ML20210D377
| ML20210D377 | |
| Person / Time | |
|---|---|
| Issue date: | 05/06/1987 |
| From: | Advisory Committee on Reactor Safeguards |
| To: | |
| References | |
| ACRS-T-1584, NUDOCS 8705080162 | |
| Download: ML20210D377 (245) | |
Text
AWW IS~g$6 ORgyp(
' UNITED STATES O NUCLEAR REGULATORY COMMISSION IN THE MATTER OF: DOCKET NO:
ADVISORY COMMITTEE ON REACTOR SAFEGUARDS SUBCOMMITTEE ON AC/DC POWER SYSTEMS RELIABILITY t
O -
LOCATION: WASHINGTON, D. C. PAGES: 1 - 180 __ _
DATE: WEDNESDAY, MAY 6, 1987 ERSOFFICECOPY
.:o1041empvgfrom.ACRS 05ce ACE-FEDERAL REPORTERS, INC.
Official Reporters sO 444 North Capitol Street Washington, D.C. 20001 g- o soeo 16 2 970506 P Drl g,[R S POR NATIONWIDE COVERAGE
() PUBLIC NOTICE BY THE UNITED STATES NUCLEAR REGULATORY COMMISSIONERS' ADVISORY COMMITTEE ON REACTOR SAFEGUARDS i
WEDNESDAY, MAY 6, 1987 The contents of this stenographic transcript of the proceedings of the United States Nuclear Regulatory Commission's Advisory Committee on Reactor Safeguards (ACRS), as reported herein, is an uncorrected record of the discussions recorded at the meeting held on the above date.
I No member of the ACRS Staff and no participant at e
() thi's meeting accepts any responsibility for errors or inaccuracies of statement or data contained in this transcript.
e l
C:)
4 4
- w r .- r e,- -m - - - - -- - - - - - - - - - , - - - - ,-,,w w., - - - - - - - - , - ~ . -------,,m.,v,,--.-. - - . - , ,_. - - . - . - - - --,,vs .-i--e--.
1 CR30853.0 COX/sjg 1 UNITED STATES OF AMERICA 2 NUCLEAR REGULATORY COMMISSION
^ ^
3 SUBCOMMITTEE ON 4
AC/DC POWER SYSTEMS RELIABILITY 5
6 Nuclear Regulatory Commission Room 1046 1717 H Street, N.W.
Washington, D. C.
8 Wednesday, May 6, 1987 9
10 The subcommittee meeting convened at 8:30 a.m.,
Mr. Charles J. Wylie, chairman of the subcommittee, presiding.
12 ACRS MEMBERS PRESENT:
13 MR. CHARLES J. WYLIE 14 MR. JESSE C. EBERSOLE 15 MR. CARLYLE MICHELSON 16 MR. GLENN A. REED 17 18 19 20 21 22
~
23 24 25 ACE. FEDERAL REPORTERS, INC.
202-347-37(W) Nationwide Coserage Nn336-6M6
l
~30853.0 cox 2 1 PROCEEDINGS O 2 MR. WYLIE: The meetipg will now come to order.
3 This is a meeting of the Advisory Committee on Reactor 4 Safeguards Subcommittee on AC/DC Power Systems Reliability.
5 I am Charlie Wiley, the subcommittee chairman.
6 The other ACRS members present today are:
7 Mr. Jesse Ebersole, Carlyle Michelson and Glenn Reed.
8 The purpose of the meeting is to review the 9 proposed Station Blackout Rule 10 CFR 50.63.
10 Dr. El-Zeftawy is the cognizant ACRS Staff member 11 for this meeting.
12 The rules for participation in today's meeting 13 have been announced as part of the notice of this meeting 14 previously published in the Federal Register on April 16, 15 1987.
16 A transcript of the open portion of the meeting is i
17 being kept and will be made available as stated in the 18 Federal Register notice. It is requested that each speaker 19 first identify himself or herself and speak with sufficient 20 clarity and volume so that he or she can be readily heard.
, 21 We have received requests from NUMARC, the Nuclear 22 Utility Management Resource Committee, to make a brief 23 statement, which we will accommodate at the end of today's
() 24 meeting. We have received no other statements from the ,
25 members of the public or requests.
ACE FEDERAL REPORTERS, INC.
202-347 3700 Nationwide Coserage W)-336-6646
i 30853.0 cox 3 1 I would like to make a brief statement and call O 2 another member of the subcommittee for comments.
3 The Staff is proposing to the Commission to 4 proceed with final resolution of USI A-44 Station Blackout by 5 amending this regulation to require that all nuclear power 6 plants be capable of coping with a station blackout of some 7 specified duration. The Staff has prepared a draft 8 regulatory guide entitled " Station Blackout" which presents a 9 method acceptable to the Staff for selecting the specified 10 duration on a plant specific basis.
11 The ACRS has previously reviewed the proposed 12 resolution to USI A-44 on two previous occasions, and wrote 13 letters to the EDO, generally supportive, with comments. I 14 believe the members of the subcommittee have copies of those 15 letters. I would just like to mention a couple of points 16 made in those letters, so we keep those in mind in today's 17 proceedings.
18 In the letter to the EDO on July 13, '83, the ACRS 19 made a statement that we believe that this proposal, for the 20 resolution of USI A-44, would result in significant 21 improvements, only for those plants that have an unusual 22 vulnerability to station blackout.
23 Also, in commenting further in that letter, made a
(]) 24 statement that actions considered should be closely l 25 coordinated with recommendations which may emerge from the
/(CE FEDERAL REPORTERS, INC.
202-347-3700 Nationwide coserage 800-336-6646
30853.0 cox 4 1 ongoing work on resolving USI A-45. Consideration should O 2 also be given to actions taken in this area by fo'eign r 3 regulatory authorities.
4 We again, in another paragraph, emphasized that .
5 coordination with USI A-45 should be done and stated that 6 further we urge that the resolution of USI A-45 be 7 accelerated and also urged that the resolution of A-44 be 8 coordinated with generic safety issues. B-56, diesel 9 reliability, and third safety adequacy of DC related power 10 supplies.
11 In the March 12, '85, letter, we stated we 12 recommend that effort be made to seek a method of G(_/ 13 implementing these recommendations that will require a 14 minimum of Staff and licensee resources. If a rule is 15 adopted, we recommend that implementing guides permit 16 analysis appropriate to each plant, and that the process be 17 kept to the minimum necessary for the appropriate resolution 18 of USI A-44. If a better alternative than rulemaking is 19 advanced, we recommend that it be given serious 20 consideration.
21 As mentioned in our letter of July 13, 1983, we 22 urge that actions considered in the resolution of this issue 23 be closely coordinated with recommendations which may emerge
() 24 from the ongoing work on resolving USI A-45.
25 I thought that we should keep those points in mind ACE-FEDERAL REPORTERS, INC.
202-347-3700 Nationwide Cmerage 800-336-6M6
4 1
30853.0
- cox 5 1 as we consider the issue today. On Friday of this week, from O 2 3:45 to 5:45, we have this issue scheduled for the full l 3 committee. At that meeting, I understand that NUMARC has 4 requested to make a short presentation.
. 5 I would request the subcommittee members to 6 consider what we should bring before the full committee on i
7 Friday, and what recommendations, if any, the subcommittee 8 wishes to make.
9 Are there any comments by subcommittee members?
10 Mr. Reed.
11 MR. REED: I am glad you mentioned A-45 and that i
12 work and how it might impact the station blackout problem. I 13 personally have a problem with what I call a piecemeal 14 approach to solving the most serious issue in nuclear power 15 generation, which I will call decay heat removal.
- 16 Now, here, I think we are into a piecemeal i
3 17 approach, where we are going to think about doing something 18 to the AC/DC reliability, when, again, if we were to be 19 talking an appropriate decay heat removal backup system, we t
20 wouldn't have to be talking about present AC/DC reliability.
j 21 I will have more to say later.
]
22 MR. WYLIE: Thank you, i'
23 Any other comments?
Q 24 MR. EBERSOLE: Yes, Charlie. I would just like to 25 comment about the matter of what we are talking about when we ACE. FEDERAL REPORTERS. INC.
1, 02-347-3700 Nationwide Coverage 8(G336-6M6 l
30853.0 cox 6 1 are talking about AC power supplies. In general, through 2 this document, it refers to the off-site power supplies, and 3 then the on-site emergency diesel generators.
4 I think we need to make a sharper definition of 5 what we mean by a loss of AC power. In part, this is done in 6 this document in related papers, by recognizing that the DC 7 driven inverters are also AC power supplies, and they are 8 excluded from this set, when one looks at failures. There is 9 another system, which is in the boilers which is the core 10 spray system, which is a dedicated diesel which goes to or 11 should go to isolated bus work which in turn feeds the spray 12 pump.
13 It seems to me that one could and should sharply 14 define what is meant by failure of AC systems when you are 15 talking about AC failure in toto and be able to exclude 16 failure from systems which are totally independent with 17 respect to bus work, the character of the engine driven power 18 supplies, whatever. They are not part of the family of the 19 AC systems we are talking about. I think we ought to be much 20 more explicit in saying what is total AC failure and what is 21 not.
22 AC failure, for the moment, I don't think it's 23 very clear here, in this paperwork here, that the power
() 24 supplies to the boiler spray system is excludable, and it may l 25 or may not be, depending on the amount of bus work, which is Acn FEDERAL REPORTERS, INC.
202-347-37(0 Nationwide Coserage 8533MM6
30853.0 cox 7 1 an AE function, not just a vendor function.
4 O 2 Another omission from this paper.here. They have 3 not recognized that, in particular, just the boiler, the BWR, 4 has a way to safety through boiling, which could be done with 5 nothing more than engine and impulse-driven instrumentation, 6 with no electric power at all.
7 It is exclusively possible with that design to 8 cool a plant, after there is no AC power failure, or for that i
9 matter, DC either, if one addresses the design, to so do it.
! 10 That's all I have to say.
, 11 MR. MICHELSON: Nothing at this time.
12 MR. WYLIE: Suppose I call on Mr. Speis. I i
1
() 13 believe Mr. Speis is going to present the first presentation.
14 MR. SPEIS: Thank you, Mr. Chairman. I see from i,
! 15 the introductory comments from the committee, we are going to l 16 be here for quite a while, and very interesting questions to j 17 address.
! 18 I think we have -- from what I have heard, we i
l 19 should be able to address most of your concerns and questions l 20 during the presentation.
21 Basically, what I will do, I will summarize our 22 proposal and provide you an outline with the picture, but the
- 23 subsequent speakers will give you a very detailed 24 presentation on a technical basis. And they will address
(])
25 some of the detailed questions that I see are hiding behind
/\CE FEDERAL REPORTERS, INC, 202 347 3700 Nitionwide Coserage 80M36-6M6
+-m, -,,---,,--,,,---,--,,,-n,-, , , - - - , , , . , _ - , , _ . , . , , - --,---w- - - ~ ~ - , _ _ , - - _ , , - , - - - - - __ ---.,.,-,-,,,v-
,e , - - . . - , , - - - - . . - - - , - , . -
30853.0 cox 8 1 the comments of the ACRS members so far.
O 2 Let me give you the outline of our presentation.
3 I will discuss the approach for the -- that we have taken to 4 resolve this issue. I will say a few things about the key 5 elements of the resolution, and I will say a few things about 6 n the status since the last ACRS meeting.
7 Baranowsky and Flack will address the technical 8 findings which are documented in NUREG-1032, which you have 9 access to it.
10 Excuse me, I have a cold.
11 Specifically, they will address the off-site power 12 reliability, the emergency power sources, and they will
( 13 provide you with the elements of the probabilistic analysis 14 that has been done in support, technical findings.
15 Al Rubin, the project manager of A-44, will go 16 into some detail of the rule itself, the regulatory guide,
, 17 and also the value impact analysis, which is a necessary 18 ingredient for resolving any issue.. Also, he will discuss i
i 19 the significant changes to the proposed resolutions -- excuse 20 me, the significant changes that have taken place from the I
21 proposed to that of the present one, the one we are proposing i
! 22 today.
23 Also, he will summarize the industry activities, i
() 24 as to understand them, based on basic meetings and 25 interactions that we have been having with NUMARC, and we
/\CE FEDERAL REPORTERS, INC.
l i ,
202 347-3700 Nationwide Cmerage 800-33MM6
J 30853.0 cox 9
- 1 will do -- to the best of our knowledge, we will try to O 2 compare our proposed resolution with their initiatives.
- 3 I will skip the first part. Mr. Chairman, you i
j 4 already talked about the previous ACRS letters, even though 5 your group supported in general, you had some comments about 6 integration of A-45. Also, there were some specific comments 7 about one of the ACRS members about considering earthquakes, 8 and we will discuss that later on as part of the technical
- 9 basis of our resolution.
10 Basically, our approach to the final resolution is i
4 11 that as the same with the proposed resolution of a year and a i 12 half ago, when the proposal went out.
13 Again, the main aim is to reduce the frequency of i 14 occurrence of core damage from the station blackout.
! 15 Our presentation will focus on changes from 16 proposed resolution, supporting technical analyses, and, as I
- 17 said already, the inclustry initiatives.
- 18 I am going to skip and go straight to Vugraph 19 Number 6 there in your package, and kind of summarize, i
i 20 Again, Mr. Chairman, you have done this already, but the l
21 elements of the final resolution. There are two parts to it, 1 22 the rule and the regulatory guide. As you have said already i
j 23 in the rule, we require that all plants be able to cope with O 24 a station blackout, and also have procedures to withstand and l 25 recover from such an event.
) ACE. FEDERAL REPORTERS, INC.
202 347-3700 Nationwide Cmerage 800-336-6M6 1
_ _ . ,,-- , _ - - - - - - - - . . _ . ~ . - - -.- . . - -
30853.0 cox 10 1 There is no specific mention of numbers or 2 anything in the rule itself. The accompanying regulatory 3 guide provides the details of coping durations and other 4 relevant information in support of the rule. Again, the 5 regulatory guide provides guidance for all plants to comply 6 with rule. Guidance on acceptable plant specific coping 7 durations. Guidance on use of alternate power sources, I am ,
8 assuming you have lost the normal on-site sources, which are 9 really the diesels, procedures and training to cope with 10 station blackout. And we addressed the issue of diesel 11 generator reliability.
12 Later on, Rubin, when he discusses the guidanca 13 and acceptable plant specific coping durations, if you recall 14 the proposed guide, discussed two hours, basically -- not two 15 hours, four-hour and eight-hour coping capability. We have 16 refind the guide and, based on the more detailed analysis, we 17 have found out there are some outlying plants that would need 18 to be able to cope with 16 hours1.851852e-4 days <br />0.00444 hours <br />2.645503e-5 weeks <br />6.088e-6 months <br />. Also, there are some other 19 plants that are much better, and two hours would be 20 sufficient in that case. Al will discuss that in detail, so .
21 that is kind of a fine-tuning of the regulatory guide.
22 I am going back now to Vugraph Number 3 in your 23 package, and address the key elements of the approach that we
() 24 have taken to address the issue.
25 Basically the first thing we did was to make an ace. FEDERAL REPORTERS, INC.
202-347-37(o Nationwide Coserage s o-336-M 46 '
30853.0 Cox 11 1 estimate of the frequency of core damage frequency due to
() 2 station blackout for a spectrum of plant designs.
1
]
3 Next, identify dominant factors affecting CDF and ,
4 cost-effective improvements. As I said earlier, it's one of 4
^
I 5 the most important elements in resolving any issue, having
- 6 the backfit rule and all the other Commission rules that we i .
7 have to follow very carefully.
, 8 In identifying the dominant factors affecting core I
9 damage frequency, the two key parts of it are the likelihood i
i 10 of the frequency and the duration of loss of AC power; and, i 11 of course, the ability to cope with extended loss of AC l
12 power. There are some plants that are able to cope more than
( 13 others. And so, those two things are the important things in ,
14 determining the factors affecting the core damage frequency.
I 15 We feel that proposed new requirements are ,
i i 16 consistent with the level of risk and cost effectiveness 1 17 considerations. ; ,
i 18 What drove us to undertake the work and be 19 concerned about this issue? ,
20 Basically, the key part was the operating 21 experience, based on our analysis of the behavior of on-site l
22 power sources as well as the behavior of the off-site plant.
]
23 We have come to the conclusion there are limits on l ({)
i 24 reliabilities of off-site and on-site AC power systems,
! 25 Also, we came to the conclusion, based on our extensive look i
/\CE. FEDERAL REPORTERS, INC.
j 202-347 370) Nationwide Cmerage 840 33MM6
,-.-.,e-.--,.--.,------~--.-%,.._.,_-.,,,_,.,m,y ,- ,.%., y, __. . , - - . . - , - - . . - , _ y. ,7__-,-m-w y ---,,, - -m..
30853.0 cox 12 1 at most of the plants, that AC power reliability varied O 2 significantly from plant to plant.
3 Also, a key consideration was the key of the 4 potentially severe consequences as a result of losing AC 5 power. There is a limited recoverability for decay heat 6 removal is limited without AC power, although I heard what 7 the distinguished member of the ACRS said earlier about 8 boilers, and we will try to address that. I think it's part 9 of the coping capabilities.
10 Of course, you have no heat containment removal if 11 you loose AC power. That's also a very important part.
12 Presently, there are no requirements for plants to t
13 cope with station blackout. All our regulations, regulatory 14 guides, the whole NRC infrastructure is keyed -- is focused 15 on reliability. There is nothing in the regulation that says 16 that you should be able to cope with such an event. That is 17 the thing that drove us, based on the advice of lawyers and 18 extensive discussions with the Staff, to recommend a rule, so 19 every plant should be able to cope with a station blackout.
20 MR. MICHELSON: Let me ask a question on that 21 point.
22 MR. SPEIS: Okay.
23 MR. MICHELSON: We have presently certain I
f () 24 requirements to be able to cope with a possible loss of a 25 control room.
i ACE-FEDERAL REPORTERS, INC.
! :o 3v.3no Nationwide cmerage sm-3w66:6
30853.0 cox 13 1 MR. SPEIS: Control room?
2 MR. MICHELSON: Yes. We have requirements for 3 that. We have to have a backup center that I think is 4 operable, get it shut down in three days, some kind of thing 5 like that. In the processing of losing the main control 6 room, I think it's quite conceivable, in a power blackout 7 situation, depending on just what the fire does, what the 8 system interaction effects are, and so forth. Even though l
9 power might even be into the switchyard, it may not be l
10 possible to get it into the plant, from loss of the control {
i 11 room.
12 So I was wondering, how did you wed that
() 13 requirement to the one we are now dealing with? Or did you 14 consider that you never have a power blackout during a loss 15 of the control room?
16 MR. SPEIS: Do you have an answer, Pat, to that?
17 MR. RUBIN: My name is Alan Rubin.
18 I am not aware of any requirement with the control 19 room being unavailable, to show that the plant had shut down 20 without any AC power available.
21 MR. MICHELSON: I don't think there was ever the 22 question of whether or not AC power was available. The 23 question was, do you have to show that if you do lose the
(} 24 control room that you have a backup point that you can go to 25 that is divorced from the control room, that will allow you ACE-FEDERAL REPORTERS, INC.
l 202-347-3700 N,itionwide Coserage 14)O-336-6M6
30853.0 cox 14 I
1 to get, with minimum equipment, one train, I am not even sure O 2 it has to be safety grade; I think you have got up to three 3 days in which to get the whole shut down.
4 MR. RUBIN: We still have available controls, I 5 think, for AC power equipment.
6 MR. MICHELSON: Do you think that you have AC 7 power available inside the plant when you lose the control 8 room?
9 MR. RUBIN: We haven't analyzed that situation, 10 loss of control room, loss of AC power.
11 MR. MICHELSON: You haven't looked at the 12 possibility that a loss of a power -- power blackouts might
() 13 be accompanying the loss of the control room?
14 MR. RUBIN: We have looked at the problem with i
15 station blackout, also having lost control of the control i
16 room.
17 MR. MICHELSON: That's what I was wondering 18 about. It appears quite conceivable that they could be wed,
, 19 depending on the nature of what destroys the control room. I 20 think we have said in the past, we can stand a fire in the 21 control room that wipes it out.
22 MR. SPEIS: I think we better go back and look 23 into this, because there are other people that have been
(} 24 involved. I don't think we are giving you a definite 25 answer. We should have some answer for a full committee.
ace-FEDERAL REPORTERS, INC.
L 202-347-37tt) Nationwide Coserage 800-336-6646
--- .. - -. ~ . - _ _- .- -- - .- _. .
30853.0 cox 15 1 MR. MICHELSON: I was philosophically wondering .
2 how we would combine the two; and if we haven't, why not, and 4 3 <so forth.
4 MR. SPEIS: Let's discuss that Friday when we come 5 back Friday for the full committee.
, 6 MR. EBERSOLE: If you have open short grounds, i 7 which would interfere with excess to off-site power, you
. 8 should be able to clear any interference from incoming power 9 whether it's from the offset or on-site power.
10 MR. MICHELSON: Jesse, I think there was no 11 requirement that you have ability to maintain power of the 12 off-site plant when you lost power of the control room.
13 MR. EBERSOLE: It was a power from wherever you 14 have it.
15 MR. MICHELSON: Right.
16 MR. EBERSOLE: That's off the emergency buses .
t 17 which have access to two sources.
18 MR. MICHELSON: But if the fire -- if it loses the 19 control room, you are not set up to automatically have the 20 diesel be operable. You will get up to your backup center; 21 hopefully you get designed so you can divorce the diesel, get 22 back on to their boards and get going. That was the plan.
23 There is a power blackout for 30 minutes to.an hour during
() 24 which you are trying to move from the old control room to the i 25 new control room.
/\CE FEDERAL REPORTERS, INC.
i 202 347 3700 Nationwide Coserage 800-336-6646 !
l
30853.0 cox 16
! 1 MR. EBERSOLE: Agreed.
I i 2 MR. MICHELSON: I wonder how that power blackout 3 relates to what is looked at here.
4 MR. EBERSOLE: There is one aspect that I think l
5 should be mentioned. As I recall, it was an old design i 6 problem. That is where you converge off-site and on-site 7 power and have a commonality, usually in two discrete boards, 8 at least. There should be no common influence that j 9 interferes with these focal points of convergence, usually 10 two boards which are shut down boards that have access to 11 both diesel and off-site power.
- 12 If you could imagine some catastrophe, and the one
() 13 that I recall was explosion of the fuel tanks, that involved
, 14 the shut down boards, they are the point of confer against 15 with both kinds of power. Then you lost all access to all
. 16 power that we characterize as AC. Any discrete accident that 17 you look at in commonality, to both shut down boards, can get l
18 this total outage indefinitely.
l 19 That should not be present in the plant design.
20 MR. MICHELSON: That's the problem you get into
- 21 when a fire engulfs one shut down board in the system fire 22 reaction protection feature shuts off, then you have lost 23 both, a power blackout and the fire is a combination.
i i
(} 24 This leads to one other question --
25 MR. SPEIS: Excuse me, Pat, you had an answer to
> /\CE-FEDERAL REPORTERS, INC.
202 347-3700 Nationwide Cmerage 800-336-6M6 i
30853.0 cox 17 1 the previous discussion.
) 2 MR. BARANOWSKY: This is Pat Baranowsky.
3 I just wanted to mention we are aware of that type 4 of concern. There is some research going on to look at 5 interacti<ns associated with fires and how-shorts that might i
6 be associated with fires in the control room could propagate
! 7 into undesirable system interactions. It's a fairly 8 complicated thing. It's very plant specific. I don't think 9 we come up with a general conclusion that would handle more i
10 than one plant if we look at one plant.
i 11 MR. MICHELSON: I also wanted to. emphasize --
12 MR. BARAN0WSKY: Separation of boards and s
! 13 everything is also very plant specific.
14 MR. MICHELSON: One must be careful in looking at 15 that system interaction question to make sure that you look i
l 16 at the safety system, which is the off-site power, it's not I
17 classified generally as safety system. It may very well be 18 lost as part of the fire in the control room. It wouldn't be I
subject to the analysis, because you say, oh, I have got the 19 20 on-site diesel.
21 MR. BARANOWSKY: The kind of analysis that is 1
22 being done is PRA-oriented, in which they would look at most 23 safety and nonsafety systems to see what the safety
() 24 interactions are.
25 MR. MICHELSON: The other question is we talk ACE FEDERAL REPORTERS, INC.
I 202-347-3700 Nationwide Coserage 800-336-6M6
30853.0 cox 18 1 about this power blackout as being a loss of all AC power.
O 2 Isn't it just about as bad to just, say, lose your 480 volt 3 level AC power, is that considered loss of all AC power even 4 though the higher level, the 60-4100 or --
5 MR. BARANOWSKY: Yes. Usually what we are talking 6 about is losing usable 481, 60 volt level AC power.
7 MR. MICHELSON: Either one or both?
8 MR. BARANOWSKY: It really depends on how the 9 plant is designed in terms of motor control centers and ESF 10 switch gear. Most plants are topping off of a 4160 volt bus, 11 and they get a fair amount of diversity and redundancy of 12 power supplies beyond that level, when you get to the 4180
) 13 volt level.
14 In particular -- on the other hand, some plants 15 are using 4180 volt levels. They write off the diesel 16 generators -- 480 volt level, excuse me. What we are looking 17 at is power available to the ESF bus level, whatever that is, 18 4160 or 480.
19 MR. MICHELSON: Are you talking about the pump 20 level or valve level?
- 21 MR. BARANOWSKY: What I am talking about is --
22 MR. MICHELSON: Depending on the plant.
23 MR. BARANOWSKY: Yes. Reason you don't go down to
,])
( 24 a lower level is you have to have numerous failures to fail 25 the numbers of motor control centers, typically, at lower ACE-FEDERAL RtronTras. INc.
g 202-347-3700 Nationwide Coserage 800-336-6646
30853.0 cox 19 1 voltage levels.
2 MR. MICHELSON: The reason I asked is that 3 although, in the plants, we have generally been careful to 4 keep the 4160 boards separated and careful to keep the 480 5 separated, we haven't been so careful about the routing of 6 480 volt cabling. There are plants where a certain 7 conflagration will get all of the 4880 volt level even though 8 the 4160 is still working. It gets kind of sticky then.
9 MR. BARANOWSKY: I would agree. The place that is 10 looked at is through fire protection.
11 MR. MICHELSON: Unfortunately, the fire protection 12 people just look at it and say, okay, we have to put some O
/ 13 sprinklers in. That isn't the answer on single failure, the 14 sprinklers, for instance, whatever happens that activates the 15 activation systems and on and on. Do we consider a 480 volt 16 AC loss across the board as a power blackout?
17 MR. BARANOWSKY: No, not as you are talking about 18 it. We are recognizing the possibilities of fires causing 19 such a situation, but we have felt that the' treatment of fire 20 is a lot different than the kind of issues associated with 21 off-site power and diesel generator reliability, which are 22 the main issues associated with what we are talking about 23 today.
(-)s
( 24 MR. MICHELSON: Although it's quite realistic, 25 it's a different problem.
ACE FEDERAL REPORTERS, INC.
202-347-3700 Nationwide Coserage 800-336-6646
30853.0 cox 20 l 1 MR. BARANOWSKY: It's a different problem.
2 MR. EBERSOLE: How do you regard, as just a case 3 in point, the AC power supplies derived from the diesel 4 generator that supports the core spray system for boilers, 4
5 recognizing that those systems don't remove decay heat from 6 the containment, they just provide makeup to the core.
7 MR. RUBIN: We are going to get into that.
8 MR. EBERSOLE: Mr. Chairman, we have a very .
9 extensive detailed presentation. If you give us a chance to 10 go through it, then later on I can address more questions 11 dealing with integration of A-44 and A-45 as well as discuss 12 more the considerations that led us into the rule.
1
() 13 MR. WYLIE: Okay. Let's proceed.
14 MR. SPEIS: That doesn't mean I am not telling you 15 to not ask questions; I can never do that.
i 16 MR. WYLIE: Hardly.
17 MR. SPEIS: I have listed here the last part on l
18 this Vugraph, the core damage frequency reduction. We feel 19 that you can do that by being able to cope with a station 20 blackout for a specified duration. If you recall also, we j 21 have been talking and using a guideline to determine as kind 22 of a goal, what do we want the sequence to be, and we have 23 set us a goal that the core damage frequency from station 24 blackout should be a small contributor to total core damage
[])
25 frequency of the order of 10 minus 5 reactor year. These 1
^
ACE FEDERAL REPORTERS, INC.
202-347-3700 Nationwide Coserage 800-336-6M6
30853.0 cox 21 1 numbers are kind of rough goals and no more than that, O 2 basically. I think you can talk about that forever and 4
3 ever.
- 4 A summary of findings. I think we kind of touch 5 cx1 them already, reliability of on-site emergency AC power 6 varies considerably. It's dependent on configuration and, of 1
- 7 course, the reliability of the diesels themselves. What I ,
8 mean by configuration is how many diesels do you have and how ,
9 many do you need for decay heat removal.
, 10 If you have two and you need one, it's not as good 11 as if you have three and you only need one. So that's -- so 12 these two are important factors in setting up this matrix of i
13 how much coping capability is required to meet this goal of 14 10 minus 5, as I said earlier.
15 Likewise, the importance of the off-site power, 16 there are two important parts to it. The site
! 17 characteristics, the weather, the grid, and, of course, plant 18 specific factors, switchyard design and transmission lines.
19 Based on these four factors, we have kind of set up this 20 matrix of what the coping capabilities should be to bring all 21 plants to some average level of reliability as far as the 22 contribution of the station blackout and the core melt 23 frequency.
1 l
a
() 24 I said already that the damage frequency varies 25 considerably from plant to plant. Again, it's both the
/\CEJFEDERAL REPORTERS, INC.
202 347-3700 Nationwide Coserage 800 336-M46
_ _ . ~ _ _ , _ . _ __ , _ _ _ _ _ . _ _ ._ _ .,__ ___ ,,,___. _ , , _ . _ _ _ _ , - - _ _ _ . , _ _ _ _ _ _ _ _ _ _ _ - - _
30853.0 cox 22 1 susceptibility to station blackout as well as the ability to
() 2 cope with loss of all AC power. We don't think that there is l
i 3 a single fix that can be applied uniformly to all plants that 4 will resolve this issue in a cost effective manner. That's 5 why we have attempted to arrive at kind of a graded i 6 approach. That is this matrix that we will talk later on.
7 The schedule that we are talking about is provided 8 in this slide here. I think I should have gone back and said 9 that the first time we went to CRGR on this issue was back in 10 March of 1984, and here we are three years later. We will be 11 going to CRGR again at the end of this month, May.
12 The proposed rule went to the Commission in May of C
13 '85. The rule stayed there for a while until we provided to 14 them an evaluation of the NUGSBO proposal; this is the 15 predecessor to NUMARC that we will be talking to you later 16 on.
! 17 Also during that time the new backfit rule went l
l 18 into effect, and the proposed rule was delayed for a while
! 19 until we recast the regulatory analysis into the elements of 20 the rule itself.
21 The Commission voted five to zero to get the 22 proposed rule out. That was back in May '86. End of public i
23 comment, June of 1986. So we are here talking to you kind of l
24 a year later, basically. Our schedule proposed to complete i
(}
25 the ACRS, CRGR review by June and then go to the Commission ACE FEDERAL REPORTERS, INC.
202-347-3HX) Nationwide Coserage 800-336-(646
1 30853.0 23 cox t
1 in August of '87.
2 With those brief introductory remarks, I want to 1
3 call out on Rubin, or I guess Baranowsky is next, to start 4 talking about the technical braces.
5 Again, it will give us a chance to tell you our 6 story. There will be a lot of time to ask questions.
7 MR. MICHELSON: One brief question, is it normal 8 that it takes over a year from the end of public comment .
9 until the final resolution as presented?
10 MR. RUBIN: That depends on the extent of the 11 comments and the kind of work we need to do to address them.
. 12 In this case, there were significant updating of analysis
) 13 that you will hear during the presentation, and we made some 14 modifications to the proposed rule.
15 There were a few comments, or very little changes, 16 it would not take so long.
17 MR. MICHELSON: Were they what we would call 18 significant changes to the rule?
f l
19 MR. RUBIN: I would say there were some changes
- 20 that were significant, in terms of their impact, and I would i
. 21 say many more of the changes were -- some of the changes here 5
22 from Pat Baranowsky, updated to support.
I 23 MR. MICHELSON: How does one decide whether or not ,
24 you have to go out again for public comment when you make
(])
25 significant changes after a public comment, and plus a long l /\CEJFEDERAL REPORTERS, INC.
202-347-3700 Nationwide Coverage 800-336-6M6
30853.0 cox 24 1 period of delay.
2 MR. RUBIN: I think I will get into the changes 3 that were made as a result of the public comment. You will 4 see they are significant. They are changes, I would say, 5 nore in the direction of clarification, one, and reducing 6 some of the requirements that were included in the proposed 7 resolution.
8 MR. MICHELSON: You might have had commenters who
! 9 were satisfied on what you said a year and a half ago, but 10 might not be satisfied with what you are now going to put 11 into --
12 MR. RUBIN: We always have people that are not O
(_f 13 satisfied.
14 MR. MICHELSON: They may not have commented 15 because they didn't realize what you were doing. Ordinarily, 16 I think it's required to go out again for public comment if 17 you make truly significant changes.
18 MR. RUBIN: We will discuss the changes later on. ,
19 You will have a chance, I think, to see where those changes-20 are made.
21 MR. SPEIS: I might want to add, it's not easy to 22 get things going fast these days.
23 MR. MICHELSON: I just wondered if a year was a
(} 24 norm or an unusual case.
25 MR. SPEIS: It is an important rule, and we ACE FEDERAL REPORTERS, INC.
202-347-3700 Nationwide Coserage 800-336-6M6
30853.0 cox 25 1 brought into consideration everything, carefully, that was
! ( ): '
l 2 said. We will talk about the changes later on.
i 3 MR. RUBIN: I would also like to say that during i
4 this period, we have had extensive discussions here with 5 NUMARC. So considerations of alternatives was one of the i 6 recommendations that the ACRS has made.
7 We needed to come to grips and a mutual ,
8 understanding of what NUMARC is proposing as an alternative 9 resolution. That's part of the time .--- we spent an extensive i 10 period of time in discussions on this position.
11 MR. SPEIS: Okay, Pat.
12 MR. BARANOWSKY: I am Pat Baranowsky. I have been O 13 involved in doing the technical analyses for the NRC Staff on i 14 station blackout since late 1978.
I 15 We gave a briefing to the ACRS on our technical 16 work several times between then and now. The most recent one 17 that I can recall, I believe, was in February of 1985. What 18 we are going to do today is summarize some of the technical 19 work we have done briefly, and point out some of the i
20 enhancements we have made to those technical analyses.
21 I will be talking about various portions of the 22 technical program, and my colleague, John Flack, will also be I 23 talking. He will be primarily talking about the loss of 4
4 l
() 24 on-site power frequency and duration analyses.
25 We had a program which started out with looking at
/\CE. FEDERAL REPORTERS, INC. l j 202-347-3700 Nationwide Coserage 800-336-6646 j
30853.0 cox 26 1 operating experience, relevant to station blackout. We look O 2 at off-site power, operating experience, in terms of losses ,
3 of off-site power, why, how frequent, how long, and we looked 4 at emergency AC power reliability considerations. They are 5 also in terms of performance, as well as design factors, that 6 influenced the reliability of AC power systems. We combined 7 these aspects together through modeling, using actual 8 operating experience for failure rates, to generate estimated 9 frequencies of station blackout and durations of station 10 blackout, so that we could look at different plant designs, 11 and extend our knowledge from what happened at one plant to 12 another.
13 We used PRA-type analyses to look at potential 14 accident sequences which could follow loss of AC power, 15 determine the characteristics of those sequences, which ones 16 would result in core melt, how soon they might result in core 17 melt, and what the frequency of those core melt sequences i
18 might be.
i i 19 Based on this, we identified technical factors
! 20 that are, could be used in proposed resolution to reduce the l 21 likelihood of station blackout to some level that might be i
22 consistent with levels of risk that the NRC has found 23 acceptable and plants are thought to have had.
l l
() 24 So, what I am going to do is give a brief 25 introduction and summary, particularly talking about the l /\CE FEDERAL REeoRTERS, INC.
202-M7-3700 Nationwide Courage 800-336-6646
30853.0 cox 27 1 operating experience that got us started into looking at 2 station blackout. Then I will turn it over to John Flack to 3 talk about loss of off-site power.
4 I might say that the station blackout issue, as we 5 have looked at it, has been real, as opposed to 6 hypothetical.
7 In other words, we have looked at the operating 8 experience, and we have tried to construct reactor accident 9 sequences and scenarios that are relevant to what we have 10 seen in operating experience.
11 I should point out that we looked extensively at 12 diesel generator operating experience and off-site power
() 13 operating experience and events that involve what I would 14 call precursors to station blackout. ,
I 15 From 1968 through 1985, we have observed 64 total 16 losses of off-site power. These range in duration from a few 17 minutes up to 26 hours3.009259e-4 days <br />0.00722 hours <br />4.298942e-5 weeks <br />9.893e-6 months <br />. The one that lasted up to 26 hours3.009259e-4 days <br />0.00722 hours <br />4.298942e-5 weeks <br />9.893e-6 months <br /> i 18 was caused by a hurricane, and the licensee for that plant f 19 feels that certain corrections and repairs could have been l
l 20 made in about five or so hours, although they waited 26 hours3.009259e-4 days <br />0.00722 hours <br />4.298942e-5 weeks <br />9.893e-6 months <br /> 21 to make the connections, because they wanted to be sure that 22 there weren't any problems associated with that event.
i 23 Also, from 1976 through 1985, we have seen over l
600 emergency diesel generator failures. These are failures
(} 24 25 of emergency diesel generators to perform an intended mission ACE. FEDERAL REronTEns, INC.
l 202-347-3700 Nationwide Coserage 800-336-6M6
- . _ _ _ _ _ . _ _ . ~ . . , , - . _ . . _ _ _ _ _ _ _ . . _ _ _ , _ _ . _ _ _ _ _ . _ _ . , _ - . , . , , , _ . , _ , _ _ _ _ _ . . . _ _ _ . _ _ - . . , . - _ ,
I 30853.0 i cox 28 1 which might be associated with a station blackout. In other 2 words, they couldn't start and they couldn't run. These are 3 not just auto start failures, where for some reason the auto
- 4 start circuitry doesn't work, but an operator in the control
\ n 5 room can manually start the diesel. e 6 Yes, sir.
! 7 MR. REED: How many are total loss of off-site AC, 8 you made mention of the fact that at some plant they could 9 have reenergized but didn't because they wanted to check out c 10 equipment, perhaps causing more damage to equipment. That's
- 11 a very standard thing that happens. I have been through a 12 total loss so-called label total loss of off-site AC power.
() 13 A transformer was not reenergized, because an indicator on
(
14 the transformer was ou*,. It turned out to be insignificant 15 and something related to the transient. But a lot of experts 16 come into the picture and take samples of gas, this, that and
! 17 the other, decide not to do that, because everything is under 18 control.
i 19 I have watched these accumulations of loss of 20 off-site power, and I think they are probably magnified above 21 real emergency action. In the case that I am familiar with, i
i
- 22 the plant -- it was an ice storm, weather-related, and the
! 23 plant kept coming from the west, and the dispatchers and all .
() 24 the people in the transmission business, highlighted 25 so-and-so's data, this is down, start to protect yourself.
ACE. FEDERAL REPORTERS, INC.
202 347-3700 Nationwide Coserage !M336-6M6
30852.0 cox 29 1 We were entirely shut down as far as nuclear power was 2 concerned, and then gas turbine, outside gas turbine, put 3 that in action to handle the situation.
4 So, really, there was no decay heat removal 5 emergency whatsoever at all, even though total loss of ,
6 off-site power occurred.
7 I think generally in these statistics, loss of 8 off-site power, total loss, that they are probably 50 percent 9 inflated, really, with respect to action that could have been 10 taken or was taken.
11 But the point is that you will lose off-site 12 power. There are electrical engineers, and there were two or
() 13 40 years ago, that would say it wouldn't happen but once in a 14 million years and then it happened two years later. I might 15 say it was in two plants. So it's a fact of life, it has to 16 be considered, and perhaps its risk number is a little i 17 inflated.
18 MR. BARANOWSKY Let me point out, just in case
- 19 John Flack doesn't point this out, that what we generally use t 20 in our analyses are the best estimates of the utilities in l
21 terms of when they think they could have restored off-site 22 power. You won't see any inflation in our numbers. We i 23 didn't use 26 hours3.009259e-4 days <br />0.00722 hours <br />4.298942e-5 weeks <br />9.893e-6 months <br /> here. We generally used the five-plus 24 hour number in our estimates. When we use a number like
(]}
25 that, we try and get some NRC verification that that is a ACE FEDERAL REPORTERS, INC.
i l 202 347-37(O Nationwide CoseraFe 8(X)-33MM6 i
i
30853.0 cox 30 7- 1 realistic estimate.
(- In this situation I am talking about over here, 1
i 2
3 there was substantial salt spray due to a hurricane at the l 4 switchyard. They wert continually washing down the 5 insulators, and some : f them were getting resprayed on .
6 again. I don't know exactly how soon the wind has died down 7 to the point where they could have actually washed them down 8 and cleaned them off, so there wouldn't be arcing and 9 faulting, but it looked like about five hours in this case.
10 MR. REED: You have adjusted, you say.
11 MR. BARANOWSKY: We have tried to be as realistic 12 as we can. Our estimates aren't inflated. They are as O 13 realistic as we can make them.
14 MR. WYLIE: The 26 hours3.009259e-4 days <br />0.00722 hours <br />4.298942e-5 weeks <br />9.893e-6 months <br /> was the salt spray?
15 MR. BARANOWSKY: That particular one was. That 16 wasn't the longest loss. I just point out for this 26-hour 17 case, which is the longest time which a nuclear unit kept 18 off-site power unavailable, when it was originally a forced 19 outage, was actually a forced outage of on the order of five i
20 to six hours.
21 MR. WYLIE: I know several years ago, I forget the 22 plant, it was in the Midwest, where a tornado took out 23 everything around the plant.
() 24 MR. REED: Dresden, I think it was.
25 MR. WYLIE: Dresden. How long did it take to i
Ace FnonRAL REPORTERS, INC.
g 202-347-3700 Nationwide Cmerage mb336-(M6
r 30853.0 cox 31 l
1 restore that one?
2 MR. BARANOWSKY: In that situation, the tornado, 3 if I recall correctly, took out not all the lines from the 4 switchyard out into the grid, but it took out several lines 5 off -- that were takeoffs from those lines, which ended up 6 tripping protective relaying and circuit breakers which 7 basically laid out the line. It took them about four hours 8 to clear those faults so they could get power back. It 9 wasn't a situation where the tornado actually damaged the 10 main lines coming into the plant. When tornados do damage 11 the fair lines coming into the plant, you are talking a fair 12 amount of time, although there are some temporary things that
() 13 can be done. You can string temporary lines up, but pretty 14 difficult.
15 MR. EBERSOLE: Are you going to get to 16 earthquakes?
17 MR. BARANOWSKY: Sure, at the end.
18 MR. WYLIE: Before you do that, on your 600 diesel 19 failures, how are they reported, LERs?
20 MR. BARANOWSKY: Yes. A lot of times they are 21 reported by LERs. I will get into the details of where all 22 this information came to. I just wanted to summarize to give -
23 you a little perspective of the whole picture before we 24 rolled through everything.
25 MR. WYLIE: Okay.
ACE-FEDERAL REPORTERS, INC.
202-347 3700 Nationwide Coserage MC 13MM6
30853.0 cox 32
~
1 MR. BARANOWSKY: Let me just go on here, and then 2 you will get additional information from the rest of the 3 presentation.
4 Also, I don't recall if I showed this slide or not 5 the last time I was here. But these are what we call 6 precursors to station blackout. That is to say, we have 7 actually had situations where AC power was unavailable for 8 something like five minutes to 25 minutes at some plants.
9 The first one that we were able to uncover occurred in 1968, 10 in which there was some maintenance being performed on the 11 off-site power supply, and a switching error resulted in the .
12 loss of all off-site power when the plant was at 100 percent
() 13 power.
14 The diesels initially started and loaded, but they 15 tripped off. They were without AC power for four minutes.
16 It took them 25 minutes to restore their off-site power.
17 They restored diesel generators within five minutes.
18 In 1976, there was an under-voltage problem on a l
19 grid, and the design of the plants under-voltage protection 20 for the ESF buses was such that it allowed that degraded 21 voltage condition to appear on the ESF buses, so that even .
22 though the diesel generators were given a signal to start, 23 because of low voltage, they couldn't -- the bus did not l
() 24 unload, and they couldn't load the diesel generators on the 25 bus. They maintained a degraded voltage situation there. It ace. FEDERAL REponTEns, INC.
! 202 347-370i> N.irionaide coserage sa L3 4 6646
30853.0 cox 33 1 tripped off basically power to -- I believe it was to all the 2 ESF sources. ESF equipment.
3 In 1983, we had a snowstorm at a site which caused 4 a loss of off-site power. In that situation, there was one 5 diesel generator down for maintenance, and a second diesel 6 generator was being paralleled with the grid. There was a 7 failure of a load shed relay on that particular diesel 8 generator.
9 It ended up causing a tripping of the bus, and 10 there was a total loss of off-site power for 20 minutes.
11 Then in 1984 an interesting event occurred, it's 12 one in which we have to be careful about; that is to say that O 13 the nac ne certeia reautre eats for dotas ereoverettoa 1 14 testing, verification of capability. There was a test going 15 on to show capabilities with regard to the loss of off-site 16 power. In this case, operators mistakenly switched off DC 17 control buses, which prevented emergency AC power supplies 18 coming on when they then -- when they had tripped off the 19 off-site power and do their off-site loss of power test.
l 20 The operators smartly and promptly traced back 21 their s ops, found out which circuit breakers they had 22 mistakenly tripped, reconnected the AC power supplies, and 23 they had AC unavailable for 10 minutes in that case.
24 So those are the cases that we have identified in ,
25 which there has been essentially a total loss of AC power.
Acti FnonnAr. RneonTuns. INC.
202 347,37m Nationwide Cmcrage mn316 (M6
30853.0 cox 34 7s 1 There have been numerous instances in which we have had loss 4 U 2 of off-site power and one diesel generator either fails or is 3 unavailable.
4 We have taken all this information and come up 5 with some summary statistics, just to give you an idea where 6 we stand. These numbers haven't changed too much since the
- 7 'last time we were here. The average loss of off-site power 8 frequency is about .1 per year. It varies. Some plants have 9 had zero. Others go up as high as about .4 per year. There 10 has been some improvement over the last few years. John 11 Flack will show you what that improvement has looked like.
12 We also think it's important to look at the O 13 restoration time, because short losses of AC power aren't a 14 big problem, it's the longer ones that are important. Median 15 here is about half an hour.
16 MR. EBERSOLE: I guess this is the point where I 17 begin to have trouble. Because you were using averages 18 across the spectrum of contributions, in which there are huge 1
19 lumps at various ends of the spectrum, and this process tends 20 to diminish the view into where the problem really is, which 4
21 is, in the worst, say, 10 percent of the distribution.
22 MR. BARANOWSKY: Right.
23 MR. EBERSOLE: So this acts almost universally as l () 24 an obscuring function to where the problems are.
l 25 MR. DARAN0WSKY: Well, this information is truly Aci-FunnRai. Riti>oRiiiRs. INC.
l 202 347.17to Nationwide cmcrage so 3twa6
30853.0 cox 35 1 summary and average information. I am glad you think that we i
2 need to look for this worst 10 percent, because the balance 3 of our presentation is going to show you how we are looking 4 for that. You are really helping me out.
l 5 MR. EBERSOLE: Okay.
6 MR. BARANOWSKY: Just to set the stage for what 7 you will see. Emergency diesel generator reliability, about 8 98 to 99 percent. It varies, it also has shown some 9 improvement over the last few years. Repair time for 10 emergency diesel generators.
11 Remember, I have thrown out the auto start 1 Now we are talking about failures that are real 12 failuros.
() 13 failure. Median repair time has been about eight hours for
! 14 real failures; auto starts are a couple of minutes. This is 15 the kind of information that we used with models to generate 16 AC power unavailability estimates. That's why I have a range 4
17 here, 10 to the minus 4 and 10 to the minus 2 on demand, 18 given loss of off-sito power, botter to worse kind of thing.
19 Same with station blackout frequency, station 20 blackout being loss of off-sito power for at least a few 21 minutes, and diesel generators available for at least a few 22 minutos, could be lower. 10 to the minus 5, 10 to the minus 23 3. And then coro damage frequoncios, wo did tho same thing, see ranges in the area of 10 to the minus 6 up to 10 to the
(} 24 25 minus 4 for plants with typical capabilities that we have ACliI711DI!RAl IlliPORTI!RS, INC.
L 20214717m Nanonwide emeuge m pu ,u,,
30853.0 cox 36 1 seen up through 1985. There are things you can do to plants i
2 that have the higher frequencies, but we can discuss that at i
1 3 the appropriate time.
j 4 So let me just mention that since the last time we 5 were here, we did updating on analyses in terms of loss of ,
6 off-site power. We added events through 1985; we revised the 7 way we treated plant design groups so we could differentiate 8 between plants with higher and lower frequency of loss of 9 off-sito power. We improved our praatment of weather hazard i
10 data, and we added considerations associated with salt sprays 11 and hurricanos at coastal sitos to better capture what 12 operating experience was showing us.
() 13 For emergency AC power reliability, we updated 14 omorgency diosol generator failure data through 1985. We 15 revised data and analysos. And just to fill in the spectrum, 16 two diesels required out of four configuration. This 17 information was used to updato loss of off-site power and 18 omorgency AC power rollability analyses that feed into 19 station blackout and core damago, 20 So the next thing that we would like to provido l
21 somo details to you on are the loss of off-site power 22 frequency and duration analyses, and John Flack will present 23 those, if I don't have too many questions.
Already in your analyses, beyond and
)
{} 24 MR. REED:
- 25 behind what you have in the updating hero, is, of course, the ACli FitnenAi. Iliti>ontiins, INC.
- o: 3MSm Nationude Gnnage No34fM,
.,-..._-._,__...,._._m - . _ _ _ _ , . . _ . _ _ _____ __ __,_ _ ___.___ _- _ ..- ___,,.._ _ . __.._, _ ._ -- _ _ _
30853.0
, cox 37 1 planning and procedures that a site may have for 2 weather-related and other longer-term developing types of j 3 incidents. That's already in there?
4 MR. BARANOWSKY: I think you will see how we tried 5 to account for that in the next discussion.
6 MR. FLACK: Good morning, I am John Flack, risk d
7 assessment engineer from the offico of research, advance 8 reactor session. I am here top present a loss of off-site ,
9 power technical analysis.
10 As Pat said, we focused heavily on past operating 11 oxperience. We analyzed the causes, frequency and duration i
12 of those events, loss of power, relationship of design and
() 13 location to frequency and duration.
14 With regard to Mr. Rood's comment, I believe on 15 that event wo gavo you six minutos to recover the plant.
16 So on the event you were talking about?
17 MR. REED: Yes, on the ice storm.
18 MR. FLACK: Wo gave it the most realistic data as
] 19 wo could find.
I 20 This Vugraph shows tho annual frequoney of loss of 21 off-sito powor. We soo that it fluctuatos from year to year
- 22 with an average of about one ovont ovary 10 years. So as 23 Jonso pointed out, it doos chango. The dashed lino is a
() 24 throo-year traveling average. The typical duration is of
! 25 half an hour, but again, that varios from very short
! Acti 17tiininal. Illii>onTiins. INC.
J '
202 term Nanonwide rmerage miwuas
30853.0 cox 38 1 durations, to, again, very long durations, rather long 2 durations.
3 Okay, all the events that we observed broken down 4 into three categories, these categories, we found, had a 5 significant offect on the likelihood and duration of off-site 6 power.
7 The plant-centered category, grid and weather 8 related ovents. On the plant-contored categories, those are 9 ovents that are recoverable at the switchyard on the sito.
10 Wo soo the bulk of ovents aro in that area, 46, bst :nly 15 t 11 of the 46 oxtend for greater than half an hour in length.
12 The grid, we have observed 12 ovents, seven of
() 13 which havo extended for a half an hour or greater. Of 14 weathor, there were six events and all six had extended for 15 half an hour or greator.
16 So, again, the frequency and the durations are 17 important in those classifications.
18 I will go into each one of those classifications 19 now.
20 This curvo shows the rostoration and frequency of 21 '
the various categorios, plant contorod, grid and weathor. Wo 22 ano highor plant contored ovents but decreasing because of i
23 the fast recovery. We have a low frequency grid and weather 24 ovents, which have a longer recovery. So, again, we oro l 25 showing the offect of the likollhood and duration on loss of Acit.1;ituitnAI. Iliti>on r Ens, INC.
( I :ounsm Nanonwiac rmer.in smmmar, t
30853.0 cox 39 1 off-site power.
2 On the plant-centered events, here the design and 3 operational characteristics of the plant are important. We 4 found no frequency correlation with design. However, we did 5 find a modest corrolation between design and restoration.
6 Okay, so we used a generic value for the plant centered 7 ovents of .087 events por year frequency. These events 8 generally, again, are recoverable at the sito. They are due 9 to hardware failures, design deficiencies, human errors as in 10 switching and maintenance, and local weather conditions such 11 as lightening or a combination of thoso causes. Durations 4
12 are typically loss, as we have seen, and factors affecting i 13 the frequency and durations e o redundancy and the 1
{
14 indopondence of the switchyard design, how many redundant
~I 15 lines coming into the switchyard, and the number of the 1
16 circuits assessable on trip, automatically transferred of the 17 proforrod power sources. How many are available.
i 18 Again, wo saw a modest corrolation between design 19 and duration. Hero wo broke the plants into three f 20 categorios, based on design. I-1, I-2 and I-3. Wo 800 the
! 21 mean timo to rostoro as somewhat different for each of tho j 22 groups. I-1, the boat group, having redundancy in lines i
j 23 coming into the sito, or having soveral availablo sourcos on O 24 trin, automatic traneter.
, 25 In the I-3 group, we would have no rodundancy l Acti I;iiniinai Ittironitins, INC.
! h 202 m.1?m Nanon%Ie cmcrage sma(M6
_ .. -m _ _.
30853.0 cox 40 1 coming into the site, no independent line coming into the 2 site, just one line coming in. Only one source available on 3 transfer on trip.
4 MR. EBERSOLE: John, it has been, in recent years, 5 recognized that it's hardly -- hardly makes much sense to 6 connect critical loads to a generator which is guaranteed to 7 fail, when you have trouble with a reactor and thus lose 8 power, and makes more sense to connect it to the off-site 9 power supplies. But it requires the presence of these great 10 big generator disconnects to make it work.
11 If you looked at that degree of fine structure to 12 see if that gives you great availability of AC power to O 13 design it with a second generator circuit breaker, and then 14 have really the normal critical loads supplied by off-site 15 power. Do you follow me?
16 MR. FLACK: We haven't looked at that 17 specifically, but I believe units are putting in some 18 disconnects. I believe that was mentioned. But we haven't 19 looked at that specifically. As to the benefit of that, you 20 mean bringing the power back in through the unit bus?
21 MR. EDERSOLE: It would appear that that would 22 make tho ideal combination.
23 MR. FLACK: Way to sot up, yes.
() 24 MR. EBERSOLE: Which was really an extension of 25 historical practico in the coal burning, and rather take the Aci!.FitniiRAi. Riil'ORTliRS, INC.
) 202 347 37m Nationwide Omcrage *n3% ud6
/
30853.0 cox 41 1 toads off the grid for critical surfaces, and let the 2 generator go ahead and fail.
3 MR. FLACK: We were modeling the data, and we can 4 look into that specifically.
5 MR. BARANOWSKY: John, I am not quite sure I 6 understand the question.
7 MR. EBERSOLE: Let me explain it. Historically, 8 one always connected the supporting loads to the generator 9 that was prodacing the juice, main plant generator. In a 10 coat-burning plant, you didn't need att those circumstances; 11 it tripped out anyway. So the power supplies to the plant 12 auxiliaries went dead. That almost, I guess, I don't know, O 13 sust was eutometice117 adopted as a routine des 19n gractice 14 when the nuclear plants came along and failed to recognize 15 you are going to have a tot of critical AC toad you need to 16 keep supp11ed if the generator was gone.
17 Anyway, it persisted. So you hook the critica1 18 auxiliary loads of generator outputs, which is guaranteed to 19 fail; are you with me?
20 MR. BARANOWSKY: Yes, I understand.
21 MR. EBERSOLE: Then, in the last few years, it's 22 been recognized a better process may be to not connect the 23 aux 111arios to the piant output itself, but connect it to the 24 grid. But then that requires the interposition, or, 25 certainly, it's possible to have the interposition of the Aci: FitniinAi. Riti>onTiins, INC.
202 m .n m Nationwide unerage 8mnwM6
- i
30853.0 cox 42 l
1 circuit breakers so that you can disconnect the generator O 2 without affecting grid stability, and even have, then, access j 3 to two sources.
4 MR. FLACK: Back feed.
5 MR. BARANOWSKY: I guess that's not necessarily 6 the situation that I have seen. I have seen plants which run 7 their power right from the switchyard, referred power supply, 8 to it is ESF buses, as opposed to taking off of the main 9 generator, and plants that take off the main generator 10 usually have some kind of fast transfer to preferred power 11 supply, given that the main generator is down.
, 12 What you are talking about, I think, is the
() 13 ability to use the off-site power source -- if the off-site 14 power supply is dead --
15 MR. BARANOWSKY: Let me see --
16 MR. EBERSOLE: The off-site power source will be 17 dead. You do know that. ,
18 MR. BARANOWSKY: Off-site, yes. If the on-site 19 power source is gone, you are going to disconnect that power 20 supply at the switchyard from the rest of the grid. You are 21 going to have another way of bringing power into the plant.
22 That source of power is available, just because the plant 23 trips doesn't mean you lose off-site power.
(} 24 MR. REED: I am trying to figure out what Jesse is
! 25 objecting to, too. Are you objecting to -- I think it's very Acn-Fnot: Rat RiteoRTt:Rs, INC.
l 202 w r oo Nanonwide nnerage m 33u6u,
30853.0 cox 43 1 standard practice in nuclear plants. I am not too familiar 2 with coal plants, but it's been the practice that you have 3 service station transformers comparable of carrying all the 4 AC, or you have unit auxiliary transfers. When you are up 5 and firm and running, a lot of people go on'the unit 6 auxiliaries and rely on fast transfer. Are you objecting to 7 fast transfer of the unit auxiliaries?
8 MR. EBERSOLE: No, I am objecting to going to unit 9 outputs and forcing yourself to have to make a transfer when 10 trouble starts.
11 MR. REED: You are objecting to the transfer.
12 MR. EBERSOLE: The way the standard design is,
() 13 when you have trouble with a reactor, you invite a transient, 14 and you also -- with the design that connects the critical 15 loads to the generator output, you also require a transfer to 16 the grid.
17 MR. REED: Reliabilities of fast transfer. I 18 thought it was very good.
19 MR. BARAN0WSKY: Very high.
20 MR. EBERSOLE: It's low enough to require that 21 Westinghouse require that you don't undertake it. Therefore, 22 they have a 30-odd second delay of not longer than that to 23 avoid transfer to protect their pins against over temperature 24 for certain conditions preceding transfer. That's a standard
}
25 Westinghouse feature. Don't transfer, for heaven's sak.>,
Acn-FnonnAn RnvonTens. Isc.
202-347 3700 Nationwide Coserage 80L33MM6
30853.0 cox 44 1 hold onto the generator, for X seconds. I don't know what it 2 X is. Don't invite failure to transfer, on the grounds that 3 that, apparently, statistically, offers a risk to fuel pin 4 damage that they would rather not incur. Look it up; it's in I 5 the books.
{ 6 MR. REED: I don't understand that at all.
I 7 MR. EBERSOLE: It's a standard Westinghouse design 8 feature.
9 MR. BARANOWSKY: Well, all I can say is that data 10 that we have on total loss of off-site power doesn't show 11 that particular scenario in which the fast transfer failures 12 is a contributor to total losses of off-site power. There i
() 13 may be some partial failures that don't allow fully redundant ,
j 14 electric power sources to be available, say to both l 15 divisions.
16 But from what I have seen, I can't think of any 1
l 17 cases in which a fast transfer on a trip of the reactor i 18 resulted in a loss of off-site power. I can't think of any.
19 MR. EBERSOLE: Charlie, I am going to invoke your i
20 judgment that I think the best way to do this is to use
- 21 auxiliary power off the generator but with a circuit breaker; 22 correct?
23 MR. REED: He knows I will disagree with him.
i 1
(} 24 MR. EBERSOLE: In that case, you get the stability j 25 of the turbine generator output over and above that of the ace FEDERAL. REPORTERS, INC.
! ' 202 347 37(n) N.itionwide Coserage MIL 33MM6
..,- ,.._~,~._ _ _. ....-.... -.-._ _ - -, -
30853.0 cox 45 1 grid.
2 MR. WYLIE: I think it probably hasn't got a lot 3 of value for this meeting, but I think what you are talking 4 about is tripping the turbine. -
5 MR. EBERSOLE: Yes. It seems a little irrational 6 to guarantee you are going to lose your AC power when you 7 trip the reactor in trouble.
8 MR. RUBIN: Can I make a comment that if the AC 9 power were lost because of a plant trip and subsequent 10 failure to auto transfer, that will have appeared in data on 11 the total loss, would have been considered total loss of 12 off-site power. As mentioned, we haven't seen that occur in
() 13 the data. We have seen a number of numerous other events, 14 which John will discuss a little further, those do occur.
15 MR. EBERSOLE: I certainly invite you to look at 16 the Westinghouse standard design features.
17 MR. REED: The only thing I can relate to what we I 18 are talking about, Jesse, is some Westinghouse plants, when 19 the boiler feed plants are electric-driven, they leave the 20 generator field energized for one minute or so, to crank the 21 feed pumps. This other thing, I don't even know what it is.
22 MR. EBERSOLE: It's not my recollection that a 23 subordinate requirement is you have AC driven feed pumps for
{) 24 this requirement.
25 MR. REED: They have AC-driven --
ACE-Fnonnan Revoninns, INC.
202 347 3700 Nan .nwide cmcrage N 0-3.wou6
30853.0 cox 46 1 MR. EBERSOLE: No, no, turbine-driven feed pumps.
2 MR. REED: I don't know what you are talking 3 about.
4 MR. EBERSOLE: It's in place.
5 MR. WYLIE: Maybe we can talk about that later.
6 MR. EBERSOLE: I think it's below the level of 7 generalities.
8 MR. FLACK: Okay. So the modest differences in 9 design, having its impact on duration, as John has clarified, 1
10 one, faster recovery than I-3, modest change in the plant 11 centered category.
12 For grid,, we found no correlation between design ,
O 13 location and frequency. If there was a -- essentially, if 14 there was a weakness and it was detected, it was corrected.
15 So it was really the unidentified weaknesses that resulted in 16 grid failures. We couldn't associate a certain frequency 17 with a specific grid and look at different grids and say they 18 have this frequency because of their design. So we looked at 19 site, plus utility, and get to a larger area, but -- council 20 will get to a larger area. But we give it a site frequency 21 on an historical site, but we couldn't say there was flaws on 22 a specific grid or that there was a problem with the design 23 of that grid specifically.
() 24 Again, since the recovery --
25 MR. REED: Weren't there a lot of problems with l ACi! Fl!DliRAI. Rlti>oRTiins, INC.
I :o:.uM7m Nationwide Cmcrage m.Hto tMi i l
. _ _ _ . . _ _ _ - - - _ - _ _ _ - _ _ - _ _ - _ - _ . _ - - _ - . . . - - _ ___-_. - _ . - a
30853.0 cox 47 1 the switchyard and grid at Haddam in the early days?
2 MR. FLACK: Haddam Neck?
3 MR. REED: Haddam Neck, yes.
4 MR. BARANOWSKY: I think there were switchyard 5 problems which we tried to treat separate from the grid. The 6 one place we identified grid problems was in the Florida 7 Power & Light system, that particular utility in their 8 sites.
9 For other sites, it was kind of hard to generalize 10 expectations in terms of what is based on the past. For 11 instance, plants that were subject to the great northeast 12 blackout wouldn't be subject to the same type of failure in 13 the future. We couldn't use that as a predictor of the 14 future because of substantial design changes that were made 15 to that whole grid and all the interconnections.
16 MR. EBERSOLE: You are still talking about plant 17 contered problems.
18 MR. FLACK: This is the grid now; this is the 19 second category.
20 MR. EDERSOLE: Power floor and light --
21 MR. FLACK: Three categorios, we moved on to 22 grid. We are looking at grid now.
23 MR. BARAN0WSKY: Although we said we couldn't find
() 24 correlations that allowed us to go and look at a design or 25 design factors, reserves and things liko that, that would Aci 17 tintinAi. RitronTitas, INC, L 00:44747m Neionwide cmenge mawwo
30853.0 cox 48 1 allow us to predict grid frequency, what we felt was that the 2 identified a problem, and so did the utility, of course, at 3 Florida Power & Light, so we tried to separate that 4 particular system from the balance of plants in the United 5 States, and we used more or less of an average for the 6 balance of plants in the United States; and for Florida Power 7 & Light we tried to use their operating experience, giving 8 some credit for the fact that they have made some 9 improvements in their system and their frequency is coming 10 down.
11 MR. EBERSOLE: You are saying this entity is 12 exclusive with the Florida system?
13 MR. BARAN0WSKY: No. This is just the approach 14 that was taken in terms of trying to find correlations in 15 which we could, say, reserve margins or numbers of 16 interconnections and things like that might contribute to 17 grid reliability. We felt like that would be such an 18 oxtensivo and questionable analysis. We used a more general 19 approach in just looking at utilities and sites to see what 20 their experience has been.
21 MR. FLACK: Giving them a sito frequency, quoting 22 them.
23 Wo did allow for recovery, the sito had tho
() 24 capability of a black start, power source available. There 25 is an onhanced recovery curvo, which gives 80 porcent Acii Frii)iinai. IttironTrias. INC.
202 147. n o N.monwwe cmcrue mu lw uw,
1 l
30853.0 l cox 49 1 recovery in half an hour. That could be shown on this slide O 2 here.
3 So here we have the data for grid failures, 4 events. Our standard recovery curve, and then an enhance 5 recovery curve which allows 80 percent recovery in half an 6 hour, or .2.
7 The next category we looked into was weather 8 events, or weather related events.
9 Here we identified several hazards, events, or 10 losses of off-site power due to snow and ice, tornados, 11 hurricanes, and another category which is not on here, salt 12 spray.
13 We have associated the hazards at a specific site, 14 according to the storm frequencies.
15 MR. EBERSOLE: Was extreme weather a major part of 16 the Florida problem, or some other factor?
17 MR. FLACK: Grid played the largest role. Weather 18 also plays a larger role. Florida Power has been lucky the 19 last few years. So it's a combination. It's really a 20 combination.
21 Again, the recovery is off-site, so duration is 22 going to be dependent upon procedures. We also allowed for -
23 recovery in the weather, which is not listed here, of 80 i
() 24 percent recovery in two hours.
! 25 We also have in our model, extremely severe ace FilonnAi. REvonnins, INC.
j 1 202 m 37m Nationwide Gncrage 8m3 %u46 L __
30853.0 cox 50
- 1 weather which is wind speeds greater than 25 miles an hour O 2 which we do not allow recovery. We are assuming the failures 3 are so gross they cannot recover from them. Our recovery 4 curve for that type of an event is fit into the data shown.
5 Our standard recovery curve and our enhanced recovery curve.
i 6 MR. BARANOWSKY: Could I add in here, Mr. Reed, I i 7 think, asked about do we give credit for people or sites that 8 have certain recovery procedures and things like that. This 9 enhanced recovery curve you see there, which shows, at least 10 from a probabilistic point of view, a significant 4
j 11 improvement, is based on considerations for sites which would 12 have the capability and procedures available for recovering 13 from the types of weather hazards that we have identified 14 that have been shown to be important.
l 15 What we tried to do was find a typical type of
! 16 recovery capability that would make a difference in our 17 probabilistic analyses, because there's a spectrum of l 18 possibilities and availabilities of sources of power to 19 recover from severe weather hazards.
I 20 In this case, we would associate that type of
- 21 improved recovery for a site that has procedures and has 22 equipment that could be protected or shown to be operable for l 1
23 the type of weather hazard we are looking at. !
i O 24 xa. aseo, oc those procedures deal with both 25 anticipatory as well as after the fact, versus, in one case i ace FEDERAI. REPORTERS, INC.
202 147-3700 Nationwide Cmcrage mn.33MM6
30853.0 .
cox- 51 1 in which I was involved in, total loss of AC, it was hours O 2 before all the lines were out that signals, telephone 3 communications are coming into the control room, saying, hey, 4 get ready for it, get ready for it, the ice is bringing 5 everything down.
6 Another case, which was the great northeast 7 blackout, things were fairly quick. So, that is from a local 8 realization. If somebody way out there knew that things were 9 getting into trouble, but they were not communicating with 10 all of the plants around through New England, New York, area, 11 and saying, hey, something is going to hit you. Anticipation 12 is part of the procedure; is that the question?
13 MR. BARANOWSKY: I think anticipation would be 14 shown by reducing the frequency so that the contribution 15 associated with long weather outages would be essentially 16 minimized. We could account for, but we would have to know 17 that that anticipation could result in successful execution 18 of alternate power supplies. I think in your case you are 19 talking about a gas turbine generator. If, for instance, one 20 could verify that it was operable and had the thing ready to 21 go, then the expected frequency for losses of power under 22 severe weather conditions, which wouldn't affect the gas 23 turbine generator, would drop down. That's how credit for
() 24 that type of thing would be taken.
25 MR. FLACK: To summarize the model, the analysis ACE-FEDERAL. REPORTERS, INC.
20 wnm Nationwide cmerage 8a> 33w,56
30853.0 cox 52 1 that was made, we found the following factors that affected 2 the off-site power availability. Redundancy and independence 3 of off-site power, circuits, alternate power sources 4 available in the plant centered events that was a factor in 5 the duration.
6 Operational procedures, restoring power at the 7 site and coordination with the dispatch system off-site for 8 grid and weather events; and location plays a role, 9 especially with regard to the weather hazards at the site.
10 We compared our model to PRAs. For example, our 11 model and Limerick model compare very well, with frequency 12 curves. In that case it agreed very well. In the Millstone 13 case we had somewhat different results. Our model was based 14 on historical data at the site.
15 MR. BARAN0WSKY: Let me add, Jesse, you were 16 talking abnut averaging things. It's my feeling that the 17 Millstone results here would show a fairly low loss of 18 off-site power frequency and duration curve, have to do a lot 19 of averaging in, occurs at other sites that have rather 20 prompt restoration. Millstone analysis that we did involved 21 getting much heavier credit for the longer duration outages 22 associated with the hurricanes that they have had at that 23 site. That's the way we tried to differentiate.
~( ) 24 Now, for every plant that we predict a high loss 25 of off-site power frequency, obviously, the average is going ACE FEDERAL REPORTERS, INC.
- 202-347-3700 Nationwide Coserage 800-336-6M6 h
30853.0 cox 53 1 to be maintained the same. So there will be plants that have 2 a low loss of off-site power frequency. That's how we 3 attempt to find the so-called bad actors.
4 MR. FLACK: The next analysis that was performed 5 was the cluster analysis. This was to allow the results to 6 be used within a regulatory framework. What we did here was 7 find a way of grouping together sites with similar losses of 8 frequencies and durations of off-site powers, although they 9 may have different factors in grouping them, so that we had 10 one recovery curve in duration and one frequency associated 11 with that group of plants.
12 This way we would end up with a few curves that 13 would represent the whole spectrum of possibilities.
14 Those curves, five curves, look like this. So all 15 our plants would fall on one of these curves.
16 Below are two curves and say most.
17 And most of the units. There are a couple of 18 plants up here, I believe five plants on the second line.
19 In summary, we found loss of off-site power to be 20 about one out of every 10 years, nationwide average. Again,
, 21 we developed a model, loss of off-site power model that 22 discriminates between plant design, operational 23 characteristics, and location factors. This model, combined O
~
. 24 with the emergency AC reliability, provides the station 25 blackout analysis.
ACE FEDERAL REPORTERS, INC. '
202- m - M OO Nationwide Coverage 800-336-6M6
. a
30853.0 cox 54
~ l Now, I would like to turn it back over to Pat V
2 Baranowsky, who will present the emergency AC reliability f 3 part of the model. l l
4 MR. BARANOWSKY: I will give a brief picture of l l
5 what we did on the emergency AC reliability analyses, look at l 6 design features, reliability data associated with the 7 performance of diesel generators, look at the common cause 8 failures associated with emergency AC power, systems design, 9 and we performed the reliability analyses for those differert 10 designs associated with the different operational data.
11 I just want to make one point with this Vugraph, 12 and that we tried to look at the redundancy considerations in O 13 terms of diesel generator, success criteria, for a loss of 14 off-site power, as opposed to a large LOCA. You need less i 15 emergency AC power supplies or power at units for loss of l
l 16 off-site power system than you do for the large LOCA. So in 17 terms of looking at system reliability, we had success 18 criteria which were based on loss of off-site power, not 19 design basis load to accident.
20 In that regard, we felt that the plants could be 21 subdivided into power supply redundancy groups, as I have 22 identified here, ranking from one old and smaller plant with 23 one diesel, requiring one diesel, not really of significance
() 24 here; and typically involving plants, one unit plants with 25 diesel generators dedicated to that unit with two diesel ACE FEDERAL REPORTERS, INC.
ll 202 347-3700 Nationwide Coserage 800-336-6M6
30853.0 cox 55 i l
1 generators and one required, some with three diesel 2 generators, one required for loss of off-site power, and one ,
3 or two, perhaps, with four diesel generators, of which they 4 might require two.
5 MR. REED: There is a size aspect here that 6 relates to reliability too. Is that figured in?
7 MR. BARANOWSKY: You mean in terms of smaller 8 diesels might be a little bit more reliabilities than 9 larger?
10 MR. REED: I think if certainly diesel 1000 11 kilowatt to 3000 kilowatt ranged probably a hell of a lot 12 better reliability record than those to get'up toward 13 10,000.
14 MR. BARANOWSKY: Okay. I think maybe this is a 15 time to talk about factors that are associated with the i 16 reliability of diesel generators. We looked at the l 17 subsystems that were involved in failures, and, to some 18 extent, causal factors, including things that might be i
19 related to size, or inherent characteristics of diesel i
f 20 generators. We found a slight correlation associated with I
- 21 size of diesel generator.
l 22 But what we really noticed, more than anything 23 else, was that it's possible to have extremely high
() 24 reliabilities on diesel generators, no matter what size it 25 is. It's also possible to have fairly unreliable ones, no ACE FEDERAL REPORTERS, INC.
l l 202 347-3700 Nationwide Coserage 804 336-6646 L
1
30853.0 cox 56 1 matter what size it is, and has more to do with knowledge, I O 2 think, about the diesel generators, how they are maintained, 3 and working the bugs out in particular for the newer ones, 4 than it does in just talking about size or other aspects 5 associated with design.
6 What we felt, after having looked at the data and 7 doing some statistical analyses, was that we couldn't 8 identify particular design factors or operational factors 9 that one could talk about and make, let's say, fixes to 10 improve diesel generator, that there were too many unique 11 aspects to the failures that we saw that were related more to 12 programmatic things like the quality of maintenance and the 13 quality of initial check out, than the to the particulars of 14 design.
15 MR. EBERSOLE: One of the more interesting recent l
l 16 experiences has been that of the trans-America valves. What 17 did you find that you want to comment on that?
18 MR. BARANOWSKY: I think that's a good example of 19 -- size is probably not as important there as the fact that i 20 there was a design flaw and probably failure to properly 21 check the machine out before it was installed in nuclear 22 power plants. When those things were corrected, the 23 trans-America generators, to the best of my knowledge, have
() 24 reasonably reliable operation. I have seen other diesel 25 generators with problems that are associated with design that ACE FEDERAL REPORTERS, INC.
) 202-347-3700 Nationwide Coserage 800-336-6M6
30853.0 cox 57 1 were latent and showed up after several years of operation.
2 That's why you can't make those types of generalities. It's 3 the type of thing that you have to learn by past operating 4 experiences and have good practices in order to assure high 5 diesel reliability as opposed to a specific element of 6 design.
7 In general, the simpler the machine, the more 8 chance you have for maintaining it reliably. Smaller ones 9 have some simplicity, pluses associated with them. But we 10 couldn't find anything major in terms of correlating factors 11 associated with design. ,
12 MR. EBERSOLE: You are saying you find no spread i 13 in what I would call operational quality or reliability, from i
14 expensive to cheap diesels?
15 MR. BARAN0WSKY: I don't know about cost. You can 16 get what you can get in terms of price you pay for it.
17 Certainly the -- I don't know that they skimped on
- 18 the cost of the trans-America deal of valve diesels. They
,' 19 probably spent some money there, but there was a breakdown in 20 the process of dt;ign and commissioning those units. I don't 21 think it was the fact that it was a cheap diesel. I bet the 22 utility will tell you it's expensive.
- 23 MR. EBERSOLE
- I guess not.
I
() 24 MR. BARAN0WSKY: That's what we ended up looking 25 at was the operating reliability at sites versus design ACE FEDERAL REPORTERS, INC.
- 202-347-3700 Nationwide Coverage 8%336-6M6
- . , m___. , _ - _ _ _ _ , _ . , _ . _ _ _ _ . _ , _ _ _ _ _ _ . _ _ , _ _ _ _
30853.0 cox 58 l
1 characteristics. We think the better indicator is to go
[h
~
2 directly to performance as opposed to talking about design 3 factors. Certainly, you can learn from practices and 4 improvements in design that have been made at other units.
5 But, in general, we couldn't identify the types of design 6 features that would give you almost automatically high 7 reliability. It's just not going to happen.
8 MR. MICHELSON: Just for my own edification, to 9 your knowledge, were there any kinds of changes made to 10 diesels to qualify them for nuclear service that then led to 11 the reduction in the reliability of the machine as a 12 consequence?
13 MR. BARANOWSKY: Probably the biggest problem is 14 the way we use them in nuclear power plants. They are all 15 designed for continuous operation, either as peaking units or 16 engines and lobbying motives, and we have a different type of 17 operation in a nuclear power plant. As far as 18 precommissioning goes, we have a lot of fast starts on the 19 diesel generators, which probably builds up stresses and 20 causes an earlier loss in terms of the expected lifetime, 21 than one would normally expect from diesel generators, which 22 should be reliability. We have a requirement to start them 23 up promptly, so we have a lot of things hung on the diesel
() 24 generators in terms of the controls to bring the diesel 25 generator up to speed quickly and allow rather rapid loading ACE FEDERAL REPORTERS, INC. I 202-347-37W Nationwide Coserage 800-336-6M6 1
30853.0 cox 59 1 of the diesel generator. You will see that you look at O 2 operating data, those types of things contribute to failure 3 as much as anything else.
4 MR. MICHELSON: I guess there weren't any design 5 changes that were required for nuclear application, as far as 6 you know?
7 MR. BARANOWSKY: I would say mainly in terms of 8 the fast start requirement, those are the types of design 9 changes, but not internal to the engines that I know of.
10 MR. MICHELSON: It's a fast start. Your machine 11 is not designed for fast start versus slow start necessarily, 12 in the sense of changing the bearings or the crank shafts or
( 13 anything like that?
14 MR. BARANOWSKY: Not that I know of. These 15 machines are designed for other commercial applications.
16 Nuclear is just a small percentage of what they sell them 17 for.
18 MR. RUBIN: I just want to comment, the one design 19 change that in order to accommodate the accommodate for fast 20 starts, many of the machines have continuous prewarming. So 21 there is one design change to try to alleviate the stress 22 effects.
23 MR. MICHELSON: Wouldn't lead to a loss of
() 24 reliability in any way?
25 MR. RUBIN: No, just relieve the problem of cold ACE FEDERAL REPORTERS, INC.
202 347-3700 Nationwide Cmerage 800-336-6M6
- .. - . ~. ._ .- _ _. _- - .
30853.0 cox 60 1 -- fast starting.
2 MR. MICHELSON: It didn't become a problem in 3 itself from fast starting it.
i 4 MR. EBERSOLE: Prelubing can bring its own 5 problem. If you lubricate the valves and they have leakage, l 6 it can blow the engine.
7 There is a little contradiction here in what we i 8 heard from the Japanese about a year, year and a half ago.
o
- 9 We asked them to give us the same sorts of statistics on the i 10 diesels. They said they had no statistical record of any 4 11 failures on their diesels, diesel starts. I wondered if you 12 had probed into that.
l 13 MR. BARANOWSKY: Depends on how you do your I
14 counting.
15 MR. EBERSOLE: In further questioning, it seems I
16 they have diesels which are designed for the service rather 17 than simply bought off the market from some other service or 18 supplier. They evidently must do something to the design to 19 make this business of fast start amenable to their design. I I 20 don't know what it is.
21 MR. BARANOWSKY: They, first of all, don't have as 22 many diesel generators operating as we do. They probaSly 23 have a higher uniform quality of operations and maintena: ca
() 24 associated with the diesels than we have in the United 3
25 States. I can tell that you there are plants in the United ACE FEDERAL REPORTERS, INC.
j 202-347-3700 Nationwide Coverage 800-336-(486
r
]
1 30853.0 cox 61 l l
s 1 States that have diesel performance that will rival anyplace b 2 in the world. We have plants that will have diesel 3 performance that will embarrass the United States.
4 MR. EBERSOLE: You have this spread.
5 MR. BARAN0WSKY: Yes.
6 MR. EBERSOLE: We have to look at the bottom end 7 of the spectrum and fix it.
8 MR. BARANOWSKY: You will see that part of our 9 proposal for resolving this issue is to try to cut that 10 spread down.
11 MR. EBERSOLE: Right.
12 MR. BARANOWSKY: Let me move along. The only D) k- 13 reason I am showing this slide is to indicate that back when 14 we originally talked about diesel generator performance, we 15 had a data base for most of the power plants which involved 16 some 13- or 14,000 demands on diesel generators, and 17 performance that indicated about one failure, in every 50 18 demands. Subsequent to that, we had some data compiled by 19 EPRI, their NSAC people, which involves some 22,000 demands 20 over about a three-year period.
21 They found diesel generator failure rates have 22 dropped slightly from.what we observed back in the time frame 23 through 1982.
() 24 But they still found the spread in diesel 25 generator reliability. On the average, you.get numbers like ACE FEDERAL REPORTERS, INC.
202-347-3700 Nationwide Coserage 8 % 336-6646
30853.0 cox 62 1 99, 98 percent reliability, but we have seen units with O 2 reliability of -- on any given year, let's say, or any given 3 100 demands, 94 percent or something like that, or 100 4 percent.
5 MR. WYLIE: What is the definition of failure?
6 MR. BARANOWSKY: This involves failure to either 7 start or load under conditions that would be associated with 8 a loss of off-site power. It could be starting loads --
~
9 MR. WYLIE: This is not the definition, say, for 10 example, if a diesel is tested, and it fails to come up in 11 the allotted 10 seconds, it's called a failure?
12 MR. BARAN0WSKY: No, this is not the reg guide 13 1.108 failure definition. If the diesel generator started in 14 15 seconds, we call it a success.
15 MR. WYLIE: In these numbers?
16 MR. BARANOWSKY: Yes. If the diesel generator 17 didn't auto start, but the operator could start it from the 18 control room, we call it a success. Those are not the large 19 contributors to failures, by the way. People get excited 20 when they are included in these numbers, because you want i
21 them to be as precise as possible. But they represent 22 something like a 5 percent of all failures, those types of 23 things.
() 24 Any hour, they are not included here.
25 MR. MICHELSON: Are you going to tell us what are ACE. FEDERAL REPORTERS. INC.
202-347-3700 Nationwide Coserage feb336-(M6
_ _ _ _ _ _ _ _ _ _ _ . . _ ~ _ . _ _ _ _ _ _ _ , _ _ _ _ . _ _ _ _ _ _. _ _ - . . . _ _ . _ . _ . , _ , _ _ _ _ . _ _ _ .
.. . . . . ._ =. -. . - . -
30853.0 cox 63 1
3 1 the big contributors?
- C:) 2 MR. BARANOWSKY: I have to say the big 3 contributors are spread all over the place. They cover all 4 the subsystems of the diesel generator. There is a picture 5 in NUREG-1032 which shows all the subsystems. In terms of
, 6 problems, we see things like inattention to detail during 7 maintenance is an important consideration. Latent design 8 flaws --
9 MR. MICHELSON: Is there a big contributor. Are 10 there two or three big contributors? Are there 100 11 contributors?
i 12 MR. BARANOWSKY: There are 10 or 20 things that 4
1 13 can contribute. It's hard to say, until you see it occur, 14 whether you are susceptible to that type of problem.
15 MR. MICHELSON: Let me ask my question one more 4
16 time. Are there any big contributors, or are there just a 17 large number of small contributors?
18 MR. BARANOWSKY: I would say less is -- there's a 19 large number of small computers.
20 MR. EBERSOLE: Is there any personalized attention 21 that's given to diesels. I am saying you take one or two or 22 X-diesels, and say to somebody at the plant, they are your 23 baby, I want them to run when they have to run. Do you have '
() 24 a focused, personalized attention, to the condition of the 25 diesels at the working plant, or is it just an impersonal, ace. FEDERAL REPORTERS. INC.
202-347 3700 Nationwide Coserage 80)-336-6M6
30853.0 cox 64 1 run-of-the-mill, care-taking function?
O 2 MR. BARANOWSKY: You can't generalize.
3 MR. EBERSOLE: You can't?
4 MR. BARANOWSKY: No. I would say I have been to 5 several sites, and at some sites --
6 MR. EBERSOLE: There is no unity of 7 responsibilities with respect to diesel reliability at the 8 plants?
9 MR. BARANOWSKY: No.
10 MR. EBERSOLE: None at all. So they are just 11 another feature of the plant.
12 MR. BARANOWSKY: I am not saying that. What I am 13 saying is that the treatment of the diesel generators, in .
l 14 terms of concern for its reliability, seems to vary, just l
15 like the treatment of any kind of pump or valve or circuit 16 breaker reliability would vary.
17 MR. EBERSOLE: There is no focus of particular 18 interest on the diesels.
19 MR. BARANOWSKY: Depends on the site. If you 20 haven't had problems with it, perhaps your programs are 21 adequate, the way they are. But those sites that have had 22 problems, hopefully they make some changes. In fact, we did 23 some time analyses, and we were, at one time, showing things 24
(]) like cyclical reliability, reliability peaks and valleys 25 ranging from two to four years, which is interesting, because ACE FEDERAL REPORTERS, INC.
202 347-3700 Nationwide Cmerage MM)-33M686
30853.0 cox 65 ,
i
_ 1 when we do reliability analyses, we are generally talking
+
%/
2 about demands that cover two or three years in order to get 3 100 demands in there. That's enough time to have a turnover 4 in personnel or a learning or forgetting factor come in 5 there.
6 MR. EBERSOLE: You find nobody, then you can't go
- 7 into a plant and say who is in charge of diesels.
8 MR. BARANOWSKY: I suspect you can. But the 9 Japanese site that you brought up, there is a person who 10 takes it very personally there.
11 MR. EBERSOLE: That's what I meant.
12 MR. BARANOWSKY: It could vary at nuclear power .
13 plants.
14 Let me just point out that we looked at causal 15 information, because we wanted to do an update to our common 16 cause analysis. All I am showing here are the areas in which 17 we observed either actual common cause failures, or events, 18 degradations in performance, but not necessarily failures, 19 that looked like they could be common cause failures if 20 things were uncorrected. It really covers the spectrum of 21 design, operations, plus we see things like'the dependence on 22 support systems associated with DC power or cooling or 23 ventilation systems. .
() 24 In fact, there were a couple of reports recently 25 in which the failure, I think it was, of nonsafety air l
/\CE. FEDERAL REPORTERS, INC.
( ,
202 347-3700 Nationwide Cmerage 800-336-6646 1
30853.0 cox 66 1 systems, could cause loss of ventilation to diesel generator O 2 rooms, which would result in the temperature in those rooms 3 exceeding the point at which the diesels were qualified for 4 nuclear service. That's just another interaction associated 5 with dependent systems. We have seen things that are 6 associated with potential fires or floods that can cause 7 multiple diesel generator failures.
8 There have been events involving severe weather, 9 either too hot or too cold, and the HVA systems either 10 weren't working or weren't adequately sized to prevent diesel 11 generator rooms from reaching temperatures that caused 12 problems with diesel generators.
13 Then we get down to things like seismic. Diesel 14 generators in general are designed for seismic category 1, 15 design basis, requirements, and this is where I think maybe 16 we can address the question on seismic t, hat you had raised, 17 and let me just tell you what we have done on this project.
18 We have looked at PRAs and the potential for seismic events 19 causing loss of off-site power and loss of on-site power. We 20 note that in most instances, in fact in every instance that I l
21 could identify, a seismic event well beyond the design basis 22 of the plant is necessary in order to cause failure of the 23 diesel generators that have been designed to normal seismic O 24 cete9orv 1 reamirements- rvvice117, two or three times the l
25 SSE is required. l l
ace. FEDERAL REPORTERS, INC. l 202 347-3700 Nationwide Coserage 80)-336-6M6 l
30853.0 cox 67 1 MR. MICHELSON: Does that include the auxiliaries O 2 needed for that engine?
3 MR. BARANOWSKY: Yes. It's not just the engine 4 that would fail, it could be anything that is associated with i
5 the support systems --
4 6 MR. MICHELSON: You are saying you have looked at 7 the battery banks and so forth, and they can stand several --
8 they can stand well beyond the SSE for that particular --
J 9 MR. BARANOWSKY: I am not saying'I have looked at j 10 it. I am saying it's been looked at by other people.
11 MR. MICHELSON: I am just asking. They really did
- 12 look at all the supporting auxiliaries; they look at all the 13 fire protection features to make sure they don't go off.
14 They have looked at everything that might affect engine 15 operation and they are making a statement they can go well 4 16 beyond the SSE.
17 MR. BARANOWSKY: I suspect there are things that 18 could be found that weren't looked at.
j 19 MR. MICHELSON: I would suspect so. I would be j 20 very careful about making that broad of statement. The 21 engine -- I think that you are probably pretty comfortable 22 making that statement. Go into the auxiliaries and you may 23 be better plant specific.
() 24 MR. BARAN0WSKY: Let me just make my point. That 25 is that whether they can or cannot stand two times SSE are i
/\CE-FEDERAL REPORTERS, INC.
202-347 3700 Nationwide Coserage 80tk336-6646
30853.0 cox 68 1 supposed to be designed and capable of withstanding at least O 2 the SSE.
3 MR. MICHELSON: For instance in the room, fire 4 protection.
5 MR. BARANOWSKY: Well, my understanding is that 6 features that share common locations with engineered safety 7 features that are designed to save shutdown earthquake or 8 seismic category 1, are to be designed such that their 9 failure under those loading conditions will not cause adverse 10 interactions with --
11 MR. MICHELSON: You have a misunderstanding then.
12 As to fire protection, if there are fire protection rooms in
( 13 the diesel room, they will certainly will not be activated by 14 an SSE, for instance.
15 MR. BARANOWSKY: I suspect that's a possible 16 problem.
17 MR. EBERSOLE: As a case in point, I was found 18 early on, that the fire protection relays were mercuroid 19 types, and they successfully closed the engine generators in 20 their own box in an attempt to respond to the need for carbon 21 dioxide submerging.
22 MR. BARANOWSKY: I am not trying to belittle the 23 seismic risk, by the way. What I am trying to say is that Q 24 plants should be designed to handle a safe shutdown 25 earthquake. Whether they are or are not designed, I think, ACE FEDERAL REPORTERS, INC.
202-347-3700 Nationwide Coserage 800-336-6M6
. -. ._= - ._ . .- - . - .-- _ . _ .
30853.0 cox 69 1 is an issue beyond station blackout. That's really what I 2 want to say.
3 The contribution to seismic, to loss of off-site 4 power, is fairly small. But the contribution to risk could l 5 be large, and it could involve things other than diesel 6 generators. So that really what I am trying to say is we t 7 know it's a possible contributor, but we haven't dwelled on 8 this one, just like we haven't dwelled on fire. They are an 9 issue unto themselves. They could impact things other than 10 the emergency AC power system.
i 11 MR. MICHELSON: As a simple-minded example that i 12 would give you a flavor for the problem, I visited one plant 13 that, of course, diesel room required extensive ventilation i 14 to keep it cool. But they use a little mouse trap
- 15 arrangement to close the ventilation of an arrangement in the i 16 unlikely event they had a fire. That little mouse trap was 17 actually a trigger which was, I am sure -- we went back to 18 look -- but I am sure it would have been very easy to shut 19 off by just shaking the equipment slightly. It was just that l
20 trigger sensitive. The arrangement was wrong. It was the 21 wrong kind of a trigger.
- 22 It was fine from the thermal linkage viewpoint.
s 23 Sure, all you had to do was heat up the room and the thermal f
i
() 24 link point would melt and slam that thing shut. But it was
! 25 like a mouse trap, it was easily set off, with temperature 4
l /\CE-FEDERAL REPORTERS, INC.
202-347-3700 Nationwide Coserage 800-336-6686 1
30853.0 cox 70 1 and seismic disturbance as well.
2 You have to be very careful to look at the plant 3 specific arrangements in such things as ventilation of the 4 room. Not just the fire protection, but look at other things 5 fire protection is doing, like shutting off ventilation ducts 6 and whatever to the room. If they are set with hairtriggers 7 on thermal links, you may be in bad shape. This one was a 8 very interecting trigger arrangement. Great from the fire 9 protection viewpoint, not so good from seismic.
10 MR. BARANOWSKY: In fact, what you are really 11 saying is there can be a lot of hidden problems at nuclear 12 power plants that we may have trouble knowing about today or 13 uncovering, until we see evidence. I would tend to agree, if 14 that's your point.
15 MR. MICHELSON: There is one little thing, I can 16 derive very little comforts from the PRAs in this regard, 17 because the PRA analyst is really unaware of these details.
18 Unless he goes in and does a very careful on-site detailed 19 examination, he won't pick up on such details.
20 MR. BARANOWSKY: Depending on the money you want 21 to spend, you can do that.
22 MR. MICHELSON: There is no question about being 23 able to do it.
() 24 MR. BARANOWSKY: What we did was we looked at 25 these things, and we incorporated factors in our analyses in ACE FEDERAL REPORTERS, INC.
202-347 3700 Nationwide Coserage MU.33MM6
l 30853.0 cox 71 1 which we were trying to do a quantitative reliability O 2 analyses, to account for these possibilities. Obviously, if 3 one does a thorough examination of plant design and 4 operations, there are ways to rule these things out, but not 5 having done that, one must account for it in your reliability 6 analysis. And we tried to account for it by putting in what 7 operating experience has shown us over the last 15 years in 8 terms of these types of failures in our common cause failure 9 analysis model. So that they are incorporated in there in 10 general, but it's not a plant specific consideration.
11 Because you have to go to the specific plant to really deal 12 with these things.
13 So using diesel generator reliability estimates 14 and common cause failure estimates that are derived from the 15 operating experience, we are able to construct models and 16 results. Is this thing blurry to everybody? It sure looks 17 blurry to me.
18 At any rate, all I am showing here is we looked at 19 results for a spectrum of diesel generator reliability 20 estimates, this would be system availability over here for, 21 say, loss of off-site power, versus emergency diesel 22 generator reliability, and we looked at variations and common 23 cause failure, depending on how well one would screen these
() 24 things out.
25 We looked at the different configurations. Here ACE-FEDERAL REPORTERS, INC.
202 347-3XV) Nationwide Coserage 800-336-6646
'30853.0 cox 72 1 we have three diesel generators, of which two were required, O 2 to cope with a loss of AC power, one out of two 3 configurations, one out of four and two out of three. What 4 one can see by the interplay of all these considerations, 5 it's possible to have system reliability levels that cover a 6 fairly wide range. These are the kinds of things that we 7 have to take into consideration as we formulate ideas on how 8 to put some sort of box on the station blackout issue, since 9 the possibilities for system reliability can vary
- 10 considerably.
11 MR. MICHELSON: Can I go back for just a moment 12 and ask a question about your previous slide. Why did you 13 leave out the fuel oil system as a common cause or possible 14 source?
15 MR. BARANOWSKY: I didn't intend to leave it out.
16 MR. MICHELSON: There have been a few examples of 17 late of water getting into fuel oil tanks and getting into 18 all the diesel tanks. Fortunately, they discovered it before 19 when they started the engine they found the water and got it 20 all cleaned up. But if they had had a demand at that point 21 for all the engine, there might have been a little 22 difficulty.
23 From the fire protection people, they discovered
() 24 one fuel oil tank around the country which has a fire sparger 25 inside of it for fire protection, not too fond of the idea.
ACE FEDERAL REPORTERS, INC.
202-347-3700 Nationwide Coverage 800-336-6M6
- . _ = _ . . .-- - _ _ . _ __
i 1
1 l
30853.0 cox 73 1 If actuated in an earthquake or other event, you might have O 2 lost all of your diesel engines from water in the fuel oil. j l
3 There are lots of clever ways to put in a potential problem j 4 that even common PRA doesn't pick up on. l 1
5 MR. BARAN0WSKY: We have identified that. It's in 6 our data base and incorporated in our analyses. We have 7 about 100 events, maybe 80, 100, I am not exactly sure of the 8 numbers, that involve either common cause failure or 9 potential common cause failure, like the one you just 10 identified. I didn't feel like listing every single one of
- 11 them. I just tried to pick the general areas out over here.
12 But I think we did a fairly extensive look at operating i 13 experience based both on LERs and getting information 14 directly from utilities to try to get as many as we could.
15 If we missed one or two out of 100, I would say 16 that from a probabilistic estimate, we have only a 1 or 2 17 percent error in the rates of estimation. What you miss, 4
18 though, is the cause and stability factors that people ought 19 to be aware of.
(
20 Just as a point of interest, there have been 21 events in which there were actual losses of AC power, not 22 just diesel generators, but two or more diesel generators, 23 whether they are due to common cause failure, or sometimes we f(])
i 24 have cases that involve one diesel down for maintenance, and i 25 there is a demand placed on the site, the second one fails, ACE-FEDERAL REPORTERS, INC.
202-347 3700 Nationwide Coserage 800-336-6M6
30853.0 cox 74
. 1 or possibly two of them fail, but for different reasons.
2 What I have shown here is some data points that 3 compare to our somewhat generic and hypothetical results, 4 data points from actual operating experience at nuclear power 5 plants.
6 In general, they seem to be falling around our 7 best estimates. If you recall, on the prior slide, I showed 8 that there was variability that was possible because of 9 uncertainties and pecularities associated with common cause 10 failure, which would put bands around these best estimates.
11 As far as we can tell, the models that we have are 12 reasonably well predictors or estimators of actual operating
/~)
5/
13 experience. It's a little bit cheating, I guess, to use 14 models that are based on operating experience to predict 15 them, but at least it shows some consistency.
16 To wrap up, what we can say is that the important 17 factors that we identified affecting on-site AC power 18 reliability have to do with the configuration numbers of 19 diesels versus numbers required for loss of off-site power, 20 reliabilities of diesels, vulnerability to common cause and 21 the various categories associated with design, human error, 22 environmental conditions and external hazards, and also 23 supporting systems, their dependence and potential
() 24 interactions. Particularly they noted things associated with 25 cooling, activation and control, as examples.
ACE-FEDERAL REPORTERS. INC.
202-347-370t) Nationwide Coserage 800-336 M46
30853.0 cox 75 1 MR. BARANOWSKY: We are going to take a little O-2 break here.
3 (Discussion off the record.)
4 MR. BARANOWSKY: So we take the off-site power 5 loss that we have and the emergency AC power models that we 6 have, emergency AC power includes failure to start, failure 7 to run, cause failure, system dependencies. We integrate 8 these things together to integrate loss of off-site power and 9 loss of emergency AC power as a function of time, for certain 10 design and location characteristics that we have been able to 11 identify. Therefore, we can start doing some discrimination, 12 at least on the loss of off-site power and on-site power or 13 station blackout as a function of plant differences.
14 When we do this, we are able to get -- essentially 15 assess sensitivity analyses which show the likelihood of 16 station blackout versus certain durations for different 17 characteristics of diesel configurations or diesel 18 reliability, and for off-site power characteristics. If you 19 will recall, John Flack, one of the last things he showed, 20 was some clumping or clustering together of off-site power, 21 design and location characteristics into these clusters. And 22 here we have, for instance, clusters 2, 3, 4 and 5 23 represented on this curve, combined with a 1 out of 2
() 24 emergency diesel generator configuration and a .975 emergency 25 diesel generator reliability. We can get some ideas on the ACE. FEDERAL REPORTERS, INC.
202-347-370) Nationwide Coserage 800-336-fM6
30853.0 cox 76 1 likelihood of losing all off-site power and on-site emergency 0_ 2 AC power for different durations and time.
3 MR. EBERSOLE: Were these clusters equal in 4 numerical number?
5 MR. BARANOWSKY: No. They were derived based on 6 combining characteristics that produce similar off-site power 7 reliability. So that when we say we have an off-site power 8 curve for cluster 2, what we mean is plants with a certain 9 set of design and location factors should have a similar loss 10 of off-site power frequency.
11 Because we have so many different things that we 12 are considering in terms of the plant switchyard design,
( 13 whether or not they are in a high or low grid group, whether 14 there is recovery capabilities, all of these things, there's 15 hundreds or even thousands of possible curves.
16 MR. EBERSOLE: What that will do'-- eventually, I 17 guess you are going to use that information to put a plant in 18 a cluster.
19 MR. BARANOWSKY: Right. In fact, the reg guide 20 has the cluster design characteristics reflected in it, so 21 that the requirements which you will hear about of X-hours of 22 coping and certain diesel reliability, reflect whether or not 23 the plant would be located in one, two, three, four or
() 24 five-type cluster.
25 So we do these kinds of analyses because we have ACE FEDERAL REPORTERS INC.
2C347-3Nm) Nationwide Coserage Muh33MM6
.- . . . .. . = - - -.- - -
I I
30853.0 cox 77 1 1 to have a capability of doing regulatory analyses and O 2 trade-offs. You can imagine, if we do a trade-off here, if j 3 you are able to cope with a blackout of eight hours versus 4 four versus two, we can see what the potential risk t 5 redundancy or at least get an idea for it. I wouldn't want 6 to say this is rigorous.
7 So we did that for all the different 8 characteristics. I won't show any more curves; I just 9 provided a few to you so you could get some ideas.
10 Having done that, looked at all the different 11 trade-offs that are possible, we could draw some conclusions 12 in terms of what effect station blackout likelihood and
!' 13 duration.
14 What we found is that the -- as you have heard 15 before, the independence of the plant's switchyard design, or 16 having backup power supplies available, like an on-site or 17 nearby gas turbine generator, important considerations.
I 18 The susceptibility to severe weather and the 19 potential for severe weather to be at that site. You have to l
)
l 20 have a high hazard factor for the site, and you have to be
! 21 susceptible, both of them. Those conditions are associated l
22 with the site. That's important for loss of off-site power, i
23 contributes to station blackout in an important way.
() 24 We found that the reliability of the redundancy in 25 common cause failure potential, not surprisingly, of the ACEJFEDERAL REPORTERS, INC.
j 202-347-3700 Nationwide Coserage 800-336-6M6
30853.0 cox 78 1 emergency AC power system, is an important factor, and the O 2 reliability of the emergency AC power supplies themselves as 3 can be measured by looking at failures per demand or 4 unavailability, are important factors.
5 So these are the kinds of things that seem to 6 differentiate based on the analyses that we have done. These 7 are the factors that would allow us to screen out plants that 8 seem to have higher or lower station blackout frequencies.
9 Now, this information was used as a starting point 10 for performing severe core damage accident analyses.
i 11 Our accident analyses involved a look at potential 12 accident sequences that would follow loss of all AC power, 13 what are the possibilities, what is the progression of events 14 in terms of timing. What are the important factors s 15 associated with plant coping and capability and damage to the 16 plant. What is the likelihood of these sequences, given that 17 other failures must occur in addition to the loss of AC 18 power, before the plant is in a core damage situation with 19 loss of cooling.
20 Lastly, we did some looking at containment 21 performance, which has been superseded, I would say, to a 22 large extent, by the NUREG-ll50 work. Our main thrust has 23 been to look down through accident sequence likelihood. I
() 24 will talk a little bit about that now.
25 Although we have used entries to look at accident ACE FEDERAL REPORTERS, INC.
202 347-3700 Nationwide Coserage 800-336-6M6
30853.0 cox 79 1 sequences, what we did was we looked at the kinds of factors 2 that would limit the capability of dealing with the station 3 blackout. If the blackout capability is severely limited 4 then short duration station blackouts will result in some 5 plant core damage and the more capability that exists, the 6 longer the plant can cope with the loss of off-site power and 7 on-site power. As you have seen the longer one can cope, the 8 more chance one has of returning power supplies and thus 9 having recovery from a station blackout successfully without 10 core damage.
11 We noted similarities between certain groups of 12 PWRs and BWRs. We noted that, for instance, on PWRs, O 13 reactant coolant pump seal leakage and don't have that as 14 part of their decay heat removal on loss of AC power, those 15 plants are susceptible to core damage if they have 16 significant pump seal leakage or any leakage. But in 17 particular that one.
18 MR. MICHELSON: Are you saying that the pump seals 19 on BWR 4, 5 and 6 are somehow cooled by the system?
20 MR. BARANOWSKY: No.
21 MR. MICHELSON: Are you saying it protects the 22 pump seals?
23 MR. BARANOWSKY: Let me just say on the BWR 2s and
() 24 3s, the pump seal leakage problem, even without the makeup, 25 is much less than on the PWRs.
ACE FEDERAL REPORTERS, INC.
202-347-3700 Nationwide Coserage 804336-(M6
30853.0 cox 80 1 MR. MICHELSON: Makeup of what, to the reactor O 2 vessel?
3 MR. BARANOWSKY: Yes, to the reactor vessel.
4 MR. MICHELSON: What does that have to do with 5 potential for pump seal leakage?
6 MR. BARANOWSKY: It has something to do with 7 dealing with that loss.
8 MR. MICHELSON: You are saying that it's a 9 non-problem because you go ahead and let them leak, you can 10 make it up. Is that what you are saying?
11 MR. BARANOWSKY: Yes.
12 MR, SPEIS: Abilities to pump, now.
13 MR. MICHELSON: How big a pump seal leakage are 14 you dealing with on 4, 5 and 6?
i 15 MR. BARANOWSKY: The information I have indicates 16 the pump seal leaks would be tens of gallons per minute.
17 MR. MICHELSON: Per pump?
18 MR. BARANOWSKY: Yes. With RCIC alone, you can 19 handle that.
20 MR. MICHELSON: You are satisfied now with the 21 test evidence on the type of reactor cool ability pumps for i
22 BWRs that says that you can go ahead and let it leak as a 23 nonproblem because you have got a lot of makeup. That's I
(]) 24 basically the story?
, 25 MR. BARANOWSKY: That's my feeling.
ACE FEDERAL REPORTERS, INC.
- 202-347-3700 Nationwide Coserage 800-336-6M6 l
30853.0 cox 81 1 MR. MICHELSON: I know if you let them keep n
v leaking over long periods of time, you may be dealing with 2
3 even larger leakage. But if you think you have plenty of 4 makeup, then you are all right.
5 MR. BARANOWSKY: When we talk a large period of 6 time here we are only talking a few hours.
7 MR. MICHELSON: That's a large period of time. -
8 MR. BARANOWSKY: The maximum leak rate that is 9 possible on the BWRs is significantly different than the PWRs 10 because of differences in design. Jerry, if you want to 11 chime in on that, go ahead.
12 MR. EBERSOLE: Let me ask a question before he 13 does that.
14 MR. MICHELSON: Is it on pump seal leakage yet?
15 If it is, let him answer the question first.
16 MR. JACKSON: Jerry Jackson for Staff.
17 When we were looking at the seal leakage problem 18 with respect to BWRs, we did decide to, after looking at some 19 of the data, the leak rates were less for the BWRs because 20 you had the steam driven makeup capability on some of the 21 plants. That was another reason that we decided it wasn't as 22 much of a problem with the BWRs. But in addition, they have
! 23 a lower pressure across the seals so you have a lower driving f
i
() 24 force in addition to that capability and ability to bring 25 pressure down.
ACE FEDERAL REPORTERS, INC.
202-347-3700 Nationwide Coserage 800-336-(M6
30853.0 cox 82 1 MR. MICHELSON: And can deal with small breaks in O 2 a mc e straightforward reaction than on a water pressure 3 reactor.
4 MR. JACKSON: Right.
5 MR. MICHELSON: I was thrown a little bit by the 6 absence of the X. I thought no leaks, and it doesn't mean 7 that. That takes care of my question, thank you.
8 MR. EBERSOLE: Let me ask a question. If you are 9 talking about the newer boilers, instead of the old condenser 10 types, heat removal is, of course, one way of rejection of 11 makeup water by evaporation. I would suspect that the seal 12 leakage rate, even if it was as part of the small percentage, 13 what sort of percentage are we talking about?
14 MR. BARANOWSKY: Let me make sure I understand 15 what you are saying.
16 MR. EBERSOLE: You will have to evaporate a lot of 17 water which has to go someplace to core the cool. You have 18 to have a lot of makeup just to keep the core cool. 2, 300, 19 5-or 600 early on.
20 MR. BARANOWSKY: Seal leakage is not a good way to 21 cool the core.
22 MR. EBERSOLE: Seal leakage is a minor part of the 23 makeup you have got to put in to cool the core.
O 4 MR. BARANOWSKY: On a BWR that's true.
25 MR. EBERSOLE: You have the privilege of ACE FEDERAL REPORTERS, INC.
202-347-3700 Nationwide Coverage 800-336-6M6 ,
30853.0 cox 83 1 depressurization down to any level, which progressively runs 7_
C/ I would think 2 the seal leakage down to an even less factor.
3 that seal leakage on a boiler is virtually ignored, because 4 it's a major operational problem.
5 MR. BARANOWSKY: That's what we did, ignored it.
6 It's really a minor problem, to the best of our knowledge, on 7 boilers, and that's why you don't see any X over here.
8 Let me just say, what we did is we went through 9 all these things. These are the things we could identify 10 that might cause problems during loss of all AC power.
11 Things that involve losses of reactor coolant on 12 PWRs and BWR 2, 3 designs are important because they don't 0)
'- 13 have the makeup. So we have several of those there, 14 including relief valves. Loss of batteries and compressed 15 air, things like that, are important, because you need those 16 to operate your steam-driven systems. We have to be 17 concerned about operating environment considerations which 18 might affect control rooms or things within containment for 19 their operability.
20 So, all these kinds of considerations go into our 21 analysis of defining where the potential sequence is in terms 22 of failures, systems or equipment, that could follow a loss 23 of all AC power.
() 24 MR. MICHELSON: Let me ask a quick question and 25 make sure this hasn't slipped into a crack. On boiling water ACE FEDERAL REPORTERS, INC.
! 202-347-37(U Nationwide Coserage MU-336-6M6
l 30853.0 cox 84 1 reactors there is a lot of low pressure equipment. You get 2 the pressure down through pressure reduction devices, which 3 in the dynamic sense, reduced pressure. But if you were to 4 let them sit at static conditions, the whole system would 5 come up for whatever the source pressure was.
6 Have you looked at those systems to make sure that 7 they are designed for full reactor pressure during a power 8 blackout. How do you assure yourself that the lower pressure 9 portions of reactor water cleanup remain at low pressure 10 during the power blackout, since isolation valves and so 11 forth don't necessarily function anymore, since you don't 12 have power to maneuver them.
13 I think they are reducing orifices and so forth to 14 break down the pressure from full pressure down to the lower 15 pressure.
16 You suddenly just stopped -- pressure will slowly 17 rise.
18 MR. BARANOWSKY: Normally, we would assume that 19 there are procedures to isolate --
20 MR. MICHELSON: You can't isolate though if you 21 don't have power.
22 MR. BARANOWSKY: There are manual valves in that 23 system.
() 24 MR. MICHELSON: Yes, but you don't walk near that 25 system. I am just asking, have you made sure that the ACE FEDERAL REPORTERS, INC.
202 347 3700 Nationwide Coserage 8(X).336-6M6 3
30853.0 cox 85 1 utilities have considered what the reactor water cleanup does 2 during power blackout and make sure they have procedures, 3 make sure they have procedures that call for things that they 4 can physically do, and do they have a time to do it before 5 the pressure reaches a high value. I am sure there are 6 relief valves. I know there are. But you don't want to sit 7 there relieving a lot of water out of the reactor water 8 cleanup because you haven't been able to keep the pressure 9 down.
10 MR. REED: A similar condition with PWRs, you-11 would have air operating valves, several series of lines, and 12 would fail to close loss of air which is loss of AC.
13 MR. MICHELSON: You have to make sure you lose the 14 air, in the long term you loose it in a power blackout, 15 short-term you don't, in the long term though you would 16 eventually lose the air pressure.
17 MR. BARANOWSKY: This is part of the consideration
! 18 that goes into our resolution of the issue. That is to say 19 that looking at those things that can results in degradations f
! 20 and complicating your ability to cope with and recover from a i
l 21 loss of all AC power. Alan Rubin will tell you exactly the 22 approach that we are taking to get that type of concern f
j 23 covered.
1
() 24 MR. MICHELSON: One of the problems is, these are 25 motor operated valves. One is DC, and I assume that that
/\CE. FEDERAL REPORTERS, INC.
20L347 3700 Nationwide Coserage 800-336-6646
l l
30853.0 cox 86 1 one, I guess, would be functional, if they have the right 2 actuation statements, or if you isolated it. Eventually, you 3 could close the DC valve on the line if the AC can't be 4 operated.
5 The other thing, you want to make sure you really 6 think it for --
7 MR. EBERSOLE: If you fall into a total AC power 8 case, which requires a lot of failures, are you considering 9 it allowable then to have one single discrete diverse face to 10 put the plant in an unequivocable state? .
11 MR. BARANOWSKY: What do you mean " allowable"?
12 MR. EBERSOLE: He was talking about reactor waters 13 clean up. Let's say you have this cleanup and you cannot 14 close the AC isolation valve to that system.
15 MR. BARANOUSKY: Yes.
16 MR. EBERSOLE: That system is a pressure bleeddown 17 system, cannot tolerate full primary pressure. If you have 18 then a discrete singular failure, the DC valves are shut, you 19 will have some problems.
20 MR. BARANOWSKY: Right. That's why we do 21 probabilistic analyses. On things like that, those are 22 mechanistically connected to the initiating event. For 23 mechanistically connected thing, the probabilities of failure O 24 is one. So we look at those types of things and we try to 25 determine whether the likelihood of sequences that involve ACE FEDERAL REPORTERS, INC.
L 202-347-370) Nationwide Coserage MC-33M646
30853.0 cox 87 1 failures that are mechanistically versus nonmechanistically O
v 2 particularly connected.
3 MR. EBERSOLE: Does that mean you blow up the 4 water reactor coolant system and then put water all over the 5 place --
6 MR. BARANOWSKY: It can mean anything. Loss of DC 7 power which may have a certain capacity, capabilities, in 8 terms of battery sizing. After a period of time, it's going 9 to drain down.
10 MR. EBERSOLE: That's a less likely thing than I 11 am talking about. I am talking about DC valve didn't work, 12 and it was dependent on the AC valve, but you didn't have any A
\I 13 AC.
14 MR. BARANOWSKY: That's a single failure on top --
15 MR. EBERSOLE: On top of this blackout.
16 MR. BARANOWSKY: We haven't seen anything to 17 indicate that that DC valve is so unreliable that it would 18 change the accident likelihood. If it's failure rate was as 19 bad as 1 in 10, it wouldn't be an extremely'large 20 contributor. It would have to be like 1 and 2. Generally, 21 they are 1 in 100 to 1 in 1000.
22 MR. SPEIS: I want to express the point that you 23 mado earlier in your conversation, he had listed some
() 24 important factors that were important. I don't think he 25 listed all of them.
I ACE FEDERAL REPORTERS, INC. '
202-347-3700 Nationwide cmerage a x)-336-6646 1
1
30853.0 cox 88 1 MR. MICHELSON: I didn't mean that you left it 2 out.
3 MR. SPEIS: This is part of our proposed 4 resolution, that a plan should undertake in an analysis to 5 see what is limiting, what is important in coping with this 6 blackout. As you will hear later on, it is one of the issues 7 that separates us and NUMARC. For some reason, they will 8 explain to you they don't want to go through this analysis to 9 assess the coping capabilities of the plant for some duration 10 that we are talking about. There could be some other unique 11 things that only a very full assessment of the plant would 12 show up to be constraining or more important basis. This is 13 not exhaustive.
14 MR. MICHELSON: No, I didn't mean to infer in any 15 way that it was. I just wondered if you had watched the 16 reactor water cleanup, because it is -- it doesn't get an 17 isolation signal from a power blackout. It doesn't 18 automatically isolate on loss of power. It only isolates on 19 pipe break signals, which eventually it might get as the 20 rooms heats up or whatever. It wouldn't be early on.
21 MR. RUBIN: We have some specific guidance going 22 on, looking at isolation capability systems for BWRs or 23 PWRs. Timeliness of procedures that may or may not be
() 24 necessary to isolate.
25 MR. BARAN0WSKY: I am almost through. So we take ACE FEDERAL REPORTERS, INC.
! 202-347 3701) Nationwide Coserage 8m336-6M6
30853.0 cox 89 1 that information, we generated accident sequences. What I am O- 2 showing here are the accident sequences that showed up as 3 being potentially important and the contributors to those 4 accident sequences. Lastly, the time in which one should 5 recover AC power in order to not result in core damage for 6 the type of failure that is identified in this contributor 7 column here.
8 Let me point out that those sequences that 9 involved the recovery of AC power on one to two hours, 10 although we can see their likelihood showing up in the 11 calculations, their probability is lower than those sequences 12 that allow recovery in a longer time frame, typically four to 13 16 hours1.851852e-4 days <br />0.00444 hours <br />2.645503e-5 weeks <br />6.088e-6 months <br />.
14 Again, we have broken down the plants into PWRs, 15 which have similar susceptibilities and accident sequences of 16 concern, BWRs with isolation condensors, but without a HPCI 17 or RCIC. Another class, BWRs with high pressure core spray 18 system and RCIC or, this should be high pressure coolant
! 19 injection, not core spray -- excuse me, make that I. BWR f 20 with high pressure core spray system and RCIC are also a j 21 different class. The reason they are different from the i
{ 22 HPCI, they have this diesel driven HPCS system which Jesse
)
i 23 mentioned earlier. In our analyses, we have generally 1
() 24 assumed that that system is not coupled to the diesel
! 25 generators that are used to provide emergency AC power system ACE. FEDERAL REPORTERS. INC.
202 347 3700 Nationwide Coserage 800 336-6M6
l l
30853.0 J cox 90 1 for the rest of the engineered safety feature.
O 2 MR. EBERSOLE: Don't you need, when you take that 3 view, to discriminate between those that just supply unit i 4 load of bus work across the ties, that is truly isolated and 5 is an independence of power plant. Usually they are 6 smaller. Then you have to look at that curious thing that 7 came up at River Bend, which you found out that diesel 8 generator was really cooled by water that was pumped from the 9 other diesel generators. I think they fixed that.
10 MR. BARANOWSKY: Right. Those are the kinds of 11 things that we would have to look at to make sure that this 12 type of credit that was given in the analysis is valid.
13 MR. EBERSOLE: That's variable, isn't it?
14 MR. BARANOWSKY: I think you will find that the 15 HPCS can be separated with cooling or perhaps ventilation and i
16 things like that you would have to look at.
17 MR. REED: I am surprised that BWRs appeared to be f
18 all and PWRs get differentiation. I am not sure I would like 19 to see PWRs classified with Babcock & Wilcox PWRs, which have l
20 little inventory. On that first line up there, you talk l
l 21 about steam-driven auxiliary boiler feeds pump unavailable, 22 one to two hours, I don't think that your BWRs are going to l 23 be in that category at all. I don't think you ought to have 24 all PWRs thrown in the same bucket.
l l 25 MR. BARANOWSKY: There is actually some slightly
! ace FEDERAL REPORTERS, INC.
I 202 347-3XW) Nationwide Cogerage M433M686
! 30853.0 cox 91 1 noticeable difference in the accident sequence likelihood for 2 B&W, Westinghouse and CE. It has to do with this particular 3 sequence that you pointed out. This particular sequence, l 4 though, is about a 15 to 20 percent contribution. Even if i
j 5 there was a factor of two difference between B&W and 6 Westinghouse CE plants, what we would see generally is for 7 that sequence, when one looks at the whole picture.
8 The B&W plants would be 10 percent higher in i 9 likelihood for all the station blackout sequences, if they.
l 10 had comparable capability for these other items. When plants l 11 don't have comparable capability, that really 12 differentiates. So what I am saying is the nature of the i
O 13 sequences. This is a little bit questionable on B&W plants.
i t 14 But the nature of the sequences that are potentially 15 important, and the time frames, are approximately in the ball
- 16 park.
17 MR. REED: I don't like your first line at all. I 18 think you are talking a matter of just a few minutes. I have i
l 19 seen it go all the way for two minutes to 12 minutes, versus,
, 20 on the other hand, something, a minimum of an hour to as many 21 as eight hours, depending on whether it's an old thing like i 22 Yankee Rowe.
23 MR. BARANOWSKY: I don't know about eight hours.
O 24 What we are talking about in that sequence is loss of 25 off-site and on-site AC power. The steam-driven auxiliary ACE-FEoERAL REPORTERS, INC.
l 202 347 37th) Nationwide Coverage MWL3346(46
30853.0 cox 92 1 feed water system is unavailable, and then all we have 2 basically is boiling from the steam generators. Eight hours 3 is surprising capabilities.
4 MR. REED: You think Yankee Rowe got eight hours; 5 that's what the original calculations were.
6 MR. BARANOWSKY: That's a two loop --
7 MR. REED: Or cam pumps.
8 MR. BARANOWSKY: It's an old four loop.
9 MR. EBERSOLE: Aren't there really two subsets 10 under the BWRs with isolation condensation, those that are -
11 makeup capability with to the isolation, and isn't there a 12 strong difference first one extending to the level of BWR?
13 MR. BARANOWSKY: Yes.
14 MR. REED: What is the statement here?
15 MR. EBERSOLE: You have no primary makeup.
16 MR. BARANOWSKY: That's right.
17 MR. EBERSOLE: Even BWRs have primary makeup, but 18 it's not steam-driven. Don't you need two sets there?
19 MR. BARANOWSKY: Actually, again, this is a 20 simplification. We are aware of the fact that there are some 21 isolation condenser plants that have RCIC. In fact, from a 22 station blackout core damage point of view, they look a lot 23 like these plants.
() 24 MR. EBERSOLE: Or better.
25 MR. BARANOWSKY: Not better in station blackout --
ACE. FEDERAL REPORTERS, INC.
! 202-347-37(O Nationwide Cmerage 8(U-336-6646 1
30853.0 CoX 93 1 well, let me take that back, isolation condenser is !1 very 2 good system when there are not substantial leaks in the 3 primary coolant system. In particular, one would be 4 concerned about stuck-open relief valves, which we have seen 5 on a few BWRs. If the reliabilities of the relief valves to 6 close under blackout conditions is good, then I would say the 1 7 isolation condenser with RCIC is a super setup.
8 MR. EBERSOLE: Where do those relief valves 9 discharge to, to a containment?
10 MR. BARANOWSKY: They always discharge to the 11 pool.
12 MR. EBERSOLE: Is there any way to condense that?
O 13 MR. BARANOWSKY: Right in the pool. One of the 14 limitations is pool heatup.
15 MR. EBERSOLE: In all of the BWRs you don't 16 reflect what I thought was a standing orders, almost, or 17 ability on the part of the boilers to do open-cycle boiling 18 atmosphere to a containment. In that case, the water can 19 almost come from a fire pump.
20 MR. BARANOWSKY: We would be willing to give 21 credit for those kinds of things. .
22 MR. EBERSOLE: Is it wrong that the current BWR 23 operators are not making available to them vessels that low
() 24 cycle boiling?
25 MR. BARANOWSKY: That's one issue that needs to be ACE.FEoEnai. REPORTERS, INC.
202 347 3RX) Nationwide Coserage 804346M6
30853.0 cox 94 1 looked at on a plant specific case, that's part of coping t0 2 abilities. What they have to do and what they can do is 3 something that needs to be determined on a plant-specific 4 basis. You have given me six different things that say why 5 plants are different copingwise. I am not going to disagree; 6 I agree.
7 MR. EBERSOLE: Is there eventually going to be 8 produced a table that says this plant will use that method?
9 MR. BARANOWSKY: No. They are going to be able to 10 look at their own plant and make use of whatever systems and 11 capabilities they have available.
12 MR. EBERSOLE: Then it may emerge from this study 13 that they will do that in buying a new bunch of diesels or 14 whatever.
15 MR. BARAN0WSKY: I can imagine this as being a 16 top-level breakdown of subsets occur. You get down to the 17 plant and there is nothing else the same.
18 MR. EBERSOLE: This may be a major incentive to 19 get with it.
l 20 MR. MICHELSON: In case of a power blackout, how 21 do you cool the condenser if it doesn't have a huge reservoir 22 of water up there for decay heat, I believe. ,
23 MR. BARANOWSKY: The isolation condenser has j
1
() 24 sufficient reservoir for one to two hours, depending on a 25 plant. It can be replenished by a diesel fire pump.
Acii-FnounAi. IttironTens INc.
3 202 m-Um Nationwide Cmcrage h1%N46
30853.0 cox 95 1 MR. MICHELSON: Inventory of water is enough for O 2 two hours of decay heat.
3 MR. BARANOWSKY: Depends on the isolation --
4 MR. MICHELSON: Full power.
I 5 MR. BARANOWSKY: There is no full power --
6 MR. MICHELSON: You were from a full power initial 7 condition, and when you get the power blackout, it has to 8 start functioning immediately, and it has to handle decay 9 heat for two hours thereafter.
10 MR. BARANOWSKY: Right.
11 MR. MICHELSON: That's a lot of water, Jesse.
12 MR. BARANOWSKY: To the best of my knowledge, all 13 the plants with isolation condensors have in their procedures 14 the calling for replenishing of the isolation condenser --
15 MR. MICHELSON: That's what I was wondering about, 16 whether you had it in the inventory or makeup capability for 17 blackout.
l 18 MR. BARANOWSKY: Initial inventory is adequate to 19 allow you timo to do that, but you best have procedures.
20 MR. MICHELSON: One other question relating to the 21 human intervention to help spread out these times, which I 22 assume are based on allowable interventions, is there a 23 requirement that the normal plant lighting systems be on the
() 24 diesel engines as well, or have you looked at how good your 25 emergency lighting is for doing some of these?
ace.FEDERAt. REPORTERS, INC.
202 347-370i> Nationwide coverage so).33 w 6*6
I
, 30853.0 1 cox 96 4-1 MR. RUBIN: Let me respond. We will get into what !
! 2 the recommendations are in terms of coping assessments and j I 3 what kinds of things you looked at.
4 MR. MICHELSON: I didn't pick it up in the guide, I
5 but I guess it's in there.
6 MR. RUBIN: Yes, it's in there.
7 MR. MICHELSON: We will talk about it later then.
8 MR. BARANOWSKY: Thank you. At any rate, this 9 gives you an idea of the kinds of considerations that can go 10 into the accident sequences for different plant designs.
i 11 Obviously, they vary from plant to plant. We have used those
] 12 types of considerations to come up with analyses that give us
]
(:) 13 a spectrum of possible station blackout, core damage li 14 frequency estimates, as a function of the time that a plant 3
15 can cope with a station blackout.
1 l 16 Essentially, what we are saying is that as you can 1
l 17 cope for longer periods of time, the possibilities for having l
18 lower core damage frequency range from somewhat of an i 19 improvement to a large improvement and all these factors that a
! 20 have been mentioned over the last 15 or 20 minutes, go into l 21 determining whether the improvement is on the higher or the l 22 lower end and also in consideration is the consideration of 23 the likelihood of the blackout itself.
() 24 So we have to integrate the likelihood of blackout
- 25 and plant design and location characteristics with its Acn. FEDERAL reporters, INC.
202-147 37(O Nationwide Coserage Mn33MM6
30853.0 cox 97 1 capability to cope to get a proper loss of station blackout 2 core damage perspective, and these factors are the ones that 3 from the loss of off-site power and on-site power, tend to 4 cause a change in core damage frequency estimate, which make 5 coping capability more or less important.
6 We are able and have done analyses to get an 7 estimate of what frequencies are for different factors. Here 8 is an example. There are tables in the NUREG-1032 that you 9 have that show core damage frequency estimates for 1 out of 2 10 AC configurationsa. We have off-site power clusters 1 11 through 5, which we talked about earlier. We have diesel 12 reliabilities running from -- these are failure rates, excuse 13 me, running from 1 on up to .01678. These are coping times 14 which means if the plant can cope for zero hours.
15 In other words, instantly there is a problem; 16 there is a fairly high probability that there will be damage 17 to the core. Our estimate is that all plants can go on the 18 order of two hours right now, and -- or greater. With this 19 type of table, what it allows us to do is looking at the 20 trade-offs associated with the different figures, diesel 21 reliability, configuration, off-site power characteristics 22 and coping capability, to see how one can work with the 23 interplay of these factors in deriving conclusions regarding
([) 24 what might be effective and what might not be effective in 25 terms of requirements to resolve the issue.
ace FEDERAL. REPORTERS, INC.
202-347-370) Nationwide Coserage 800-33MM6
30853.0 cox 98 1 Which gets me to my last slide, which is really a O 2 restatement of the slide that was shown earlier by Dr. Speis, 3 and only summarizes to say that the important findings are 4 that reliability of the emergency AC power systems have to be 5 taken into consideration. We need to consider configuration 6 redundancy and diesel reliability.
7 In particular, the losses of off-site power, since 8 they vary, we need to consider site characteristics that are 9 associates with weather and grid, as well as design factors 10 associated at the switchyard, and we found that coping with 11 blackouts of greater than two hours can be significant 12 contributor to risk, depending on the factors associated with 13 on-site and off-site power reliability. Given all this 14 information and differences in plant design, things vary so 15 much that it's impossible to identify a few key specific {
16 fixes that make the issue go away and would be necessarily 17 cost effective for all plants.
18 So we recognize the fact that susceptibility to 19 station blackout and ability to cope with station blackouts l 20 varies from plant to plant and thus the resolution of this 21 issue should take these factors into consideration and the 22 variability associated with it.
23 That actuality completes my discussion of the
() 24 technical bases. This information is used then by us to 25 derive the so-called resolution requirements for station ACE FEDERAL REPORTERS, INC.
l 202-347-3700 Nationwide Coserage 800-336 6M6
30853.0 cox 99 1 blackout. Alan Rubin is the next presenter to discuss that 2 topic.
3 MR. WYLIE: We will take a break at this time.
4 MR. MICHELSON: Do you have questions on this or 5 do you want to do that later?
6 MR. WYLIE: We can go ahead and do it now.
7 One question on this part is, how did you treat 8 the loss of heating and ventilating from the viewpoint of 9 buildup and consequential effects on equipment that is still 10 being poured through the DC power solid state and control 11 devices? That's one question that you have to answer.
12 Another question has to do with, as power is O 13 restored, have you looked carefully at the state of the 14 control systems which were energized throughout the event, 15 which have been actually going through certain logics as 16 conditions have been changing their state as a consequence, 17 and now are lined up to start the minute power comes back, 18 and a whole lot of interesting things are going to happen if 19 you haven't thought about it --
20 MR. RUBIN: Those will both be addressed in my 21 presentation. It's part of the coping ability.
22 MR. BARANOWSKY: The way we treated that in these 23 analyses was on a parametric basis in which we said if your
() 24 heating and ventilation system has either a low reliability 25 or fails as a result of the station blackout within X-time, ACE FEDERAL REPORTERS, INC. f 202 347 37(X) Nationwide Cmerage MMM36-6M6
30853.0
] cox 100 1 here are the consequences in terms of risk.
2 MR. MICHELSON: Your reliability numbers did not 3 necessarily incorporate elevated temperature effects, because 4 you don't have any --
5 MR. BARANOWSKY: If elevated temperature effects
, 6 cause failures within a certain time frame, say, four hours, 1
7 I can tell you with the likelihood of core damages within 8 four hours if that's the period of time which you have some 9 sort of a degradation in your ability to cope with a loss of a
10 AC power.
11 MR. MICHELSON: I am not sure that answers the 12 question, but that's all right.
13 MR. RUBIN: Let me try to clarify a point --
14 MR. MICHELSON: If you are going to get to it 15 later --
F 16 MR. RUBIN: Just to set the stage for the next 17 presentation that I will be making. Pat's analysis on h 18 estimates of core damage frequency is okay if you have a loss 19 of all AC power for this period of time, and you have the i
20 capability at the plant to withstand the loss of AC power for 21 that period of time. Here are some estimates of core damage 22 frequency.
23 It doesn't mean that he has looked at every plant 2
l l
() 24 or analysis, he has looked at every single plant to see that 25 they have that capability, or what particular plant specific ace FEDERAL REPORTERS, INC.
l Nationwide Cmerage 8(X) 33666:6 l 202 347 37(X)
I 30853.0 cox 101 l 1
1 features might make that plant vulnerable. That's an area 2 where we get in the recommendations and coping capabilities 3 of the plant, the recommendations that we put forward in 4 terms of -- not specific design requirements, but rather 5 performance or target goals that we have. It would require 6 some plant specific analyses and the kinds of questions you 7 are in.
8 MR. MICHELSON: In doing the PRA, you also have 9 some human factor reliabilities. I mean, if you are going to 10 try to end up estimating core melt, you will have to have 11 some probability about people doing things right along the 12 way.
13 How did you determine what those probabilities 14 might be on the adverse human condition of a power blackout?
15 MR. BARANOWSKY: We looked at essentially the 16 sensitivity of the results to high and low human 17 reliability. What we found was that if the human element has 18 appropriate procedures and training, that the kinds of 19 reliability associated with taking proper procedures, even 20 conservative estimates, would be a small contributor to 21 failures in the scenarios we looked at. If the human is not 22 prepared and doesn't have adequate procedures, then the 23 contribution to human failure particularly could go any
() 24 active types of errors, would be a significant contributor.
25 Therefore, as a sensitivity, we looked at human Acn FEDERAL. RneonTEns. INC.
I 202 347 3700 Nationwide Cmcrage mhDMM6
30853.0 cox 102 1 error with and without training and procedures. Since O 2 procedures are required, or are going to be a requirement and 3 associated with this issue, we felt that that was the way we 4 would take care of the human reliability.
5 MR. MICHELSON: I guess you are saying, though, 6 that the anticipated performance is the same with or without 7 the blackout if he has proper procedures; is that what you 8 are saying?
9 MR. BARAN0WSKY: Yes, if you look at -- it's not 10 just blackout, it's anything, humans are expected to make and 11 perform and what we found was that with the type of training i 12 and procedures that the NRC is requiring, the human 13 reliability is adequate whether it's the sequence or --
14 MR. MICIIELSON: Is there a good technical l
15 foundation to believe that a person performs the same, you 16 believed, adverse conditions as he does under ideal 17 conditions?
18 MR. BARAN0WSKY: I wouldn't make that claim. I 19 don't think there is any technical basis for it.
20 MR. MICl!ELSON: There is some adjustment needed.
21 I was searching for how much adjustment you might have put 22 into your numbers.
23 f tR . BARAN0WSKY: Let's say that the human error
() 24 rate for a noncorrectable error was once overy 10 sequences.
25 We know these operators go through simulator training and Act:.Fitotinai. Illii>onitins, INC.
I :o: m t u m meionwide rmerage sm. m te,
30853.0 cox 103 1 their failure rate is much better than that.
O 2 MR. MICHELSON: I am talking about out in the 3 plant. This isn't control room control anymore. We already
- 4 admitted we don't have valve power and so forth. We have to i
j 5 try to, in the darkness, find the valve in the darkness, or i
6 whatever model light you provided and so on. You know, you 7 have got to adjust these numbers a little bit for the fact 8 that these are very --
i 9 MR. BARANOWSKY: That's why I say if they are not 10 prepared, they don't have procedures, they need adequate 11 lighting, all these things going along with it.
12 MR. MICl!ELSON: Are we requiring they go inside a
() 13 containment during blackout?
14 MR. BARANOWSKY: No.
1 15 MR. MICHELSON: Is it required to require that as 16 part of your responso procedures?
17 MR. BARANOWSKY: It's not prohibited, but I can't 18 think of any reason why someono would go insido.
19 MR. MICl!ELSON: I am thinking of isolation valves, 20 which sometimos is a little nicor to go insido a containment 21 and some of the other locations whoro they aro located.
4 22 Rapid water cleanup is a good examplo; ono is hot as a ,
, 23 pistol, the other insido of containment is not quito as bad, i () 24 MR. BARAN0WSKY: I would be inside if they could 25 got even containment in blackout conditions.
Acit Fiti)iinAi. Illii>onTiins. INC.
20244WW NahonwMe Cmcrne Emmt, (Mi
30853.0 cox 104 1 MR. MICHELSON: I think it would be rather O 2 impracticable. There is no emergency lighting inside a 3 containment, is there?
4 MR. BARAN0WSKY: No. We haven't seen anything 5 that indicates a requirement to go inside containment. There 6 are other ways around those problems. If the reactor water 7 cleanup system has to be isolated, there are ways to isolate 8 the system with manual valvos and one can provido appropriate 9 radiation shielding, if that is necessary, so that those 10 manual valvos can be accessed and isolated.
11 MR. MICHELSON: If they are identified and they 12 are found not to be adequately shielded, you would say we 13 have to go back and put in adequato shiolding ahead of time.
14 MR. BARANOWSKY: Right.
15 MR. WYLIE: Let's break till 11:30.
16 (Rocess.)
17 MR. RUDIN: My namo is Alan Rubin. I am with the 10 Office of Energy Research, currently. Last month, I was with 19 the Office of Nuclear Reactor Regulation. Pat and I have 20 switched offices as of last month in this reorganization.
21 I am going to cover a number of areas. As you can i
22 800, we have dono an extensivo amount of technical analyses 23 and ovaluation on the ntation blackout issue. What I am
() 24 going to cover are soveral topics, 25 I am going to go over the resolution of the issuo ActiFitoiRai. Rema rras. INC.
l, 202 mmm Nanonode nneuge m un ua6
1 30853.0 cox 105 1 in terms of what are the highlights or high points of the 2 proposed rule that has been issued for comment and our rule; 3 the current status of the rule. The associated regulatory 4 guide on station blackout, a very, very brief summary of the 5 value impact analysis that we needed to perform, in terms of 6 complying with the backfit rule that currently exists.
7 I will discuss briefly the categorization of 8 public comments received on the proposed rule and whether or 9 not there were changes made in response to those comments.
10 I will discuss the significant changes to the 11 proposed -- from the proposed resolution to our current 12 resolution, and I will also discuss some of the industry O 13 initiatives that are under way or have been under way on the 14 station blackout issue.
15 Let me first give a schematic diagram that tries 16 to portray what the station blackout rule itself entails.
17 There are several steps along the way. The first one would 18 be a requirement that plants determine an acceptable period 19 of time that they should be able to cope with a station 20 blackout. That time would depend on the factors that were 21 dircussed earlier, in terms of AC power reliability. That's 22 the reliability of the diesel generators, the configuration 23 or number of diesel generators available to provide cooling,
() 24 decay heat removal, the frequency or likelihood of losses of 25 off-site power, and the probability or likely time to restore ACli-Fimi!Rai. Rl!PORTl!RS, INC.
l, 202 mmm Nanon%te romage m3we6m
30853.0 cox 106 1 off-site power.
2 The rule doesn't specify itself how long a period 3 of time that should be. The regulatory guide, which I will 4 get into next, does give guidance on acceptable durations.
l 5 The second step in the rule would be for plants to l
l 6 determine or assess their actual capability to cope with a l 7 blackout. In other words, there is some target duration; we 8 have talked about four hours and eight hours before. The 9 second step would be for the plants to go through a plant i 10 specific analysis to assure that they actually do have this 11 capability. That was determined earlier.
12 If the plants have the capability to cope with the 13 blackout for the period of time, be it four hours or eight 14 hours, depending on the plant specific case, then there is 15 nothing else required other than that there would be 16 procedures and training in place for operators to cope with a 17 blackout for that period of time.
18 If there is some limiting factor in the plant, in 19 terms of its characteristics, in terms of water availability 20 or battery capability as examples, to limit the capability of 21 the plant to cope for the target period of time, then it 22 might be some plant modifications that would be required, 23 and, in addition, also procedures would need to be in place.
() 24 The rule itself is a modification addition to the 25 part 50, 10 CFR part 50. The definition of loss of off-site ACE. FEDERAL. REPORTERS, INC.
202447 37W Nationwide Canerage MXb33MM6
30853.0 cox 107 l
1 power is included, with the requirements that plants should 2 be able to cope with the station blackout for a period of 3 time to be determined in accordance with the four factors 4 that I mentioned earlier.
5 We have in the rule a schedule for implementation i
i 6 that once a rule, if it were issued, that licensees would 7 submit to NRC within nine months a proposed duration that the 8 plant should be able to cope with the blackout and the 9 justification for that time. An identification of the 10 factors that would limit the plant to cope for the period of 11 time determined above, a description of the procedures for 12 coping with the station blackout; and, if any equipment 13 modifications were needed to achieve this coping duration, F
14 that should also be provided.
15 NRC would review these submittals, which would not 4 16 be necessarily be extensive in detail. Just looking for 17 determination of basic compliance with the rule and how it
- 18 compares with guidance we would be issuing an associated l 19 regulatory guide.
4 j 20 After this review, those licensees that would need 21 equipment modifications would primarily submit schedules for
) 22 completing those modifications with a target of completing i 23 them within two years of an NRC review. But the final
() 24 schedule would be determined on a plant specific basis l
25 mutually agreed upon by the Staff and the licensee.
ACE FEDERAL REPORTERS, INC.
- f 202 347-3700 Nationwide Cmerage 80lb3EtM6
30853.0 cox 108 7 ,s 1 We don't expect that there would be -- the d 2 majority of plants, our expectation, would be, fall above 3 this dotted line, would not necessarily need hardware 4 modifications.
5 The rule is not quantitative, it's not very 6 detailed. It provides the basis for the recommendations and 7 guidance in the regulatory guide. There are several aspects 8 to the guidance in the regulatory guide.
9 One, does it focus on the on-site emergency AC 10 power sources. The diesel generators for a majority or 11 almost all of the plants. The licensees should have the 12 reliability of program to ensure that the diesels are m
'd 13 maintained at some minimally acceptable rc'lable level; that 14 there be target values for diesel generator reliability; and 15 that plants have procedures for restoring emergency AC power 16 when those sources were lost.
17 There is also guidance relating to off-site power 18 systems that utilities should have procedures to restore 19 off-site power or use nearby power sources when off-site 20 power is unavailable.
21 These two items focus on reliability of the AC 22 power systems. We have guidance in the regulatory guide 23 that, I recall, focus on more on the in-depth approach of
() 24 ability to cope with station blackout.
25 MR. WYLIE: Let me ask you about that, on the ACE FEDERAL REPORTERS, INC.
I 202 347-3700 Nationwide Coserage 800 336-(646
J 30853.0 cox 109 1 off-site -- procedures to restore off-site power, nearby 2 power source, suppose it's not available. Suppose there is 3 no nearby power source.
] 4 MR. RUBIN: Then they can't have those 5 procedures. We are not requiring that there be any 6 additional sources nearby. But if they are available, if 7 they are nearby, ; hen they should be considered in the 8 plant's enhancements of their AC power reliability.
9 MR. WYLIE: You are not requiring that?
10 MR. RUBIN: That's right. This is guidance. The 11 ability to cope with a station blackout in the regulatory 12 guide.
l 13 First of all, there is some guidance on 14 determining these acceptable coping durations. I will show 15 you on a later slide how that is proposed, this guidance on 16 evaluating or how licensees should evaluate their capability ,
17 for coping with a blackout; what features they need to look 18 at. They need to look at their capability to cope with a i
! 19 blackout, their capacity of systems, be it batteries or
- 20 storage tank or equipment operability during a station 21 blackout and environmental conditions that are associated 22 with a station blackout.
23 For example, without the availability of their
() 24 HVAC systems. We also have guidance, I will focus on this a 25 little bit, on giving credit for alternate power sources to
/\CEJFEDERAL REPORTERS, INC.
l 202-347-37(o Nationwide Coserage 8(0-336-6646
30853.0 cox 110 s 1 cope with a station blackout. We have had comments on the 2 proposed rule from utility groups and individual utilities, 3 that there ought to be some credit given or how should credit 4 be given in a plant has an alternate power source -- this 5 relates to a question earlier -- that can be used. It's 6 separate; it's not part of their off-site power system or 7 their emergency diesel generator systems. But they may have 8 a gas turbine on-site. They may have an appendix R diesel 9 generator on-site that can be used to provide power 10 independent or separate from their emergency or off-site 11 power systems.
12 We never precluded this from a possibility in our
(~s kJ 13 proposed rule, but we have clarified the acceptability of 14 alternate AC power sources that could be used in the 15 eventuality of loss of the emergency diesel generators and 16 the off-site power sources.
17 We have provided guidance.in the regulatory guide 18 in a general way, that these power sources should be separate 19 from the emergency and off-site power sources. They should 20 be operable during station blackout conditions. They should 21 not be subject to the same common mode failures that could 22 cause failures of the on-site or off-site power systems.
23 They should have the capability to provide the AC power D)
(, 24 loads, and they should have a testing and maintenance program 25 associated with those power sources to assure their
/\CEJFEDERAL REPORTERS, INC. '
] 202-347-3700 Nationwide Cmerage 8 2 3366646 l 1
30853.0 cox 111 l i
7s 1 availability and reliability. l V 2 If these criteria are met for the alternate power 3 sources, that could, we call it, cope with the station 4 blackout, an advantage of this approach is that it would be 5 likely that the extent of time that the plant could cope with 6 the blackout would be very much in excess of the four- or 7 eight-hour time periods that we were talking about and 8 wouldn't require the extensive analysis for plants to 9 determine that they can cope with a blackout independent of 10 any AC power.
11 MR. EBERSOLS: What is the discrete difference 12 between this and just another diesel engine?
\ 13 MR. RUBIN: Difference should be they are diverse 14 enough so it's not subject to common mode failures. We don't 15 want, for example, at a multi-unit site, a swing diesel 16 generator that can go between one unit and another, to be 17 used as credit for an alternate power source.
18 MR. EBERSOLE: You are going to have to get into, 19 I guess, what constitutes diversity.
20 MR. RUBIN: We have had extensive meetings with 21 utility groups on this issue just within the past couple of 22 weeks. They are attempting to develop detailed guidance to 23 go along with what criteria and guidance should be met.
() 24 MR. EBERSOLE: Are any of them claiming the 25 ability to run back to house load?
ACE FEDERAL REPORTERS, INC.
l 202-347-3700 Nationwide Coserage 800-336-6616
30853.0 cox 112 1 MR. RUBIN: As a coping capability?
2 MR. EBERSOLE: Yes.
3 MR. RUBIN: I hope not. I haven't seen it yet. I J
4 MR. EBERSOLE: I hope not too. I think I would 5 reject it automatically.
6 MR. RUBIN: That is an issue that is still being 7 discussed with the utility group. We don't have a point of 8 agreement right now with the utility group's guidance that 9 they are developing. But we feel, in princ5ple, that is an 10 acceptable approach. We didn't want to preclude it, because 11 there are certainly some benefits and it reduces the risk.
12 MR. EBERSOLE: This business of hanging on to 13 excitation compounds the problem of turbine generator run 14 away. If you carry up the parallel loads with the turbine in 15 a runaway load of up to 80 percent, the whole pipe comes off 16 as to speed. That's one of the problems, generator sees l
l 17 turbine excitation and hangs on it to on a trip. You may c 18 carry the parallel load. Normally, we consider turbine 19 run-away to involve the turbine because presumably it's 1
20 disconnected from the parallel loads. If you attempt to hang 21 on to a tripped turbine by continuing excitation, it's no 22 longer a single turbine run away, it's connected to all the 23 AC loads.
() 24 MR. RUBIN: If that were the approach the 25 utilities would choose to take, would have to meet the same ACE FEDERAL REPORTERS, INC.
l 202 347 3700 Nationwide Coverage 800-336-6646
30853.0 cox 113 1 kinds of objectives we are trying to achieve with the ability 2 to cope with the station blackouts. There should not be 3 dependent failures that result with the probability of one of 4 its failure, given a station blackout occurs.
5 Let me finish a couple of points on here. The 6 regulatory guide gives additional guidance on if 7 modifications were needed to cope with the station blackout.
8 This guidance is needed, because there have been some i
9 questions of whether we are trying to make station blackout a 10 design basis event, which may mean, in that context, that 11 equipment needs to be seismically qualified, redundant, needs 12 to meet single failure criterion. Guidance that we have 13 developed that the equipment for coping with a station 14 blackout need not be seismic. It should be operable, 15 however, in station blackout environments, and be reliable.
16 We also have guidance that procedure and training 17 are important in assuring that the operators know what 18 actions that they should take, if a station blackout were to 19 occur.
20 MR. WYLIE: What about your guidance for the 21 alternate power source, does that require that it be a 22 permanent on-site source?
23 MR. RUBIN: When we talk about coping with the
() 24 station blackout, you are talking about a portable generator 25 or so?
ACE FEDERAL REPORTERS, INC.
202-347-3700 Nationwide Coverage 800-336-6646
I 30853.0 cox 114 1 MR. WYLIE: Yes.
1
() 2 MR. RUBIN: Utility could provide some assurance.
3 We are not saying this is an approach to take, but in coping 4 with a blackout for four or eight hours, which is the general 5 duration we are talking about, it might be difficult to show 6 with assurance you could get that portable alternate on the 7 site and having it hooked up in time without having to
- 8 provide the independent AC capability.
9 If equipment were to be used, and were taken 10 off-site, it should -- there should be some guidance of a 11 period of longer than eight hours to take credit for it, 12 unless there can be justification for shorter times.
( 13 MR. WYLIE: You are not closing the door on it?
14 MR. RUBIN: We haven't closed the door on it. We 15 are trying to provide performance rather than design 16 requirements or guidance. There are a lot of alternatives 17 here. The questions came up earlier on what about this
- 18 aspect of coping capability. What about the isolation 19 valves, what about room cooling equipment operability. Those 20 are all aspects of ability to cope with a station blackout 21 that would need to be addressed. ]
l 22 I would just going to present one matrix, which is ,l 1
4 23 how we view plants, could come up with acceptable durations
() 24 for coping with a station blackout. You wouldn't be
- 25 surprised at the factors that are on here, because you heard
< / ACE FEDERAL REPORTERS, lNC.
l 202-347-3700 Nationwide Coserage 800-336-6M6
30853.0 cox 115 l 1
- 1 about them -- you heard them discussed repeatedly this O 2 morning. Emergency AC power configuration groups. These are 2
3 the groups of numbers of diesel generators compared to the s
4 numbers required. Going from A through D, the A systems are
'l 5 the more redundant systems, three or four diesel generators i
6 for a plant, where one would be required.
7 Going down, C is a typical plant, for a reference 8 point, which has one or two diesel generators, which many 9 plants have that configuration. D are multi-unit sites which 10 would require two or three diesel generators or perhaps one 11 of two diesel generators.
12 We mentioned earlier the diesel generator J
13 reliability, in addition to the configuration, is a 14 significant factor in station blackout frequency and risk.
I 15 We have developed guidelines that target j
16 reliabilities for diesel generators should remain at 95
- 17 percent or better; and for some situations, for example, the i 18 plants that have the least redundancy, were aiming for 19 targets of better reliabilities than the .05.
20 For a frame of reference, Pat showed earlier that l
i 21 the diesel failure rates are, on industry average, about .01 22 or .02 failures per demand, with variants from site to site.
23 We are trying to address those sites that have the
() 24 reliabilities that should be improved to make sure they could
- 25 be achieved, some minimum level.
ACE FEDERAL REPORTERS, INC.
202-347-3700 Nationwide Coserage 800-336-6M6
.- __ - _ - _ . . _ . ~ . _ _ . - - ____ _ _---._.. _ _ _.._.-__ --.__-._--,._ __._, _ -. -
30853.0 cox 116 1 MR. EBERSOLE: Let me ask a question, about two or 2 three years ago, didn't you have a table that resembled this 3 somewhat, this cluster-type regulation, or whatever, and 4 wasn't it stalled because it was just that?
5 MR. RUBIN: No.
6 MR. EBERSOLE: Is this table substantially -
7 different from the original proposition, other than the fact 8 you don't require a detailed analysis of how long you could 9 survive?
10 MR. RUBIN: The difference between this table and 11 the earlier proposed table and the draft guide several years 12 ago, we had done the initial analyses. The approach is the O 13 same. We have added another group, two or three diesel 14 generators. We have looked at the off-site power 15 characteristics, new data, and have gone through this 16 reclustering analysis that you are seeing.
17 Through that determination, taking the same 18 approach that we had before, we have come up with this 19 separate matrix. The reason we have this is what was 20 mentioned earlier, is that we don't see that there is one 21 requirement that can apply to all plants, because there are 22 some better ones, and there are some average ones and there .
23 are some less than average.
() 24 MR. EBERSOLE: Wasn't there a rejection of this 25 process simply because you don't have enough support data?
ACE FEDERAL REPORTERS, INC.
202-347-3700 Nationwide Coverage 800-336- % 86
_ _ _ _ . . _ _. _ - . _ . - _ _ . _ , _ _ _ . __~ _ -._ __ , _ - - - - . , . _ _ _ - -
30853.0 cox 117 1 MR. RUBIN: I don't agree that it was rejected.
7-)
V 2 We went through ACRS. In fact, ACRS endorsed our approach.
l 3 MR. EBERSOLE: Didn't it go up to CRGR?
4 MR. RUBIN: They also supported our approach.
l 5 However, they had a recommendation that in addition to times l l
6 on this matrix, that there should be plants that have a zero 1
7 coping time. The approach was not rejected by anyone. There 8 was a rejection by CRGR for zero hour duration and, in 9 essence, when you look at this alternate approach, that's 10 what that is. You don't need the extensive or analysis to 11 assure we have four hours, because if you have an acceptable 12 alternate AC approach, as long as you have fuel and other 13 systems available, you can go for more than four hours.
14 MR. EBERSOLE: Are you going to give them some 15 guidance as to what to pursue in some detail to be sure they 16 will be able to survive these hours?
17 MR. RUBIN: We have guidance in a regulatory
, 18 guide, l
19 MR. EBERSOLE: Is it in kind of a singular 20 document and will not go down into the HAV considerations and 21 the auxiliary features on turbine driven pumps, et cetera?
22 Is it scattered or integrated?
23 MR. RUBIN: Right now, let me mention three
() 24 documents. We have a draft regulatory guide that provides 25 general guidance on what should be looked at, in terms of ACE FEDERAL REPORTERS, INC.
202-347-3700 Nationwide Coverage 800-336-6646
30853.0 cox 118 1 assuring that this coping capability exists. We have -- not O 2 we have, but industry, ANS has developed a draft standard on 3 station blackout that goes through in more detail than a 4 regulatory guide, what should be looked at. That draft 5 standard had been -- has been put on, I say, the back burner, 6 or on hold, until industry can see in which direction NRS is 7 going to resolve this issue.
8 MR. EBERSOLE: Just to take a shot in the dark.
9 Does it have consideration of environmental temperature 10 controls on solid state control devices on which the 11 shut-down heat process is dependent in duration?
12 MR. RUBIN: The general approach.in the 13 environmental conditions is that it says that the equipment 14 that you are counting on to cope with a station blackout, 15 should be operable in the environmental conditions that exiet 16 during the station blackout. That means without HVAC. There 17 is guidance provided that if there is some determination that 18 a control cabinet, for example, is going to overheat.
19 First of all, the loads are much lower, as you 20 know, during station blackout. But there still may be some 21 areas of concern, either in small rooms, pump rooms, 22 electrical components, and if there is a concern, there can i 23 be approaches to resolve that concern. One might be what was
() 24 needed was opening some cabinet doors to get some natural l
25 circulation going, providing some portable blowers or fans.
ACE-FEDERAL REPORTERS, INC.
4 202-347-3700 Nationwide Coserage 800-336-6646
l 30853.0 cox 119 i 1 It has to be looked at in advance so an operator knows what 7s b 2 to do and is identified.
3 MR. EBERSOLE: I hope the plugs have some juice in 4 them.
5 MR. RUBIN: DC fans. If their definition of 6 station blackout does not include failures of the batteries.
7 In other words, AC power from converters from the batteries 8 is assumed to be available. But the duration that that DC 9 battery has would need to be addressed.
10 MR. EBERSOLE: That should be reflect in the life 11 of the battery.
12 MR. MICHELSON: If they plug them into the vital 13 power supply, they better wonder what the instruments are 14 going to do when they put that motor load on the instruments, 15 power supply.
16 MR. RUBIN: We can't, on a generic basis, say all 17 plants have to have this as a requirement. There are too 18 many plant differences.
19 MR. WYLIE: Were you going to discuss Table 3.
20 MR. RUBIN: I have Table 3. Wasn't planning on 21 it. If you have a question on it, I could discuss it. I 22 think this, for the -- the discussion would have to be very 23 brief. I can tell you what the basis of this table is.
() 24 By the way, I didn't finish this slide off. Over 25 here are the off-site power design characteristic groups. We ACE FEDERAL REPORTERS, INC.
202-347-3700 Nationwide Coserage 800-336-6M6
30853.0 cox 120 1 have three groups on here. Earlier you saw, in the cluster 2 analysis that John Flack prepared, there were five clusters.
3 These off-site power groups are directly related to those 4 clusters. This P-1 group was cluster 5, P-2 group was 5 cluster 4, this P-3 group were the remaining three clusters, 6 1, 2 and 3.
7 The characteristics of the clusters is that you 8 get up here, these have the best characteristics in terms of 9 lower likelihood of losses of off-site power, either because 10 of weather factors or switchyard characteristics that can be 11 less likely to lose off-site power.
12 Average group might be in the the P-2 group, lower i 13 than average or higher likelihood of losses of off-site power 14 in the P-3 group. Therefore, you see the longer durations 15 for coping with a station blackout in the P-1 group, you see
. 16 that the plant again.
17 Plants generally have lower durations, lower 18 target levels for coping with the station blackout.
19 The Table 3 that you referred to, Mr. Chairman, is 20 further defining how these P-1, P-2, P-3 groups are to be 21 determined. They are based -- it looks fairly complex. The 22 basis for these groupings is the cluster analysis that you 23 heard about earlier, is the characteristics of the site, the
() 24 switchyard, severe weather, extremely severe weather and the 25 likelihood of recovering from severe weather, affects the l ACE FEDERAL REPORTERS, INC.
202-347-3700 Nationwide Coserage 800-336-6646
30853.0 cox 121 1 likely who the of loss of off-site power and the likelihood 7-
's_/ -
2 of restoring off-site power.
3 I should mention -- I am getting ahead of myself 4 a little bit. But one of the industry initiatives, through 5 NUMARC, was to try to go identify which plants fell into 6 which categories. In other words, they have gone through --
7 not this exact analysis. They went through an analysis based 8 on the draft regulatory guide that was issued for comment.
9 These characteristics are the same, with some modifications 10 as to how they grouped, based on our reclustering. So the 11 utility group has gone through a characterization of each of 12 their sites in terms of the switchyard and whether to design A
13 weather characteristics.
1 14 MR. WYLIE: On the switchyard configuration, you 15 didn't try to make any assessment with regard to the design .
16 on the switchyard itself?
17 MR. RUBIN: An assessment of the design of the 18 switchyard?
19 MR. WYLIE: Yes.
20 MR. BARAN0WSKY: In terms of what design aspects?
21 We did look at, for instance, the number of independent 22 circuits that are available to provide off-site power to the 23 ESF buses.
() 24 MR. WYLIE: You didn't assess whether or not it
!, 25 was a breaker and a half configuration, double breaker
! /\CEJPEDERAL REPORTERS, INC.
202-347-3700 Nationwide Coserage 800-336-6646
30853.0 cox 122 1 configuration, or single bus, single breaker configuration in O 2 the switchyard; is that correct?
3 MR. BARANOWSKY: We didn't go down to that level 4 of detail in the reliable analysis, nor did we think that the 5 data was available on the total loss of off-site power data, 6 to distinguish differences between ring bus, breaker and a 7 half.
8 MR. WYLIE: Or single bus.
9 MR. BARANOWSKY: Whatever. A single bus, I can't 10 figure out exactly how that would work.
11 MR. WYLIE: There are stations that are designed 12 with a single bus and single breakers. There are breaker and 13 a half-ring buses that are a much more reliable station.
14 MR. BARANOWSKY: In terms of the likelihood of i 15 losing off-site power for a duration of significance on this 16 issue, we were able to identify the characteristics that had 17 to do with the numbers of pathways or circuits that could be 18 available for providing power and whether they were 19 automatically or annually actuated. Other factors, we tried 20 to look at. We just couldn't distinguish from the data, even 21 though in theory one might think those things are more 22 reliable.
23 MR. WYLIE: For all practical purposes, what you
() 24 have done, really, is you have considered them all as single 25 bus stations, and whether or not there were other independent ACE FEDERAL REPORTERS, INC.
202-347-3700 Nationwide Coverage 800-336-6646
__ _ . _ , _ .__.4 __._. - , _ _
30853.0 cox 123
,7 1 lines into the station or not.
V 2 MR. BARANOWSKY: We differentiated when they had 3 either large or multiple switch yards, as weil as whether 4 they had independent lines coming in.
5 MR. WYLIE: Doesn't mean whether or not you had 6 more than one line or not.
7 MR. BARANOWSKY: In particular, where you have two 8 different voltage level switch yards, that's certainly -- I 9 wouldn't call that a single bus.
10 MR. WYLIE: That's two yards connected by an auto 11 transformer, or something of that nature. But a large 12 switchyard could have multiple lines. But as far as that k 13 particular station is concerned, you considered them all a 14 single bus station.
15 MR. BARANOWSKY: That's true. There is some 16 grouping together of design factors like that.
17 MR. RUBIN: To continue on, this slide attempts to 18 identify some of the important features, items that are in 19 the regulatory guide, that would need to be addressed in 20 terms of assuring the capability of a plant to cope with a 21 station blackout.
22 We have gone through analysis of the accident 23 sequences, following a station blackout, and have identified A
() 24 a number of dominant factors that need to be addressed in 25 terms of systems and components and functions that are ACE FEDERAL REPORTERS, INC.
I 202-347-3700 Nationwide Coserage 800-336-6M6
30853.0 cox 124 1 necessary to cope with the loss of AC power.
2 The capability of the capacity of the station 3 batteries for a certain period of time, availabilities of 4 water to continue to move decay heat, to condensate storage 5 capability.
i 6 Batteries, for example, if there were a battery 7 capability that would not meet minimum coping durations, 8 there were alternatives that were possible. We don't specify 9 which ones a utility has to take. They may be anywhere from
. 10 procedural, which is shedding nonessential loads, to adding 1
11 an independent battery charger, or adding additional 12 batteries. But some way to meet the functional requirements 1 13 for batteries for a certain period of time. Similar to 14 condensate storage tank.
15 MR. MICHELSON: Excuse me, on the station 1
16 batteries, you say you talk about charge herself, which 17 infers that you mean to charge the batteries during the i 18 blackout; is that right?
19 MR. RUBIN: Yes.
20 MR. MICHELSON: How do you handle the battery room 21 ventilation during that period of time?
l 22 MR. RUBIN: That would need to be addressed.
23 MR. MICHELSON: It would have to be poured i ;
- () 24 appropriately and ventilation would not just have to go into
{ 25 some duct, but there have to be exhaustors at the other end ACE FEDERAL REPORTERS, INC.
202-347-3700 Nationwide Coserage 800-336-6M6
30853.0 cox 125 1 of the duct work where you fill the duct work.
2 MR. RUBIN: It would be considered a hydrogen 3 build-up.
4 MR. MICHELSON: Quite complicated when you start 5 talking about charging that.
6 MR. BARANOWSKY: That's true. Probably isn't a 7 large accumulation of hydrogen in a few others. On the other 8 hand, one has to look at the consequences of any fix 9 completely. That would be one thing you have to consider. I 10 don't think for the types of two, four- and eight-hour 11 durations we are talking about the accumulations would be 12 significant.
l
( 13 MR. MICHELSON: If you are charging them, they l 14 could be quite significant. As I recall, you don't get any 15 hydrogen on the discharge, you have to be charging, which 16 might mean you would be charging one bank -- in other words, 17 you are supplying part of the energy to charging part of it 18 to the load. Whatever charging rates you induce into the t
19 battery bank is developing a calculatable amount of hydrogen, l 20 or you don't want it to accumulate in a battery room; maybe 21 you could exhaust it out of the battery room into the duct 22 work. Unless it goes directly into atmosphere, you have to 23 worry about a quick checking of the duct work.
() 24 They have to have the induction equipment out on 25 the other end of the line; otherwise, you just pull it out of ace FEDERAL REPORTERS, INC.
l 202-347-3700 Nationwide Cmerage 800-336-6M6
30853.0 cox 126 1 the room into the duct and that's not enough.
2 MR. EBERSOLE: Another matter is as there is a 3 standing hydrogen leak on supply systems, hydrogen 4 concentration may just be maintained by ventilation 5 capability. You lose it, you will get an accumulation, 6 but --
7 MR. MICHELSON: That is another problem, that is, 8 of course, leakage from hydrogen systems that you might have 9 in the building.
10 MR. EBERSOLE: They will accumulate. When you get 11 power back, you will have a surprise.
12 MR. RUBIN: Other items are compressed air 13 capability for operating the necessary valves for controlling 14 decay heat removal --
15 MR. MICHELSON: On the question of compressed air, 16 in determining the effect of the power blackout, is there 17 some kind of an analysis done that determines the effec: of a 18 slow bleed down in air pressure, which would be the case 19 unless you provide some kind of a dump valve to kill the air 20 pressure. You are getting a very slow bleeddown as you use 21 up the air in the accumulator tanks and so forth. Is that 22 analyzed in terms of the effect on plant -- things that are 23 happening in the plant during that period of time?
() 24 MR. RUBIN: During the station blackout?
25 MR. MICHELSON: Yes.
/\CE FEDERAL REPORTERS, INC.
202-347-3700 Nationwide Coverage 800-336-6646
I 30853.0 cox 127 1 MR. RUBIN: If there were a known loss or 2 bleeddown?
3 MR. MICHELSON: Yes. It will clearly bleed down 4 with time, in some manner.
5 MR. RUBIN: One of the reasons to look at that 6 issue is to find out roughly how long the air condition will 7 provide the capabilities to operate valves.
8 MR. MICHELSON: To operate essential equipment, 9 but it's also operating nonessential equipment by the fact 10 it's bleeding down pressure even nonsafety-related valves.
11 Some of them open, close, whatever, on loss of air pressure.-
12 They will be doing their thing, and that will be the state, 13 during the blackout. And then when power comes back on, that 14 will also be the state of the system, that you are suddenly 15 repouring -- you might have to go through and do some manual 16 operation before you dare repower some of the system.
17 MR. RUBIN: Part of the analysis --
18 MR. MICHELSON: Discharge valves may be closed 19 that you thought were open and pumps start up because the 20 power comes back on, and depending on what all the logic was
- 21 doing. Interesting problem.
22 MR. BARANOWSKY: Especially air-operated valves.
- 23 MR. MICHELSON
- Yes. They will slowly move or do
() 24 something as the air pressure bleeds off the system. It's 25 just a matter of thinking ahead so that you don't get a lot ACE-FEDERAL REPORTERS, INC.
202-347-3700 Nationwide Coserage 800-334 6646
._. _ . - ~ , , _, _ _ . - _ _ _ , , _ _ _ _ _ - . _ _ _ _ _ _ _ . . _ . . _ _ . _ _ _
30853.0 l cox 128 1 of surprises when the power comes back on.
2 MR. RUBIN: Part of the coping capabilities 3 includes the period during the blackout and the recovery 4 time. If there are loads that would come on automatically 5 and overload the essential buses, diesel generators, when 6 power was available, that should be looked at.
7 MR. MICHELSON: I would think the study of 8 previous more extensive blackouts would probably give you all 9 the data you need to kind of determine how plants behave when 10 the power comes back on. It may be that the utilities took a 11 number of precautions before they allowed the breakers to be 12 reclosed even. This is all part of what you have to look at
/~1
\/ 13 in terms of the recovery plant.
~
14 MR. RUBIN: As Pat mentioned earlier, there have 15 only been a few actual total losses of all AC power for 16 relatively short periods of time.
17 MR. MICHELSON: Maybe they weren't long enough to 18 see some of the effects. How do you handle this problem of 19 the RCIC, which I think you want to use during the blackout, 20 and the fact that its steam isolation provisions are on 21 temperature and the chase -- the piping chases are no longer 22 ventilated, so they start getting hot from the thermal energy 23 from the steam line? Even in cases where ventilation has
() 24 gotten a little bit sloppy, they have hit isolation of RCIC, 25 temperature detector, says I think I have got a steam line ACE FEDERAL REPORTERS, INC.
202-347-3700 Nationwide Coserage 800-334 6646
30853.0 cox 129
,~ 1 break, even though there wasn't any. How was that done on 2 this case. Is there a handy switch to cut out the automatic 3 isolation or something?
4 MR. RUBIN: If the temperatures got enough.
> 5 MR. MICHELSON: There should be no doubt about 6 that. Those are rather confined and, obviously, the RCIC 7 steam line is still hot.
8 MR. BARANOWSKY: The analyses we have seen 9 indicate that the RCIC rooms might reach the environmental 10 temperature for isolation in four to eight hours, depending 11 on openness of the room and size. There are things you can 12 do in terms of opening up doors to provide ventilation, if 13 you know beforehand you have to do it, and you can also 14 defeat that isolation interlock. It's the kind of thing that 15 might have to be addressed if you had a rather rapid room 16 heatup around the RCIC.
17 MR. MICHELSON: How about the fire protection 18 around the RCIC and HPCI pumps, how it's arranged. Hopefully 19 around something besides temperature. If they put in a spray 20 system with thermal links, you are sunk, of course. Even 21 then the question of products and combustion, I don't know 22 how those detectors work if they get hot enough. They may 23 also actuate even though there are no products of l
l
() 24 combustion.
25 MR. BARANOWSKY: Possibilities. .I don't recall
/\CE FEDERAL REPORTERS, INC.
202-347 3700 Nationwide Coserage 800-336-6M6
l 1
l 30853.0 cox 130 1 whether this is addressed or not, Alan, in the ANS guide.
(
2 Things like that are addressed. I 3 MR. RUBIN: We have identified the need to look at 4 heat loads where there is either power or steam in areas and 5 what the effects of that would be.
6 MR. MICHELSON: We know in the case of HPCI, if 7 you turn the fire protection on which has happened 8 inadvertently, already, it kills the HPCI, because it gets 9 into some of the electrical equipment like the motors on the 10 land seal condensors, that sort of thing. So you just really 11 don't ever want to spray that system unless there is a real 12 fire.
13 MR. BARANOWSKY: I agree. Room heatup would be an 14 item that one would have to look at, and the consequences of 15 room heat.
16 MR. RUBIN: In other areas where other equipment 17 is necessary.
18 MR. MICHELSON: I am sure if you question the 19 entire building in terms of if there is an area that can get 20 hot, even though you are not depending on the equipment. If 21 the fire protection comes on and sprays the equipment, are 22 you really going to get yourself in further difficulty, 23 including difficulty when power does come back on, can you
() 24 now recover like you thought you could, because you only 25 designed for two hours of loss, you may find when the power ACE FEDERAL REPORTERS, INC.
202-347-37(U Nationwide Coverage 800-336-6M6
30853.0 cox 131 1 comes back on, in the meantime you have lost your boards from 2 being wet down by fire protection.
3 MR. RUBIN: I think we are talking about a limited 4 period of time of generally four or eight hours. Some of the 5 concerns that you bring up are important, I agree. I think 6 when you hear later on, one of the modifications we made, 7 based on public comments, was that we take out the 8 requirement that plants evaluate their maximum capability to 9 cope with the station breakout. I will talk about that later 10 on.
11 I think this slide has been discussed enough 12 already, with other points made earlier on developing O 13 procedures and maintaining diesel generator. reliability. So j 14 let me go on to the next subject that I said earlier we will
- 15 talk about, which is a very brief summary of the value impact i 16 analysis. Just putting this up for purposes to illustrate i
j 17 that when we develop recommendations and resolutions of these l 18 issues, we need to show a benefit analysis and estimates of 19 the cost to comply with the Commission's backfit rule. It's l.
20 nice to postulate a lot of nice requirements that might be f 21 coming down the road. But if the Staff can show the cost
! 22 effectiveness, it's not likely to get very far.
l 23 Summary for the station blackout issue, based on l
() 24 the analyses we showed earlier, the mean core damage t 25 frequencies for 100 or so plants that we evaluated was about l
l /\CE.FeoERAL REPORTERS, INC.
- o:-m 37m Nationwide coserage 8m-336.ua6 l
i . - - - - __ _ _____ _ ___ _ _ _ _
f 30853.0 cox 132 em 1 4 times 10 to the minus 5th reactor year.
U 2 I should point out these analyses, these 3 estimates, are not to provide the requirements, but to 4 provide some characteristics that could show what the 5 benefits are to the rule. They are based on certain 6 assumptions. The assumptions, one of the assumptions for 7 this core damage frequency is that plants are able to cope 8 with a blackout for two hours, before the rule, if we do 9 nothing.
10 If we carry out and implement the recommendations 11 for the recommendations of this issue, as we discussed 12 earlier, we see an estimated mean core damage frequency for (m
k- 13 all plants of around between 1 and 2 times 10 to the minus 14 5th for a reactor year. These are means. The point is there 15 are differences between the plants how many plants may have 16 higher benefits, in terms of reduced core damage, have also 17 been estimated. This is an overall average.
18 The cost side, average cost of around half million 19 or $600,000 per reactor, total industry cost around $60 20 million. We have ranges of these values given in the 21 detailed documents that you have. Our estimates of risk 22 reduction, in terms of minimizing or reducing consequences of 23 this event, are 145,000 person-REM over the life of 100
() 24 reactors for the next 25 years, which comes down to a value 25 impact ratio on the bottora line of 2400 person-REM per ACE FEDERAL REPORTERS, INC.
202 347 3700 Nationwide Coserage 80tb336-6616
30853.0 cox 133 1 million dollars.
l 2 We have looked at ranges, higher cost estimates, 3 lower cost estimates for the one given here, as well as
( 4 higher and lower consequences, or estimated risk reduction, 5 and have come up with a range for this ratio which does not 6 include on-site costs of between 700 to 5000 person-REM per I
7 $1 million.
8 If we included on-site cost -- the word from the 9 Staff now is we did not include these costs, these were 10 averted damages by use of reducing the cost for repair, 11 replacement power, in the event of an accident at a nuclear 1
12 plant -- those costs are estimated to be present valued at I
( 13 $38 million, and that would make a significant change in the 14 value impact ratio in terms of additional benefits. So that !
( l 15 number is provided just for information.
16 I would like to move on in consideration of the 17 time, and skip one slide, and go on to, I think, an important 18 area, which is response to public comments. The slide I am 19 skipping shows the distribution of estimated core damage 20 frequencies for plants. But I think there are certain areas 21 that I do want to cover, and I think from the comments 22 earlier, need to be addressed.
23 I am getting to public comments on the proposed
() 24 rule. Proposed rule was issued for comment in March, 1986.
25 90-day comment period ended in June. We received 53 letters ACE-FEDERAL REPORTERS, INC.
202-347-3700 Nationwide Cmerage 800-336-(646
l 1
l 30853.0 cox 134 l l
1 responding to the proposal. 45 of the letters were from !
) '
2 utility industry, either nuclear utility groups, such as 3 NUMARC or individual utilities. In general, characterize the 4 comments from the nuclear industry that they are opposed to 5 the proposed rulemaking, and NUMARC has proposed some 6 initiatives as an alternative to the rulemaking, which I will 7 discuss later on, and I am sure you will hear more about at 8 the full committee meeting on Friday, i
9 Received eight other letters from -- anywhere from 10 the state to individuals to consultants to UCS, that in 11 general supported the objectives of the rule, but didn't feel 12 that the Staff went far enough. In other words, they wanted
' 13 some additional coping time or recommendations that we 14 consider approaches, such as the French have, for long 15 duration coping capabilities. That was the' gist of the 16 comments that were not from the utility industry.
17 The 53 letters.were all reviewed, and we 18 characterized the comments. Before I discuss that l 19 characterization, I am going to discuss some of the changes 20 that we made in this resolution that were based on the public
, 21 comments, and, in part, on a reanalysis that was done in 22 response to the public comments.
23 The question came up earlier whether or not we
() 24 need to go out for additional comments because there were 25 some significant changes in the rule itself. Let me ACE FEDERAL REPORTERS, INC.
202-347-3X10 Nationwide Coserage 800-336-6646
30853.0 cox 135 3 1 characterize the changes to the rule and characterize them in l
v 2 two items.
3 One, I mentioned earlier that in relation to the 4 maximum coping capability, there was a requirement that 5 plants analyze the maximum extent of time that they could 6 cope with a station blackout. We had comments that that was 7 not well defined, it was very difficult to determine when you 8 stop this analysis, and could be very, very costly. We have 9 taken out the requirements to determine this maximum coping 10 time, but we still have a requirement that is in the rule 11 that plants should be able to analyze and evaluate their 12 plants for this acceptable duration of four hours or eight O'
'- 13 hours1.50463e-4 days <br />0.00361 hours <br />2.149471e-5 weeks <br />4.9465e-6 months <br />.
14 MR. EBERSOLE: Let me ask a question. Suppose a 15 plant, and I will take the case of the boiler as the case in 16 point, can show that with modest cost increments, they can 17 show an indefinite survival time without any AC power. What 18 kind of credit would you give them, if any? Nothing?
19 MR. RUBIN: Give them all the credit in the 20 world. We give them credit for complying with the rule.
21 They can cope with it --
22 MR. EBERSOLE: That's not an unbelievable 23 situation.
() 24 MR. RUBIN: The approach that the utilities -- I 25 mentioned this earlier, alternate AC approach, which gives ACE-FEDERAL REPORTERS, INC.
202-347-3700 Nationwide Coserage 800-336-6M6
30853.0 cox 136 s 1 them -- it may not be indefinite, but a long period of time.
2 We would give them credit for it. They would meet 3 the rule, then. Plants that had an indefinite coping 4 ability, and we don't know there are any --
5 MR. EBERSOLE: It might be just as well to say we 6 can get along without AC, period, might be the least 7 expensive way.
8 MR. RUBIN: We said that, they should be able to 9 get along without AC. It does not require that. It could be 10 expensive. Certainly more expensive backfits and 11 modifications than we have indicated in our value impact 12 analysis.
's) 13 MR. EBERSOLE: You don't know that. It might 14 not.
15 MR. RUBIN: To our best estimates. If there are 16 less costly alternatives, we certainly would endorse those 17 other approaches. But this is a change from the proposed 18 rule.
19 Another change, Which is really a minor -- we 20 clarified the definiti4 cf station blackout. There were 21 questions whether or not station blackout meant the loss of 22 AC power from inverters. Specifically said, no, it does 23 not. That was a clarification. The intent was the same in
() 24 the proposed rule, we clarified that.
25 MR. EBERSOLE: It's not only that, but in certain
/\CEJFEDERAL REPORTERS, INC.
202-347-3700 Nationwide Coverage 800-336-6M6
30853.0 cox 137
)
l 1 cases for the core spray system; right?
2 MR. RUBIN: That's right. Those are discussed in 3 terms of credit for these other power sources. They are 4 discussed as guidance in the regulatory guide. In other 5 words, the coping capability, to comply with this rule, can 6 be shown or demonstrated that there's an alternate AC 7 approach that NUMARC has discussed. Changes to the 8 regulatory guide, you mentioned earlier, what we have 9 clarified guidance, giving credit for alternate AC. It was 10 never intended that that be excluded. We didn't focus on 11 that.
I- 12 We tried to clarify guidance on giving credit for
.' () 13 alternate AC power. We clarified guidance on diesel 14 generator failure rates. Wasn't specified earlier. We got 15 comments. What do you do about auto start failures that came I Should that count as a failure in the 16 up at this meeting.
- 17 diesel generator reliability targets? Our regulatory guide i
18 1.108 would count auto start failures as failures. We have I 19 excluded that, which is consistent with the analysis that was 20 down.
l 21 We modified guidance to reflect maximum coping i 22 time. We have added regrouped plants and added categories of l
23 two hours and 16 hours1.851852e-4 days <br />0.00444 hours <br />2.645503e-5 weeks <br />6.088e-6 months <br /> in the matrix.
i f
() 24 In the regulatory guide we deleted any reference l 25 to determining maximum coping capabilities that have been ACE FEDERAL REPORTERS, INC.
202-347-3700 Nationwide Coserage 800-336 6646
30853.0 cox 138 l 1 included in there to reflect the rule.
2 We have revised and relooked at the value impact 3 analysis to summarize in NUREG-1109. It includes 100 plants,
'4 whereas previously we looked at 67 operating reactors a 5 couple of years ago. Accounts for results of source term 6 work, recent NRC work done on source terms, discussing NUMARC 7 initiatives, which I will get into. You already heard about 8 the updated analysis that was done for NUREG-1032 technical 9 basis.
10 In my judgment, these changes are, in my own view, 11 are rather secondary in nature in the rule, and would not 12 require that we go out for additional public comments.
13 MR. MICHELSON: Let me talk to you, you talked in 14 the regulatory guide about the plant being able to recover 15 and withstand from a station blackout lasting a specified 16 minimum duration. Then the utility comes back and tells you 17 what their minimum duration is going to be. Then the Staff 18 looks at that and decides whether or not that is okay. What '
19 is your rationale or approach in the decision of whether that 20 is the particular number they come back with would be okay?
21 MR. RUBIN: If a utility says okay, we are using 22 this as our basis --
23 MR. MICHELSON: One of the guidances you give is -
() 24 it says the probable time needed to restore the off-site 25 power is one of the guides. That one seems to be a funny ACE FEDERAL REPORTERS, INC.
202-347-3700 Nationwide Coverage 800-336-6M6
30853.0 cox 139
- 1 one. It isn't -- I don't think it makes any difference how 2 long, but rather how long can the plant remain in a safe 3 state. How long does it take to get power-back. So I don't 4 know that that would help you much. The frequency of
! 5 off-site power, I guess, helps you to calculate the risk, of 6 course. But once having decided that the risk is enough that 7 they must do something, how do you use that. Do you do a 8 sensitivity study or something to see how the risk varies?
9 MR. BARANOWSKY: Could we go back to your 1
10 statement on the probability of loss of off-site power not 11 being important. I would say that in general we have to i 12 disagree with that comment, because we spent a fair amount of 13 time looking at how important it is to recover power, whether 1 14 it's off-site power or alternate AC power, and we see that 15 that can be a significant factor in causing either low or 16 reduced risks associated with station blackout.
17 MR. MICHELSON: I think you have probably taken my 18 words out of context. Having decided that there must be a 19 minimum duration, and the utility comes back in four hours, 20 they say we can do it in four hours, you must decide that 21 four hours is okay. One of the factors apparently you are 22 going to use is the probable time needed to restore off-site 23 power.
t
() 24 MR. BARAN0WSKY: That's how you get into a NUMARC l 25 group.
JACE FEDERAL REPORTERS, INC.
} 202-347-3700 Nationwide Coserage 800-336-6646
. - -~. . . . _ , . - , . - - - - - . _ . - . . . ~ . . . . - - , . . - _ - - - . - , .-. .- - - _ - - . - - , . - - .
30853.0 cox 140 1 MR. MICHELSON: You didn't put it in a group, the 2 utility comes back and says our answer is four hours. We 3 think our minimum time should be four hours.
4 MR. BARANOWSKY: There are criteria that say how 5 you derive whether or not you are in a four-hour group. One 6 of the criteria has to do with looking at your off-site or 7 alternate AC power syhtems to determine whether or not they q 8 are sufficiently reliable or capable of being restored in a 9 four-hour time frame. So it's not just you pick a number and 10 then see what the probability is; that's all pre-thought 11 out.
1 12 MR. MICHELSON: You are saying that the first i O 13 thing you will do is see if they categorize themselves 14 correctly.
15 MR. RUBIN: That's right.
16 MR. MICHELSON: If you don't think they have, 17 that's where you start.
- 18 MR. SPEIS
- Yes.
19 MR. MICHELSON: This is a factor in deciding 20 whether they have categorized themselves correctly. Okay.
21 Now, if they think they are a four-hour plant, 22 then really, the only review you do is to determine that 23 they, indeed, have categorized themselves correctly. If they
() 24 have, then nothing more need be done.
25 MR. BARANOWSKY: On that classification.
ACE FEDERAL REPORTERS, INC.
202-347-3700 Nationwide Coserage 800-336-6M6
30853.0 cox 141
- 1 MR. SPEIS: But then the coping is the next step.
\_/
2 MR. MICHELSON: Yes. Then they have to 3 demonstrate by whatever means they can handle the four-hour 4 duration like they said. If you find in that process that 5 they really can't cope with four hours, then you have to --
6 there's some kind of an agreement, or you will just specify 7 that it shall be only two hours because of a particular 8 problem. Is that the way it works?
9 MR. BARANOWSKY: In general, they would either 10 propose modifications to allow their plant to cope for four 11 hours1.273148e-4 days <br />0.00306 hours <br />1.818783e-5 weeks <br />4.1855e-6 months <br />, or they would come in with another technical basis, 12 which would be allowed, that shows that for their plant, some
')
13 of the generalities that were used in deriving that 14 requirement are not quite applicable, and they are actually 15 better from a reliability point of view than our criteria 16 might suggest.
17 MR. MICHELSON: We are getting closer to what I 18 was leading to. That is, let's say I come in and I can cope 19 for a half hour. Now what do you do. First of all, you 20 verify that I am right.
21 MR. BARANOWSKY: I think this may help, 22 hopefully.
23 MR. MICHELSON: That was probably one I missed, C)
(_ 24 since I had to be gone for a while.
25 MR. RUBIN: First step, which is what we are 1 ACE FEDERAL REPORTERS, INC. l 202-347-3700 Nationwide Coserage 800-336-6M6
30853.0 cox 142 l
l 1 discussing right now, let's call it an acceptable coping 2 duration. What is acceptable, in terms of the likelihood of 3 loss of station blackout occurring? What is an acceptable
- 4 duration so that the probability of recovery of AC power is 5 high? This has nothing to do with what the plant's actual 6 capability is yet.
7 MR. MICHELSON: Does the licensee do that or do 8 you do?
9 MR. RUBIN: They come back in and say we have 10 determined what our coping duration should be. NRC, do you 11 agree, here is what it is. They may use regulatory guide or 12 some other arguments.
( 13 Then, once that is determined, is the question, 14 can the plants actually cope for that period of time. I 15 think this is where your questions are. Suppose utility 16 comes in and says I can only cope for a half hour, or I can 17 only cope for two hours. That's a decision down here, or can 18 you cope for as long as you should be able to, target 19 values. If you can, fine, you are done. You have some 20 procedures to do it. If you can't, you may need to make some 21 modifications. The question is, if you are within a short 22 period of time, 15 minutes, suppose you are in a four-hour 23 group and the licensee plants can cope for three hours 45
() 24 minutes. It's not likely that it will be cost effective to 25 make any modifications.
ACE FEDERAL REPORTERS, INC.
202-347-3700 Nationwide Coserage 800 336-6646
30853.0 cox 143 1 MR. MICHELSON: You are saying the two to four 2 hour is your range of acceptability, predetermined range of 3 acceptability; is that right, for most plants, that's the 4 target you have in mind?
5 MR. BARAN0WSKY: We feel that most plants would 6 fall in the four to eight hour categories, both in the four.
7 MR. SPEIS: We have assumed in this cost benefit 8 analysis, every plant called for two hours. That's an 9 assumption.
10 MR. MICHELSON: Yes.
11 MR. RUBIN: We have not verified that for every 12 plant.
13 Let me go through quickly the summary of the 14 public comments.
15 The Commission requested comments on four areas of 16 the proposed rule. The first one being the quality 17 classification of modifications. What I am showing on the 18 slide is the category of comments. In parentheses, whether
, 19 there was a change from the proposed rule to what we i 20 currently have in front of you. The question was should the 21 equipment that would be required to cope with a station 22 blackout be safety grade, be seismically qualified, et i
23 cetera. We have reviewed the seismic question before,
() 24 recommendation, which is similar to earlier, that the 25 equipment need not be seismically qualified. So, therefore, j ACE. FEDERAL REPORTERS. INC.
202-347-3700 Nationwide Coverage 800-336-6M6
30853.0 cox 144 1 it should meet some reliability criteria, maintenance, O 2 testing. It doesn't have to be seismically qualified or be 3 classified 1-E.
4 MR. MICHELSON: It's pretty clear that you are 1
5 likely to get a power blackout external grid, at least, 6 during a seismic event. I guess there's no argument that 7 that probability is 1 or very close to it, or am I wrong?
8 MR. BARANOWSY.Y: Not just for any seismic event.
9 MR. MICHELSON: I am talking about SSE.
10 MR. BARANOWSKY: SSE, reasonable to expect a loss 11 of off-site power.
12 MR. MICHELSON: You are using 1 in your analyses, 13 I assume.
14 MR. BARANOWSKY: That's what we assumed in looking 4
15 at seismic to see how big of an issue it would be.
16 MR. MICHELSON: Unless somebody can demonstrate 17 it's less than 1, 1 is the number you expected to use.
i 18 MR. BARANOWSKY: When we used 1, we didn't find i
19 that it was a large contributor to the probability of loss of 20 off-site power.
21 MR. MICHELSON: Next question is, having --
22 working on that basis, then, in doing your seismic -- your 23 risk analysis from the seismic event for the plant itself,
() 24 you are dealing now with certain probability that the diesel 25 won't start, for the seismic event, which I guess is the same
/\CE FEDERAL REPORTERS, INC.
202-347-3700 Nationwide Coserage 800-346M6
30853.0 cox 145 1 probability as normal. You are not differentiating because 2 it's a seismic.
3 MR. BARANOWSKY: That's correct. As long as it 4 was within safe shutdown earthquake or equivalent.
5 MR. MICHELSON: There is a finite probability that 6 the diesels don't start. Now you have the loss of all AC 7 power. How mucit credit are you taking for nonseismically 8 qualified equipment during the seismic event to keep yourself 9 out of trouble?
10 MR. BARANOWSKY: None that I know of.
11 MR. MICHELSON: Aren't these -- other features 12 that take care of the power blackout case are not necessarily O 13 seismically qualified.
14 MR. BARANOWSKY: I see. If you are talking about 15 some of the add-on coping capability.
16 MR. MICHELSON: Yes. That's assuming you didn't 17 have to have some and they weren't seismically qualified.
18 MR. BARANOWSKY: Correct.
I 19 MR. MICHELSON: In your analysis, how much credit 20 are you going to take?
21 MR. BARANOWSKY: Our analysis doesn't take any 22 credit for those things, because we find that the probability 23 of a seismic event causing loss of off-site power at the SSE
() 24 level, combined with the nonmechanistic or nondependent 25 failure of the diesel generators, is low enough that it's l
ACE. FEDERAL REPORTERS, INC.
202-347 3700 Nationwide Coserage 80)-336-6M6 l
30853.0 i
cox 146 1 acceptable.
2 MR. MICHELSON: What you do now is to make sure 3 that the nonmechanistically pertains, you ask the utility to 4 look at this plant from the viewpoint of thsre aren't going
{ 5 to be any spurious accusations of fire protection that will l 6 mess up the equipment, things of that sort. So that, indeed,
- 7 there's no mechanistic connection between the seismic event 4
8 and the loss of off-site power.
i 9 MR. RUBIN: What we do, having the guidance, any I
10 equipment that might be added to meet the requirements to 11 comply with his rule, although it need not be seismically 12 qualified, there should be assurance that it's a seismic
! } 13 event would not impact -- have a systems interaction impacts j 14 of the edit equipment on the safety systems.
i 15 MR. MICHELSON: I didn't find that requirement in 4
16 the regulatory guide. Did I miss it?
i 17 MR. RUBIN: Should be there.
18 MR. MICHELSON: Where in the regulatory guide does
< 19 it say there won't be any systems interaction effects?
I
- 20 MR. RUBIN: Doesn't say systems interaction, it 21 says that the -- I will come back. I don't have it in front i
! 22 of me right now.
t 23 MR. MICHELSON: In a little bit if you could cite
() 24 it. I want to read it so I make sure I understand.
25 MR. RUBIN: By the way, the range of comments on ACE. FEDERAL REPORTERS. INC.
202-347-3700 Nationwide Coserage 800-336-6M6
30853.0 cox 147
,_ 1 this one was nuclear industry saying no, they should not be V 2 quality classified 1-E. Some of the non-utility comments was 3 they should be. Without a lot of justification, just 4 opinions were expressed.
5 Second Commission question was does backfit 6 analysis implement the backfit rule. Comments that we 7 received in this were discussed in a number of categories, 8 from considerations that we have overestimated the benefits 9 of value impact analysis in terms of risk reduction, and we 10 have underestimated the cost to implement the rule. We havo 11 taken a relook at both of those points and have revised our 12 analysis and revised our backfit analysis. We think we have O
\/ 13 included that in the package that you have received. No 14 changes to the rule based on that. Other than revising the 15 cost benefit analysis.
16 Would the modification substantially increase 17 overall protection of public health and safety? We feel on 18 the value impact analysis we can't answer that question as 19 yes.
20 Fourth is point: should NRC require improvements 21 beyond those proposed? Comments from the utility industry 22 says no. Comments from some of the nonutility industry says 23 yes, with, again, opinions from the nonutility industry but
() 24 not adequate justification. We have not made comments based 25 upon the public comments on that aspect. In fact, we have ACE FEDERAL REPORTERS, INC.
202 347-3700 Nationwide Coserage MO-3%f616
30853.0 cox 148 1 eliminated some of the requirements, in particular maximum 7-V 2 coping duration.
3 Other public comments, again categorized, they are 4 extensive. The extent of coping studies. In determining the 5 maximum coping capability, there was a lot of concern from 6 utilities that the specificity of the requirements and when 7 it ends wasn't clear. What we have done to respond to that 8 requirement, we have eliminated the requirement for a 9 determination of the maximum coping capability of the plant.
10 That's been taken out of the rule.
11 Credit for alternate power AC sources, what about 12 power sources we have on-site that we should be able to use O 13 and take credit for? We discussed that earlier, we would 14 provide additional guidance in the regulatory guide for using 15 both AC power sources. We have accepted, in principle, that 16 approach.
I 17 Acceptable duration to cope with station l '
5 18 blackout. There were comments earlier, as from CRGR, they
- 19 should be zero coping capabilities, plants should have i
! 20 features that show them less than the guidance we have 21 addressed. These comments are more on the regulatory guide 22 than the rule itself, by the way. We have revised the 23 guidance in the regulatory guide. You have seen the matrix
() 24 that I have prepared earlier.
25 The need for generic rulemaking. This was a ACE FEDERAL REPORTERS, INC.
202-347-3700 Nationwide Coserage Mn336-6M6 I
30853.0 cox 149 l s 1 comment made by industry that station blackout is very b There are different l
2 plant-specific in nature.
3 characteristics at the site, different levels of risk.
4 Therefore, it's very -- shouldn't be a generic requirement.
5 We felt that the requirement for all plants being able to 6 cope with a blackout, we are still proposing that, we have 7 not changed our requirement. We do take into account 8 plant-specific, nature of the problem, and the regulatory 9 guide.
10 These two comments are related, plant-specific 11 features and capabilities. Some utilities came in and 12 comment that we have this or that feature, can we take credit
\ ') 13 for it. Yes, in the plant-specific analyses we would.
14 We have looked at source term, we have revised 15 that and considered ranges of source term based on current ,
16 information that's been prepared in preparation of ,
17 NUREG-1150.
18 I should say that just to clarify where we are on 19 the source term work, based on comments, we have taken ranges 20 in reduction of what was an earlier SST source term in 21 WASH-1400, reduction of that source term, release fractions, 22 from ranges of 1/3 to 1/30 of the SST-1 releases, so there 23 are significant reductions in the risk analysis.
() 24 Comments on the trends and reliability of AC power 25 sources, things are always getting better, diesels are ACE-FEDERAL REPORTERS, INC.
202-347-3700 Nationwide Coserage MWh3364M6
30853.0 cox 150 s 1 getting better, off-site power insurance is showing 2 improvements in their reduced frequency of losses of off-site 3 power. We feel we have already taken these trends into 4 account in our analyses, as you have heard earlier.
5 Specificity and clarification of requirements. We 6 have tried to clarify in the rule and in the regulatory 7 guide, addressed some of the specific comments and where 8 there were questions in these areas.
9 Another comment that was made by ACRS was made 10 earlier, which we addressed in a Commission paper, the 11 relationship to other issues. Previous discussions two years 12 ago, the ACRS felt that we had done a capable job in O
k/ 13 addressing related issues, and we have discussed that in our 14 package in response to public comments. We have not made any 15 changes from the proposed rule.
16 There is one last topic. It has to do with 17 industry initiatives. I am going to cover this briefly. You 18 will have an opportunity to discuss this more with utility 19 group and NUMARC.
20 Early on, we were working with this issue, Pat 21 Baranowsky mentioned earlier there was some significant work 22 that was done by EPRI in checking data on loss of off-site 23 power experience and diesel generators reliability. We have
() 24 made good use of this information in developing our 25 recommendations and doing the analysis.
ace FEDERAL REPORTERS, INC.
202-347-3700 Nationwide Cmcrage 8m-3 W6M6
30853.0 cox 151 1 We mentioned earlier an ANS standard that had been 2 worked on, work started around 1984 and 1985, a draft has 3 been proposed, gone out for comment within the ANS group.
4 Work on that standard has been put on hold pending NRC's 4
5 resolution of this issue.
6 NUMARC, in responding to public comments, has 7 proposed initiatives that you will hear about tomorrow. I 8 will briefly discuss them, showing perspectives and where
- 9 there are some differences and agreements between the 10 industry and the Staff's proposal.
4 11 We have been meeting with NUMARC a number of times 12 from July to the present to discuss for NUMARC to come in and 13 discuss their guidance, their guidelines and initiatives with 14 us. The written NUMARC initiatives were very brief in 15 nature, and they wanted to discuss with us more details of s
16 their initiatives.
17 You may be hearing about this tomorrow; I am sure 18 you will. There were four initiatives. I will just briefly 19 touch on those initiatives and tell you whether we are in 20 agreement or difference.
21 One issue was to categorize plants into these 1 22 categories, four- or eight-hour groups. Utility group has 23 sent out a guidance for each utility to evaluate their plant (j 24 characteristics to see how they fall in this matrix, four or 25 eight hours. They have completed this analysis. They may be
}
ACE FEDERAL REPORTERS, INC.
202-347 3700 Nationwide Coserage 800-33HM6
30853.0 cox 152 1 making some modifications because of the revisions we have O 2 developed, which are relatively small in the guidance, and 3 they say for the plants in the eight-hour group, they will 4 fix those plants, do something to the AC power system either 5 in terms of reliability or alternate sources, to get them 6 into a four-hour group. Their focus is on the plants in the 7 eight-hour group.
8 I must say that the guidance that the NUMARC used 9 was consistent with the guidance we had in the draft 10 regulatory guide to categorize the plants. We have not seen 11 the results of this analysis in terms of which plants fall 12 into this group. Just a general summary right now.
13 Second initiative has to do with procedures; 14 NUMARC is proposing that plants have procedures to cope with 15 a station blackout, which we are in agreement with. Where 16 there are some differences is that the specific duration of 17 how long a plant should be able to cope with a blackout, and 18 assessment of the actual capability is not included right now 19 as far as we are aware in the NUMARC initiatives.
20 There are several points in the NUMARC initiatives 21 which we think are steps in the right direction. We are not 22 on 100 percent agreement, though, yet, at least NUMARC and
- 23 the Staff's position have some differences.
i
() 24 Procedures for restoring AC power following j 25 station blackout, which is consistent with the guideline, we ACE. FEDERAL REPORTERS, INC, 202 347-3700 Nationwide Coserage 800-336-6M6
30853.0 cox 153 1 also have procedures for those events.
2 Utilities should have procedures to prepare a 3 plant for severe weather, in terms of shutting down in 4 advance of hurricanes, storms that might be likely, causes of 5 long duration of off-site power. We think that's a good 6 initiative.
7 Diesel generators, area which relate to the third 8 and fourth initiatives. Utilities are following up on a 9 generic letter that NRC sent out, which would reduce or 10 eliminate the need for cold fast starts of diesel 11 generators. We believe that the utilities are convinced that 12 those cold fast starts are limited or reduced at most to 13 perhaps once every 18 months.
14 Fourth initiative NUMARC has proposed, which is to l
l 15 monitor the emergency diesel generator unavailability, which l
l 16 we think is an important step. We have that as part of our 17 resolution as well. We also have additional guidance that 18 reliability programs should be in place to assure that if 19 there are problems with a diesel generator, there are 20 procedures in place and that the reliability can be assured 21 and be maintained above certain levels.
22 Status of the NUMARC initiatives, general, they 23 are utility group is preparing specific guidelines to go
() 24 along with these initiatives, and some are in different 25 states in readiness.
ACE FEDERAL REPORTERS, INC.
L 202-347-3700 Nationwide Coserage 800-3346M6
(
_ _ _ _ _ _ _ _ . _ _ _ _ _ _ _ _ _ _ _ _ . _ _ . i
30853.0 cox 154 g~ 1 A conceptual comparison of the NUMARC and the V) 2 slide you have seen earlier, where we had these four steps in 3 the proposed rule. NUMARC initiatives, they have the 4 determination of what groups plants fall into, either four-5 or eight-hour groups. They say plants should have procedures ,
6 for coping with station blackout. The middle steps, which is 7 determining that there is insurance that the capability 8 really exists, is not currently part of the NUMARC 9 initiatives. Let me clarify this one point. With the 10 alternate AC approach that we discussed earlier, these steps 11 could likely be bypassed. So if you have a qualified 12 available alternate AC power approach, you would not
( 13 necessarily have a limitation on your coping capability.
l 14 MR. EBERSOLE: What is the significance of that 15 line around there? Is it supposed to be in a fog or cloud or 16 whatever? ,
17 MR. RUBIN: These are not part of the current 18 NUMARC initiatives.
l 19 MR. MICHELSON: That's the hard part, Jesse.
20 MR. RUBIN: That's the hard part. We feel in 21 order to develop procedures, you really need to do some kind 22 of an assessment with respect to where we stand right now.
23 In conclusion, we have raised some questions for
() 24 you which are considerations relating to whether or not there 25 should be a rule to resolve this issue. First question, ACE-FEDERAL REPORTERS, INC.
202-m-rm NationwMe coserage sm 33sw6
30853.0 cox 155
- 1 first point is that we have identified, to our best 2 understanding, differences between technical positions 3 between NUMARC initiatives and the Staff's resolution.
4 If we were to reach an agreement with industry, 5 there are still some questions that would need to be 6 addressed on whether a rule would be necessary or not. I 7 just put them down here for your own thoughts. One is a 8 question on enforceability, how can the NRC insure that all 9 plants will actually implement the resolution. Second item 10 is that could NRC issue guidance as in a regulatory guide, 11 coping with blackout and guidance in requirements for reactor 12 coolant pump seal integrity, without a rule.
G
- 13 The third point, which is significant, which is 14 that the rule in the regulatory guide provides guidance not 15 just for industry, but for the NRC Staff as well, so that 16 there are specific areas of understanding that the Staff has 17 in terms of reviewing these requirements.
18 That concludes my various parts of the 19 presentation.
20 MR. REED: Supposing, just suppose now, that you 21 go ahead and you get NUMARC coordinated and all these things, 22 you get out, you are doing modifications, you are down the 23 trail about a year from now, with lots of money having been
() 24 spent to do this, what I will call piecemeal activity, which 25 Jesse Ebersole calls patching.
ACE. FEDERAL REPORTERS, INC.
202 347 3700 Nationwide Coserage Mo-336-M46
30853.0 cox 156 1 Suppose, at the end of this year of activity and 2 expenditures and engineering and work, the A-45 comes to 3 fruition, and the A-45 says all right, the only way out, 4 really,-to provide the protection for these dozen aspects 5 that are involved, like blackout, fire, security, small 6 LOCAs, interaction, steam-generated tube failures, steam 7 generator overfill, seismic, emergency auxiliary feed 8 systems, beyond SSE, severe accident; the only real way to to 9 cope with all of those on a core melt frequency, the number 10 that somebody wants to have, is to add a backup, particular 11 type of system for a PWR. In some cases, some PWRs seem to 12 look like they need it more than others. Suppose, what do 13 you do then, after the utilities have got halfway down the 14 pike. Now you come along and say from A-45, let's do this, i
15 isn't this a rather embarrassing situation, and people have 16 spent money that they shouldn't have?
4 17 MR. EBERSOLE: It's a prolongation of what we 18 already do. We redefine it to death, i
19 MR. RUBIN: I have a couple of comments to make on 20 your points, they are good points. We consider the 21 resolution of A-45 as well as other issues, we focus on this
- 22 one. First of all, you said what the utility decides l
l 23 voluntarily to put in the separate system. We don't know
() 24 that. They still have that option. There is nothing in this i
25 rule that would preclude a utility from doing that. They l Acn. FEDERAL RneonTEns, INC.
! 202 347 3700 Nationwide Coverage 800-336-6M6
30853.0 cox 157 1 would be given all the credit in the world on the station 2 blackout issue if they were to take this approach. That 3 would basically provide them with the capability to cope with 4 the blackout for a long period of time. If the system had 5 its own cooling capability separate from the normal cooling 6 system with its own AC power source.
7 MR. EBERSOLE: When you said a long period of 8 time, you could have said indefinitely.
9 MR. RUBIN: We don't see that will be the case.
10 We mentioned earlier, if NRC were to require that kind of a 11 system for all plants, it would have to be justified on a 12 cost-benefit value impact analysis.
O 13 MR. REED: I know how utilities will react. They 14 will react to this rule and its time frame. Out comes this 15 rule, and utilities say, okay, I will go around and patch 16 here, patch here, patch here, now I meet this particular new 17 rule. Next month, next year, we, the regulatory, will come 18 along with another need, be it Michelson's interactions or 19 upgrading on fire protection or whatnot. Rather than 20 leapfrogging to the real issue, the hard issue, that is, 21 decay heat removal on a more assured basis, we keep going 22 around from issue to issue. Shouldn't we be thinking further 23 out on the decay heat removal, better assurance, which
() 24 reduces a lot of other things, like containment vent 25 considerations perhaps, and that's what you wi.1 be weaving ACE FEDERAL RevonTEns, INC.
202-347-3700 Nationwide Coserage M4336-6M6
30853.0 cox 158 1 into this fabric in another year or two.
2 MR. RUBIN: I think Dr. Speis wanted to respond.
3 MR. SPEIS: I appreciate what you are saying, I 4 don't want to characterize them as Utopia, but we have 5 thought carefully about all of these issues, and also we have 6 thought carefully, or more importantly we are living with the 7 constraints that exist. We have rules, we have backfit 8 rules, we have guidance for the backfit rules. We have a 9 safety goal, even though it's not some elements that are 10 finally qualified.
11 When we take all of these things'into 12 consideration, I can tell you that the A-45 issue that we O
k# 13 concerned concerns you and concerns some of us, I don't think 14 you will have on a value impact or the other criteria that if 15 you mentioned, would be justified in doing any big fixes of 16 the types that you are thinking about. For example, 17 dedicated systems, either hot shutdown or cold shutdown, the 18 only thing that might be justified to be able to do, based on 19 these considerations that we are talking about, are fine 20 tuning of certain things.
21 Therefore, I don't think there would be any 22 overlap or any reason to delay A-44 right now. I think it is 23 an important issue. If we wait to arrive at the perfect way
() 24 to integrate everything, that would probably take forever.
25 But, again, you know, we know what A-45 is looking ace FEDERAt. REPORTERS, INC.
202 m .3700 Nationwide Cmerage M m 31 W 46
30853.0 cox 159 1 like right now. We want to come talk to you the next few 2 months. We are fine tuning the regulatory analysis, we have 3 evaluated a number of options, five or six of them. Some of 4 them do include the things that you are talking about. We 5 don't see anything down the horizon, unless somebiay at the 6 Commission level or level of the United States says that 7 because of some other factors that are not part of our 8 infrastructure right now, are going to impose a $200 million 9 or $50 million heat decay removal system.
10 MR. REED: Charlie opened up this session by 11 talking about an ACRS letter that talked about you have got 12 to leave in the A-45 thing. You should look at the foreign 13 trends. I think if you look at the foreign trends, and that 14 the more astute of the foreign trends, you will find that 15 have dedicated decay heat removal, as you call it, is coming 16 into the picture. You know, our nation used to be the leader 17 in nuclear figures generation and nuclear power reactors. It 18 seems to me what we have done is we have fallen out of 19 leadership position from the early days, creation of systems 20 and so on, and we have let the systems be sold to France and 21 sold to other places. Then they are seeing where they should 22 leapfrog to better assurance of core melt against core melt.
23 We are not doing that. We are going around and
() 24 patching the designs and concepts that we had 20, 30 years 25 ago, rather than, say, hey, let's move out conceptually and Ace FEonRAL REPORTERS, INC.
4 :02.mmm Nationwide cmerage smaw6u6
- , -----. - ... - __ _-- . - . . . . . ~
. ~ - . - _ - _- - - - - _ - .- - - - - . _ - . .
30853.0 cox 160 1 do something like this, and then forget about a lot of the 2 surveillance, monitoring, quality assurance, inspection and 3 so on and so forth, all through the woven complex systems 4 that you find out there, auxiliary feed, cables, control 5 rooms, access and security and emergency operation.
6 I disagree with the trends, and I understand what 7 you are saying. You are going with, I guess what you might 8 call the art of the possible. You analyze a piece of the 9 pie, and then you come through in a very good detailed 10 analysis, you come through with saying, okay, now we are 11 going to push, nudge in this direction. I guess perhaps 12 that's possible. The utility people, having been a utility 13 person, will see it as their answer, okay, we will take this 14 rule and take this rule, along will come another one, okay, 15 we will move on this one. We keep building this topsy thing, 16 rather than moving to another concept.
17 You don't have to defend what you are doing.
18 MR. SPEIS: I am not defending. As I say, we have 19 constraints and we have our perceptions of what is acceptable 20 as far as risk is concerned. The Commission has spoken to 21 that. Some of us, I work for the Commission, maybe you guys 22 are more independent than I am.
23 MR. REED: I work for core melt protection.
() 24 That's all I work for.
25 MR. SPEIS: I think we all work for that. I think ace FEDERAL REPORTERS. INC.
202 347-3700 Nationwide Cmcrage 80k33MM6
l 30853.0 cox 161 l
1 it's fair to say that the issues that we have under 2 consideration now, we have looked at them very carefully for 3 duplication. We have to be realistic. As I said, we don't 4 see down the pike this super system that is going to be put 5 together and be able to sell it to be able to resolve all the 6 issues you are talking about.
7 MR. REED: Let me go way out in philosophy. I 8 said I worked for a nuclear reactor core melt protection, 9 that basic one. Basic two, I work to try to reduce the 10 number of burning hydrocarbon fires, because I think that we (
11 are beginning to see the ultimate consequences of too many 12 burning hydrocarbon fires in carbon dioxide increase in the O 13 atmosphere, in weather extremes that we are getting lots of.
14 In Wisconsin we had summer all winter this winter. In New 15 England they had winter like they had never seen before. All 16 kinds of extremes.
17 So how in the hell, as we look way out there, are 18 we going to get the public to understand that they ought to 19 be reducing hydrocarbon fires and inducing fission fires. I 20 don't think we are going to do it with patching.
21 End of speech, thank you.
22 MR. EBERSOLE: I would just like to say that I am 23 sympathetic to Glenn's big pitch here about integration. I
() 24 am fearful, Tom, if we now look at the entire field of our 25 patch work, that that will deny us the privilege of what we ACE-FEDERAL REPORTERS, INC.
! 202 347-37(O Nationwide Coserage 800-336-6646
30853.0 cox 162 1 should be doing right in the first place. It's sort of an b
ss 2 organizational problem. You get orders to work on a 3 compartmentalized basis, and that's the bounds of your 4 activity. So you do the best you can. You are at cc power.
5 In some other case it's fire, some other case it's small 6 LOCAs. I wonder if it doesn't tend to step back and look at 7 the cost of these compartmentalized efforts to put in patches 8 in a totalitarian sense, rather than just one piece at a 9 time, as you look at it here, against the cost of going home 10 and fixing the route problem, as Glenn described it, as the 11 A-45 should. ,
12 If we look at the A-45 against the background of O 13 all the patches we have put in, and include the interval cost 14 of all of them, not just that one, I think we will easily see 15 a picture that we started down the road long ago, and all we.
16 are doing is perpetuating it.
17 MR. SPEIS: I want to add two things to that.
18 First of all, Dirks stopped us from proceeding with that 19 issue about four or five months, and gave the industry a 20 chance to go out and fix it up, and they didn't come.
21 Number two, those two issues are phased together, 22 44 and 45 will be coming due. We have a number of 23 recommendations. It's possible, somebody else, maybe you
() 24 guys will be more effective than us and overcome our 25 constraints we have to work with.
ace FEDERAL REPORTERS, INC.
2024 47 3700 Nationwide Cmerage Ms336-ud6
30853.0 cox 163 73 1 MR. EBERSOLE: We are just a little bit out of
(_)
2 phase here.
3 MR. SPEIS: I don't think we are not.
4 MR. EBERSOLE: 45 is close enough --
5 MR. SPEIS: They are very close. There is no 6 question we are out of phase.
7 MR. EBERSOLE: Even if 45 comes out negative, I 8 certainly could hope it comes out in recognition of the 9 innumerable patches and the cost of each individual one, and 10 we now must sustain through all time, through 50 years of 11 plant life, supporting the quality levels of all these 12 peripheral and interdependent systems. I guess the O
\'
13 background of not having to do that.
14 MR. SPEIS: I shouldn't be getting into a 15 philosophical discussion, but I don't think if all of us had 16 the ultimate system that will assure us will never challenge 17 the public by an activity, that we will be happy, we will be 18 trying to make sure that the course don't make, that the 19 challenges to engineering safety features are reduced. You 20 always have to make a judgment how far are you going in one 21 direction versus the other, and cost effectiveness, it 22 shouldn't be a part of this somehow.
23 Again, we are working for the present plan. I i
() 24 think the speeches Dr. Reed made are too future for relevant 25 plans.
Acn FEDERAL RunonTnns, INC.
202 347-3700 Nationwide Coserage N n31MM6
30853.0 cox 164 gg 1 MR. REED: I resist being called Doctor, you may G'
2 call me activist, Mr. , mechanic --
3 MR. MICHELSON: Or philosopher.
4 MR. SPEISs I apologize for making that error. I 5 won't make it again.
6 MR. EBERSOLE: It's working in reverse here.
7 MR. WYLIE: Does that complete your presentation?
8 MR. SPEIS: Yes.
9 MR. MICHELSON: We have a term, 4 on page 11 of 10 the regulatory guide. I will tell you why I have a problem 11 with that item. This item addresses the fact that we may add 12 some extra components to take care of power blackout, for O
\2 13 whatever reason. I assume here following a seismic event is 14 the reason.
15 Now, what they do in item 4 is caution you that if 16 these added features in any way cause a loss of a 17 safety-related equipment, you have got to do something. Of 18 course, that's not my problem at all. My problem is the 19 added features that I have got to have to handle the blackout 20 until I can get back into operation in X number of hours are, 21 in themselves, not seismically qualified. In fact, it 22 addresses the fact that they may even fall over here. My 23 concern is if they fall over, they can't handle the power
() 24 blackout problem. How do you address that in your PRAs or
- 25. rationale or whatever?
/\CEJFEDERAL REPORTERS, INC.
202-347-3700 Nationwide Coserage 800-336-6M6
30853.0 cox 165 1 MR. BARANOWSKY: We are not addressing it in the 2 PRAs, but I have to admit that that consideration was not 3 part of the guidance we had in the regulatory guide.
I 4 Failures that are not expected during design basis 5 seismic events, and their impact on emergency AC power, has 6 not been included in guidance that we have nor in our I
7 analyses. l 1
8 MR. MICHELSON: I can't say, of course, whether I i 9 would expect those failures or not, because it's a l 10 nonseismically qualified equipment. I don't know how it's l 11 being supported and so forth.
12 There is no requirement to even support them for i s 13 the seismic event in the regulatory guide. It's a nonseismic l 14 design. I shouldn't say unknown seismic design. It's known 15 quality. We address words about we want good equipment 16 because we want high reliability.
i l 17 But we remain totally silent on the seismic 18 question. In fact, here we do the opposite; we point out if 19 you do have to add equipment, make sure it doesn't jeopardize 20 equipment that we need that performs safety-related 21 functions.
22 MR. BARANOWSKY: Let me point out that these are 23 things that could be corrected if necessary in the future.
() 24 MR. MICHELSON: Then the regulatory guidance --
25 what is your guidance to begin with. What are you trying to ACE-FEDERAL REPORTERS, INC.
202-347-3700 Nationwide Cowrage 800-336-6646
30853.0 cox 166 1 tell the utility to do here.
2 MR. BARANOWSKY: In that area, there's not any 3 guidance on it or any proposed requirements. What I am 4 saying is if it turns out that that could be shown to be an 5 important area, or even after rethinking about it, that it 6 was important, it could be added to, it could be done in tune 7 with the rulemaking or it could be done subsequent.
8 MR. REED: What he is saying is this is a 9 nonseismic patch. It's really only half a patch. The 10 Regulatory Commission has done these kinds of things before.
11 It says okay, do this to your control room. You don't have 12 to have this seismic or this or that. But a year or two O 13 later, along will come the next thinker, the next idea, and 14 the next beginning to a rule.
15 You know, I used to hate to go through Logan 16 Airport 30 or 40 years ago or 20 years ago, because every
, 17 time I went to Logan Airport it was being ripped apart, 18 reconstructed. I could never figure out where the hell I was 19 going. I might be going there every month.
l l 20 It seems to me that this is the way of the 21 government, the way of the regulatory, the way of the
! 22 bureaucracy, rather then to think out and you do, you think 23 for the day and patch. I am sorry, I just get all steamed up l
() 24 about this.
25 MR. MICHELSON: This particular one ought to be ACE FEDERAL REPORTERS, INC.
l 202 347-3700 Nationwide Coverage 800-336-6646
! 30853.0 cox 167 1 thought through, because we are conceding that there is a ,
2 front of approaching unity that if we do get an SSE, we will
' 3 lose our off-site power. Now we are dependent totally upon i
4 the probabilities that the diesels under these circumstances 5 will start a function. Now we are comfortable if we don't 6 have a problem on the diesels, I guess you could argue we 7 don't need this. It will never happen. An earthquake will 8 never come -- pardon me. We will never get a total loss of 9 AC power from an earthquake. That's what we are saying.
i 10 MR. BARANOWSKY: That's not really what we are i
11 saying. In fact, let me clarify something.
l 12 MR. MICHELSON: That's what we are designing for.
! (:) 13 MR. BARANOWSKY: We have looked on this thing.
14 Based on operating experience, we have tried to address the 4
15 issue that is relevant to what we have observed and not
- j. 16 address every hypothetical possibility. That as these 17 hypothetical possibilities are identified, we will take a 18 look at them and determine whether or not they are 19 sufficiently likely that we should take additional action.
l 20 MR. MICHELSON: I think we ought to be careful 21 then not to talk about two or three SSE, when we don't even l 22 know what SSE will do to us, except in the realm of normal m
23 response, not earthquake effects.
(
l
() 24 MR. BARAN0WSKY: I will agree there's enough i
25 uncertainty in these analyses of effects and likelihood, once ACE. FEDERAL REPORTERS, INC.
202-347-3700 Nationwide Coserage 800-336-6 4 6
. . - - _ _ = . . - - . . . _ _ _ _ . , _ - - - , . . _ _ . - . - - , . - - - - - . _ - , , , - , _ . .
30853.0 f cox 168 f
1 you take all of these things into consideration carefully.
(
2 MR. MICHELSON: Is there big money involved in 3 doing a seismic-type qualification of this equipment if it is l
4 needed? Would that be a big imposition on the part of the 5 industry? It would sure remove some objections that I have, 6 at least. I don't know if it's big bucks we are talking 7 about.
8 MR. BARANOWSKY: I can't answer that question.
9 MR. RUBIN: Let me just comment, the utility chose 10 the route to go separate AC, it could be big bucks.
11 MR. MICHELSON: The big bucks we are talking about 12 is the addition of the seismic features.
'T (V 13 MR. RUBIN: That's right.
14 MR. EBERSOLE: I want to go back and reinforce 15 something that I mentioned before. My impression now is that 16 the boiler owners have ways of coping with containment over 17 pressure, this could be one of the causes of that, prior to i 18 core damage, by containment opening and venting, and pursuing 19 the cooling process by depressurizing and furnish water from 20 the nearest creek if they have to, from whatever source they 21 need. And that the ultimate cooling process is no more than 22 cooling pins in an open bucket of water. These are 23 undamaged. It's not close core damage I am talking about.
() 24 It's in the field of predamage that you are looking for 25 here. Not to invoke that and be open about it, is whether or ACE FEDERAL REPORTERS, INC.
202-347 3700 Nationwide Coserage 800-336-6M6
- - - - . _ - ~ - ~ - - - -
30853.0 cox 169 1 not it's possible, practical or whatever, I think is sort of 2 clouding over. Whether that isn't a practical recourse 3 approaching that, of adequacy of an integrated process, but 4 not quite that good. I don't see that being brought to the f
l 5 circus and chewed up as well as I would rather have it be.
6 MR. BARANOWSKY: It's probably not been made as 7 visible as you would like to see it, but I can assure you 8 that any reasonably effective way of cooling the core, under 9 the circumstances that we are talking about, would be ,
10 acceptable, and if it's possible to run creek water in there 11 and keep the core cool, fine.
12 MR. EBERSOLE: I want to bring up the reason, I O 13 suspect myself, for lack of visibility. It's uniquely 14 adaptable to one of the two kinds of plants you have. You 15 almost avoid any discretionary approval of any' kind.
16 MR. REED: You mean it's uniquely adaptable to the 17 PWR.
18 MR. EBERSOLE: You are talking about the blowdown 19 system.
20 MR. MICHELSON: He is talking about a different 21 system.
22 MR. WYLIE: Is that just an observation?
23 MR. EBERSOLE: Whatever it is. I don't know what .
() 24 it is. But there is that practice. Avoiding being 25 discretionary.
ACE FEDERAL REPORTERS, INC.
202-347-3X0 Nationwide Coverage 800-336-6M6
30853.0 cox 170 s 1 MR. WYLIE: The subcommittee have requests of the
_]
2 Staff? I believe NUMARC has a statement they would like to 3 make.
4 MR. MC GARRY: My name is Mike McGarry. I am a 5 consultant to NUMARC. With me is Steve Maloney. He is also 6 a consultant to NUMARC. We are also consultants to NUGSBO.
7 MR. MICHELSON: Excuse me, what is NUGSBO?
8 MR. MC GARRY: Nuclear Utility Group on Station 9 Blackout. It preceded NUMARC. NUMARC has asked NUGSBO to be 10 the consultant to it. In the room at this time are members 11 of the NUGSBO/NUMARC effort.
12 In February of 1985 NUGSBO came to the ACRS and O 13 made a presentation indicating it was considering 14 alternatives to rulemaking. In March of 1985, the ACRS, as 15 Mr. Wiley pointed out, wrote a letter to the Commission and 16 indicated, among other things, that if there was an 17 alternative to rulemaking, that it should be given serious 18 consideration. NUGSBO continued to proceed with 19 consideration of alternatives, and in June of 1986, NUMARC 20 decided to take on the station blackout issue. Actually, it 21 was in September of 1985, when it made that decision.
22 This June of 1986, NUMARC commented on the 23 proposed rule, and in addition advanced four initiatives,
() 24 Alan Rubin has mentioned those initiatives today. Briefly, 25 they are the reduction of risk, is the first initiative.
ACE FEDERAL REPORTERS, INC.
202-347-3700 Nationwide Coserage 800-336-6646
. - -. . . _ _ ~ . .. - .- _.
30853.0 cox 171 i 1 Second initiative is procedures dealing with station
- 2 blackout, procedures, power restoration procedures and severe 3 weather procedures.
U 4 Third initiative concerns itself with the i
5 reduction and the cold fast start of the emergency diesel i 6 generators, and a fourth initiative concerns itself with 7 monitoring the availability of diesel generators. Work is 8 ongoing at this time with respect to those four initiatives.
9 Preliminary results, for example, indicate, based j
10 on information that accompanied the proposed rule, that a 11 little more than a dozen plants would' fall in an eight hour 12 category, which affirms a NUMARC position that this is not a
~
(E) 13 generic issue deserving of rulemaking. NUMARC is not in a 1 ,
14 position today to provide the specifics, the details of its 15 initiatives. As I indicated, work is ongoing. It's for that i 16 reason that NUMARC chose not to make a presentation to the 17 subcommittee. Rather, on Friday, NUMARC will appear before 18 the full committee and give a status report as to where it is i
19 on each one of those initiatives.
20 We anticipate that work will be completed on these 21 initiatives in the late summer, early fall of this year. We 22 would then be in a position to make a presentation, if 23 appropriate, to this subcommittee, concerning the details of i
() 24 the four NUMARC initiatives.
l At the very end, 25 I did want to clarify one point.
/\CE FEDERAL REPORTERS, INC.
202-347-3700 Nationwide Coverage 800-336-6646
30853.0 cox 172 1 we have talking about the integration of a variety of issues ,
k_)s 2 with the station blackout issue. The statement was made that l 3 the utilities were given the opportunity to pursue this and 4 did not. That's incorrect. Utilities provided a fairly )
5 exhaustive integration plan to the Staff that died at the 6 Staff. We have heard nothing further. I don't mean to be 7 entering into a controversy here, but I want the record to be 8 corrected on that point.
9 MR. REED: Could you clarify what you just said a 10 bit more? What was this integrated approach?
11 MR. MC GARRY: We look at a variety of power 12 issues, including decay heat removal. What we attempted to O 13 do was integrate the resolution of those issues into a 14 unified plan, so that just as you are suggesting, Mr. Reed, 15 if we were to make modifications directed to a station 16 blackout issue, and two years or three years down the pipe, 17 the decay heat removal issue is resolved, we might find 18 ourselves in a position that the fix that was made for 19 station blackout is inadequate to address the decay heat 20 removal issue. It seemed to us, let's come up with an 21 integrated plan that would bring the resolutions into a 22 logical sequence.
23 MR. REED: I must still have my utility thinking
() 24 cap and don't even communicate with these people. So it died 25 with the Staff.
ACE FEDERAL REPORTERS, INC.
202-347 3700 Nationwide Coverage 800-336-6M6
30853.0 cox 173 1 MR. MICHELSON: They are going to tell us in a b O T. SPEIS, RES A. RUBIN, RES l
P. BARAN0WSKY, NRR J. FLACK, RES MAY 6, 1987
[
~
O
OllTLINE OF PRESENTATION O
o INTRODUCTION (T. SrEIS)
APPROACH FOR PESOLIIT!0F 0F ISSUE ELEMENTS OF FINAL RESOLUTION MILESTONES AND STATl'S SINCE LAST ACRS MEETING o TECHNICAL FINDINGS (NUREG-1032) (P. BARAN0WSKY/J. FLACK) 0FFSITE POWER EMERGENCY ONSITE AC POWER PROBABILISTIC ANALYSES o FINAL RESOLilTION (A. RUBIN)
O -
RULE (950.63)
REGULATORY GilIDE VALUE/ IMPACT ANALYSIS (NUREG-1109) c SIGNIFICANT CHANGES TO PROPOSED TECHNICAL RESOLUTION (A. RUBIP)
PUBLIC COMMENTS RESPONSE TO COMMENTS AND REVISIONS TO RESOLUTION o INDllSTRY ACTIVITIES (A. Rl!EIN)
NUMARC INITIATIVES COMPARISON OF NUMARC INITIATIVES TO FINAL
{)
USI A-44 RESOLUTION 1
O INTRODUCTION o ACRS LETTER (MARCH 1985)
SUPPORTED STAFF'S PROPOSED TECHNICAL RESOLUTION CONSIDER ALTERNATIVE TO RULEMAKING CONSIDER EARTHOUAKES BEYOND SSE o APPROACH TO FINAL RESOLUTION IS SAME AS PROPOSED RESOLUTION REPUCE FREQUENCY OF OCCURRENCE OF CORE DAMAGE FROM SB0 o PRESEf!TATION FOCUSES ON s
l CHANGES FROM PROPOSED RESOLUTION SlfPPORTING TECHNICAL ANALYSIS l -
INDUSTRY INITIATIVES l
lO 2
. _ . _ _ _ _ _ _ _ . _ _ - _ _ _ _ . - ~ - - _ , - _ - . . - - _ . - - - - _
APPROACH FOR RESOLllTION OF ISSUE O
o DETERMINE CURPFFT ESTIMATED FREQUENCY OF CDF DUE TO STATION BLACK 0UT FOR A SPECTRUM 0F PLANT DESIGPS o IDENTIFY DOMINANT FACTORS AFFECTING CDF AND COST EFFECTIVE IMPROVEMENTS LIKELIHOOD OF FREQUENCY AND DURATION OF LOSS OF AC POPER ABILITY TO COPE WITH EXTENDED LOSS OF AC POWER o PP0 POSE NEW OR REVISED REQUIREMENTS CONSISTENT WITH LEVEL 0F RISK, AND COST EFFECTTVENESS O
3
CONSIDERATIONS FOR RESOLVING USI A-44 O
o OPERATING EXPERIENCE ,
LIMITS ON RELIABILITY OF 0FFSITE AND ONSITE AC PGWER SYSTEMS i -
AC POWER RELIAPILITY VARIES SIGNIFICANTLY FROP PLANT TO PLANT c POTEtlTIALLY SEVERF. CONSEQUENCES LIPITED DHR CAPABILITY W/0 AC POWER NO CONTAINMENT HEAT REMOVAL V/0 AC POWER 1
()
o PRESEFTLY NO REQUIPEMENT FOR PLANTS TO COPE WITH STATION BLACK 0UT o CDF REPUCTION ALL PLANTS SPOULD BE ABLE TO COPE WITH A STATION I BLACK 0UT FOR A SPECIFIED DURATION USE GUIDELINE TO DETERMINF DURATION THAT KEEPS AVEPAGE COPE DAMAGE FREQUENCY FROP STATION BLACK 0UT A SMALL CONTRIBUTOR TO TOTAL CDF (i.E.,cc 10-5/RY) o BACKFIT RULE (CONSIDER COST EFFECTIVENESS OF RESO.LUTION) 7_
v l
4
SUPMARY OF STAFF FINDINGS O
RELIABILITY OF ONSITE EMERGENCY AC POWER SYSTEMS VARIES CONSIDERAELY CONFIGURATION (REDUPDANCY)
EDG RELIABILITY FRE0UENCY AND DURATION OF 0FFSITE POWER LOSS VARY CONSIDERABLY
- SITE CHAPACTERISTICS (WEATHER, GPID)
- PLANT FACTORS (SWITCHYARD DESIGP, TRANSMISSION LINES)
EXTENDED DURATION STATION BLACK 0UTS (> 2 HOURS) CAN BE SIGNIFICANT CONTPIBUTORS TO RISK
(])
CORE DAMAGE FREQUENCY VARIES CONSIDERABLY FROM PLANT TO PLANT SUSCEPTIBILITY TO STATION BLACK 0UT ABILITY TO COPE WITH LOSS OF ALL AC POWER N0 SINGLE FIX APPLIED UNIFORMLY TO ALL PLANTS WILL RESOLVE THIS ISSUE IN A COST-EFFECTIVE MANNER
O ELEMENTS OF FINAL RESOLUTION
~
r o FTATION BLACK 0UT RULE REQUIRES ALL PLANTS TO BE ABLE TO COPE WITH STATION BLACKOUT FCF SPECIFIEn PURATION o REGULATORY GUIDE ON STATION BLACK 0UT IN SUPPORT OF RULE
() -
PROVIDES GUIDANCE FOR ALL PLANTS TO COMPLY WITH RULE ,
GUIDANCE ON ACCEPTABLE PLANT-SPECIFIC COPING DURATIONS GUIDANCE ON USE OF ALTERNATE POWER SOURCES l
i PROCEDUPES AND TRAINING TO COPE WITH STATION l
BLACK 0UT 1
i MIN!PUM EDG RELIABILITY
~
()
l G
MILESTONES AND STATilS OF USI A-44 O MAR 1985 ACPS MEETING ON PROPOSED RFSOLUTION MAY 1985 PROPOSED RULE TO COMMISSION (SECY-85-163)
SEPT 1985 STAFF'S EVALUATION OF NUGSB0 PROPOSAL (SECY-85-]63A)
JAN 1986 BACKFIT ANALYSIS (SECY-86-28)
MAR 1986 COMMISSION V0TES TO ISSUE PROPOSED RULE MAR / APP 1986 PROPOSED RULE / DRAFT REG GUIDE ISSUED FOR C0HMENT O
JUNE 1986 END OF PUBLIC COMMENT PERIOD (N!1 MARC INITIATIVES SUBMITTED)
FEB 1987 FINAL RESOLUTION F0P DIVISION REVIEW APR 1987 4 FINAL RESOLUTION TO ACRS/CRGR i SCHEDULE JUNE 1987 ACRS/CRGR REVIEW COMPLETED
,AUG 1987 FINAL RULE TO COMMISSION
)
l
i O O O .
l
' ELEMENTS OF STATION BLACK 0UT PROGRAM i
OPERATING EXPERIENCE RELEVANT TO STATION BLACK 00T I
l l LOSS OF 0FFSITE POWER PROPOSED
> ESTIMATED ACCIDENT SEQUENCE l FREQUENCY AND DURATION y TECil.NICAL STATION CHARACTERISTICS RESOLUTION BLACK 0UT > AND CORE MELT FREQUENCY . FREQUENCY AND (DOMINANT
~
EMERGENCY AC POWER
> DURATION ACCIDsNT SEQUENCES)
RELIABILITY ..
4 l
4 e
e 4 -$f %
i
SUMMARY
OF OPERATING EXPERIENCE RELEVANT TO STATION BLACK 0UT FROM 1968 - 1985: 64 TOTAL LOSS OF 0FFSITE POWER EVENTS OF A FEW MINUTES DURATION UP TO 26 HOURS
- FROM 1976 - 1985: OVER 600 EMERGENCY DIESEL GENERATOR FAILURES DURING TESTING AND ACTUAL DEMANDS; 22 INSTANCES INVOLVED MULTIPLE FAILURES DUE TO COMMON CAUSE, FROP 1968 - 1985: SEVERAL STATION BLACK 0UT PRECURSORS INVOLVING
(]) TOTAL LOSS OF 0FFSITE POWER AND UNAVAILAPILITY OR FAILURE OF ONE OR MORE EMERGENCY AC POWER SUPPLIES, SEVERAL OF THESE INCIDENTS INVOLVED LOSS OF ALL AC POWEP. - STATION BLACK 0UTS,
- LICENSEE STATES THAT POWER COULD HAVE BEEN RESTORED IN 5+ HOUPS.
IF NECESSARY.
~
9
O PP.ECURSORS TO STATION BLACK 0UT
~
1968 SWITCHING ERROR CAUSED LOSS OF 0FFSITE POWER (PLANT AT 100%' POWER); DIESELS STARTED AND LOADED BUT TRIPPED OFF (4 MIN. W/0 AC; 25 MIN.
TO PESTORE OFFSITE POWER).
1976 UNDERVOLTAGE TRIP CAUSED LOSS OF 0FFSITE POWER (PLANTAT100% POWER);DI$SELSSTARTEDBUT UNDERVOLTAGE CAUSED LOADS TO TRIP FROM SAFETY BUS (5 MIN. W/0 AC).
983 SEVERE WEATHER (SNOWSTORM) CAUSED LOSS OF 0FFSITE O .
POWER (PLANTSHUTDOWN);ONEDIESELOUTFORMAINTENANCE SECOND DIESEL WAS BEING PARALLELED; LOAD-SHEDDING RELAY FAILURE RESULTED IN LOSS OF ALL AC POWER FOR 25 MIN.
1984 DURING LOSS OF 0FFSITE POWER TEST, OPERATORS MISTAKENLY SWITCH OFF DC CONTROL POWER WHICH PREVENTED EMERGENCY AC POWER SUPPLY OPERABILITY OR RETURN OF NORMAL POWER. ALL AC POWER UNAVAILABLE FOR 10 MINUTES.
. l0
_ _ . _ _ _ , _ _ - - _ _ _ - = - _ _ _ - _ - - _ _ _ _ , . . . . _ . _ . . _ , . . _ - - _ . . - . - - . _ _ _ - _ - - - -
O Table 1.1 Summary of station blackout program technical results Parameter Value Operational Experience .
Loss of offsite power (occurrence per year)
Average 0.1 Range O to 0.4 Time to restore offsite power (hours)
Median 0.6 90% restored 3.0 \
Emergency diesel generator reliability (per demand)
Average 0.98 Range 0.9 to 1.0 Median emergency diesel generator repair 8 time (hours)
Analytical Results Estimated range of unavailability of 10 4 to 10 2 emergency AC power systems (per demand)
Estimated range of frequency of 10 5 - 10 8 station blackout (per year)
Estimated range of frequency of core damage 10 8 - 10 4 as a result of station blackout (per year) e
~
O 11
O REVISIONS TO ANALYSFS SINCE MARCH 1985 LOSS-OF-OFFSITE-POWER ANALYSES UPDATED AND INCORPORATED LOP EVENTS THRU 1985 REVISED PLANT DESIGN GROUPS AND DATA CLASSIFICATIONS IMPROVED WEATHER HAZARD DATA
- ADDED SALT SPPAY WEATHEP.IIAZARD FOR C0ASTAL SITES EMERGENCY AC POWER RELIABILITY UPDATED EDG FAILURE DATA THRU 1985 PEVISED CCF DATA AND CCF ANALYSES AD ED 2-0F-4 CONFIGURATION O
STATIONBLACK0UTANDCOREDAMAGEFREQUENCk USE UPDATED LOP AND EAC ANALYSES l
l l
1 1 -
l x
iz l
l LOSS OF 0FFSITE POWER Q
PAST OPERATING EXPERIENCE 1
ANALYSIS OF CAUSE, FREQUENCY, DURATION d
l RELATIONSHIP 0F DESIGN AND LOCATION TO FREQUENCY AND DURATION
. O e
l l
J J
O j l
LOSS OF OFF ' SITE POWER -
ANNUAL FREQUENCY l LEGEND nu wa ut AvenAsc ALL SITES 2s__ Tuvcuou 3 YEAR AvERA6E I
I N i
l
. \ l 20--
\ l o
~ f \
X l \ l
(
1s-- ,
\
5 h /
\ \
T
\
m s
,e_
ji .
I \
/ \ / '
N /
s-
{ \ / )
1 O . , . . , l l . l l l l . l .
1966 1968 1970 1972 1974 1976 1978 1980 1982 1984 1967 1969 1971 1973 1975 1977 1979 1981 1983 1985 YEAR l 8
O -
Table A.1 Summary of loss-of-offsite power experience No. of events Frequency per Category (),% hr) critical year *
(>_\hr)
Plant-centered 46 0.87 .
(15) (.028)
Grid 12 .014 .
(7) (.008)
O weather s
-(6)
.007
(.007)
Total 64 .108 (28) (.043)
Number of reactor-critical site years through December 1985 is 527, the number of site years is 664.
- Based on reactor-critical site years for plant-centered events; 0.8 x frequency based on site years for grid and weather events.
~
O IS
l l
i O
l Total i
Data 0 05 -
. Plant - O Tote Centered d Plant Centered 3
$ -y Grid j 0.04 ,
j g E Severe Weather .
E E
5 A 2 0.03 -
u O !
O U
x w
$ 0.02 g .
E .
m Grid l
l . 4 m 0.01 -
l O
! Severe g l Weather ,
0.00 +
f -
l 0.1 1.0 10.0 DUR ATION (Hours) l
~
O Figure A.1 Frequency of loss-of-offsite power events exceeding specified durations i
IG
O INSIGHTS REGARDING "PLAh7 CENTERED" LOSS OF 0FFSITE POWER fOST LOSSES OF 0FFSITE POWER DUE TO PLAh7 CENTERED CAUSES o HARDWARE FAILURES o DESIGN DEFICIENCIES o HLFAN ERRORS (PAlhTENANCE AND SWITCHlff3) o LOCAL WEATHER (LIGHTNING) o COMBINATION OF ABOVE DURATION TYPICALLY LESS lliAN SEVERE WEATER AND GRID RELATED O LOSSES FACTORS AFFECTING FREQUENCY AND DURATION RELATED TO o REDUNDANCY AND INDEPENDENCE OF SWITCHfARD DESIGN o NUMBER OF lttEDIATE ACCESS CIRCulTS (AUTOPATIC OR PAN'JALLY SWITCHED) i
~
O
l l O _
e r
t s
o t
e s t e R
l a o c t i
t e s m 0 9 8 i i 2 3 7 t T a
t n s a e
d M n
a r
e wg on pi p
eu ) )
to ir 3 a sg 8 2 4 B f 8 r fn og o r d o i n b es a 2 1 O t re od st
)
2 A
8
(
d B
(
d en n n ra r a a op l o
) )
t l 2 2 r A A A eo (
mf r r i r o o ts o e l l nu n 3 A A al o A ( (
ea i Mv t a
n 3 g i
A s e
e D I 2 3 l I 1 1 b p a u T o r
G _
S O
0 0
' s
- O O O -
I
==- 03 1.0 -
p.0%
ll ,
0.9 -
~
0.8 - -" Al 0 af l 4
, a.
w em t~ 0.1 - ##Y E )4 B "I.3 o
~
l .of o Tl "
?.
l E 0.s E
- n ->4 g I o Z w a
0.4 -
lg g o ,
C r3 i 5
- a 0.s -
i 4
- a .
,- ,o &
E L
- u -
i . .i
- 0.i -
1 0 8 I I I I IIII I 'III ' ' '
O.01 0.1 1.0 . 10 DURATION (hours) l Figure A.2a EstimatedfMuencyofocc~rrenceofplant-centeredlossesofoffsitepowerexceeding u
' specified durations (for offsite power groups as shown in Table A.3)
,i 3
(
($) GRID DESIGN / LOCATION FACTORS l
NONE IDENTIFIED FOR FREQUENCY EXPECTATIONS BASED ON OPERATING EXPERIENCE SITE l
. UTILITY COUNCIL DURATION DEPENDENT ON
([) PROCEDURES AVAILABILITY / CAPABILITY OF
" BLACK START" POWER SOURCES i
" ENHANCED" REC 0VERY:
i
>0.8 AVAILABILITY WITH CAPABILITY TO RESTORE Pil0ER IN sh HOUR 20 t
O 0
1
. N '
, N e
cm nr l
a
\* '
r eo N e w
idN o p
f r noyr N s of e ' e y
Csv t io
\ s t i
r e
v
%mc 0i e 4N f s
9LR '
f o
c
\ o N
R e \ N f l \ o a
m
\ N s e
r 0 N
o ~
\ '
1 s
s g \ ' o
\ l
\ \ '
d
\ ' e
)
\ \
s r
t a
' G \ u o l N ' h e g t r N -
- ,\ ' O I
d i
r O N N
a
, \- '
T A
R U
g r
D o N f
\
4 y N \ i t
l
%O - 1 i
' g dy 0 b a
' er ' b ce o nv ' r ao p hc '
ne n ER ' o i
' t a
r o
' t s
e
_ R 4
A e
1 r
- ~ - - - - - - - 0 u g
- - 0 0 9 8 7 6 5 4 3 2 1 0 i 0 F 1 0 E 0 0 0 0 0 0 mas U_ E u. EiS0EIt po2 u. $s;;$ E' O
T
.l j li l , i,1 'lli:l ll l ,i.! , i'!
l
+
l SEVERE WEATHER DESIGN / LOCATION FACTORS O
~
I HEATHER HAZARDS:
SNOW / ICE TORNADO HURRICANE /HIGH WIND
. SUSCEPTIBILITY AND DURATION DEPENDEllT ON DESIGN AND PROCEDURES
! EXTREME HEATHER CONDITIONS WHICH CAN CAUSE O STRUCTURAL FAILURE REPRESENT POTENTIAL FOR LOW FREQUENCY, VERY LONG OUTAGE (>8 HRS) 1 i
o 22
I l j' '
O 0
0 1
1 a r e
- a w a
o p
e e t
i a
s f
f E o f
a o D
t a %e s l \ e a s e \ s 8 cmr \ o n o l e \
dN 0 d i r e y
r e
f no of e y
r
\
\
\e a a
1 c
u v Csv \ d o it o
\
\ e n -
c e %mc 0i e \ e i
R l
9LR
\
\
- \
\
e
)
s r
u r
e a \ \ o h m g4 a h t r
t a
o g N e N \
\ e O .
w
\ I -
\* T A
e O \
\
a R
U D
r e
v e
g \ s
\ r o
f y
\ a a
0 1 t i
a l i
a b
derY a a b
ce nv o ao a
r p
hc
\' ne e n
ER o x i x
a t
a r
o t
\ a s
e R
7
. 1
- . *- ~ - - - - -
0 A 0 . . s. s. 3 2 t g e e e'
. e 1 . 0 0 r u
g xgew IO -
g $BI b h i F
O oW t
l 1l\l '
l l
O
~
l FACTORS AFFECTING OFFSITE POWER AVAILABILITY DESIGN - REDUNDANCY AND INDEPENDENE OF 0FFSITE F0WER
- CIRCulTS, ALTERNATE POWER SOURCES AVAILABLE OPERATIONS - PROCEDURES FOR RESTORATION AT PLAhT AND COORDINATION Willi DISPATOi SYSTEM 0FFSITE LOCATION - GRID STABILITY, WEATHER
]
'O i
i e
~
O m
O 0.04 0.03 -
pLimerick PRA T ,
3 5
i &
> 0.02 -
U
\
z i
w 3
1 O Model
! Ea C.01 -
I l
l ,
\~
i e i 0.00 1.0 2.0 4.0 8.0 16.0 DURATION fHoural Figure A.13 Estimated frequency of losses of offsite power exceeding specified durations for Limerick
~
O Pr
. i O
i uS l
I
- 0. 'l -
i j 0.3 -
E c
2
)
- y C
5 $.L -
t
- 0 Model 0.1 -
Millstone 3 l PRA M -
0.40 1.0 2.0 4.0 s.0 is.O l
DUR ATION IHours) i l
\
i Figure A.12 Estimated frequency of losses of offsite power t
exceeding specified durations for Millstone 3
~
O OG
O CtVSTER ANALYSIS PURPOSE: REGULATORY ANALYSIS OBJECTIVE: REDUCE LARGE NUMBER OF " CORRELATIONS" RETAIN FREQUENCY / DURATION DIFFERENCES METHOD: ANALYZE SIMILARITY OF DISCRETE FREQUENCY AND DURATION " CORRELATIONS" ;
O I
O n
e W - a n + o e
x o a 4 +
]
O o w '
O f m t --S LL] l :
ZH / l.l !
Ld w j y l __,
DD l lI l '
OA j l l WO e' l l ,f / ..m n
~w
? i T , ;
x k1 s
/ /e ,1/ i
/
..oo 3
O . ;
kk l ll l z m j l J l
~"*
O OZ .? j l JO_
I
/ ,/ i
/ Z o
g F- ,d
/ ? - $ i-l l <C Ld 4 l l/
/ / E T . D Il/l -*Q
^
p(
< D l
- 2O }.s,! , ./
lll
, . ..j pQ w ,./,_ .......... -
u g.,. 9
- wZ 4 d' 6g. . .o I t g g g ,
I1 I I I I
3 5 ' ' I 9
i fIfI f I t raa s i : i i t t i
Ii11 isi ,1 ie it I i
gg
,, 9, 1, , , , ,
g e e ,- , ,.
9 9 (BV3A-311S 83d) rJN300383 031VHilS3 29
EMERGENCY AC POWER RELIABILITY
- . O 1
DESIGN FEATURES i
RELIABILITY DATA .
1 COMMON CAUSE FAILURE ENERGENCY AC POWER SYSTEMS RELIABILITY lO i
e
'O ei
O DIESEL GENERATOR DESIGN CONFIGURATIONS DG Success Criterion for Initiating Events other than Large LOCA
- 1. DGs Dedicated to One Unit A. 1/1 B. 1/2 O ,,
C. 1/3 D. 2/4 II. DGs Shared Between Two Units, A. 1/2 B. 2/3 C. 2/4 D. 2/5 111 . DGs Shared Between Three Units ,
A. 3/8
~
O 30
Table 4.la Diesel generator start attempts and failures for tests and
]
/~
actual demands * (from NUREG/CR-2989)
No. of Auto-auto start Start No. of Fail- start fail-attempt No. of fail- ures per fail- ures per Unavail- Unavail-category demands ures demand ures demand able ability Test 13,665 253 0.019 55 0.004 ---
0.006 Loss of 100 5 0.05 3 0.03 3 0.03 offsite b power **
All 539 14 0.026 5 0.009 3 0.006 emergency demands Failure to run: 2.4 x 10 3/hr***
- Summarizing the responses to diesel generator reliability questionnaires based on 45 nuclear power plants, with 86 diesel generators, for operating years 1976 through 1980.
- Updated from data reported in NUREG/CR-2989.
- 8ased on 314 attempts at scheduled run time of 6 hours6.944444e-5 days <br />0.00167 hours <br />9.920635e-6 weeks <br />2.283e-6 months <br /> or more with 9 failures to run during these attempts.
Table 4.lb Diesel generator start attempts and failures for tests and actual demands (from NSAC 108)
Start Failure attempt No. of No. of per Unavail-category demands failures demand able All '
22,180 260* .012 -
Emergency 424 3 .0071 Loss of offsite 41 1 .024 1 .
power Failure to run: 3.2 x 10 3
- Includes 39 failures identified from LERs and/or categorized as non-failures in NSAC 108.
~
O 3i j
O TYPES OF ACUTAL AND POTENTIAL COMMON CAUSE FAILURES CO.'*,"Oh C AUSE FAILURE GROUP TYPES OF FAILURES DESIGN / HARDWARE MECHANICAL / ELECTRICAL DESIGN INADEOUACY SUBSYSTEMS (FUEL, COOLING, START, ACTUATION)
ENVIRONMENT (NORMAL)
OPERATIONS / MAINTENANCE INADE00 ATE PROCEDURES WRONG PROCEDURE ERRORS OF OMFISSION/ COMMISSION
() DEPENDEhCE/ SUPPORT DC CONTROL POWER'..
SYSTEMS SERVICE WATER COOLING ,
EDG ROOM HVAC ELECTRICAL INTERFACE EXTERNAL FIRE FLOOD SEVERE WEATHER SElSFIC OTHER ENVlFONMENTAL EXTREMES CE) st
i
> ny n> 'q > t ;- ;
c.
N c. N m NG O' i! x > >
nC E X
.O
> > X >
3 E nO
> X O r0
>N>
mc i
i n n nN Cs N v t t nn I C N Nc, N N N N N N N N .N ',
l
- c cs es -- - - c; ce c
- - - -
- _ i U l ! l I
i l l l l :
p H j i I i : I I i g- , , : i l : ,
->- i ,
._J __ - t-JO O_U Q t y l{ ! I t'
,c
< U- /
/ ..9, !./lI l .I (
/ / ,o I . ;
lo I i .n Z Tz UJ U
7
//' i .
l *
'l u l' l
~u l l
l.1 .llI ! M '-
e-f 5
DUO' i .
lii ' l ' i r--
b' I l l l M$
O h- ll p '!
.h,/ ! l .l ,l UU9 j*. li i E'e
.I'tl ! li l ll ll i
$ I /
l #l i .
i s. ,
?
(/) D l 'l l l l. i 'Ct e w u_
g */ /l / /
/ // '
!,j ';
i (i
' l' l i/
o l
I 1
il' ;
Q (f) // ./!'l I i/ / // ' '. !
i T4 0 11 ll l ll Qj b g. /
,/ , //
/ //
_>W I -
// I i E < '
l
, l l . l 'll 'l ,
i -
- e e -
C .
O-O- ,
kJ
- g i O x '
f JJ.n!8HYMNO B3M0d OV A0!GDd3fG 33
- j
] 1 1 Generic Estir. ate
_ Plant-
$pecific everage configuration O ,,
e o 1/2 5 Q 2/3
_ A A 1/3,2/4,2/5.3/5
~
s e
4 g 1c 7 -
3 4
E ~ o 3
E
= -
A
- Da toc
$ eA CONFIG UR ATION E
o.
N 2 cf 3 6
w w
O 8 E .
s -
~
A .-
, 1 of 2 A
l of 3 e 21 B 1, <10~4 A Il s
e e oc I
e.s2 e.M I l e se ai e.se I
.,1.00 IfolVIDUAL EDG RELIABILITY Figure 1. Co carisor, cf emergency AC power syste unreliatility results t,ase on ger,eric estimates and plant. specific ,
syste- level operating exp(rience.
O l
34
O FACTORS AFFECTING ONSITE AC POWER REllABILITY
- ONSITE AC POWER SYSTEM CONFIGURATION
- RELIABILITY OF EMERGENCY DIESEL GENERATORS
- VULNERABILITY TO COMMON CAUSE FAILURE l - DESIGN ERROR
- HUMAN ERROR
- INTERNAL ENVIRONMENT
- EXTERNAL HAZARDS l
- SUPPORT / AUXILIARY SYSTEM DEPENDENCE
- COOLING
- ACTUATION AND CONTROL 4
l j
- 35
O STATION BLACK 0UT ANALYSIS INTEGRATE LOSS OF 0FFSITE POWER j FREQUENCY AND DURATION WITH EMERGENCY AC POWER RELIABILITY MODELS INCLUDING: FTS, FTR, CCF, SYSTEM DEPENDENCIES DETERMINE SENSITIVITY TO DESIGM, LOCATION, AND EDG RELIABILITY O
1 I
l l
4
'l ~
lO 4
j i
u l.____-._..______ . . _ - . _ , - - - _ . - - , _ - _ _ _ . _ _ _ _ _ . _ , - _ _ , _ _ - _ _ . - _ - . . _
w .
e
- O
- ~2 m e c w$cs g O
V Om j l
. wp i Jj l ?
l ! .
J-E I2 ~ t DJ $$5 -p j
/ //
./t hg(f) i En F
/ /,/ l
]3 G' / // ,
0 J; j
- / // i
-20 2
l *P / /-
i M _O : / .// E i
0 -
yu / // en
>m i
i <E.
3 wE .
//
/ : ./ QC I
CD .
O! w o
/j// e/
- cc o o) 50I
~ i q
- u. -
///
i
/ . ,-
! ou
- 2 y ,-
' 'j . ; x On' t
-v<a l
v -. /
,. s/
/.i F_L
/...,.
- w
<C,Ji
. / !
l-- ,>
/
- g J ,) -- /
/ ..... /./ -
/ .. /
i c
i
! i / / . ** / io
[ '_I i l T'
- i i I
h ?
! E 2'
8 8
8
]~
8 e
' n-e s -
s
- a. e J $4 *
,g .
A0N30038 i
b 37 1- - -- - -_ _ .- __ -
i
- y 1 c :
c_.s e an N @ '
- g< e. o e. m.
O
_3 -
W l s
i 7 l 1 I a l i 1 o .
W i
T .
A i
b( :
- ,.e ~ -
f / /i l///
D . n>
LL) yT
/ /),./ 1 l.. 9 W/
l 3 0 0 ///
i i
i
.v
- \
3 l/ .f -,-
L LLiZ !
i o
w / :l
/// .
, 2 e.
u
)O: -
- Tn p- 'g i+-
/
/ /e' '
/
I, (Y_i
.?
u F--
o
~=-
q< ,/ i)i a
m T. , /
y.: ./ '
u-vD. / ..
Mo
/ .. 9 -
/ .... .
/
=
,J f
.* / /
/ . .. *..' . -
v I i
.m
..t I
l b 5 I I I' iII 6 a 6 4 7 P P
/- F m m e e
< s
- # 8"% ("% .^g 9 '-
b 'd_
W Q
^
f,o ADN3':03b'.d be 3%
. _ - .- _ ~ - - - - -. -
- - a r,x pxNN ,n :
7 N -
ca -
L,
,O i
8 ,
I i. I
- i, I
! m r3 i Wo '
W uJ r IE l s P '
!m U& '
$ _Y. j // T
>Z 3A / ./ / '
FEZ WW t
3
/
/ .// z
// ,?d
- OWO ZE- lwA /
/
//
.!/ T Q
wwk e cy
/
/l/ e
- 7 DE O
<C 4- r wO o
/
/ -
l/. i rc a ';
2.
- i / / / I ~ F- C. '
lCO.a y ]0 / ./ ./ :r
/ .-
.- f
, o, m y P L_
i
, L, u c :
/
/ ..' f
/ i
++;s A.
. i,
- Z. / **../
- r e
/
DZ Oi l
l h !
g.......-
/ -
t ,
O_O O .c /
i b
- Q
, Q 4 I l l
qE
_D l l l
! ! i
, CD Q ::: : : :::: . :::::: . :::n : : : -
I - -
e e 8 8 s - s. .m I
- C O o A .: i l
O O 6 ( ( h.
9 g y
~
-: e O (;oaA-;orotsa aaa) h ti ADN300383 031VW11S3 0 .
39
.. _ _ = .
O O o .
4 IMPORTANT FACTORS FOR DETERMINIFG i STATION BLACK 0UT LIKELIHOOD 1
1 INDEPENDENCE, FACKilP / ALTERNATIVE OFFSITE POWER SUPPLIES POTFNTIAI FOR AND SilSCFPTIBiliTY TO SFVFRE WFATilFR RFI ATFD LOSS-OF-0FFSITE-POWER 4
- EMERGENCY AC POWER REDUNDANCY / COMMEN CAUSE FAILURE POTENTIAL i
l f
PEllABILITY OF EMERGENCY AC POWER SUPPLIES l
i 3
l
O ACCIDENT SEQUENCE ANALYSIS l
ACCIDENT SEQUENCE DELINEATION ACCIDENT SEQUENCE PROGRESSION ACCIDENT SEQUENCE LIKEllH00D CONTAINMENT PERFORMANCE AND RISK O
e e
O el
. 1 Table 6.2 Possible factors limiting the ability to cope with a station blackout event Type of plant !
I Limiting factor PWR BWR 2/3 BWR 4/5/6 RCS pump seal leakage X X RCS letdown / makeup and water X X chemistry control lines Stuck-open relief valve X X DC battery capacity (instrumenta- X X X tion and control)
Compressed air (valve control) X X X Decay heat removal water supply X X X (condensate, firewater)
Operating environment (temperature) g Control room X X X (y (instrumentation and control)
Containment X (suppression pool, wetwell, drywell)
Auxiliary building X X (AFWS/ room) (HPCI/RCIC room) e n
l Table C.1 Summary of potentially dominant core damage accident sequences AC recovery Generic DHR system / component time to avoid plant type Sequence contributors core damage, hr PWR TML1B3 Steam-driven AFWS unavailable 1 to 2 (all)
TML22 B DC power or condensate exhausted 4 to 16 TMQ2B2 RCS pump seal leak 4 to 16 BWR TMU3B2 Isolation condenser unavailable 1 to 2 w/ isolation condenser TMQ3B 1 Stuck open relief valve 1 to 2 TMQ2B2 RCS pump seal leak 4 to 16 BWR TMU3B3 HPCS/RCIC unavailable 1 to 2 w/HPCS-RCIC TMU22 B DC power or condensate exhausted, 4 to 16 component operability limits Q exceeded (HPCI/RCIC)
BWR TMU2B1 HPCS/RCIC unavailable, I to 2 w/HPCS-RCIC THU22 B HPCS unavailable, DC power or 4 to 16 condensate exhausted, component operability limits exceeded (RCIC)
Notes:
DHR = decay heat removal HPCI = high pressure coolant inspection AFWS = auxiliary feedwater system RCIC = reactor core isolation cooling RCS = reactor coolant system HPCS = high pressure core spray O
e
Rsnga of Estimated Cars Demage Frequency from
. - Station Blackout Events O
/
r
$ , /
? ,
8 is Higher EDG Reliability
(] E < M' ore Redundant EDG
- j. < Configuration e ~
i
- u. , Better Offsite Power
& Characteristics
= < (e.g. susceptibility to Q
l < severe weather)
! Plant's Ability to Cope with Station Blackout (hours) i Yk k
Table C.4 Core damage frequency
(])
1/2 AC configuration t(hrs) Offsite power cluster 1 2 3 4 5 EDGR = 0.1 0 5.1E-3 3.0E-3 2.6E-3 1.4E-3 1.1E-3 2 1.9E-3 6.1E-4 3.8E-4 1.7E-4 5.1E-5 4 9.0E-4 2.9E-4 1.5E-4 6.8E-5 2.0E-5 8 2.5E-4 1.0-E-4 4.0E-5 2.2E-5 6.3E-6 16 1.2E-5-8.4E-5 9.6E-6-3.2E-5 2.4E-6-1.6E-5 2.0E-6-8.2E-6 5.0E-7-2.4E-6 EDGR = 0.05 0 2.0E-3 1.1E-3 9.5E-4 5.0E-4 4.0E-4 2 7.7E-4 2.5E-4 1.5E-4 6.9E-5 2.1E-5 4 3.8E-4 1.3E-4 6.2E-5 2.9E-5 8.7E-6 8 1.1E-4 4.5E-5 1.7E-5 1.0E-5 2.8E-6 16 6.8E-6-3.5E-5 4.4E-6-1.4E-5 1.1E-6-6.7E-6 9.1E-7-3.5E-6 2.2E-7-1.0E-6 EDGR = 0.025 6.0E-4 4.8E-4 2.5E-4 1.9E-4
. 0 0 2
4 1.1E-3 4.4E-4 2.2E-4 1.6E-4 7.9E-5 8.3E-5 3.6E-5 3.9E-5 1.8E-5 1.2E-5 5.2E-6 8 6.2E-5 2.8E-5 1.0E-5 6.1E-6 1.7E-6 16 4. 2E-6-2. 0E-5 2. 8 E- 6-8. 6 E-6 6.3E-7-3.7E-6 5.8E-7-2.0E-6 1.4E-7-5.8E-7 EDGR = 0.01 0 7.4E-4 4.0E-4 3.1E-4 1.6E-4 1.2E-4 2 3.1E-4 1.1E-4 5.7E-5 2.8E-5 8.3E-6 4 1.6E-4 5.9E-5 2.5E-5 1.3E-5 3.8E-6 8 4.6E-5 2.1E-5 7.1E-6 4.5E-6 1.3E-6 16 3.2E-6-1.5E-5 2.2E-6-6.4E-6 4.7E-7-2.6E-6 4.5E-7-1.5E-6 1.1E-7-4.1E-7 4
l l 1 i 45 l 1
If1PLEVENTATION OF STATION EL4CK0UT RULE O
PETERMINE ACCEPTABLE (1) SB0DURAT!0N(BASEDON FOUR FACTORS)
V ASSESS PLANT'S CAPABILITY (II) TO COPE WITH SB0 FOR DURATION DETERMIflEP IN (I) j i
i O
~ PLANT COPE WIT (Ill) SB0 FOR DURATION NO ,_
IMPLEMENT MODIFICATIONS DETERMINED TO EXTEND SB0 COPING IN (1) ? CAPABILITY YES V
IMPLEMENT PROCEDURES
- C (Iv) ANP TRAINING FOR COPING WITH SB0 ys
STATION BLACK 0UT RULE O
950.2 DEFINITION OF STATION BLACK 0UT 950.63 LOSS OF AC POWER o REQUIREMENTS PLANTS MUST BE ABLE TO WITHSTAND STATION BLACK 0UT PER GDC 17 o IMPLEMENTATION O APPENDIX A, GDC 17 THE REACTOR CORE AND ASSOCIATED COOLANT, CONTROL, AND PROTECTION i
SYSTEMS, INCLUDING THE STATION BATTERIES, SHALL PROVIDE SUFFICIENT CAPACITY AND CAPABILITY TO ASSURE THAT THE CORE IS l
COOLED AND CONTAINMENT INTEGRITY IS MAINTAINED IN THE EVENT OF A THE FOLLOWING FACTORS STATION BLACK 0UT FOR A SPECIFIED DURATION.
SHALL BE CONSIDERED IN SPECIFYING THE STATION PLACK 0UT DURATION:
(3) THE REDUNDANCY OF THE ONSITE EMERGENCY AC POVER SOURCES, (2)
THE RELIABILITY OF THE ONSITE EMEPGENCY AC POWER SOURCES, (3) THE EXPECTED FRE00ENCY OF LOSS OF 0FFSITE POWER, AND (4) THE PROEAPLE TIME NEEDED TO RESTORE OFFSITE POWER.
(:)
n,
IPPLEMENTATION SCHEPULE FOR 950,63 o LICENSEES SUBMIT TO NRC (WITHIN S M0tlTHS AFTER RULE IS IS
- PROPOSED ACCEPTABLE SE0 DURATION, INCLUDING JUSTIFICATION
- IDENTIFICATI0F 0F FACTORS, IF ANY, THAT LIMIT CAPABILITY OF PLANT TO COPE FOR ABOVE DL' RATION DESCRIPTION OF PROCEDURES TO COPE WITH SB0 FOR ABOVE DURATION O - LIST OF EQUIPMEt!T MODIFICATIONS, IF ANY, TO COMPLY WITH RULE AND PROPOSED IMPLEMENTATION SCHEDULE o NRC REVIEWS LICENSEES SUBMITTALS FOR PLANT'S REQUIPING EQUIPMENT MODIFICATIONS o WITHIN 6 MONTHS OF NRC REVIEW, LICENSEES SUEMIT SCHEDULE FOR COMPLETING MODIFICATIONS, WITH JUSTIFICATION IF SCHEDULE IS LONGER THAN 2 YEARS AFTER NRC REVIEW o LICENSEE AND STAFF MUTUALLY AGREE ON FINAL SCFEDULE FOR IMPLEMENTING MODIFICATIONS 1(])
Se
_ . - _ _ . . . . =_ .- _ _ _
STATION PLACK 0UT REGULATORY GUIDE O
o ONSITE ENERGENCY AC POWER SOURCES RELIABILITY PROGRAM TO MONITOR AND fiAINTAIN EDG RELIABILITY AT ACCEPTABLE LEVEL TARGET VALUES FOR MAXIMUM EDG FAILURE RATE PROCEDURES FOR RESTORING EMERGENCY AC POWER o 0FFSITE POWER PROCEDURES TO RESTORE OFFSITE POWER AND USE NEARBY POWER SOURCE WHEN OFFSITE POWER IS UNAVAILABLE o ABILITY TO COPE WITH STATION BLACKOUT GUIDANCE OF DETERMINING MINIMUM ACCEPTABLE SB0 COPING CAPABILITY GUIDANCE ON EVALUATING PLANT'S ACTUAL CAPABILITY FOR COPING WITH SB0 FOR ABOVE DURATION GUIDANCE ON CREDIT FOR AL' WATE POWER SOURCES TO COPE WITH SB0 o MODIFICATIONS TO EXTEND PLANT'S CAPABILITY TO COPE WITH SB0 PROCEDURES AFD TRAINING TO WITHSTAND AND REC 0VER ER0M SB0
() o 9,
I Table 1 Acceptable Station Blackout Duration Capability'(hours)*
Emergency AC Power Configuration Group D A B C 0 Maximum EDG Failure Rate Per Demand Offsite Power Design Characteristic. Group c 0.025 0.05 0.025 0.05 0.025 0.05 0.025 l
P1 2 2 4 4 4 4 4 P2 4 4 4 4 4 8 8 P3 4 8 4 8 8 16 8
' Variations from these times will be considered by the staff if justification, including a cost-benefit analysis, is provided by the licensee. The methodol-ogy and sensitivity studies presented in NUREG-1032 (Ref. 2) are acceptable for use in this justification.
b See Table 2 to determine emergency ac power configuration group.
c See Table 3 to determine groups P1, P2 and P3.
O So
O ine0aTAnT 11 ems T0 reeT SB0 DUaA110N GtiiDetises o ENSURE CAPABILITY OF SYSTEMS /COPP0t:ENTS FOR SB0 DURATION STATION PATTERIES CONDENSATE STORAGE TANK COMPRESSED AIR EQUIPMENT OPERABILITY UNDER SB0 CONDITIONS RCP SEAL INTEGRITY c DEVELOP PROCEDURES AND TRAINING o VAINTAIN/ IMPROVE EDG RELIABILITY I
VAlllE/ IMPACT
SUMMARY
O ESTIMATES REDUCTION IN COPF PAMAGE FREQUENCY .
MEAN SB0 CDF BEFORE RULE 4.2 x 10-5fpy MEAN SB0 CDF AFTER PULE 3.6 x 30-5/RY ESTIMATED COST AVERAGE $600,000/ REACTOR RANGE $350,000 - $4 MILLION TOTAL COST TO INDUSTRY - APPR0XIMATELY $60 MILLION PRESENT VALUE OF ONSITE COSTS (I.E., AVOIDED DAMAGE)
) $38 MILLION ESTIMATED RISK REDUCTION 145,000 PERSON-REM VALUE/ IMPACT RATIO 2,400 PERSON-REM /$1 MILLION (6,100 PERSON-REM /$1 MILLION IF ONSITE COSTS ARE INCLUDED) i i
(1) r
(A) BEFORE RULE 30-1 O 25 - MEDIAN CDF I
l iM g =EAN CDF ,
= 2.4 x 10 - 5/RY g 4.17 x 10 - 5/RY I
I
=> l i
~
R I l
I V I N l
'5-B l l 5 I sore:
10 - C 0 8 '
z l l
5-C a
a s s I.
a l
a la 8 s. 8 I $=k=8kkk
- s
!II.I. I.E a s a a a i a aa r
.u a ra y a O N o ? ; ? ?? ? i i ? ?* ? ga* - E 9
o - 9 o 9 o 9 o 9 o 9 o 9 oe e e 8 o
- n n n n w w in -
ESTIMATED CORE DAMAGE FREQUENCY ( x 10-5 PER REACTOR-YEAR')
O (B) AFTER RULE 30 MEDIAN CDF
= 1.1 x 10 - 5/RY 25 - l i
I i
7 E lMEAN CDF f$ l g 0 g l = 1.61 x 10 - 5/RY 20 - g 8 E 1 l
t; z l
l
, d [fi S l l .
a v is g
B '5- ;, l i
5 y '
I
- s s .
3 '$ l z 10 - g$ 4 ,
,h L'x
- A 9 ,
l 5- f f h 9 g i
- 2 W
2 W
n . . . . '. . , . . , , . .nr . 4 . . .
i 1 a e a e a e, aeaeaeak : O
- Q o ?T ? ? ? ?i i ??=
? ga
- g ESTIMATED COR'E DAMAGE FREQUENCY ( x 10-5 PER REACTOR-YEAR)
Figure 4 COMPARISON OF ESTIMATED STATION BLACKOUT (SBO)
CORE DAMAGE FREQUENCY (CDF) BEFORE AND AFTER RULE
PUBLIC COMMENTS ON PROPOSED RULE O
1 o PROPOSED RULE ISSUED FOR COMMENT 3/21/86 (5) FP 9829) o DRAFT REGULATORY GUIDE ISSUED FOR C0PMENT 4/3/86 (51 FR 11499) o 90-DAY COMMENTS PERIOD ENDED 6/19/86 O o 53 LETTERS COMMENTING ON PROPOSED RULE f -
45 FROM NUCLEAR INDUSTRY OPPOSED GENERIC RULEMAKING FUMARC INITIATIVES SUBMITTED 4
8 OTHER LETTERS SUPPORTED OBJECTIVE OF PROPOSED RULE 4
PROPOSED RESOLUTION DOES NOT G0 FAR EN0 UGH
~
()
[4
1 SIGNIFICANT CHANGES FROM PROPOSED RULE 8 REGULATORY GUIDE O
o RULE ELIMINATED RE0ViREMENTS TO DETERMIflE MAXIMllM COPING CAPABILITY CLARIFIED DEFINITION OF SPD o REGULATORY GUIDE CLARIFIED GUIDANCE GIVING CREDIT FOR ALTERNATE POWER SOURCES (E.G., REC 0VER ALTERNATE AC SOURCE TO COPE WITH SBO)
CLARIFIED GUIDANCE ON EDG FAILURE RATE (AUTO-START FAILURES NOT COUNTED AS FAILURES)
MODIFIED GUIDANCE TO DETERMINE ACCEPTABLE PLANT-SPECIFIC COPING DURATIONS BASED ON UPDATED ANALYSIS IN NUREG-1032 (REGROUPED PLANTS AND ADDED 2 AND 16 HOUR COPING DURATIONS)
DELETED GUIDANCE REFERRING TO MAXIMUM COPING DURATION o VALUE-IMPACT ANALYSIS IN NUREG-1109 INCLUDES 100 OPERATING REACTORS ACCOUNTS FOR RESULTS OF LATEST SOURCE TERMS DISCUSSES NUMARC INITIATIVES o NUREG-1032 UPDATED ANALYSIS BASED ON CURRENT EDG AND LOSP DATA
()
~
r,
S =
0 PUBLIC COMMENTS (PEVISIONS TO PROPOSED RESOLUTION) o F0llP COMMISSION QUESTIONS DUALITY CLASSIFICATION OF MODIFICATIONS (NC CHANGE)
- DOES BACKFIT AFALYSIS IMPLEMENT BACKFIT RULE S50,109 (BACKFIT ANALYSIS REVISED)
WOULD PROPOSED RULE SUBSTANTIALLY INCREASE OVEPALL PROTECTION OF PUBLIC HEALTH AND SAFETY? (VALUE/ IMPACT ANALYSIS REVISED)
SHOULD NRC REGUIRE IMPROVEMENTS BEYOND THOSE PROPOSED?
(N0 CH/NGE) l l
sa 1
~
OTHER PUBLIC COMMENTS o EXTENT OF REQUIRED COPING STUDIES (ELIMINATED REQUIREMENT TO DETERMINE MAXIMUM COPING CAPABILITY) o CREDIT F0P ALTERNATF POWER SOURCES (ADDED CRITERIA IN REG. GUIDE FOR ALTERflATE POWER SOURCES) o ACCEPTABLE DURATION TO COPF KITH STATION ELACK0UT (REVISED GilIDANCE IN REG. GillDE) c NEED FOR GENEPIC RULFMAKING (N0 CHANGE)
O o PLANT-SPECIFIC FEATURES AND CAPABILITIES (F0 CHANGE) o SOURCE TERN (REVISED VALUE-IMPACT ANALYSIS) e TRENDS ON RELIABILITY OF AC POWER SOUPCES (ALREADY CONSIDERED, NO CHANGE) o SPECIFICITY AND CLARIFICATION OF REQUIREMENTS (REVISED RULE AND REG. GUIDE) o RELATIONSHIP TO OTHER ISSUES (N0 CHANGE)
~
()
1 57
O V
INDUSTRY ACTIVITIES 4
I o EPRI TECHNICAL STUDIES NSAC/]O? LOSP MAY 1986 i
4
! NSAC/108 EDG RELIABILITY SEPT. 1986 i
o ANS DRAFT STANDARD 58.12 ON LOSS OF ALL AC POVER O REVISED DRAFT MARCH 1987 o NUMARC INDUSTRY INITIATIVES JUNE 1986 i
PfEETINGS WITH STAFF JULY 86-PRESFNT l *
.O 1
i t
NUMARC INITIATIVES (h,
U INITIATIVES SUBMITTED IN RESPONSE TO PilBLIC COMMENTS ON PROPOSED PlfLE .
- 1. CATEGORIZE PLANTS INTO 4 OR 8 HOUR GR0llPS, AND FIX 8-HOUR PLANTS TO BECOME 4-H0VR (USED STAFF Gl!IDANCE)
- 2. PROCEDURES A. COPE WITH SB0 (SPECIFIC DURATION AND COPING ASSESSMENT NOT SPECIFIED)
B. RESTORE AC POWER FOLLOWING SB0 C. PREPARE PLANT FOR SEVERE WEATHER
- 3. ELIMINATE EDG COLD FAST STARTS
. 4. MONITOR EDG UNAVAILABILITY O- (DOES NOT INCLUDE RELIABILITY PROGRAM) i STATUS OF NUMARC INITIATIVES
- 2. GUIDELINES FOR PROCEDURES BEING DEVELOPED GilIDELINES TO COPE WITH SB0 INCLUDE AC-INDEPENDENT AND ALTERNATE AC APPROACHES (STAFF NOT IN AGREEMENT YET)
- 3. REPORT ON COLD FAST STARTS TO BE ISSUED
- 4. EDG MONITORING TO BE BASED ON INP0 SAFETY SYSTEM MONITORING
, PROGRAM (STAFF DEVELOPING DG RELIABILITY PROGRAM)
~
()
S
NUMARC INITIATIVES COMPARED TO STATION BLACK 0VT RULE DETERMINE ACCEPTABLE (I) SB0 DURATION (BASEDON FOUR FACTORS)
M W Ws
.- V .
ASSESS PLANT'S CAPABILITY (II) TO COPE WITH SB0 FOR DURATION DETERMINED IN (I)
)
N of \
/
[ PLANT COPE WIT (
' NO IMPLEMENT MODIFICATIONS
(!!!) SB0 FOR DURATION
~ DETERMINED TO EXTEND SB0 COPING IN (I) ? CAPABILITY
\
'v, ~
' Jw f) v IMPLEMENT PROCEDURES (IV) AND TRAINING FOR COPING WITH SB0 0
CONSIDERATIONS RELATED TO NEED FOR A RULE O
o TECHNICAL DIFFERENCES BETWEEN NUMARC AND STAFF POSITIONS o OTHER CONSIDERATIONS IF TECHNICAL POSITIONS WERE IDENTICAL ENFORCEABILITY - CAN NRC ASSURE ALL PLANTS WILL IMPLEMENT RESOLUTION?
- CAN NRC ISSUE GUIDANCE WITHOUT A RULE (E.G., COPING WITH SB0, RCP SEALS)?
1 RULE AND REGULATORY GUIDE SPECIFY POSITIONS TO NRC (2) STAFF AS WELL AS TO INDUSTRY GI
_ _ _ _ _ _ _ _ _ _ _ _ _ _ - _ _ _ _ _ _ _ _ _ _ _ _ _ _ _