot plants, what 22are your expectations or what do you see their 23involvement as it relates to the generic issue from 24 the perspective of single failure and GDC-3 25 284 explosion?

1 They might not approach it, I mean, just 2 as good to go forward with the blinders on if fire 3 PRA unless we know there is some difference.

4MR. MILLER: I'm not really sure how to 5answer that, to be honest with you. Because I'm 6 still a little bit fuzzy on how this plays over on 7 the GDC side because of the single failure criteria.

8 And, by definition, these HEAF events are 9 indicative of a fault and then failed protection 10 which goes beyond single failure.

11 So, I'm not really sure how to can that 12yet. But, I guess my thought was at least being 13 aware of how this plays out and see if there's any 14issues with, for instance, some of the criteria we 15 have for physical separation between safety trains 16 and whether or not an event like this, you know, has 17 the potential to affect and alternate train.

18 So, I'm not really sure yet how that 19plays. Just know that, you know, on the surface it's 20we've got to think about independence of safety 21 trains and how this could affect that.

22MR. FUNK: I understand. What is the 23 inherent nature of doing the deterministic safety 24analysis, we know that fire can affect multiple 25 285trains, you know, as a common cause. But, the 1 Appendix R analysis --

2 MR. MILLER: Yes, fire --

3MR. FUNK: -- deals in that so, it 4 depends on while we're looking at this as this is a 5fire event or something else? Because that 6 determines whether it's a single failure impact or 7 not.8 MR. MILLER: Well, in GDC specs --

9MR. FUNK: Under the design basis 10 criteria.11MR. MILLER: In GDC specs, it's a single 12failure. I mean, that's what you talk about, you 13 know, the safety systems, what your dual train and 14 you have some fill there of safety function assuming 15 it's single failure.

16 So, I'm not sure.

17MR. FUNK: Okay. There again, the safe 18 shutdown analysis, if you separate -- and the single 19 rule specifically states you do not assume single 20 failure so, that's a rimless oxymoron.

21MR. MILLER: Right, that's -- that's kind 22 of a separate thing.

23 MR. FUNK: I've reviewed this.

24MR. MILLER: That's a separate rule. But, 25 286 yes, I'm not sure, I'm not sure.

1 MR. FUNK: I understand, sir. Thanks.

2MR. MILLER: That will be an interesting 3 observance when it comes out. And, I know that's 4 what Bob was thinking also.

5MR. GARDOCKI: To follow up with your 6 saying that, there are some mitigated measures that 7we explored during our committees. The heat shield 8was one that we kind of touched on but we didn't go 9 into depth on and I don't think industry is really, 10 well, explored that and come up with anything that's 11 qualified on how this is going to affect that type of 12 mitigated measures.

13 So, this is really going to affect some 14 of those mitigated measures.

15 There is some photo detecting, you know, 16 light sensitive tripping breakers that the industry 17 is exploring right now.

18 I'm not sure what stage industry is 19 taking credit for those as mitigated measures for 20 this kind of arc flash and tripping the breaker 21 within microseconds.

22 So, those are some exploratory mitigating 23 measures out there that we don't know exactly where 24industry stands. I don't know if anyone here can 25 287 touch on that.

1 MR. MILLER: Kenn Miller.

2 I know one thing I would say, based on 3 looking at this type of event, certainly preventative 4 measures is something I would be looking at from GDC 5 deterministic perspective, obviously, that your 6 protected schemes are proper in terms of clearing 7 faults as quick as possible within the 2 second limit 8 of the equipment which should preclude this type of 9 event in the first place.

10 So, I mean, from -- so, from that side of 11 the house, I would -- that's something I'm hoping to 12 see out of, you know, reaction to the industry and 13 they go back and look at their preventative measures 14 and getting in a position where this kind of event 15 can't occur.

16 So, that's an important lesson here, I 17 think, just because the potential of this event to do 18 damage.19MR. TAYLOR: Anything else? We're getting 20-- I think we're starting to wind down here.

21 Anything from today's or even yesterday's 22 discussion?

23 (No response) 24MR. TAYLOR: What a bout on the webinar, 25 288 Tom, do we have anything coming in through the 1 webinar?2MR. AIRD: No public comment on the 3 webinar.4MR. TAYLOR: We do have to open up for 5 public comment before we close.

6MR. AIRD: All right, let's do that right 7 now.8MR. TAYLOR: Okay, so, for those on the 9 phone line, I'm going to have Tom open up the phone 10lines which will take a few seconds here. And, if 11you do have any comments, please identify -- unmute 12 your phone, identify yourself and then make your 13 comment.14 So, I'll wait for Tom to give me the nod 15 that the phone lines have been unmuted.

16 Okay, the phone lines are open, so if you 17 have a comment, please identify yourself and make a 18 comment.19 Okay, so I don't hear any comments from 20the public. Okay, no comments during the public 21 comment period.

22 With that, we'll go ahead and move to 23 Mark Salley for some closing remarks.

24MR. SALLEY: Yes, and then, Mike Cheok is 25 289going to make a couple closing remarks here at the 1 end.2 But, thank you all for attending. Gabe 3did a good job covering the actions. This is a 4 couple I just wanted to touch on here.

5 Actually, working under the MOU, I'm 6 sorry, I just wanted -- yes, work on the MOU, two 7 things I just want to touch on.

8 Again, you saw the KEMA offered on the 9 field trip if that makes it more understandable, the 10 generator, the facility and that, please talk with 11NEI. If we have anything there, we'll be happy to 12 set something up.

13 And then, do a -- meet you up there for 14 a day trip or something along that line.

15 The other thing is, Gabe, again, touched 16 on this, too, and it's the risk analysis nuclear 17pilot plants. Again, that's an area where we could 18 work with you to get that and move the generic issue 19 along, we would like to work with EPRI on that one.

20MS. LINDEMAN: We would need to 21 understand your expectations for a pilot plant as 22 well.23MR. MELLY: We'd need a follow up meeting 24 to be sure on specifically the pilot plant questions 25 290 and actions.

1MS. LINDEMAN: Do you want me to repeat?

2 MR. SALLEY: Sure.

3MS. LINDEMAN: Yes, I think just 4 clarifying the timing of the pilot plant just seemed 5that was uncertain. And, where the intern ZOI fit in 6 to all of this, that and then, just what your 7 expectations are.

8 As you saw, there's a wide variety and 9 just tell us what you need.

10MR. SALLEY: Okay. And, working on the 11 MOU I can get with you and we can talk about that.

12 And, let's see, that's the only thing 13 I've got, so I will give it to Mr. Cheok.

14MR. CHEOK: So, thank you again for 15coming out. I know two days is a lot. And, thank 16 you all the people on the webinar and the public 17 phone lines, two days is a lot of time to spend with 18us and that's not even counting travel time. So, 19 thank you.

20 And, hopefully, you'll have learned 21 something about our process and understand it a 22 little bit more.

23 I know, I sit in here and I understand 24 your comments and your questions and concerns a 25 291little -- a lot better. So, I'm glad I showed up and 1 we had this thing.

2 And, I apologize to those people out in 3their presentations yesterday afternoon. I fully 4 intended to be here, but I made a mistake of opening 5 some emails at lunchtime. So, that kind of affected 6 my afternoon.

7So, again, thank you. But, I want to 8mention something else, too. So, the generic issue 9 process and everything else involves more than just 10 the Office of Research.

11 NRR is also a player there, you know, our 12counterparts they are totally involved in this. And, 13 I know my counterpart, Mike Franovich has been 14 following this here with EPRI now and I've had a 15 couple of emails from him, you know, actually making 16 comments and asking questions.

17So, he has been listening in. He's not 18 been here, but so we have been working fully with NRR 19 and they are on board and we will continue to have a 20 panel basically with NRR and Research as we go 21 forward.22 Again, thank you.

23MR. TAYLOR: Okay, with that, that's the 24 end.25 292MR. SALLEY: Oh, one other thing. I'm 1 sorry.2 MR. TAYLOR: One more.

3MR. SALLEY: Before Mark calls, is Kenny 4 Hamburger not here, he's in Paris, you know, he did 5 a lot of work to set this meeting up, do all the 6presentation material put together. So, I think we 7 should thank Kenny for doing all that.

8MR. MELLY: He's potentially still on the 9 phone still working at 11:00 Parisian time.

10MR. SALLEY: Mark Earley turned around 11 and he reminded me of something.

12 So, Ashley one other thing with the 13 definitions, NFPA and Research have been working and 14 we've been having a few webinars, exchanging some 15 ideas.16 Mark has a very impressive thing that 17 we've been talking to and we've been talking to and 18 we've been learning a lot about the work they've 19 done.20If you want to dial in with us on the 21 definitions, Mark, if that'd be okay, we'd get EPRI, 22 you know, you and Kenn work with us, we'd really 23appreciate that. And, I think, again, with that 24 collaboration, I think we could really get some high 25 293 quality work turned around and a lot of work.

1 Okay, with that, we'll go off the record 2 and we're done.

3 Thank you very much.

4 (Whereupon, the above-entitled matter 5 went off the record at 4:52 p.m.)

6 7 8

9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25