ML20058C781
| ML20058C781 | |
| Person / Time | |
|---|---|
| Issue date: | 09/28/1990 |
| From: | NRC |
| To: | |
| References | |
| NUDOCS 9011050086 | |
| Download: ML20058C781 (177) | |
Text
.. -
n JR 3 \\AL OFFICIAL TRANSCRIFT OF PROCEEDINGS -
x s
$$$fl%
0.3. Nuclear Pegulatory Canmission Tide:
Event Reporting W rkshop
'l Docket No.
l(
mns
. Atlanta, Georgia mm Fri& y, Septenber 28, 1990
- p. m 218 - 391 1
I t
- !?
- p -
l ANN M & MIAM, LTD.
l 3Ci 1612 K St N.W. Suite 300 g
Mbgon, D.C 20006 (202) 293-3950 9011050086 900928 PDR MIGC I
l 9011050086 PDC l I
!l e
~
(
,4 l
)
218 u.
1 sd 1
BEFORE THE
({
f'n.
h" 2
U. S. NUCLEAR REGULATORY COMMISSION l
3
'----------------------------------X I
4 In the Matter of:
)
l-
'5
)
6 EVENT REPORTING WORKSHOP
)
1' 7
X 8
Rutherford Room L
9 Omni International Hotel 7
10 Atlanta,. Georgia I:
11-Friday, September 28, 1990 14 12 3
j~ ~.13 The.above-entitled matter convened at 9:06'a.m.
-t X,_j/
.r L.
14 APPEARANCES:
15 1'
16-WORKSHOP PANEL MEMBERS:
17; TOM NOVAK, Nuclear Regulatory Commission l
18.
ERNEST ROSSI,-Nuclear Regulatory Commission
'19 EDWARD JORDAN, ~ Nuclear Regulatory Commission
'20 STU EBNETER, Nuclear Regulatory. Commission
- 21 ERIC WEISS, Nuclear Regulatory. Commission 22.
LUIS REYES, Nuclear Regulatory Commission
-23 PIERCE SKINNER, Nuclear Regulatory Commission 24.
MARK WILLIAMS, Nuclear Regulatory Commission
,f)l
'g_,/
25 JiCK CROOKS, Nuclear Regulatory Commission
V, 4
219.
/,NY),
1 INDEX l
u N'
2 Opening Remarks-Page...220 i
3 Immediate Notificaticn (10 CFR 50.72)
Page.. 233
[
- 4 NRC Panel Discussion - Industry Feedback Page.. 259 5!
LER System (10 CFR 50.7.1)
Page.. 292 6
Current Rulemaking/Guioance Revision Page.. 312
--i
- 7 NRC Panel Discussion - Industry Feedback Page.. 321
'8-Safeguards Events (10 CFR 73.71)
Page.. 352 9
NRC Panel Discussion - Industry Feedback Page.. 371-L E
10 Summary Discussion Page.. 377 2
- 11
- l'2.
l: V 14:
1 1'
15 l
16-
+
~ 17-H
- 18 19' i
. 20 1
21:
22'
" i 23 5
--y
..ry e
m wm
220
<r su 1:
PR0CEEDINGS O
\\s/-
2 (9:06 a.m.)
4 3
OPENING REMARKS 4-MR. NOVAK:
Good morning.
Welcome to the NRC's 5
workshop on event reporting.
I haven't been in Atlanta for 6
awhile.
My name is Tom Novak and I'm the Director of Safety 7
Programs in the Office of Evaluation of Operational Data.
8 It's a pleasure'to be here in Atlanta.
We've enjoyed 1
l 9
ourself.
There was a work session yesterday on backfitting l
10
-and from what I could see, it was very useful and the;e was l
11 a good exchange of-information.
I'm hoping that we can do p.
l 12
'at least that well today and perhaps better.
I think we've 13 got a little larger turnout and we hope not to lose anyone 14 too early-in the afternoon.
k 15 A couple of points that I would like to mention L
'16 very early.
There is a sign-in sheet and it will tua part of l
17-the-transcript.
So when you get an opportunity, either at y
L 18-the break, if you haven't signed in, please sign in.
19 Also, as~far.as the slides are concerned, we i
20-didn't bring enough to put in front-of every seat in the l
l211 audience; so if you're missing anything, there are copies --
l
~22 full sets of copies of slides up at the front row of chairs
- 23 and so forth.
So, maybe if you want to move up or sometime 24_
grab a complete set, please do so.
h-
.(,/
25' As I mentioned, we are keeping a transcript.
So, 1
l
t' 221
/N 1-an'we go through this meeting today, when there is an
)3 3
2 opportunity for discussion from the floor, please identify
-3 yourself and.your position with your company would'be very 1
4 helpful.
So identify yourself, your occupation so to speak 5
and your affiliation and we'll try to do the same thing when 6
people from the staff get up and comment.
We do have some
-7 people'from the NRC who are in the audience.
8 Just to move along, I would like to introduce --
u 9
and you'll probably recognize all of these people on the 7
10-panel.
Certainly you people will recognize Luis Reyes and L.
11 Stu Ebneter from'the Region II headquarters; Pierce Skinner, 12 the senior resident-at Oconee; Eric Weiss from the Office of 1'
I h.
.13 Analysis and Evaluation of Operational Data.
He's concerned o
g.
with the instant response center and you'll be hearing'from 14-1 15 him a little later this morning.
l 16 To my left.is Ed Jordan, the Office-Director of 17-AEOD.
Ed will have some remarks, just very quickly.
To his 18 left is Mark Williams, Chief of-the Trends and Analysis 19 Branch.in the office of Evaluation of Operational Data.
To 20 his left is Jack Crooks who is Chief'of the Data Management s
21
.Section in the Office of Evaluation of Operational Data.
i 22 And then finally to Jack's left is Ernie Rossi, Director in 23 the Office of NRR.
Ernie was on panel yesterday and you'll a
24 be hearing from him very quickly this morning.
- e s
?
25-Those are the brief comnents that I wanted to
,-,w
l_.'
w 222 1
-mention.
I'll be the moderator.
I guess it's my job to say 2:
it's break time and come back and try to orchestrate some of 3
-the question and answers during the discussion.
Personally,.
4
.I hope.this is a very productive workshop for you and for 5
ourselves.
We've worked in trying to get together the best 6
material we can so that you understand how we've looked at 7.
event reporting over the last several years.
The people at 8
the podium up here, you may not recognize.
They are the 9
people generally who are looking at a lot of the event 10 reporting.
So, we've brought the people who do use the 11 information day-in and day-out and I assume that the people ih-the audience are the people that provide the information 12 13 day-in and day-out.
So, we should have a good exchange of 14 information.
15 With that, I would like to turn the meeting over i
16 briefly to -- I guess Stu had some comments and I think it's 17 appropriate that we first hear from the Region regarding the
'18 Region's views on how plants in Region II are' reporting
/'
19 events of interest.
Stu.
20 MR.'EBNETER:
Thank you, Tom.
21 Just a-few comments.
I would like to welcome you 22 to: Atlanta, the city of the olympics, if you haven't heard 23 and I'm sure you have.
But this meeting on event reporting, 24 this workshop, is extremely important, as was the one 25 yesterday that AEOD also held.
These meetings are really
?
I~
?
3 L R ' :
.A Y c Y.
?.?Y L'
' ~
1 223 1
essential to you and us.
We noted in the Regulatory Impact t
2 Survey that the.*gency does not alwayc communicate our 3
requirements well and this area today, event reporting, was 4
another one of those topics that was controversial, has been 5
controversial for many years and will probably remain 6
controversial.
We do have difficulties with the industry on 7-reporting.
8 This' reporting, it is a critica? function of 9
yours.
The Agency relies on you to tell us what you're 10 doing.
We have repeatedly sent notices to the industry 11 about accurate, timely and full reporting.
So it's not a 12 new topic and unfortunately still one that we have some 13 disagreements on.
But it is required by the regulations and o
14-
.the rules of the Commission and you are obligated to comply 15' with those.
It is extremely important -- and I just want to 16' cite a little bit in relation to the enforcement policy.
17 Appendix C, 10 CFR 2 specifically has guidance for the staff 18 when licensees are not in compliance with reporting 19 requirements.
The guidance specificOly tells the staff 20 that we expect licensees to provide. full, cemplete, timely 21-and accurate information.
I particularly want to stress c
22
.that word " timely" because that is one area where we
~23 continue to have problems.
What does prompt mean, and I'm 1/
24~
not sure we're going to resolve it today but it is one that 25 we-still have difficulty with.
.,l I
224 1
Since it is no important to us with regard to 2
enforcement, when we do have issues with reporting, the 3
severity level -- the category, the azverity level that we 4
assigned to a non-compliance includes this problem of 5
reporting and it particularly -- the level of significance -
6
- in determining that, we consider whether you reported that 7
t aely or not and it's also used in there for mitigation of 8
the severity of the -- of the severity.
We can mitigate the 9
severity level by 50 percent if you have reported that 10 timely.
If you identified the problem or the event or the t
h 11 incident and you reported it timely, there's a factor in 12 there that we can mitigate the penalty if it's an escalated 13 enforcement action by 50 percent.
So, this is an incentive 14 for you to report to the NRC those items that are I
15 reportable.
I wanted to mention that.
l 16 The other side of that coin is if you don't report 17 it, and you knowingly don't report it or you should have 18 known of it to report it, then you've in a little bit deeper 19 because you're getting into the area of careless disregard 20 and willfulness of not reporting.
Now the willfulness 21 aspect, of course, not only invo.ves civil penalties but can i
22 involve -- and has in the-past involved criminal penalties.
23 So, you need to be fully aware of those consequences of not 24 meeting the reporting requirements.
25 They -- I jur quie.kly -- Part 20, for example, i
I 225 1
which we're not going to discuss today.
I think we're just 2
going to discuss Part 50 and 70.
But Part 20, severity 3
level two, enforcement action, escalated enforcement, would 4
5e imposed for failure to meet the immediate notification 5
requirement of Part 20.403.
If you fail to meet the 24 hour2.777778e-4 days <br />0.00667 hours <br />3.968254e-5 weeks <br />9.132e-6 months <br /> 6
requirement, the penalty is a severity level three, which is 7
still escalated enforcement and will probably carry with it 8
a civil penalty.
And if you fail to meet the 30 day 9
notification of 20.403, it's a severity level four.
10 Part 50 also has some guidance for us.
I'm sure 11 you're-familiar with this.
This morning the staff, the 12 headquarters staff in particular will be discussing Part 50 13 and 70 and I would encourage you, if you do have problems, 14 and I'm sure you do because of this area, to ask thc staff j
i 15 and try to clarify what the intent of these reporting i
16 requirements are.
17 I would like to compliment the Region II 18 licensees.
I think we are doing much better.
We don't 19 always agree on this business, but certainly, I think, the i
20 reporting area has improved considerably.
Probably the one 21 that's most controversial to us yet is in the Part 70 area.
22 We still seem to have some difficulties in that area.
But 23 generally it has improved.
It does help us.
We have 24 limited resources and we do depend on you to make these 25 reports as accurately as possible and on time.
I appreciate
226 1
all of the good compliance licensees.
Those of you who 7x
(
4
\\s-2 aren't compliant, we still have to work on you.
3 I think that's really all I want to say this 4
morning and I'll turn it over to -- are you up, Ed?
5 MR. JORDAN:
Yes.
6 MR. EBNETER I'll turn it over to Mr. Jordan.
7 Thank you.
l 8
MR. JORDAN:
I don't know what Tom Novak meant by l
9 quick comments, but I do have a few comments to make.
Mine l
10 are a little bit different than Stu's.
I -- in fact, I'm 11 going to downplay the enforcement side of it and put it in 12 terms of why do you report, what should you report and how f}
13 can we improve the process.
And as far as the "why report",
l V
l 14 I would say that my understanding of why there is a
(
15 reporting system is to provide information in order to 16 extract the lessons of experience.
And this is both plant 17 specific and generic.
18 You have to sort out the issues in your own 19 facility that are important to you.
The object is that you 20 also identify issues in your facility that may be important 21 to others, so that they don't end up having to learn from 22 their own bad experience what you learned and could 23 communicate to them.
l 24 There are both prompt and long-term considerations 1
-'s l
in reporting -- and I'll mention in a moment the aspect of 25
227
("
1 prompt.
From the NRC's view, we feel that there's an t
s 2
importance in order to provide an opportunity to identify 3
precursors of more serious events.
Once we have that 4
information, to feed it back to industry and to regulatory 5
so those more serious events can be prevented.
This was one 6
of the more important lessons that we all got out of the TMI 7
event; the need to do a better job of identifying precursors 8
and feeding it back.
If we identify and keep in the desk 9
drawer it doesn't help.
If we fail to identify and we re-10 examine data after the fact and find that we had it but we 11 didn't do anything with it, that's a terrible shame, our 12 system is such that we can do a much better job than that.
,25 l
I 13 It does require a systomatic and a cooperative effort O
14 between industry and the NRC and we're here to help foster 15 that cooperative effort.
16 The NRC actions in this area is -- The NRC 17 actions, I would say, following TMI -- and just a quick 18 summary of things that were done.
The entire reporting 19 system was revised.
We relied on a computer type licensee 20 event report that was a checklist item so that we could put 21 it into a computer system and sort it easily.
That was an 22 early attempt that was actually wrong because having a few 23 boxes to check and an entry of 20 or 30 items was not 24' sufficient.
We revised the system to require a written
-~
\\s 25 description of the actual event, a detailed written 2 %
., ~ -
4,--
r
228
, --w-1 description, and then from that extractions are made.
That o
i hs-2 was a great benefit and that change in the requirements, I 1
[
3 think, was very beneficial.
That was the change to 50.72 l.
4 and 50.73 from the old Reg Guide 116, Revision 4.
l 5
The other ti.ing we did was make a continuous i
6 notification system in terms of emergencies, moderately 7
important events and less important events all in one l
8 continuous spectru.m.
So you have one call to make by 9
telephone to our operar'sns center.
There is a very quick
[
10 screening of that event and a determination as to whether
- 1 there is an smergency or a prompt generic action required L
12 and then proceed from there.
Then the written report, of 1
13 course, is put into a' data base and is available for longer 14 term review and for review by summing across a large number 15
.of events.
Those things, we feel, have resulted in improved 16 generic correspondence.
17 Now, from the meeting yesterday, quite a bit of 18 the generic correspondence that comes to you as a backfit 19 comes out of that review process.
I hope you find that 20 those are benefits in terms of you being able to reduce the 21
. likelihood of having a problem at your facility that 22 somebody else has already encountered.
23 Industry actions after TMI were the establishment 24 of INPO and the overall review of operating experience that t,
25 INPO provides and then feeds back to the utilities.
There
229 1
is a methodology in the redundancy and diversity between the
~#
2 industry's process and the NRC's process and even within the 3
NRC at the regional level, at the headquarters level and 4
across offices.
So, we do intentionally have redundancy of 5
reviews.
We have some diversity through different -- people 6
with different backgrounds reviewing the same material.
The 7
object being that we have a greater likelihood of extracting 8
the lessons of experience and trying to get some of the more 9
obscure ones before they trip us all up.
10 The issues, I think, in this meeting are existing 11 reporting requirements and guidance providing the most cost 12 effective basis from which the NRC and industry can identify g~ s) 13 precursors that would reduce the frequency and the severity
(
14 of operational events.
We'll be reviewing in the course of 15 this discussion our current programs.
We expect to identify 16 areas for which reports don't help identify precursors.
In t
17 some other situations, we have identified some of those and 18 they'll come up in the discussions today.
19 We expect to consider whether revisions to rules 20 or guidance is warranted in order to do away with reports R21 that are invaluable -- not valuable, and to identify areas 22 for which we need reports that we're not currently getting 23 them.
And it is fair to say that the staff is considering 7s minor = rule making as a way of making a rather prompt fix to 24 i
'N 25 reduce some of the reporting requirements.
I won't 1
l 230
/]-
1 elaborate on them because there'll be further discussion.
]
\\"J 2
My expression is -- there is an emphasis that I 3
would like to place on reporting important stuff -- and
'4 Cindy Tulley smiled.
I made the same comment at an NUS 5
workshop.
I would like not to argue so much over the legal 6
reporting requirements but to emphasize that things that 7
your neighbor facility ought to know about that happened at 7
8 yeur facility ought to be reported.
Things that have 9
important satety significance at your own facility that are 10 unique ought to be reported.
That's important stuff.
(
11 Right after TMI, we issued a bulletin and my s
12 phrasing of it would be, it said please report important
~
13 stuff, anomalous conditions, et cetera, because Reg Guide i,%j-l l
14 116 wasn't sufficiently expressive.
It was aimed towards 15 limiting conditions for operations.
So my message today is f
16 report important stuff.
17 In many cases now the lessons are derived from 1
18 combinations or the frequency of occurrence of certain 19 events.
We find though screening and examining events over 20 a period of several years, we can sift out common themes 21 across events that we can then feed back to you and INPO is i
1 22 doing the same thing.
I think the overall plant performance 23 in the U.S.
reflects that we are learning the lessons, that 24 the industry is applying the lessons and fixing repetitive 25 problems.
l
.. _., _ _ _ - -,. ~.
231 I p}
1 The NRC is de-emphasizing bean counting of events,
{%/
2 that is, the utility that has 30 events versus the utility
[
l 3
that has 15 events, the numbers are somewhat irrelevant.
4 It's the significance of the events that we're 1.'terested 1
5 in.
Our screening p rass is all geared towards measuring 6
and weighing the significance c' events.
7 The emphasis in headquarters, certainly among my i
8 staff now, is to do more in the way of risk based 9
determinations regarding events.
We try to screen more 10 promptly and establish where does this fit as an accident 11 sequence precursor.
What is the core damage probabilit.y 12 given this event?
That's the kind of consideration that we A(,)
13 think is more important by far than any kind of a bean 14 count.
We need your cooperation in order to continue to l-15 extract these lessons of experience.
We find there is an 16 importance in having a stable data base on which to work 17 with events.
I would give the example from reviewing this 18 data base over the past five years, that it has had a 19 reasoriably consistent reporting on the part of the utilities 20 and certainly the requirements have stayed stable.
We can 21 say -- and I've said it personally in front of Congress, 22' that-the performance of the U.S. plants have improved 23 substantially in the last five years.
That the numbers of 24 safety system actuations, safety system failures, reactor
(
25 trips, all of those things have shown a substantial
232 fN 1
improvement which is reflected in safety performance.
When 2
we look at the accident sequence precursors and put it into 3
a risk space, we can say that there is a clear indication 4
4 that the actual risk relating to operating events has 5
declined over the years.
The number of higher risk events 6
has decreased.
So that's one of the benefits of having a 7
stable data base.
8 We both are spending time reporting and evaluating 9
events for which there is not a pay-back that's obtious.
10 So, it's focusing on those -- and I think tuning, the 11 regulations and the guidance that we want feedback from you 12 on.
So, we would like to convey to you our best l
q
{
t 13 understanding of the process.
We want you to understand As/'
14 what we do with the information and then we're looking for 15 an active feedback from you on things that can be done to 16 improve the process.
17 So with that, Tom -- with those quick remarks, 18 I'll turn it back to you.
Thank you.
19 MR. NOVAK:
Thanks, Ed.
20 Just one point, the way we plan to do this, so 21 that we make sure we do cover the whole agenda is, we're 22 going to have more or less the formal presentations and then 23 we'll have a special period where all questions can be 24 brought up from the floor.
I think that way we can hear the 7-ss
\\_-
25 material presented, jot down your questions and we'll make l
- =
233
s 1
sure we have time to cover them all.
So, in that manner, f
2 what we'll do first is start off with Ernie Rossi.
He'll be 3
followed by Eric Weiss and they are specifically going to 4
focus on 50.72 reporting.
5 Ernie.
6 IMMEDIATE NOTIFICATION (10 CFR 50.72) 7 MR. ROSSI Good morning.
What I'm going to do is 8
to tell you what it is we do with the 50.72 reports once 9
they get to the NRC and I'm going to talk in some detail 10 about how we review them and what they're used for and that 11 sort of thing.
Later today, there's going to be a similar 12 discussion on the reports that come in under 50.73 as to how g^N
()
13 the LERs are used.
I have one viewgraph in my package here 14 that sort of generally discusses our perception of various 15 problems that we've experienced with the 50.72 reporting.
16 The main thing I want to do is convey an 17 understanding to you of how we use the 50.72s.
I think by 18 understanding what we do with them, you'll know a little 19 better why we feel they are so important.
20 Could I.have the next viewgraph please?
21 Now as you all know, 10 CFR 50.72 requires 22 telephone notification to the NRC operations center via the 23 red phone for specified types of events at operating reactor 24 plants.
Now the 50.72' report is the primary source of 7-('
25 information for the NRC to begin its short-term evaluation
e 5
i 234
~x 1
process of an event or problem that occurs at a specific
\\ ')
t 2
plant and use them also to make an assessment of whether 2
problems or events could be generic.
i 4
10 CFR 50.73 requires a written report on events 5
and it's basically complimentary in general to the 50.72 6
report requirement.
There are some differences in the 7
reporting requirements but generally 50.72 would' be followed 8
up by a 50.73 with more information.
The 50.73 report is 9
used as the record of an event by the NRC.
It's used widely 10 by INPO, foreign governments and others and the NRC's review 11 of these reports will be discussed later today by Mark 12 Williams.
13 Could I have the next viewgraph, please?
l-14 These are the NRC organizations that are involved 15 and deal with the events assessment process.
L 16 the regions all have complimentary functions in reviewing l
17 events that are reported.
Now the NRC Operations Center, of i
18 course, first receives the event and they are a part of 19 AEOD.
The NRC Operations Center is staffed 24 hours2.777778e-4 days <br />0.00667 hours <br />3.968254e-5 weeks <br />9.132e-6 months <br /> a day 20 with trained professionals to take the 50.72 telephone calls 21 and to determine the need to notify a senior NRR manager or
.22 other agencies of the event.
23 Now, the people that work in the Operatio7s Center 24 receive an extensive amount of training.
They go to systems j.
N, 25 courses at our technical training center and they have a
l 235
~s 1
good understanding of the plants.
Now recognize that they 7
\\ -)
\\
2 are not going to have a detailed understanding of each and j
4 3
every plant specific detail but they have good extensive 4
training in the plants in general.
The region is promptly 5
informed of every event in that region that's called into 6
the operations Center.
The regions follow up on the plant l
7 specific aspects of each event.
8 The Events Assessment Branch in the office of l
9 Nuclear Reactor Regulations screens each event for the need l
10 to follow up on a plant specific basis from headquarters and l
11 for the need for any generic actions by the NRC.
- Now, I
I 12 generic actions can consist of the issuance of an
[j) l 13 information notice, a bulletin or a generic letter.
Now l
\\
14 typically the most common thin; that we do with events that l
l 15 have generic implications is to issue an information notice l
I 16 and we typically somewhere between 80 and 100 of those a
~
17 year.
The purpose of the information notice is to share 18 information on problems and events that we find in our 19 assessment with the industry and they require no response 20 back.
They are sent out to help you avoid similar problems 21' on your plants by knowing what happened at other plants.
22 NRR Projects follows the status of operation and 23 problems on each and every plant.
So they look at the 24 events so that the project manager knows exactly what's Ak,)
25 happening on a day-to-day basis on his particular plant.
L 1
']
236 1
AEOD studies trends and patteras of events on a t(N i
)
2 longer term basis for lessons learned and the need for 3
generic actions.
AEOD primarily depends and uses the LERs.
4 Now, NRR primarily depends on the 50.72s and we use the LERs 5
as a means of obtaining a more detailed and final 6
information on the events.
Basically AECD deals mostly with 7
the LERs and NRR deals with the 50.72s.
8 Could you go to viewgraph number five now?
The 9
next one I want to skip.
10 This viewgraph gives the details of the 50.72 L
11 reporting requirements and it should be in your handout.
I 12 don't intend to go through it.
What I do want to say is l
l.g
(
13 that the highest agency priority related to the 50.72 i
u 14 reports is to determine the need to activate the NRC's 15 emergency plan, staff the NRC's Operations Center with L
16 senior managers and technical r,xperts and follow a safety 1
17 significant ongoing event until effected plant is in a 18 stable condition.
Now, that's the most important thing that l
19 we have to do with the 50.72 reports.
20 Could I have the next viewgraph, please?
l 21 Now this viewgraph shows a flow-chart of the.NRC I
L 22 response to events reported to the Operations Center under 23.
50.72.
Now the licensee, of course, calls the event into i
24 the NRC Operations Center and the person there takes the p-1
\\
25 event and he has a criteria that's used to tell him or her l
-l
i 237 1
what needs to be done next with that event.
4 2
The NRR emergency officer shown on this viewgraph 3
is a senior' executive service manager who is on call 24 4
hours a day.
Now, within NRR, the emergency officer 5
rotating duty list. includes assistant division directors up 6
through division directors.
That assignment rotates weekly 7
and the emergency officer is imformed of all unusual events or higher classifications and he is the person who makes the 8
L 9
first decision on whether other managers should be informed 10,
and whether consideration should be given to staffing the l-
[
11' operations Center.
12 After the process of whether immediate 1
l (^\\ -
13 notifications need to be made to the emergency. officer or 1.
14 other government agencies is made, the person on duty in the 15 Operations Center prepares a brief description of each event 16 for a word' processing system and each event is numbered t
17 sequentially and then has a unique number for it thereafter.
18 All events that are reported since the previous 19 working day, along with follow-up information and problems 20 that are reported by each of the five NRC regional offices H
21 are screened during approximately the first two hours of
[
22
_each working day.
These events are screened by a number of 23 people and the people screening them have a variety of L
j technical and regulatory backgrounds, so you get a somewhat 24 25 different perception from the various people that screen l
238
(~~N 1
these events.
As I'll get to a little bit later, they
(
2 discuss the events and decide what needs to be done with 3
them.
So each and every event that's called in to the 4
operations Center goes through this screening process during 5
the first two-hour period of every working day.
6 Now, even though, each and every event gets read, 7
most events do not require follow-up by NRC headquarters 8
personnel; however, all of the events are followed by
[
9 personnel from the responsible NRC region to whatever degree 10 is necessary to ensure that appropriate plant specific 11 actions are taken.
12 could I have the next viewgraph?
[
' A}
(
13 This viewgraph shows what goes on on a daily basis 14 with regard to the review of the 50.72 reports.
There is a 3,
systematic 100 percent review by the Events Assessments 16 Branch in the office of Nuclear Reactor Regulation during l
17 approximately the first two hours of each working day of the 18 50.72 report write-ups from the operations center, region 19-daily reports and preliminary notifications from the regions 20 and any other information from.the regions or the project 21 managers on reactor plant problets.
22 Now, let me say a litt.'e bit more about the 23 information that we get from the regions.
Each day -- each f-
-24 working day each of the regions sends us a region daily I
l
's_
25 report and that region daily report describes in about the C
239 1.
same kind of detail as what is called into the operations 2
Center on the 50.72s, the problems that have occurred at 3
plants within the region.
Now, in many cases the regions 4
can augment and give their perspective on events that have 5
already been called in under 50.72.
6 The other thing that we get from the regions is,
)
7 we get descriptions in many cases of problems that either 8
should have been reported and weren't reported by the 4
9 licensee or in a lot of cases, we get reports from the 10 regions of problems that really didn't fall within the 11 reporting requirement of 50.72 and therefore, we didn't --
12 the licensee was not required to tell us about it.
But l
13 nonetheless, the region -- from the region's perspective, y,
1 14 it's an important enough problem that we ought to be told 15 about it within headquarters.
So, we get that information 16 each day and that information is reviewed just the same as 17 the 50.72 reports during this first two hour period of each 18, working day.
19 The division directors within NRR, or their 20 representatives are briefed by telephone at 8:15 a.m. each 21 working day on significant events, including all reactor 22 trips that have occurred at plants.
The way this briefing 23 works is, we have a conference call bridge and all of the 24 division directors within NRR are given a number to call in 25 at 8:15 on the bridge.
At 8:15 somebody from the Events
i 240 1
Assessment Branch gives about a five -- it typically is a 7
i
's 2
five to ten minute briefing on significant events and 3
problems that have occurred since the previous working day.
I 4
Now, in that briefing would be included things that have 5
been found from the review of the 50 -- a quick review of 6
the 50.72 reports that morning and also things that have 7
been reported to us by the regions in their region daily 8
reports 9
Now in addition to that, NRR -- and really agency 10 senier management is informed even separately of highly 11 significant events.
Senior management ;;iinin the agency 12 would be informed of highly significant events ao soon as
-[')
13 possible after they occur.
This would include the EDO, the
\\_ /
14 deputy of EDo and various office directors.
So that process 15 is a continuing one that can occur weekends or whatever.
16 But that briefing goes on.
But there is this routine 17 ilivision director briefing each morning at 8:15 so that all 18 of the division directors within NRR know about the-19 significant problems that have occurred since the previous 20 working day.
21 The Events Assessment Branch conducts the
~22 tilephone conference call to discuss the more significant 23-events at 8:50 a.m. each working day.
Now that's a 24 telephone conference call again with a fairly significant
,(N
(_jl 25 number of people involved.
That conference call includes
~
241 f3 1
representativos of the Events Assessment Branch and Generic
_(
J-2 Communications Branch in NRR, representatives of the Office 3
of Analysis and Evaluation of operational Data and 4
representatives of the Vander Inspection Branch and 5
Projects.
Now the purpose of that call is, they gc through 6
the more significant events and decide on whether they 7
thoroughly understand the event and they decide on whether 8
there should be additional follow-up to get more information L
9 on the event and actions in that telephone conference call 10 are assigned to individuals to obtain additional information i
L 11 where warranted.
12 Now,.the kind of thing that may happen is, for
(
13 example, there may be an event that involves a particular 14 piece of equipment and the Vendor Inspection Branch may be 15 given the job of going back to the vendor in some way to get l
L 16 additional information on what may have caused the problem 17 or to pursue whether the problem could be generic in some l
18 way.
l 19 During that 8:50 call there are discussions on the 20 need for an augmented inspection team or an incident 21 investigation team for the more complicated or serious 22 events.
23
.Now, I would again like to emphasize the fact that l
24 probably 90 percent of the 50.72 reports that we receive, 25 that there is no follow-up by NRC headquarters after the
242
./N.
1 initial screening.
These events may not even be discussed 2
in any of the telephone conference calls.
They'll be 3
screened and a decision will be made upon the initial 4
screening that there is not anything important to follow up 5
on and nothing more -- or no further discussion will occur 6
on those events.
Now, th do remain in our computer 7
systems, so later on, we can go back if something comes up 8
and find the event or do searches and that kind of thing.
9 But as far as this process within headquarters, about 90 1
10 percent of them would receive very little discussion, j
11 Again, the regions involvement is somewhat more 12 because they look at the plant specific aspects of the event j
L r~'s
{ s) 13 to whatever degree they need to to make sure that any plant l
A l
14 specific issues are addressed.
l 11 5 Could-I have the next viewgraph, please?
16 Now a major objective in all of the events i
17 screening is determining the basic factual information of 1
18 the event.
Again, that's dependent on the utility's 19 telephone notification and it's supplemented by information 20 obtained by telephone from the regional office which 21 generally would come originally faom the resident inspector.-
i 22 Then we can have confirmation and augmentation of our j
23 understanding of the event from the written report.
24 Now for complicated and more significant events,
\\s-25 we can collect the facts by having an incident investigation c
.7
..__...-.s-
243
/N 1
team or an augmented inspection team.
But, you know, we do 2
look at each and every event to be sure that we have a clear i
3 understanding of what happened and to have a clear 4
understanding of'the significance both from plant specific 5
standpoint and generically.
6 Could I have the next viewgraph, please?
7 Now, there are in addition to these daily things 8
that happen with the events, this viewgraph shows some 9
weekly briefings and discussions that take place on the 10 events.
At 1:15 each Tuesday there is an event meeting and 11 this event meeting includes reviewers, engineers, project 12-managers, section leaders and branch chiefs that happen to n'
)I 13 have an interest or a responsibility for whatever occurred, l
14 whatever probleu may have come out of the particular events 15 of that week.
16 Now in this briefing -- by this time the number of 17 events that are discussed have been thinned down to a 18 relatively few of the more significant events and the 19 reviewers, project managers and branch chiefs and so forth 20 then discuss the events, their significance.
They discuss 21 the need for any long-term follow on these events and 22 there's an assignment then for long-term follow-up where 23 that's appropriate.
3 24 Now, in addition to that, they do a dry-run in N-25 critique of the briefing that I'm going to talk about in
l 244 (N
1 just a minute which occurs on each Wednesday morning.
So L' )
2 that's also done at that 1:15 Tuesday meeting.
Now, that 3
1:15 meeting is basically a working level staff meeting on 4
the events to discuss their significance and the need for 5'
any follow-on actions.
6 Now at 11:00 a.m. each Wednesday there is an event 7
briefing of only the very significant ovents that may have 8
occurred during the wook.
That brief.ing is directed 9
primarily at agency division directors and above.
The 10 briefing is generally attended by a fairly large number of 11 people.
That briefing genera 1'
' tracts cotmissioner's 12 assistants.
It attracts ACRS stu sembers, attaf f members m
i 13 from the EDo's office and that briefing includes -- it's 14 directed at the entire NRC.
So each of the five regions 15 participate in that briefing by telephone.
They are all l
16 hooked in by telephone to the briefing and the briefing 17 generally is done by somebody from the Events Assessment
\\
18 Branch in NRR, although occasionally it may be done by 19 someone else who may have more information and expertise on 20 the details of the particular event.
The regions are 21 generally able to provide additional details or correct 22 facts based on follow-up actions that have been taken by the 23 region during that briefing, g
Now the duration of that briefing is typically 24
~
25 anywhere from -- well, I have on my viewgraph a half-hour to
245 t'~ N 1
one hour.
Sometimes it's -- you know, can be five or ten N'
2 minutes.
l; h
3 The briefing viewgraphs and the attendan'. list 1
4 from that briefing are placed in the puolic document room.
5 So the viewgraphs on the events that are briefed in that
(
6 Wednesday morning meeti.
1 to the POR and they are i
7 available to the public and frequently various news media 1
8 organization will -- I think they probably routinely get 9
those.
10 Now, if we don't have any significant events 11 during the week -- and in the last year that has frequently 12 occurred -- we just don't have either the Tuesday briefing
/'
(
13 or the Wednesday briefing.
Those briefings are only held if K
14 we have something to say.
We don't just hold them routinely 1
15 if there aren't any significant events and in the last two i
l or three months there have been several weeks in a row at 16 1
17 times when we haven't had any events or problems reported l
18 that we felt were significant enough to have that Wednesday 19 morning briefing.
20 okay, the last viewgraph that I'm going to discuss l
l 21 is the next one.
This is just'a very brief summary in 22 general of problems that we've experienced with the 50.72 23 reporting.
Now, Eric Weiss is going to follow me and he's 24 going to go into some details about the NRC's perception of O).
(s_
25 reporting problems and why we view some of these things as
- a
-.w--,
-r,-
-.c.9,,.
246 1
problems.
2 Now, we all understand that the rule currently 3
requires reports on some events of minor significance.
4 These include reactor water cleanup system actuations on 5
BWRs that are innocuous and have very little -- well, they 6
have no effect on the operation of the plant and reactor l
7 trip system actuations when the plants are shutdown and the 8
rods are already fully inserted and someone is doing 9,
maintenance.
We appreciate that the rule requires those 10 reports.
I think later one someone is probably going to 11 talk about what might be done to fix the rule with respect 12 to those.
Our problem always turns out to be.aw to word 13 the rule or the guidance in such a way as to eliminate the 14 things that are clearly not significant and don't need to be i
i 15 reported, without at the same time eliminating some reports f
16 that may have a significance that wasn't recognized when we i
17 wrote the rule.
So that's our difficulty.
Generally our 18 approach is that on close calls we had rather have the 19 report'obviously than not have the report because that way 20 it's very unlikely that we'll miss anything of significance.
l 21 We recognize that there are different definitions 22 of systems that are ESF systems at different plants.
I 23 believe that what people are doing now is depending only on 24 whether the FSAR calls something an ESF system or not.
That 25 is somewhat of a problem to us because it's my understanding
1 i
l 247
,7 3
that there may even be some plants out there where diesels 1
s_-
2 aren't referred to as ESF systems and so inadvertent 3
actuations of diesels may not get reported.
In my personal 4
view -- and I know that we don't have guidance out that says' 5
this -- I think an ESF system ought to include any system 6
that used to mitigate a transient or accident.
If we had I
7 given that kind of guidance that probably would have solved 8
most of the problems that we have in this area.
But again, 9
I'm speaking-from my own personal view on that.
In order to 10 provide that guidance, we would have to go through a 11 structured system for providing guidance.
i 12 There is a lot of judgement involved on what's a
/
13 serious degradation of plant safety systems, what an 14 unanalyzed condition and what's outside the design basis.
15 So these are problems.
Again, our general approach is that 16 when in doubt, we would rather hear about it because if we 17 hear about it, we can compara it with things that we may 18 have heard about from other plants and in many cases, we can 19 find out whether there is or is not some generic problem 20 that everybody ought to know about.
21 We are very sensitive to events or conditions that 22 could prevent fulfillment of the safety function.
One of 23 the particular things that we're quite worried about are any 24 kind of equipment problems that could lead to a common mode l
f' 25 failure.
Now equipment problems that could lead to a cor. mon
\\ _j i
e
'm um-
-m-,,
w 5
y y-p y
248 i
f~'N 1
mode failure can be problems -- well, there were some N) 2 problems awhile back with -- I think they were check valve i
3 bolts where because of the material they were made of they 4
were cracking.
That was a material selection problem for 5
particular kinds of valves and that obviously had the 6
potential to lead to a common mode failure because if it's a 7'
design problem or a material problem that could effect all 8
trains of a safety system, that ceuld clearly lead to a
]
9 common mode failure.
It's clearly generic and those are the
)
l 10 kind of things.that we worry a lot about finding and at l
l 11 least notifying everyone of the possible existence of -that 12 kind of a problem at their plant.
I 13 Degradations in equipment -- and this is a 14 timeliness issue more than anything else which could cause 15 the safety function not to be properly carried out.
The 16 timeliness has to do with the fact that in some cases 17-licensees will study the problem for a very long time before l
18
'they decide whether it is or is not enough of a degradation i
19 to report.
I guess in my view, we would prefer the report 20 if there's a reasonable chance that it is indeed that kind l
l 21 of a degradation.
But Eric Weiss is going to talk a little l
l 22 bit more about these.
23 Now, you have in your handouts a number of other l
24 viewgraphs that describes some of the criteria that are used
,-sx 25 for event follow-up.
Needless to say there is a lot of l
l
249 1
judgement involved in everything that we do.
There are some 73
(
I x_ /
2 more viewgraphs there and they are in your handouts.
I'm I
3 not going to go through and discuss them.
4 I'm now going to turn things over to Eric Weiss 5
and he's going to talk in more detail about the problems 6
that we see with the reporting.
t 7
MR. WEISSt Good morning.
Dr. Rossi told you how 8
important your 50.72 reports are and what some of the 9
problems were.
I'll try and identify some additional 10 problem areas, specifically those things that we thought wo 11 were going to get when we wrote 50.72 and are somewhat i
12 surprised that are not getting consistently reported now.
l
- (~*g 13 My focus will be primarily on 50.72, but as you know, 50.72 1
V 14 and 50.73 are similar, so much of what I say will be 15 applicable to 50.73 as well.
16 Bofore launching into this, I want to put it in 17 perspective.
We get about 3,000 calls a year to the 18 operations Center from operating reactors.
About 2,400 of 19 those are under 50.72 and we only have a few serious events 20 a year that go unreported, where we are really surprised.
21 This slide has got a lot of information on it.
22 This talk ordinarily has 30 slides and takes 40 minutes but l
23
'I'm going to do it very quickly, in about 15.
So, we've got l
24 a lot of information here.
(~)s
(
25 I would say consistency is our biggest problem
=*wt
,-w-r
~-
,.,,g
--we w
w p-,---
,y,p p,
e
250
}
/
1 wiuh 50.72.
It's interesting to note that of the things
'~'
2 that I've got listed here that are not consistently 3
reported, only one or at most two of these are explicitly 4
listed in 50.72.
5 To begin with, let's talk about the anticipated 6
emergencies.
I was in Ernie's office some years ago and 7
someone came in saying that someone else in the Commission 8
had heard that a plant was in a state of emergency.
They 9
were about to be flooded.
We called the plant and sure 10 enough they were.
They knew for some days, as a matter of 11 fact, because they had detailed hydrographic information 12 that they would enter an emergency condition and they hadn't i
fs
)
13 called us yet.
They were waiting for the emergency to be 14 declared and for the one hour time clock to expire.
- Well, 15 the regulation says specifically that we should be notified 16 as soon as possible and in no case later than an hour and we 17 would very much appreciate a call if you know that you are 18 going to be in an emergency, to call us then.
The reason 19 for this is that the NRC takes some time to get prepared and 20 we have responsibilities to notify other federal agencies.
21 They have things to do to prepare for an emergency should it 22 get worse.
So, we very much appreciate a call as soon as 23 you know you are going to be in an emergency.
-'s 24 The second bullet I have listed is rather i
\\-
25
.important.
If I had to choose one criteria over the past
251 1
year that would characterize events that have gone 2
unreported, this would be it.
Spill is somewhat of a loaded 3
term.
It almost connotes something that's insignificant but 4
spills are often more serious than they seem.
A large spill 5
has implications for environmental qualification.
Equipment 6
can be wetted or submerged that was not designed to be 7
wetted or submerged.
As Ernie pointed out, thero are a i
8 number of people that study precursor events, they are 9
interest in intersystem LCCAs and those sorts of things.
10 Very often those large spills have implications.
If we're 11 talking of thousands of gallons of reactor water, I'm very 12 much surprised when we don't hear about that type of event.
L 13 Even some of these events have implications for 14 fuel uncovering.
There were events some years ago where 15 plants in refueling status lost large quantities of water i
16 from their refueling poolc and if there wasn't the potential 17 for fuel uncovering in the pool or the vessel, there was at 18 least the theoretical. potential for fuel in transit to have 19 been uncovered which is a very serious accident, very 20 difficult to recover from.
l 21 Inadvertent criticalities is another thing that 22 has caused us some surprise.
There have been plants that 23 have had serious events that have launched big 24 investigations, very serious investigations, and they 4
25 haven't considered the event reportable.
Now, I should put
I i
252 1
this in some perspective.
I've heard people say that all j
-criticalities are inadvertent to some degree or another.
2 3
You never hit your estimated critical position.
That's not
=
4 quite what I had in mind when I said inadvertent 5.
criticality.
What I had in mind was where we have a trainee 6
at the panel and he's not being properly supervised or the
~
7 rods are pulled according to the pull sheet or you're so far x
8 off from your est-imated critical position that a reasonable 9
person would say that really is an inadvertent criticality.
10 Another category of things that surprises us are h
11 small waterhammers and fires.
Why?
well often because 12 these events have wide spread implications that aren't jr 13' obv'ious to the people that are calling them in.
There may 14.
be a new mechanism for a waterhammer or a fire.
There may 15 be generic implications for what caused the fire or what the j
i 16 fire could do or the waterhammer.
Perhaps it wasn't very 17 serious at your facility in this particular instance, but if-the waterhammer had occurred at another facility it would 18 i
i 19 have been much more-significant.
Perhaps you had the small1 20 waterhammer or the small fire and in a different system
~
21 alignment or with a different design it could have been much 22 worse.
g 23 The riext category, over-pressurization events --
24 you can also add over-temperature there -- have been very 25 serious events.
We've had AITs on these types of events.
=
1 1
f 253 1
It's not enough -- I should say it is enough that the FSAR 2
value is exceeded on an over-pressurization or over-3
. temperature for us to be interested.
We've had some 4
licensees come back and say, well, that wasn't reported.
We 5
hired a contractor and his analysis shows that we had enough 6
margin in the design spec for this pipe that it really could 7
handle it.
But at the time the event occurs, the licensee 8
didn't know that.
In any case, we're very much interested 9
in this.
There are people who study intersystem LOCAs and 10.
if you have reactor water pressurizing a system outside of 11 containment, we would very much like to hear about that.
My 12 argument has always been that 50.72 requires the reporting 13 of a degradation-of a primary system pressure boundary and I 14 personally have always had a problem with interpretations of 15 such events that find them not reportable.
If you've got 16 reactor water pressurizing a water fountain outside of
]
i 17 containment,-that -- I'm interested.
It doesn't matter that 18 it's a non-safety system.
19 There has also been a number of potentially 20 generic events.
I mean, I think anyone would agree that 1
21 they were potentially generic and yet they haven't been 22 reported.
I have in mind,-for example, an event that 23 occurrred at a reactor where the vendor was called within 24 24 hours2.777778e-4 days <br />0.00667 hours <br />3.968254e-5 weeks <br />9.132e-6 months <br />; the state was notified; the licensee management was 25 notified and the last outfit to be notified some four days
t 254 m
1 later was the Nuclear Regulatory Commission.
l
.f
's-).
L1 2
Of course, Ernie has dealt with the last criteria H
L 3
to some extent, ESF actuations.
We not only have the l-4' problem that Ernia outlined where some plants don't define 5
certain systems as ESF's, but there is also some controversy 6
about what constitutes an ESF actuation.
For us, if a 7
device actuates that's for the purpose of containment 7
l 8
isolation or for the purpose of mitigating an accident for 9
us, we would expect that to be reported.
It wouldn't be i
10 necessary that all components function properly in the 11 system.
12 Next slide, please, f
n L['%_ ]\\,
13 Notifications for NRC response to media public --
{
14 and.I guess you can add states and locals here--- have often 15 been untimely and the threshold has.often been off.
We need l
l.-
16 to know about any event where thc public, the media, the 1
l 17
' ate or local government. perceives a safety problem.
Even 18
- . that perception is wrong, no one's interest is served, 19 not the licensee's, not the public's, not the NRC's if the l
l 20 NRC is not aware of events that cause public concern..The 21 public, the Congress, other federal agencies depend upon the 22 NRC to know what's going on and we can only hold the=
23 public's trust when we can address their concerns.
24 I'm always amazed at the irony of us getting calls
/
Y
(,/
25 about some. things and not about others.
We're told about l
l I
205 l;
u-~s 1
sea' turtles in traveling screens and cooling tower lights
(\\m-)
2 being off.
There was on licensee that had a steam generator
[
3 tilbe leak.
They thought it was below tech spec and they 4
held a press conference and they held a second press-l:
5 conference on the steam generator tube leak.
It happened to 6
be in a locality where people were very nervous about 7
nuclear power and we were getting calls in Washington being 8
asked what we knew about it and we knew absolutely nothing; 9
which, of course did nothing for our credibility with the j
10 public or the locals.
11 If you've got time to hold a press conference, I 12 would think the NRC should be called.
When we say as soon
~'\\
13 as practical in the regulations, what we mean is that your L[d 14 primary responsibility, of course, is to maintain the~
15 reactor in a safe condition.
We don't want notific'ation to 16 interfere with that.
That's why we allow an hour.
But that
.17 doesn't mean that you can hold press conferences and issue 18 press releases on steam generator tube leaks and then four 19 days later'say,' whoops, well, we did exceed tech specs and 20 there was a release to the environment and the calls you 21 were getting were' valid and we're sorry that you didn't know 22 about it.
23 Next slide, please.
24 Deficiencies are not always reported when found by
./~s 25 NRC personnel.
Sometimes our inspection teams or residents
}
256 1
find _ things other than the licensee, but the regulation
,r y -
2 doesn't make any distinction.
If it's reportable, it should 3
come in over the red phone.
It should come in under 50.72.
j l
4 The reason for that is because we have a complex set of J
5 responsibilities to execute in the NRC, including notifying 6
other federal agencies, getting our own complex organization 7
to look at precursors to events.
In order to fulfill those 8
responsibilities things have to happen in the proper way.
9 We need that red phone call.
- 10 I might add here, once an event is reportable, we 11 do expect that it be described completely, includir.g 12
-description of circumstances and events that we,uld not
. [ I, 13 otherwise be reportable.
That may include status of non-
\\ /
14 safety systems, such as if you're using a non-safety grade 15 pump to mitigate the course of an accident or if for some
'16 reason a bunch of non-safety grade equipment is not 17 available.
I might also say that 50.72 paragraph C requires 18 a_ call-back should you discover any additional 19 complications.
20 The evening before I flew down here there was a 21_
plant-that.had a reactor trip with complications and we did.
22 not receive the call-back despite the operations officer's 23 request for any additional-information should it develop.
24 It turned out that following that reactor trip some MSIVs
(()N 25 stuck partially open and an IRM failed down-scale and
257 1
there's some other things, and as a consequence, teams were
,.Y
[\\s )
2 dispatched from both the Region and headquarters.
3 Next slide please.
4 Required oral reports are sometimes made to other 5
NRC personnel other than the Operations Center.
As I 6
indicated before, in order for things to work right, we 7
really need the red phone call-and it's come to my attention 8
recently that some licensees have been using 50.9 as a 9
substitute for 50.72 and they really are not 10 interchangeable.
50.72 stands as it is and 50.9 is icing on 11 the cake.
If there is something that you're sure is 12 absolutely not reportable under 50.72, then 50.9 may be ET 13 operative.
But if it's covered by both regulations then we b
14 should get the red phone call under 50.72.
15 With those very-brief remarks, I think I'll end my 16 discussion and say thank you very much.
17 MR. ROSSI:
Let me add just one other thing.
I 18 think Eric covered it pretty well but I want to just re-19 emphasize it and that is that the Operations Center is set 20 up to handle the events.
They have procedures on who to 21-notify when they are called in and they have the equipment
'22 available to do those notifications correctly.
This, I 23 think, is particularly important on an event that is likely 24 to attract a lot of public attention, even though it may not
,r%
l., )
25 be highly safety significant.
l l-L
1 258 1
If there's an event that's going to attract a lot
-f-~\\v
(_/
2 of public attention, it's to your advantage to call that in 3
-to the.NRC Operations Center.
The reason is, if it's called 4
into the NRC Operations Center, we will screen it and 5
recognize generally that it's going to get a lot of public 6
attention.
What we will do -- and we're set up to do this I
7 right -- is, we will brief all of the managers and senior 8
people within the NRC on that event so they know the facts.
9 That will include the commissioner's assistants; it includes 10 the executive director of Operations Office; it includes our 11 Public Affairs and so forth.
We are set up to be able to do 12 that briefing in one briefing with the best available facts
(h 13
.that'we have at that time, so that if any of those people
\\
14 get a call from outside.the Agency, they will~have the best i
15
.available information on the significance of what happened.
16
.Obviously if the event isn't very significant but it's being 17 blown out of proportion by states or the press, it's even 18 probably more important that we have the right facts and 19 that our' senior management within the Agency has an 20 opportunity to know those facts.
t 21' MR. NOVAK:
Thank you, Ernie.
22 We are running a little bit.behind.
Actually the 23 dry-run only took half as long, so it must have been-the 24 number of questions.
O j.
25 (Laughter. )
Im m
m y
w Y-
259 1
MR. NOVAK:- Just one quick point.
When Ernie
,s.
)
(_,/
2 mentioned the weekly events, it's not exclusive to just our 3
reactors.- If we know about a foreign reactor event, that
-4 will also be brought up.
Obviously it didn't occur 5
necessarily the week before, but we do look at international 6
experience for lessons that we can learn.
So those are 1
7.
brought up as well.
8 We'll take our creak now and then we'll try to get 9
back on schedule.
We're going to mh!ntain the amount of 10 time tnat we need for q2estions though.
11 Thank you.
12 (A short recess was taken.)
jN
.13 14 NRC PANEL DISCUSSION - INDUSTRY FEEDBACK ON 50.72 REPORTING 15 MR. NOVAK:
We're going to try to get back on 16 schedule.
This is an opportunity now for some questions and 17 answers between us.
We've set enough' time, we hope, to get 18 a good portion of the questions out.
If there are more 19 questions than we can' handle, we'll pass out some three by 20 five cards and let you write them down and they'll get in 21 the record and we'll have a hance to consider them.
22 What we want to do now is just give you an 23 opportunity to ask questions and to bring up things that you 24 think are relevant with regard to 50.72 reporting.
It j
/~'N.
()
25 doesn't necessarily have to be a question.
If you have an
1 260
/~'
1 idea-about what should or can be done with 50.72 reporting, Q,h) 2 I think that's fair game as well.
This workshop is to try 3
to come away with the best understanding of how to report 4
events that need to be reported for the kind of work that i
5 we're in.
I think you can agree that we do try to wring out 6
50.72s.
When you heard Ernie's report, there's a wide l
7 understanding of the event.
It's discussed very openly and 8
we try to do the best we can with the information.
9 So let's start out now with any questions.
- Again, L
10
. when you do come up to one of the speakers, identify l
11 yourself so that the record will show who is asking the 12 question..
I
/
13 KR. CURTIS:
I'm Tom Curtis, Compliance Manager at j
s.
14 Oconee.
I have some comments in regard to Eric's 15
. discussion.
I think the industry agrees that NRC is 16 definitely vital in your role to be aware of events that are-l L
17 taking place at the facilities, particularly safety 18 significant events.
In fact, Duke Power and Oconee, a large l
19 fraction of the events that we report through the LER system j
i 1-20 are voluntary or courtesy type reports which aren't
'21 reportable by the regulations.
About 10 to 2'O percent of 22 the LIR type reports from Oconee are of that nature.
p 23 Eric's concerns about events which are not L
7" g 24 reported seems to me could be broken into two categories.
25 Reportable events which are not reported -- I think a lot of
I 261 s
1 those you mentioned were clearly in that category.
To me p;s
\\
2
'that is an enforcement issue and the NRC should take the 3
necessary action to ensure that the regulations are being 4
followed.
5 Events that are not reportable which the NRC will 6
be interested in -- and I think no matter what form the 7
regulations are in, there will always be events of this 8
type.
I think that those should be reported through the LER 9
system in a manner that I've just described where in the 10 cases in which the licensees feel that's appropriate.
But 11 in general, the role of the resident inspector seems to havo 12 been overlooked in that discussion.
It seems incredible to
('~)
13 me that a facility would have an impending emergency but the j
U 14 NRC would.be unaware of that situation with the resident 15 inspector program in place.
I know at Oconee all events of 16 safety significance -- really routine events that we think 17 the NRC might have some interest in, we're in full 18 discussion with our residents and we find that a very 19 helpful way to communicate with the NRC.
20 MR._NOVAK:
Thank you.
21 MR. EBNETER:
Could I comment on that, please?
I 22-appreciate your comments and you're absolutely right and I 23 do commend Oconee.
Oconee-is one of the stations that has 24 an excellent relationship with the resident and that
(N
( j' 25 communication channel is absolutely essential.
He commented
l l
l 262 l
(N 1
-- and that happens to be Pierce Skinner --
k\\~
2 MR. SKINNER:
And I didn't pay him to say that.
3 MR. EBNETER:
-- and he didn't pay him.
4
( Laughter. )
l l
5 MR. EBNETER:
The one that Eric mentioned, the l
6 spills; there's always a debate on whether a spill should be reported or not.
That's the type of event that you should 7
l l
8 always use a conservative judgement in whether you report or l'
9 not and Oconee does do a lot of voluntary reporting.
But l
10 even if they don't report it under the LER system, they go 11 and see_ Pierce and he knows what's going on at the station, 12 ar.d when he knows, we know.
Usually the media will find out A
\\
13 about it and we're all prepared.
We're all saying the same (d
14 thing.
We know what the circumstances are.
But the role of 15 the resident is extremely important in these issues that are l
16 debateable whether they should be reported or not and I 17 would encourage all'of you to work with.the residents to the 18 same degree that Oconee does.
It's essential.
19 Now, you may not report an event -- it may not be 20 reportable under the rules but we would expect you to follow 21 up on that and. provide the resident with the results of it
-22 and root-cause analysis which may be important to us.
We l
23 may want the results of your root-cause analysis to send to 24 Ernie Rossi or AEOD for their experience base.
So, I would i
4 i,j/
25 certainly encourage you -- and you're certainly right, s
l l
L 263 1
oconee does work very well with our resident.
s l
)
l=(j 2
MR. JORDAN:
I would like to maybe add a couple of 3
comments to that.
The role of the inspector is very 4
important.
The NRC does receive a lot of input.
I think as 5
Ernie and Eric indicated, the NRC's morning report or daily l~
6 report is not only what you report but what the regions send l
7 in that the residents have collected.
The residents are at 8
your sites.40 plus hours a week.
Somehow the events seem to 9
happen in the middle of the night when the resident is not L
10 there.
So you call him -- we call him when there's a 11 significant event.
But we don't rely, and shouldn't rely, 12 on the resident being there 24 hours2.777778e-4 days <br />0.00667 hours <br />3.968254e-5 weeks <br />9.132e-6 months <br /> a day, seven days a 13 week through the entire year.
So that's why there is the g
14 telephone notification and a way for us then to call in the 15 right people.
So, we urge you to make the call 16 conservatively.
17 I suspect too that we don't give the industry as 18 much credit as we should for the voluntary reports that 19 didn't quite -- weren't clearly required by the regulations
=20 but they were submitted and we certainly benefit from them.
21 We have a better memory for.the ones that were we feel 22 clearly required and clearly necessary that weren't reported 23 than we do for the ones that you voluntarily reported.
We 24 do appreciate those and that's where I said earlier, report jmi
- .(j 25 the important stuff and then help us revise the requirements
264 1
so that it does in fact fit more nearly what we all agree is
' mportant stuff.
i 2
3 I would kind of like to pose a question to you.
f 4
Do'you feel constrained because of an NRC or a media bean 5
count to have as few reportable. notifications as possible 6
because in our business we're trying to de-emphasize numbers
[
7 and emphasize quality and significance.
We hope that's 1_
l 8
getting'to you as well.
9 Thank you.
10 MR. BETHAY:
I'm Steve Bethay, the Manager of
-11 Licensing for-Plant Hatch.. Just to respond to your 12 statement, Mr. Jordan, people do use those numbers.
One of l
O 13 the things that we have to do is provide the quarterly l
14 performance indicators to the Georgia Public Service 15 Commission.
They don't' understand all of the events, so l
l.
16-they do count the beans.
It's important to us then that we 1
17
-report those things that are significant and those things 18 that aren't.
We need to jointly work for a way to reduce 19 those things because it.is used by outside agencies.
l l
20 The other question I think Mr. Rossi brought up 21 earlier was -- and this is not a quote, but the effect of 22 what you said was that sometimes it takes licensees a long time to determine if a condition is reportable.
That's 23 l-24 true.
Sometimes it's not immediately apparent that an event
/ 3 lj 25 meets a criteria, particularly unanalyzed conditions and w
w
=
--g w
-y
.+
I 265
/N 1
outside design basis type questions.
With the one and four y
'\\_sf 2
hour clock running for those types of things, how would you 3
give us guidance on what constitutes a long time?' What sort 4
of guidance could you give on that?
5 MR. ROSSI:
Well, I guess what I would say is, if g
L 6
there is a reasonable belief that it is likely to be outside 7
the design base or in an unanalyzed condition, that you 8
ought to report it as soon as you have-that reasonable 9
belief.
Certainly, you know, if you go a month before you 1
10 report something, that's not timely.
There was recently an l '.
L 11 event at one plant where they had some kind of problem with 12-the emergency diesel generators, some of the controls
-13 overheating and that was indeed -- eventually turned out to p
14 be really a common mode failure problem and it didn't get L
L 15 reported to us for two or three months.
It appeared from 16 the_ report that was finally given to us that they could have-l 17 come to the conclusion that it was likely to be either.
18 beyond the design basis or an unanalyzed condition much, L
L 19 much sooner than that.
So, I would say if -- you know, if l
20 there's a reasonable belief that it's likely to be, you 21-ought to tell us about it and then if you study it and 22 decide that it isn't, then you can come back and tell us i
23 that also.
24 MR. NOVAK:
Ernie, let me just ask, just for your
.I l
25 information, Eric Weiss -- all the ops officers report to i
1-
266 l
0
,r~Nf.
1 Eric.
Maybe Eric could give an example of what he felt was
/)
2 a' good dialogue where something happens within this one to 3
four-hour period.
There may be a couple of calls and-when
]
4 we look back at it, we said, that was really a good way of l
]
5 transmitting the information.
I think that's in the general 6
theme of the question.
7 (Speaking to Mr. Weiss)
Do you have any examples 1
8 that'would come to mind where you feel people came forward
-9 and described the event and then came back and again told l
10 you some more about it?
11 MR. WEISS:
Well nothing specifically comes to 12 mind.
As a matter of fact, I think in the vast majority of i
?
(
13 cases that's what happens.
The operations officers speak to
\\m_
14-people in the control room and they are forthcoming and if 1
15 they don't know the answer they get the shif t supervisor on l
. the phone who answers the question.
When'they find.out 16 17' three hours later what the cause of the scram was, they call 18 back and they mention any additional complications.
We l
l 19 didn't realize it at the time of the call but we had the l
>20 following additional problems with whatever..And when they l
l 21 make the initial call, they give a complete description, 22 including anything that's wrong with the balance of plant or l^
23 non-safety systems.
They tell you about the equipment' 24 that's out of service at the time of the event.
I think 25 that's more the routine.
Maybe I've given the wrong 1
L 267 1
impression.
But, I think, on the whole, it's really a rare 2
exception that we have one of these major problems go 3
unreported.
It's really only about a dozen a year but I'm 4
nevertheless mystified when it happens.
5 MR. ROSSI:
Yeah, we see a lot of reports that 6
come in on being outside the design basis or an unanalyzed 7
condition that come-in fairly promptly and then when they 8
analyze it, they come back and tell us that they really were 9
within -- you'know, it really wasn't a problem.
We do see a 10 lot of those.
So, I think a lot of licensees are indeed 11 doing exactly what we would like to see done and I think 12 that's -- from our standpoint, that's good because it gets 1
13 us involved early on.
In some cases, we can pursue things 14 with vendors if they look like they are generic.
It gives 15 us an opportunity to get information out to the rest of the 16 industry.
17 MR. LEXIC (ph):
I'm Chuck Lexic from Carolina 18' Power and Light.
I'm the Manager of Regulatory Compliance 19 at the Harris Nuclear Project.
My question concerns the 20-
~very last item in 10 CFR 50.72 concerning the establishment 21 of continuous communications with the operations control 12 2 center.
Basically, what are the ramifications if during an 23 event and the critical time is within the first hour or so, 24 the licensee has a limited number of resources on site, a 25 limited number of licensed operators in the control room and
268 y
1 a limited number of auxiliary operators; if continuous 2
communications are requested by the op con center, typically 3
the licensee will make an attempt to provide the most 4
knowledgeable individual possible to communicate with the 5
Commission during that period of time.
However, also, 6
during that period of time they need their experienced 7
people to do the things that need to be done up front.
In 8
the event the op con center says we would like to establish 9
continuous communication and the licensee makes the initial 10 report and makes a follow-up report and says, look, I've got
]
11
-- I would like to use this operator who is on the phone 12 right now and do something with him and I'll give you an A
13 update'as soon as I can and they hang up the phone and 14 continue on.
What's the potential ramifications of that 15 type of approach?
-l 16 MR. JORDAN:
I would-be pleased to answer that.
17 It's a question that has come up many times.
Our. view is 18 that the utility must be staffed to be able to provide that 19 continuous communication to us.
We will not accept speaking 20 to a guard at the plant during that time frame, but you may 1
21 pass,it to an equipment operator that will, in fact, j
22 maintain the open line and we can get back any prompt 23 changes in the status of the emergency.
It has to be a 24 person that is technical that can speak to us and keep us V
25 appraised of the condition becaust we're also in the front l
l
c 269
,-~3 1
end of an emergency trying to decide should we go into our -
\\_-
2
- what we call initial actuation.
Should we send people 3
from the regional office by fastest transportation possible 4
to the site?
Should we staff up our operations center now?
5 We're trying to make those decisions and the roll that's l
6 even perhaps more important that we both have to work better
{
7 with, I think, is communicating to the state and locals and l.
L 8
being able to agree and support the utility's 9
recommendations for any protective actions.
We've probably 10 done that less well in exercises because of non-play of the l
11
.NRC than we should.
We have to be sufficiently 12 knowledgeable so that we can say, yes, we understand that I
13 there is a steam generator tube leak; that there is an
[l A-s 14 ongoing release, that it is at very low levels and no 15 protective actions are. warranted at this time; or on the 16 other hand, there has been fuel damage and there is a need 17 for a close-in sheltering or a close-in initial evacuation.
18 We have to be able to support you with the state and locals 19 because the whole purpose of having a fast moving capability 20 is to advise about protective action.
So having a dead 21 phone is.not acceptable.
22 MR. LEXIC (ph):
I have another question, more or 23 less a statement.
Understanding the very nature of 911 24 calls and everyone having monitors nowadays, I think
/%
25 utilities -- and I concur with what Steve said earlier about y_,
l
l' l
270 l
L 1
being counted by the Georgia Public Utilities Commission.
, ~y
.\\,, y i
2 Whenever a 911 call goes out over the airwaves, it's picked L
i 3
up by people.
The news media.gets involved, even if it's' 4
something like an individual falling and breaking a leg or 5
something like that that has no direct application to the l
6 nuclear aspect of our business.
I think we would be much
.7 more inclined to call in those type of calls that aren't 8
required as a courtesy call if we weren't being counted by 9
other regulatory agencies and other oversight groups that 10 provide feedback to us via direct paths and also the media 11 indirectly.
12 MR. NOVAK:
Thank you.
[ I
13'-
MR. JORDAN:
And that is a matter of, I think,
)
14 great concern.
We have tried to develop performance 15 indicators in order to. assess significance of events, to 1^
l-16-sort the events and say among the events here's the set that 17 are worthy of count.
That is a public document, we issue it 18 quarterly; we send the utilities a copy of their cards,
'19 scorecards so to speak, in terms of that set of events.
20 It's a'similar system with four identical indicatorsothat i
21-the INPO performance indicators list.
We hope it is 22 beneficial to you and to the public in terms of having some 23 objective indication of significance of events.
So, my 24 suggestion would be to -- if you have a problem with a local
/'"
(
,Y[~
25 media or with a rate commission, perhaps that is something
271
y 1
you might identify to them in terms of your prrformance QY 1
2 compared to.the industry average and to indicate that the 3
Commission has published in the front of that quarterly 4
report the conditions and the limitations on that kind of 5
data that helped put it in context because we really are 6
talking about statistics, very small numbers at most plants.
7 The trends are important over long periods of time and over 8
collected data.
Certainly the industry average has been 9
shown to be improving over the past five years.
As I 10 mentioned earlier, that's all very useful information for l
11 the NRC and perhaps the utilities can take a little better 12 advantage of it to show that just the sheer numbers of y-~~
(5f 13 events-that you call in or are reported are somewhat 14 irrevelant.
It is really the significance of those events l
15 that is individually important.
Five events that have a l
l 16 risk associated with them -- and those are the only events 17 that occurred at a facility-during a year, is much more 18 important than 20 events that had essentially no safety 19 significance that were reported by another facility.
So, I 20 want to weigh very heavily the significance.
21 MS. TULLEY:
I'm Cindy Tulley and I work in 22 licensing for Plant Hatch.
I'm probably wearing two hats 23 today in that I also chair the BWR owner's group committee 24-on LERs which is attempting -- as many of you know, to 7-~
(
):
\\~~/
25 develop some further utility guidance in that area and we l
1 277 97 3 1
have presented some concepts in that area to members of t'te
\\
M' 2
staff that are here.
l-3 Two of the' things that I was hearing expressed my 4
Mr. Rossi and Mr. Weiss were the theme of timeliness in 5
terms of 72 reports and also in Mr. Weiss' presentation the 6
consistency of reporting.
I just wanted to note that I 7
think the two issues are somewhat linked in that part of our 8
problem -- and I can speak for Plant Hatch -- especially 9
when you're trying to determine whether you need to make a L
10 one-hour report under an unanalyzed condition or outside 11 design basis'is -- one, as Steve mentioned, the fact that it l
l 12 does take a lot of time a lot of times to try to understand l^
l [~h.
13 exactly what situation you're in.
But on the other side of r
V 14 the coin is lack of guidance that's out there in terms of i
15 what an outside design basis situation is or an unanalyzed 16 condition.
Th'at also flows in and causes us to take longer 17 than we would like, and I'm sure you would like, in terms of i
18 making a determination, especially with regard to the one-i 19 hour2.199074e-4 days <br />0.00528 hours <br />3.141534e-5 weeks <br />7.2295e-6 months <br /> report.
20 The other side of the coin, or the other concern.
l 21-in that area, is also -- and I'll speak more from the 22 committee's frame of reference on this -- is a number of 1
23
' utilities in the committee are very concerned when they do l
fN).
24 make a one-hour report that generally they are going to have t - (_,
25 phone calls from the region and the staff.
They are going
l U*
273 L
/~43.;
1 to have a whole lot of interest in that event.
So, they
}
l:
2 want to make sure when they make that one-hour phone call 3
that they are on top of it, number one.
And number two, 4
that all of the interest and the level of resources that are 5
going to be concentrated are warranted.
So those are some 6
'of the concerns.
~
7 I guess my question is, given these two areas of 1
8 concern, timeliness and consistency of reporting, does the 9
staff envision any specific efforts under 72 to try to put l
10 together more guidance?
11 MR. JORDAN:
Well, Cindy, from our previous 12 discussions, we would like to be cooperative with industry
. [%)
to establish joint guidance within the existing requirements
(
13-
\\, J 14 and we would be very receptive to that.
I think there is
]
15 enough experience now since the rule was issued that we L
l 16 ought to be able to identify the hard spots and come up with l
17 useful guidance we could.both agree on.
So, we're very 18 interested in that.
19' MR. EBNETER:
Could I comment on that a little l
20 bit?
The lack of guidance will never -- or any guidance we i.
21 provide is not going to substitute for good judgement.
You 22 heard earlier that somebody doesn't clascify the emergency 23 diesel as an ESF' system.
You don't have to be very smart to 24 realize you have to report a problem with a diesel whether
,-s 1
Y
\\s,/
25 it fits a certain classification or not.
That's just common
274 1
sense.
If you don't do it, it's guaranteed that we're not 2
going to be happy about it.
3 But with regard to the design basis, we're not in 4
a very good position with that right now, but most utilities 5
are working on design basis reconstitution things which 6
should help you.
But the utilities that are out in the 7
front today have good technical staff available to them on a 8
24 hour2.777778e-4 days <br />0.00667 hours <br />3.968254e-5 weeks <br />9.132e-6 months <br /> basis in the forms of systems engineers and tech 9
support departments who do provide pretty prompt response in 10 these areas.
If you do have a problem with that -- and I 11 think-the staff has t sn pretty reasonable in extending the 12 time, at least my staff has when it's reasonable.
13 I think Ern'.e Rossi hit on it before, if you have 14 a reasonable belief -- I think that's the way Ernie 15 characterized it, and he can elaborate on it.
If you have a 16 reasonable belief -- and that means you have some technical 17 data that supports it, then you should report it within the 18 one hour.
19 MR. WILLIAMS:
I just want to add one point to 20 what Stu mentioned-about design basis reconstitution.
I 21 think the NUMARC guidelines for design basis reconstitution 22 has been reviewed at least once by the staff.
In those 23 guidelines, it does mention specific as[.ects about reporting 24 and some of the topics that you discussed, Cindy.
So 25 pending those guidelines -- and we have looked at those
275 1
guidelines, there's information that will be available, I im
{
}
2 think, for outside design basis and those kinds of 3
engineering judgments that are of ten very tedious.
4 KR. CURTIS:
Tom Curtis with Duke Power at Oconee 5
again.
I feel compelled to make a couple of more comments.
6 In regard to the question about guidance, I think the most L
l 7
helpful guidance that's written for 10 CFR 50.73 is the 8
NUREG 10.22 document.
Because 50.72 generally parallels p
l 9
50.73 that guidance is very useful.
It might be possible to 10 expand that particular document to ensure that all the 11 aspects of 50.72 are also addressed.
12 I also have a comment in regard to the need for 13 consistency of reporting and the statements that have been 7s.
I i
x/'
14 made concerning reporting important stuff.
I feel that --
s l
15 when we make reports, LERs, red phone calls,.we make those l?
16 reports according to specific reporting criteria in the 17 regulations.
When we make voluntary reports, we use our own 18 judgement and also interact with the residents and other-NRC l'
19
. representatives.
Basically, it's a non-regulatory matter.
1:
20 Reporting is important stuff now, it's not a regulatory 21 requirement.
So, I think we have to be careful in the way l
[
22~
in which we say that we're using the required reporting 23 systems.
24 KR. JORDAN:
Okay, since I'm the one that keeps
[m}-
25 saying important stuff, I would respond to that first.
We w/
l
{
wi-,
-n-,
276
,s 1
wrote the rule to try to describe important stuf f and we
)
/
\\- /
2 became very legalistic in our writing and you in your 3
reading.
I think that's the mistake I'm trying to correct, 4
to say, hey, I'd like to improve the rules so that it's not 5
quite so legalistic and the basic purpose behind it is not 6
really obscured by detailed reporting requirements.
I think 7
there are two ways we can go in this whola thing.
We can 1
L 8
become more legalistic and we can add lots more conditions 9
and specific items to be reported or we can become less 10 legalistic and we can improve the guidance and the basic q
11 understanding and the purpose for having gone through 12 reporting.
I would like to do the latter rather than the l
l' 11 3 former.
You know, I really don't want to have to describe V
~
-14 in sufficient detail that we will be debating with our 15 lawyers over whether or not it was reportable, but that we l
16 would be debating as engineers whether or not it's important-17 from your standpoint in terms of safety of your specific 18 plant and from the standpoint of feeding back generic 19 information and-having the potential for digging out these 20 rather obscure lessons because they are buried in there.
21 The next event that happens that is upsetting to
-22 the industry and to.the NRC, we're going to find somewhere 23 in the database yours and ours, the precursor that would 24 have likely prevented that from recurring.
I think it's
.O,)
25 important that we do the very best we can to try to prevent
277 l
- f 3 1
that next event from occurring by using this kind of
-(\\~<') 1 2
information.
So, that's my clear basis.
3 MS. TULLEY:
I would just like te respond to Mr.
(.
Jordan's comment to say that what we're trying to do in the 4
5 owner's group committee, we share your concern greatly about 6
not trying to get excessively legalistic in terms of what j
7 we're putting together and argue over legal details.
We too 8
Want to report what the quote important stuff is.
I think l
[
9 what we constantly see is diversity of opinion on what the l
10 important stuff is.
Unfortunately, I think, one of the j
{
l, 11 things we tend to see sometimes though when we do ask for l
12 interpretations from the staff in terms of what's reportable
\\'
'l 13=
is -- is we're seeing interpretations come back -- and I 14-think the key area in which we've seen it lately is on the 15 SF actuations -- that do appear to be concentrating only on 16 the legal framework of the rule and not necessarily really lj 17 looking at what do you really want to hear about ESF's.
l 18 I call attention to Mr. Weiss' comment in his 1;
19 presentation -- I'm not sure if I heard it right, number 20 one, but I thought what I heard was, sie want to hear about i
21 when ESF's actuate in response to a plant transient or to 22 mitigate a plant transient.
If that was what I was hearing, 23 I like that.
I think a lot of people in the industry would 24 say that's very much -- that's the same thing we want to j_
25 report.
But I might note that we've gotten some other
l-1 278 n
1 interpretations from various people that have said that
(
)-
V 2
anytime anything in the ESF system actuates for whatever 3
reason, say you have a relay that fails and as a result of L
4 that, you have an ESF actuation.
Now, I can compound it 5
even further.
You go in and you try to fix that relay and 6
your fuse -- somehow you blow a fuse or something else and i
L 7
you get another ESF detuation.
There are some 8
interpretations from the staff that say you want to hear 9
about those things.
10 MR._ WEISS:
There is a specific exemption in 50.72 11 that is there.
Maybe not everything is exempted that we i
12 would like to have exempted, and you'll hear more about that t
13 later.
But the exemption refers to those things that are V
14 part of. pre-planned operation or maintenance activities.
So 15 if the piece of-equipment.is out for maintenance and part of 16
'the trouble shooting process is to start the diesel again, 17 then that's exempted.
As a matter of fact, the statements 18 of consideration for the rule, I believe, even mention some 19 scrams that -- I think the example we used were boilers have 20
.it in their procedure to speed a shutdown and it's part of 21 their pre-planned operational sequence to get down to 30 22 percent power or so and then to bypass the RSCS, they 23 perform a scram.
24 Well, that's one thing, but on the other hand, if 25 in tae course of this maintenance or this operation, you
279 1
have an actuation that you didn't anticipate, there may be a 9
2-lesson from it and that's why we want to hear about those 3
sorts of things.
You may be telling us a message that you 4
don't even realize.
I don't mean to trivialize the subject, 5
make it sound simple, but there are some specific exemptions 6
that were written into the rule and you'll be hearing more 7
about what we're doing to try and add others.
8-MR. ROSSI:
Whenever we're asked for guidance on 9
reporting -- and I'm familiar with the particular example 10 you just raised because I think I signed the letter giving 11 the guidance.
If you -- if we are asked for guidance and we 12 give it in writing, we're obliged to go back, and what we do 13 is, we'll look at the rule and we'll look at the guidance 14 that's already out there and we'll give you a written 15 response that reflects our best understanding of what the 16-guidance is that_we've already given.
And when we do that, 17 we.will gerarslly -- and we did in this case -- get the 18 concurrence cJ the offices that are involved in both the 19 50.73 and the 50.72.
Obviously we're going to be obliged to 20-give you -- I mean there's not much choice but to give you a 21 fairly legalistic answer because we have to follow our own-22 rules.
We can't give you an exemption by giving you 23 guidance that we think is giving you an exemption and on the 24 other hand, we can't go beyond the guidance that we already 25 think we've given, because if we do, we have a process to go
280 1
through on backfit.
2 The particular case that you discussed, the staff 3
researched as best they could what we had said in the NUREG 4
and we think m.:st people are doing that by the way.
We gave 5
you, you know, the guidance that we think follows the rule 6
and doesn't go beyond what we had already given.
We do also 7
recognize that there is probably room for actually refining 8
the rule.
I mean, you know, some of these things require 9
refinements to the rule and I think later on somebody is 10 going to talk about that, I believe.
Il MR. NOVAK Ernie, not intended to you, but to all 12 of us, I think we'll try to make our responses a little bit 13 more brief so we can get the questions out on the table.
14 Starting over there.
15 MR. REPKA:
My name is David Repka and I'm an 16 attorney with the firm of Winston and Strawn.
I just wanted 17 to make a comment.
I think this discussion of reporting 18 important stuff is an impcrtant philosophical point and I 19 think it brings up an issue of the interrelationship between 20 50.72 and 50.9.
Mr. Weiss earlier today said that 50.9 is 21 not a substitute for 50.72 and that's a statement that I 22 concur with.
However, the inverse or the converse of that 23 ir also true, 50.9 is not a substitute for -- or 50.72 is 24 not a substitute for~50.9.
What I get the sense here is, as 25 we talk about various things that are quote important -- and l
ig u
i i i
i 281 7'~x 1
I think Mr. Weiss gave a list of about eight or nine of
(
)
4' 2
those things that are important and should be reported.
It 3
puts the licensee in a bit of a dilemma if it does not fit 4
really into the narrow legalistic criteria of 50.72.
We 5
don't want to force them into 50.72 and I don't think the l
6 NRC would want to either for all of the reasons of bean 7
cointing and public exposure, et cetera.
I think a lot of 8
that important stuff should be reported perhaps under 50.9 9
and not under 50.72.
That's the way I think thn regula~ ion 10 should be structured, not to force these things into 50.72 11 unnecessarily.
12 A related point really is something to be e 'x
)
13 considered in the context of, you know, the future of the
'v 14 reporting requirements and perhaps rulemaking, is when you 15 evaluate the philosophy of the report?, requirement, the 16 important stuff concept really is already in the regulations 17 in 50.9 and, you know, maybe that's a more general approach 18 you want to take in lieu of the bean counting of 50.72.
19 That's e comment that I -- I don't think you want to force
)
20 more things into 50.72 than are already there.
l 21 KR. NOVAK:
Do you have a question?
22 MR. BETRAY:
I just have a comment for future 23 consideration as you look at providing more guidance or 24 possibly revising the rule.
I think one of the things that ps I
b
(_,/
25 would help us the most -- and certainly would help me the 1
i
282
~'s 1
most, as an engineer to an engineer, toll me why you want to 7
\\ -)
i
~
2 know, as you've done to some extent in this workshop and 3
others.
You know, if you would tell me why an ESF actuation 4
is important maybe with respect to an ESF failure; you know, l
5 Why -- hos is that information used, whnt are you going to 1
l 6
do with it.
That helps me to determine what the important 7
stuff is.
If you tell me why you want to know about outside 8
design basis, it helps me decide whether I need to tell you l
9 or not.
So, as you think about that, I would appreciate it 10 if you would look at it in that context of telling us why, 11 not just in a regulatory -- here are some words on a CFR I
12 page.
<-'s 13 MR. NOVAKt That's a good comment.
14 We have some time for a few more questions.
15 (Speaking to Mr. Tunstill)
Go ahead before you lose your 16 opportunity.
17 MR. TUNSTILL:
I'm Jack Tunstill with Florida 18 Power Corporation.
The issue that seems to trouble a lot of 19 us at Florida Power has to do with the outside design basis.
I 20 We seem to spend a considerable amount of time trying to 21 decide if something is, and again, we've got the Lime 22 limits, the one and the four hours.
We also have two other 23 groups kind of watching what we're doing that aren't here 24 and that's the Public Service Commission and the general
\\
(,/
25 public.
And what we're trying -- I think as a utility, is,
283
's 1
-- you know, we may be reporting more than is necessary but f
(
)
x-2 we're trying to stay away from that embarrassing question, 3
how come you didn't report it.
The regulations say that you 4
should.
In other words, you've got other people out there 5
interpreting the regulations besides the two of us and 6
you've got to respond to the public's comment when that 7
reporter asks your senior V.P. how come you didn't make this 8
report.
They can read that Code of Federal Regulations just l
9 as well as we can.
So, we may err on the side of just l
l 10 giving you too much stuff because we're exercising that 11 judgement you just told us about that we should do.
We're 12 trying to march to several different political leaders and
,m
/
13 influences on the outside.
L 14 Now, that kind of leads me into the question --
15 one of the dilemmas that we have is, we look back at 16 problems that have happ6ned in the past in enforcement 17 activities in things like environmental effects; a lot of 18 the things that we're running up against now have to do with 19 whether something is inside or outsido environmental limits 20 and that kind of runs against reporting.
Because right now, 21 Florida Power is reporting everything that effects 22 environmental qualification of equipment that's essentially 23 outside design basis under 50.72.
This results in giving 24 you minor reports that probably aren't safety significant.
k,,,l 25 I would like'to give you an example.
You've got a pressure
284 1
transmitter in the reactor building, that pressure
)
j 2
transmitter signal is sent to the reactor protection system 3
for input as a trip function and it's buffered out for 4
indication on the control board as opposed to accident 5
monitoring instrument.
If surveillance finds this 6
transmitter is inoperative, then a technical specification 7
action statement becomes applicabin but the event is not
)
8 reportable.
9 However, if we do an environmental inspection on l
l 10 that same transmitter and you find that the Raychem splice 11 is half of what the qualification report had for it, then 12 we're outside the design basis.
The deficient splice didn't j
G 13 keep the transmitter from operating and the worst case
./
\\
i l
14 failure is that the transmitter does not operate, which is I
l 15 right back to the technical specification action statement.
l 16 You know, here we are catch 22.
17 MR. ROSSI Well let me just say something about 18 the Raychem splice problem.
If you have a problem with the 19 Raychem splice environmental qualification that would effect l
1 20 the operability of that transmitter during an accident, q
l 21 that's fairly significant.
In that case, if you've got that 22 problem, probably a lot of other people in the industry have 23 the problem also.
So that's the reason we would want to 1
24 hear about that.
If a transmitter is inc7erable for another (m
25 reason and you go into the tech spec and it's inoperable for I
285 1
a short period of time because of some random failure,
,s l
)
(,/
2 that's not likely to affect -- that random failure is not 3
likely to affect other plants.
There's a real reason there 4
why we probably don't need to hear about that one but we do 5
need to hear about the other one.
6 I don't know whether anybody else up here wants to 7
augment what I said or not.
8 MR. REYES Ernie, can I add to that?
This is 9
Luis Reyes with Region II.
The other thing is that if the 10 other redundant transmitter in the other train was also on a 11-RayChem splice, you probably have a problem with both 12 trains.
If you had both trtnsmitters inoperable by tech j
(}
13 specs, you're probably going to have to take more severo L.)
14 action.
So there is other consideration about how wide 15 spread the deficiencies at the reactor.
16 MR. TUNSTILL:
okay, but at this point we're not 17 more than one hour beyond when we found it maybe and we 18 haven't had a chance to go out there because maybe there are 19 a lot of reasons we can't get to that other transmitter 20 right away.
We make a judgement.
It's reportable, okay, 21 and we don't feel like we should be criticized later for s
22 reporting it -- you know, it's not necessary that you -- you 23 should have told us about that.
We just ask for a little 24 bit of fairness -- I guess is the word here.
If we make a
^g
(
25 report and then we later retract it, that it not be looked
386 1
upon as a licensee has not exercised prudent judgement in
)
,/
2 doing something, especially when we've got these short time s
3 limits.
4 KR. ROSSI:
I don't think anybody would criticize 5
you -- at least that I know of, for reporting something that 6
later turns out to be not reportable.
Just make sure that it is 8
retracted.
I mean, we have -- we do compare 50.72s and 9
50.73s routinely.
We have a lot of 50.73s on safety system 10 failures or outside design basis for which there is no 50.72 11 when we look at those and conversely, we have 50.72s for 12 which there is no 50.73s.
So, in many of those cases there f'~*)s 13 was not a retraction.
There was not a second call made to
\\'~'
14 take the report back but we have the capability to do that.
15 MR. NOVAK:
Question?
16 MR. MOORE:
Brad Moore with Alabataa Power Company.
17 I'm the Licensing Manager for Plant Parley.
I heard wnat I 18 perceive as a contradiction and I just want to get a 19 clarification; and that was you say you receive about 3,000 20 calls per year, of which I heard 90 percent of them get 21 screened out as requiring no action.
With regard to Mr.
22 WelsL' comments about -- I thought I perceived you saying 23 that you wanted to get increased sensitivity toward calling 24 in things like spills or small fires, waterhammers, things (v) 25 which I wouldn't perceive as tripping into the four-hour
287 1
report criteria.
Do you really want us to report these
,ss
(
)
(_,/
2 types of things via four-hour report or use the LER or the 3
special report?
4 KR. WEISS:
Well, my point is that very often 5
these events are more significant than you would first 6
suspect.
You don't -- you choose not to call in this small 7
fire or this waterhammer and it turns out that you've 8
discovered a new mechanism for waterhammer that you were 9
unaware of.
It turns out that if under a given set of 10 circumstances this had occurreo st another plant or in a 11 different operating mode, you would have had a very severe 12 accident.
My purpose in giving the talk was to tell you
[ 'N 13 about those things that I'm surprised to hear about that N
14 aren't reported.
I don't wish to trivialize it but if you 15 have one of these events, I think -- my point is, you should 16 give careful consideration as to whether you have something l
17 that should be reported.
I personally led an AIT at a plant 18 that had a small waterhammer and it was a new mechanism and 1
19 even after we left the site there were more waterhammers 20 that were occurring there.
There was a debate even within 21 my team as to whether more waterhammers would occur, whether l
?. 2 what occurred in the opposite unit, the hanger damage was 23 caused by a waterhammer.
I can't change the regulations by 1
24 telling us this.
I just mean to sensitize you to the fact C's
()
25 that if one of these things occurs at your plant, you should l
l
288 7-s 1
give extra care as to whether you should report it or not.
\\ _,-
2 The statement of consideration for the rule says 3
that when in doubt, report it.
I do wish to concur with Mr.
l 4
Jordan's remarks.
We don't want to get legalistic but we 5
can't just be callus about these types of events.
6 Historically -- they've already shown us that we're 7
launching AIT's and IT's and special inspections for these 8
types of events and getting into arguments after the fact 0
whether they were reportable or not.
Rather than have that 10 happening, we came here to share with you what our L
11 experience has been; namely, these categories of things have 12 shown up as being very important and they are not be 13 reported.
14 MR. MOORE:
Well, I guess my point was that -- I 15 agree that a lot of these things need to be reported for r
16 generic implications but is really the red phone the proper 17 mechanism or, you know, if there's questionability about it, 18 wouldn't it be better to go ahead and do thorough research 19 on it and give an LER or special report in some cases 20 instead of picking up the red phone and making a report when 21 you may not have all of the facts you really need to know l
22 whether it has generic implications or not?
l 23 -
MR. JORDAN:
Yeah, I think what we want are both.
24 You make the preliminary notification on the red phone and k,)
25 while you're assessing the prompt significance, we're also l
l
289 1
assessing the prompt significance, and then make a 30 day
-p i
(s_,/
2 written report when you've completed your review.
We'll use 3
that 30 day report for our detailed review.
4 There was also a comment you made about the 5
numbers out of the 3,000 events that are reported that are 6
significant.
We do significant screening in a number of 7
different ways and as was discussed earlier, we have a 8
morning call that the events of the last 24 hours2.777778e-4 days <br />0.00667 hours <br />3.968254e-5 weeks <br />9.132e-6 months <br /> are 9
reviewed and screened for prompt significance and those are 10 followed up in real-time.
In many cases however, a longer 11 term but generic significance is determined from an event i
12 that happened one, two or five years ago by summing various
(N 13 events from the events data base.
So the information is 14 never lost and, in fact, is cumulative.
An event that l'
t 15 happened at your plant several years ago may, in conjunction 16 with other events, show a problem that we only today can eke 17 out by summing across them.
So the longer term studies of 18 events are, we think, beneficial both from INPO and NRC's 19 viewpoint in giving you back feadback that can prevent you l
20 from getting into a situation where a more complicated event 21 occurs.
22 MR. EBNETER:
If I can comment just quickly, Ed, l
23 and it's in your area.
We don't hear much about it but Ed l
24 Jordan's group does an annual study now on precursors.
-s v) 25 That's one of the most important studies we use and I can s
1
(
~
290
,/}
tell you that the senior managers and Dr. Murley almost 1
A' 2
always ask Ed where is that report and that comes from these 3
types of studies.
It's a very risk based approach and it's 4
valuable to us.
5 Just one other comment.
These problems that we're 6
having in mid-loop operations.
Many of these things were 7
exempt from reporting years ago and if we had known those 8
years ago, we would have probably been much better off to l
9 eliminate some of those earlier in the game.
10 Ed, that's all I wanted to say about precursors.
11 MR. NOVAKt I don't mean to cut off the questions 12 but I do want to keep us sort of on schedule and then we can (ys_)
come back and if there are questions on 50.72, we'll have 13 y
14 chances yet this afternoon.
15 Let's have one more question.
The gentleman has 16 been very patient and then we're going to go on with our 17 formal presentation.
18 MR. CONAR:
This should be a simple one.
My name 19 is Jim Conar from Florida Power and Light.
I'm the 20 Regulation Compliance Supervisor from Turkey Point.
We have 21 an ESF system that's used and -- some components in that 22 system are used for non-emergency situations.
CCW is a good 23 example.
We have a start sequence -- actually a start relay 24 that will start these pumps, that is bypassed during any EST
~
(-)
25 actuation, for instance, SI or something like that.
We
291 r^'-
1 periodically will get automatic starts of a component, one
(
\\--
2 CCW pump because of low discharge pressure.
That relay is 3
bypassed totally on an autostart of the diesels and the l
4 sequencing on of these pumps.
We have in the past been 5
reporting that ESF component start as an ESF actuation, even 6
though the start reason is not an ESF reason.
Are we off in 7
left field or is this what you want us to be doing?
8 (Pause.)
l 9
MR. NOVAK:
You see us all running to the 10 microphone, don't you?
11 (Laughter.)
12 MR. WILLIAMS:
We can go on with the fm l
)
12 presentations.
x
/
l 14 (Laughter.)
I think the answer to your question 16 has in part been answered by our discussions with the owners 17 group and our responses -- at least in one plant specific 18 case that Ernie discussed earlier.
We're getting into the 19 real details of what constitutes an actuation, not the ESF, 20 but what is an actuation of an ESF and dual function 21 components.
On each one of these cases, we've gone cases by 22 case.
If we get a generic question as to whether the 23 component alone starts without the ESP signal or you have 24 the signal without the component starting, in generic l
,)
25 questions, we have to give you generic answers.
So the
{.
292 1
generic answer might be, yes, you should report it.
7~
As/
2 Case by case we go through these things, j
3 especially with dual function components and we've decided 4
in giving guidance either verbally or written.
So that's 5
one of the thirgs that we would like to take up in this 6
workshop.
Jack Crooks will get to that later on this 7
afternoon.
We'll be issuing more guidance as a result of 8
this workshop.
So, if we can pin that down a little more 9
generically, we'll try to do that.
Sometimes it's just too 10 tough and you have to go case by case.
l 11 MR. NOVAK:
We will have chances for some more l
12 questions later on.
Right now, we would like to have our
)
13 presentation on 50.73 and to do that is Mark Williams, Chief (V
14 of Trends and Patterns Branch in AEOD.
Mark.
j 15 LER SYSTEM (10 CFR 50.73) 16 MR. WILLIAMS:
Okay, can we have the first slide, 17 Jack?
18 One of the things I wanted to start out with was a I
19 discussion briefly of the background as to how we got where 20 we are with the original rule.
I think we've talked about 21 it several times but we here today are part of a team.
Up j
22 at the table in AEOD, we wade through all of the information 23 that you send in.
The team was really set up by the TMI 24 action plan, Item 1E(6) of the action plan.
Specifically r;'s.
i Y
(,j the operating experience feedback programs that all of your 25
293 1
facilities, and our existence, were set up to avoid serious
,~}
'u '
2 reactor accidents in this country.
3 The work that we're doing, I think it's important 4
to keep a focus on the national goal and our national goal 5
is to avoid serious reactor accidents.
The programs in 6
recent years, I think, have had an obvious payoff, reactor 7
scrams are an example where the reports used in the system 8
that we've set up have reduced challenges to reactor safety 9
systems markedly in the last few years.
So, we argue about 10 what constitutes an ESF actuation or specifics about dual 11 function components, we think it's very important to keep l
12 our eye on the national goal and what our commitment is and l
,(
13 why we're doing the job we're doing.
v 14 Initially, the rule was set up in response to L
15 reports from the Regovin Commission and the Kemeny 16 Commission.
There was also a report by the ACRS and there 17 were GAO reports on the current reporting system post TMI.
18 I would like to read the -- some of the initial goals of the l
19 rule.
Some of the initial goals of the rulemaking were to 20 upgrade reporting to include all events.
It was eliminate 1
21 reporting of insignificant events, achieve reperting 22 consistency across licensees and include reporting on all 23 systems and components that may have safety significance.
I
,q think we've made a good try at that with the new rule.
I 24 25 think we're in the same place we were back then with the
294 l
7'~'3 1
same goals to upgrade the rule we have today.
The staff has j
\\
)
2 over the last couple of years been looking at the rule to 3
see what improvements we could make.
As a result of these 4
workshops, we intend to try to refine the existing rule and 5
we also plan to try to look at longer term revisions to the 6
rule and we encourage you to work with us.
7 Prior to 1984, I think we recognized the reports 8
that were coming in under LCO entries and set point drift 9
and some of the other things that were reported we not 10 significant in our eyes now and we did make some i
11 improvements in the rule but there is a lot more to come.
l l
12 I wanted to visit the IOERS system just because
/
\\
(
)
13 it's one of the benchmarks that seems to be coming back i
%J l
14 throughout historical reviews.
The 10ERS system was the 15 integrated operating experience reporting system and it was 16 first published in an advance notice of public rulemaking 17 back in 1980 and it was our first attempt to modify the 18 rule.
That system -- for those of us who weren't involved 19 with it then, had reporting of safety significant component 20 failures in kind of a tabular format.
Then it had safety 21 related events reported with an engineering evaluation that 22 described the complete sequence of events that occurred.
l 23 very much today, we still us the same kind of 24 reporting aspects.
We use component failures and we use p-~
\\s /
25 event level information.
Back then, INPO volunteered to 1
l
295 CN 1
manage the component level reporting and the staff wound up
\\
\\
2 with a rule that really focused on event level reporting and l
3 that's how we *"st to where we are today with 50.73 events l
4 and the system level of reporting.
The staff still does 5
rely on NPRDS for component level reporting very heavily.
l 6
Next slide, Jack.
l 7
50.73 really is our primary source for events 8
data.
The NPRDS is supplemental of that, of course, but the 9
LERs really get a very wide dissemination nationally and 10 internationally.
We use the LERs and the NRC -- and I'll go l
11 into that in some detail.
INPO obviously uses all of the 12 LERs.
The public; I think you're familiar with rx 13 organizations such as Public Citizen and others who use the 14 licensee reports.
Our contractors and the industry's l
15 contractors use them as really a primary source of L
l 16 information and so does the academic community.
17 We give these reports international use.
One of 18 the things that we do is, we issue reports to the Nuclear 19 Energy Agency in Paris.
We issue about 55 reports a year, 20 most of which are based on 50.73s, either grouped together 21 as generic evaluations or occasionally individually.
So 22 your events are making it to the international community, 23 post analysis by AEOD in many cases.
Most of the 24 information notices that we issue or some of the significant p
(
25 inspection findings that we have are issued to the
296 1
international community.
In turn, we get back their events.
(,~w)
N- '
2 So on AEOD studies -- and I hope most of you have seen some 3
AEOD studies by now, you'll find some international events.
4 They share a similar number of events with us each year and 5
the number that we get from the international community is 6
growing.
7 As far as our use of these events, we analyze the 8
LERs for individual and combined significance.
Mr. Ebneter 9
discussed the accident sequence precursor program.
This is l
l 10 one of the programs where we look for potentially combined
(
11 events.
We may see events that, if combined, increase the l
12 additional probability of core damage substantially.
We
'G
)
13 have set up a system -- some of you may be aware of the l
\\m /
14 sequence and coding search system.
We have two people from 15 Oak Ridge here at this meeting that are clesely involved 16 with SCSS.
17 Ont, of the things we do with your LERs is we break 18 each LER down into sequences and we code all of the 19 sequences and each LER into a computer searchable format and 20 we put them into a mainframe down at Oak Ridge.
So we can 21 trend and pattern across the industry or MS system 22 activations or whatever kinds of sequences that we want.
23 occasionally we do searches where we will take sequences 24 that actually didn't occur together at the same plant and
~;
i
\\
\\,/
25 the same event and look for the cumulative impact if they l
297
~3 1
had occurred at one plant.
That's some of the things that
(
)
(/
2 AEOD was chartered to look into in trending and patterning 3
events.
4 The SCS system has about 30,000 -- 33,000 LERs in 5
it today and it's available to you.
It's available on a 6
cost recovery basis from Oak Ridge.
So in generic studies 7
that we do, we inevitably turn to the SCSS as the first i
8 source of information.
We would suggest that it's available 9
to you and you can do the same thing for a very small 10 charge, just on-line time.
11 The threshold of the rule was originally set very l
12 low to capture all of the sequences.
One of the questions
['
13 is, why are you interested in only this particularly small L/
14 sequence; and one of the reason was, we had intended to go 15 across sequences at different units and look for serious l
l-16 events -- hypothesized serious events and look for actions 1
17 to mitigate those events.
We still do that today.
We still 18 look at the risk significance of combined events.
So that 19 is on area where you will see information coming from the 20 NRC.
You'll see that discussed in our reports.
l 21 Other NRC products that use SCSS is background.
22 The LER system is background that you see all of the time, 23 information notices bulletins, NUREG's aiid generic letters.
24 The last bullet on this slide is licensee
/m
,)
25 perspective.
This goes back to keeping our eye on the goal w
,w
298 o
1 of why we're reporting what we're reporting and kind of
[
J' V
2 rising above some of the detailed questions that we do in l
3 our day-to-day activities.
We have seen a change of 4
perspective in licensee reporting over the last few years.
5 One of the causes of this -- the primary cause, I think, is 6
using the licensee event reporting system for performance 7
feedback or performance monitoring and it raises a double 8
edge sword aspect of LER reporting.
9 We do look at the significance of events in our 10 performance indicator program.
We don't guarantee that the 11 public and all other do that but we encourage everyone to 12 look at the significance of the events.
{'
13 Some utilities -- many utilities, have goals on l
14 the numbers of LERs.
We do not eticourage goals on the l
15 number of LERs.
We do review LERs carefully -- and we'll 16 talk about some problems on attempts to minimize the number 17 of LERs that we've noticed, such as bunch reporting, putting 18 a number of events in one LER and so forth.
But in the j
19 spirit of the overall operating experience feedback and 20 making us of that to avoid serious sequences in the future, 21 we would not encourage goals on the number of LERs by 22 different utilities.
We would like to maintain an 23 engi'neering perspective on this and we continuously support 24 the view that if we maintain an engineering perspective and
(~
25 not really a legal perspective, we'll hit the goal that
r 299
, ~N 1
we're all trying to get to.
'v']
2 Next slide, Jack.
3 There is some guidance -- a couple of slides from 4
now, I'll talk about Supplement 3 to NUREG 10.22.
We would 5
like to issue a Supplement 3 to NUREG 10.22.
We want your 6
questions and answers to be answered in that supplement as 7
best as we can.
In the past, we have issued 10.22 3
Supplement 1 which has questions and answers.
That's really 9
a primary document.
That and a statement of consideration 10 of the rule which is in 10.22 itself to some extent.
We 11 always go back to those documents for interpretations to i
12 make sure we have some consistency on interpretations across l
j 13 licensees and acroso regions.
V 14 The primary source of feedback for you for l
t 15 reporting questions.
The first source is always the l
16 resident inspector in our view and then the region.
The 17 regions do call us, they call NRR, they call AEOD to get l
l 18 interpretations and we work directly with the regions and we l
19 work directly with the residents and that's in our day-to-20 day activities.
21 occasionally, we'll discuss an item or compliance 22 issue with a licensee on reporting, NRR will, generally my 23 branch will and we provide verbal feedback.
You're free to 24 call us up just to discuss 10.22 or any of the guidance
._\\
/
\\
)
25 documents or the statement of consideration rule or the
300
^f' 'N 1
intent of the rule as far as 50.73 goes.
('
2 AEOD and NRR have occasionally provided written 3
guidance.
The latest case that Cindy talked about here 4
earlier, there was written guidance provided.
That was 5
concurred in by all of the involved offices within the NRC.
6 So, when we provide guidance, it's coordinated among the 7
offices.' We'll provide guidance generically in writing or 8
on a plant specific basis.
There have been other cases l
9 where particular events were noted at a utility and the NRC 10 took the initiative to go ahead and write them a letter and 11 say that those events should have been reported.
12 One of the things we used to do that people might n\\
l (b miss today is the feedback of the quality of the LERs.
We 13 14 used to have an LER quality evaluation program.
Some of you 15 may remember it.
It was from 1985 through fiscal year 1987.
16 We took a sample of 15 LERs and numerically evaluated the 17 quality of the LERs; how well it described the causes of the 18 event, the sequence of the event, the failures that occurred 19 during the event; we evaluated the text, the abstract.
20 The next slide has a population of how the 21 industry came out based on about 70 plants.
It turned out 22 that there was a rapid improvement in LERs.
After the rule 23 was issued, there was a dramatic improvement in the quality 24 the second time we did this evaluation.
Some of you got n
[
)
V 25 this evaluation back attached to your SALP reports.
I guess
301 1
we would like to know whether this provided any useful
,s
(_j 2
infotsation to you.
We can take comments on that during *,he 3
panel sessions.
4 cak Ridge is here, they have some comments that 5
they've made to me on the quality of LERs.
I think in 6
general today, we're really pleased with the quality of the 7
LERs.
Just as this population shows, there's maybe ten 8
percent of the population still on the low end of the 9
spectrum in LER quality as there was then.
But in general, 10 the industry has improved over time and we're satisfied with 11 the quality of the LERs.
12 You may get coding questions from either our
/}
13 branch or from Oak Ridge contractors -- Oak Ridge National 14 Laboratory who maintains the sequence and coding search 15 system for us or from Idaho National Engineering Laboratory.
16 The reason they're calling you up is, they're trying to code 17 the information you're presenting in the LER into a sequence 18 of events.
So they need to know the function of each 19 component that's discussed in those events to accurately l
20 make the information available for the future for us and for.
21 you.
22 Briefly, the NRC use of LERs has been discussed.
23 One of the things that's coming up -- and is upon us right i
24 now, is plant aging and life extension, license renewal.
As g
/
\\
25 far as the licensee event report goes, it's probably -- if
( j y
y
302 1
not the most, it's one of the most important documents for
,x
-fv) 2 life extension.
I think it creates the historical record 3
for the industry in our country for all plant designs and we 4
are using that -- research and it's contractors are using it 5
heavily for plant life extension reviews.
We turn to it for 6
different approaches to plant life extension.
The 7
Commission is considering what level of detail is needed in 8
areas such as life extension.
The operating record created 9
by the LERs that you write are really one of the primary 10 sources that we'll be turning to for aging and life 11 extension reviews.
12 I think we discussed in detail the generic 13 communications and how the LERs are used.
They provide
, -s w-14 other events that we compare with the current events 15 received under 50.72 or we do technical studies.
A lot of 16 the information notices that you've received come from 17 technical studies of the LER database.
18 I want to spend a minute on the operating 19 experience feedback program in AEOD to answer one of the 20 questions we received.
How do you use the data?
If you 21 will tell me how you're using it, then I can better provide 22 the information you need.
That'll be on the next slide.
23 The last bullet on this chart is performance 24 assessment and monitoring.
This is an area that we have to f}
25 recognize and deal with.
That's why we've issued guidance V
~v
303 1
for our performance indicator program and that's why we're
?,, )
\\,/
2 very careful on our SALP reports and we're very careful in 3
how we use the information that we receive in LERs.
I think 4
we recognize the sensitivity in using LERs inappropriately 5
and I think you can rely on the staff or the management of 6
the staff to respond ta those kinds of problems and remedy 7
them.
8 Next slide, Jack.
l 9
This slide is in combination with the next picture 10 and it really has three components of how AEOD uses the 11 information.
We are constantly conducting the operating 12 experience feedback program.
It really has three elements, f"'}
13 events screening, events analysis and events feedback.
In 14 the next slide, I would like to talk about some of the 15 aspects of that program.
16 First of all, when your LER comes into AEOD it's 17 read by a relatively senior engineer or two and it also l
l 18 receives a management review.
And the management review and 19 the screening of that LER evaluates the significance of the 20 LER.
We categorize LERs with a significance rating of one 21 through four and we encode them into a computer system 22 that's within AEOD.
The next thing that happens to that LER l
23 is it's coded into all of our databases, sequence and coding 24 search system database and our databases at Idaho National CN (v) 25 Engineering Laboratory.
304 1
50.72 -- in the morning phone call, Ernie
(
2 discussed how 50.72s are reviewed.
Any one of those 50.72
%./
3 events could trigger an AEOD study.
One hundred percent of 4
the time, whenever one of those events does trigger a study, 5
the LER datab. e is interrogated.
It's interrogated for 6
similar events and we try to find the generic significance 7
and other related events.
In addition to that, LERs coming 8
in by themselves in many cases, probably the majority of 9
cases, trigger a generic study.
Then we turn to the SCSS 10 and find related events.
11 Once the study is initiated -- this picture down 12 on the icwer left shows you the different kinds of databases
<-'s 13 that we turn to.
For the components involved, we turn to 5\\~ ')
14 the NPRDS, INPo's databased for component level reporting.
l 15 That is really an integral part of the current reporting 16 requirements, although it's not required.
It's a voluntary l
17 system, but the staff has formally evaluated that.
When the j'
18 NRC agreed to go with the event level reporting and the 19 system level reporting, the Commission directed the staff to 20 evaluate the NPRDS semi-annually and report to the 21 ~
Commission.
Now we're doing that annually but we've issued 22 13 formal reports to the Commission on the status and 23 qualify of the NPRDS.
So the staff does pay a lot of 24 attention to the NPRDS.
We look at it in the maintenance
/~'v
(
)
25 team inspections and in a lot of other areas.
We consider v
m
~
-9
,-...y
303 1
it something the NRC uses and relies on.
/
)
(,/
2 We also 1cok at something you may not be aware of 3
called the generic communications index.
The GCI is an 4
index -- a computer searchable index of all the information 5
notices and all the bulletins and all the generic letters 6
that the NRC has issued for some years and it's pretty 1
7 complete.
We look back through all of those notices and 8
bulletins to see if problems are recurring; to see if a 9
situation has crept up that we thought was resolved and we 10 address that issue in the study.
11 We also look to our foreign events databases.
We 12 receive information -- as I mentioned, from the
(' N 13 International Reporting System and from the Nuclear Energy i
l 14 Agency.
We also receive information on bilateral agreements
\\
i 15 with other countries and that's in our foreign events 16 database.
That's the FEF that we have there.
It will 17 include events of this nature.
A recent one that comes to 18 mind-is the intersystem LOCA event at one foreign plant that 19 caused us a lot of study.
1 20 We also make plant visits, we look at INPO SERs 21 and SOERs.
We conduct interviews.
Then if an issue is l
l 22 identified, we write a report that has recommendations.
l 23 Generally, we have -* commendations or suggestions and you'll i
24 see these reports as AEOD case studies or trends and p
/(v) patterns reports or engineering evaluations.
25
306 1
If it's a major report, a case study or a trends a
1
(',/
2 in patterns report, the utilities involved will see it, INPO 3
will see it, NSAC or EPRI will see it.
The NSSS vendors or 4
equipment vendors for the solenoid operated valve report, 5
ASCO saw it, and they'll provide peer review comments.
So 6
any of our major reports go back to you and the major 7
vendors for peer review.
Then ultimately, if we come up 8
with solid recommendations, we may present the report to the 9
ACRS and in some cases the Commission.
There was a report a 10 couple of years ago on new plant experience that we 11 presented to the Commission and then sent out to new plants 12 for start-up consideration.
Or you might see an information 73 13 notice, a bulletin or another NRC document issued on it.
ki 14 A show of hands of anybody who has seen an AEOD 15 report.
16 (A show of hands.)
17 MR. WILLIAMS:
That's good.
That's over half.
18 Next slide, Jack.
19 One of the things we va.nted to talk about here is 20 current issues.
I think current issues have come already in 21 some of the questions.
One of the issues is missing 22 reports.
The missing reports that.we see most of the time 23 involve engineering judgement.
And one of the biggest areas 24 that we see is the potential common mode failure report.
An
'(
)
25 example might be where several relief valves individually
1 1
307 1
failed a test with set points that are outside tolerance but 2
the common mode failure aspect of that event is never 3
recognized or reported to the NRC.
4 There are some other potential common mode failure 5
events.
One of the aspects of 50.73 is, if there is a 6
potential common mode failure, it has to be reported.
It's 7
not one that has to have occurred, but it's engineering 8
judgement that shows there is a potential for a common mode i
9 failure in moro than one train of safety system.
10 We've seen some liberal rather than conservative 11 judgments.
Again, this is one area we focus on.
It's not a 12 legalistic area.
It's an area where we think the operating 13 experience that has led us to avoid possibly some accidents 14 over the past ten years has paid off in this area of 15 potential common mode failure.
These are the ones that 16 we're interested in.
1 17 Some significant events are not being reported.
18 We see significant events on the daily reports that we get 19.
from our regions.
In the perfot
,ee indicator program, we 20 have a lot of significant events that do not have 50.73s 21 associated with them.
So we look at those and sometimes we 22 may contact the utility to see what the case was.
23 We see differences on the interpretations of the 3,
24 requirements.
Some -- Ernie mentioned -- with scram 25 breakers with the plant shut down but in general equipment
i I,
308 j
1 that doesn't. operate but it's not required to operate.
In 2:
other words, the plant is otherwise in compliance with the
-l 3-tech specs of-the LCo's and a condition exists where it is-
{
s 4-not in compliance with the design and it would be reportable 5
if the' plant wes Meerating but since the plant is shut down 6
it's not reportaole.
50.73 really doesn't rect,nize mode of 7
the plant.
That 's one of the things you might work on and
'8 the guidance,, we might need some rule changes for it.
9 We see widespread differences on-what constitutes
{
10 an ESF actuation.
A little bit on what constitutes an ESF.
i 11 Certainly a lot on what constitutes an actuation of an ESF.
12 Some of the-other areas where we don't have 13' missing reports.
Well, one I should mention that is
-y i
j
-14 especially important is where the NRC knows about the 4
'15 report.
In many cases, if it's in an inspection report and 1
16-a' notice of' violation or even if thera was a major I
~ 1'7
-inspection, an AIT say, or if_the resident inspector found
- l q
E18 the problem, Lit's viewed that it's not reportable to the NRC
- 19:
in an LER and we just have to emphasize that those kinds-of.
20 events really are reportable.
Because we're part of a team l
21'
' working with you to capture that operating experience and-22:
_unless we get the paperwork, none of that ever goes in to 4
123 the operating record of the historical databases and it's 24-not available to the other plant who may need it'and may get q
25 the results of a generic study that would help then avoid an
'tt.
l 309 1
event.
So please report those.
9 '.
2 We see -- although not missing reports, some 1
3 problems with bunched reporting and then we have an issue on 4
voluntary versus required reporting.
Most of the reports we i
5 see as voluntary we find are required reports.
That's
[
l 6
something we might deal with in the guidance also.
7 We are receiving some reports that we consider of 8
low safety significance.
I would just like not to over-E 9
focus on the reports of low safety significance.
We are really doing this job since the action plan required AEOD to 10 L
=
i 11 be in existence and the OE feedback programs to be in T
12, existence.
We're doing this job to get the events of safety l
l 13 significance.
So let's not focus too much of our time on i
14 events that are not of safety significance.
We'll provide 15' guidance, we might modify the rule and we'll do our best to c
16-eliminate the ones that we believe should be eliminated and 17-I think we can come to agreement on those.
18' Again, in this area, I wanted to mention that 19 there are some things we believe should be eliminated from a
.20 the current-reporting.
There are some. things we believe 21 should be added and Jack will talk about those this
'22 Lafternoon.
But the' threshold was deliberately set low.
It 23 vasn't accidental.
It was set to capture sequences of
-24 y
events, not necessary an entire significant event.
25 Our improvement approach is really to try to do
310 1
some short-term work and then some long-term work.
The
,m
.k,.
2-short-term work, we're talking -- Ed will try-to' pen me down 3!
here, a series of months.. We would like to get out some 4
guidance for the existing rule.
We would like to possibly
]
5 do something about a couple of reporting aspects that Ernie 6
mentioned earlier; namely control room HVAC actuations and
+
7 RWCU isolations.
We think we can deal with those in short 8
order and it won't be too much trouble.
If we do need a 9
rule changed to deal with those, then we might sweep some 1
10 other things into it but we'll certainly be able to get the l
11-short-term guidance out.
12 In the afternoon session -- in the aftsrnoon panel-1 13-session, I think what we would like to do would be to focus fs e-
):
'\\'
14' on the short-term and the long-term approach for 15' improvement.
Jack will talk about some past staff O
16; initiatives and then if we could get your comments on what 17-
-you think we could solve with guidance and what long-term 18
, improvements do you'think would be needed in the LER rule, 19 recognizing that'would involve a rule change, which is a.
201 Ecouple of years at a minimum.
21 So with that,' that ends our presentation on 50.73 and I'll turn it bac'2 over to Tom.
'22
- 23 MR. NOVAK:
Thank you, Mark.
24.
According to'my watch, we're about ten minutes
'l 4
- (A) 25-
'before twelve.or so.
Why don't we try to restart it at-y
311 1
about 1:15 and I think we'll be pretty much on schedule.
pw L '(
f 2-Again, we're: going to pick up just as the agenda suggests.
. ' ~ <
3 We'll have a panel discussion and try to be responsive, to 4
your questions.
1 L
5 One final -- just a point.
You may not all be i
L 6
familiar.with the accident sequence precursor program.
7 Actually it's a program that started out more than ten years 8
ago.
We have supported it, it was work that was done at oak 9
Ridge., It's changed in some ways but we believe now by 10 looking at events and ranking them in significance, not 11 worrying about.the absolute magnitude of the core damage 12-probability, you get some idea of the kinds of evente that i
13-seem to tua of most concern.
And what I recall over the last fx
(
)
\\~ /
14
.two years, I'think, if you looked at the most significant
'15 events, five'out of seven.or four out of seven were due to 16 common mode failure and I think that's a point'that's. worth 17 picking up.
That our operating experience is suggesting 18 that: common mode failures are real, they happen at our l'
19 plants and we have to be cognizant of it.
So I think our s
20 concern about making sure we recognize common mode failure 21-is founded on those kinds of analysis, i
i 22:
All right, we'll see you at 1:15 and have a good 1
23 lunch.
'24' (Whereupon a luncheon recess was taken at 11:55
,K i
. r%I
- (
25 a.m., the workshop to resume at 1:18 p.m., the same day. )-
V.
l
312-l 1
AFTERNOON SESSION p-r r
' (,,T 2
MR. NOVAK:. Let's get started again. We can't 3.
wait for everybody, but we have enough to get started.
4 We are' going to make one small change in the order 5
of presentation.
I think it will provide for a more 6
meaningful panel discussion. I'm going to ask Jack Crooks to 7
give his brief discussion since a number of points that he 8
was going to cover have been touched on in earlier 9
presentations on what we are doing in current rule making 1
P 10 and guidance revisions.
Jack is the Chief of the Data 11 Management Section in the Office of Analysis and Evaluation
-12 of Operational Data.
So, Jack, if you don't mind, why don't I
('~
13.
you getLatarted.-
14' CURRENT RULEMAKING/ GUIDANCE REVISION s
11 5 MR. CROOKS:
Good afternoon.
I notice I've run 16' into a little problem here.
I either need to take my 17 glasses off to look at my notes and not see.the audience --
18 and I thinkIthat's the way I'm going to have to.do it
'19 because I guess I'm reaching the bifocal stage.
20 I'm here to talk about rulemaking and possible 21 guidance revisions that we have under consideration.
There 22 are a few general comments that I need to'make first.
What 23-we are going to do is present three slides that will give t
24 you some background information on what currently is being O
.t 25 reported.
313 1
L 1
Raji, If you would put the first slide up.
)-
1(_,/ -
2 Now, a couple of points that I want to make is --
3 again, we're talking about -- we're 'ot counting things, but n
4 these'obviously have some implication of that, but I think 5
it's needed to give everyone a broad overview in this pie
[
6 chart format of what has recently been reported in 1989.
l' L
7 This information can also be found in the AEOD annual report H
1 L
8 and there are some other supplemental comments on these l-
[
9 areas in that report, h
10 We are all, aware, too, that what's being presented 11 is not based on safety significance of events on these 12 particular pie charts.
Okay, what we're -- and the other Tr~'N 13 thing we are doing is we are presenting an agglomeration of 14~
things-that involve transient events, plant conditions that 15-aren't necessarily related:to an event.
There could be 16 potential problems presented through design basis p
17 documentation review or through design reviews-.
It'could be L
18 s'ituations that are identified as a result of studies of p
~19-events.
L 20-So, keeping that in mind, we'll move on, and-what p
21 I. plan to show.in the three slides'is just a pie chart on 22~
the reporting by basic category of reports under 50.73.
l 23 This will strictly be the LER information.
Ie
.1 cover 1',
another pie chart that shows a rough breakdown of the ESF L
24
\\
py-
-(
):
25 actuation reports being receliad by general area.
v i
I i
4 v
-e' W
w?
'w wr 314 1
I should mention that this is an area that's been
,c y,
{ )--
2 previously mentioned where there are potential changes being-
~-
3-considered.
We -- well, I'll get into the details later.
4 Then the other thing I will addr64s is just the general pie b'
5 chart on what the broad areas are.under tech spec violation 6
-reporting.
L 7
Raji, if I can have the next slide.
8 okay, as you can see from the pie chart in your v<
i 9
handout, that in 1989 about 80 percent of what's been 10 reported has come under two general reporting criteria.
The f
L 11 tech spec violation criteria 50.73(a) (2) (i) and the ESF r
12 actuation criteria. They both brought in about 39 percent.
L f-s 13 The other areas, common mode failure, single events are 5\\ /
14 conditions that could lead-to potential system failures, and 15 unanalyzed conditions are also highlighted there as three, 16 ten and nine percent.
And these are the areas that --
17'
'again, as has been~ mentioned earlier, that involve a fair o
18 amount of engineering judgment.
This is an area where, 19 again, we hope in our guidance.to try and' clarify what we 20 were looking for in the original rules.-
21-Raji, go to the next slide.
22:
This'shows a rough cut of the systems.
Generally 23:
it had been reported'under the ESF actuation reporting.
24l We're currently focusing on two areas, the HVAC and RWCU ac 2-s.
[
i 25-reporting which accounts for almost 40 percent of what's
\\m /
i 315 1
being reported in this area. Now, the information that you
. ?,-4 -Y
(_,,/'
2 are looking at too, I should mention, is based on the total 3
number of actuations rather than the number of LERs.
It's 4
like 1.7 actuations on average per LER.
You know, there may 5
be five in one.
There may be an individual reactor water 6
clean up system event that was reported, so that's the only 7
thing in the particular LER.
8 We have looked at when these types of things are 9
happening, and based on that look, we find that between 50
+
10 and 60 percent of the ESF actuations are occurring during 11 operations, but the trend over the last few years has been 12 down in this area.
About 30 percent of what's being
(}
13 reported is occurring during testing and the remainder
()
14 during maintenance.
15 There was some discussion earlier of emergency 16 diesel power, ESF actuations and an inconsistency in 17 reporting there.
We think that influences somewhat the i
18.
percentage -- well, influences the percentage of what's
-19 being reported.
So we know that's probably on the low side.
120 In tais particular area, again, as was mentioned earlier, 21 there is staff guidance that was provided through Region III 22 by letter to one cf the licensees, and if people are 23 interested in that, why, that can.tua made. available to As Ernie mentioned, staff used all the background 24
-others.
i V'
(
j 25 for the rule and put forth a position based on that look.
l
316 L
1
.Go the'next slide.
- c;,
2' This particular pie chart just shows you, again, a 3
broad breakdown of the areas of what's being reported under 4
-the technical specification violation area, and it just 5
simply shows that about three-quarters of what's coming in
'l 6
involved exceeding limiting conditions for operations. The l:
7.
majority of the others involved surveillance -- surveillance 1
I.-j 8
testing results, and a few percent involved variances from 1
9-the administrative section, section 6 of the technical y
1.
10-specifications.
These are things such as high radiation 1.
11.
area' doors being left open and some other administrative i
12 requirements.
'1
. /
'Y 13 Now, in this particular area too, everyone is T
)~
31 14 aware that the technical specification improvement program l
j.
15 is underway and that there are considerations being_given in 16-this program to changing some of the LCO requirements and 17 also to moving.some of the surveillance tests out of-the jj 18.
tech specs.
So, we know that there will be an impact on LER l\\
19 reporting from these particular efforts.
~
l' 20
. Raji, will you put up the next one.
l-
. 21-okay, a couple of points before I get into the i
i_
22-
' staff initiatives here.
We did focus. earlier this morning
'23 and'some.of the discussion was based on.a reporting of H
24' individual events that are of low significance.
What we j fN,
- N 25 have found is that this information is being used to l'
L l
317 l'
supplement some studies -- at least some of this information
-s-l(_,/
2 is being used to supplement. case studies.
For example, some g
1 i
3 of it was used in the emergency service water study, air 1
4 system studies, and solenoid valve report.
The problem that J
5 I see is where do we draw the line?. We all would ideally i
L L
6-like to have some fine lines.
Unfortunately, I think-the --
L
'7 there are no fine -- there are some fine lines, but E
8 generally we are talking about bands in which we are making I
9 judgment.and our-guidance changes will be to try and narrow 10 that-band so we will be more in concert-with the judgments
-11 being made'.
g 112
'Okay, getting onto past staff initiatives.
I 13-think we've already covered -- Mark covered NUREG 10.22 and 14l its supplements,'so I-guess'I would just like'to re-l 15
. emphasize that chey'still are the governing documents.
He 16 talked -- you know, 10.22 has the background from the rule, 17
.from the' Federal Register notices and.it provides examples.
18 Again, you~are all familiar with 10.22 Supp'lement 1,'which i
19-pretty much packaged'theLquestions and answers'from previous
'20 seminars..And~I'think in that regard, one thing we will try 721 Jand do this time, based on-these'four workshops, is get the
- i 22 essence of the, concerns and the questions and.the specific
'23 ;
concerns on specific criteria and try and address the 24-issues. - There is --'with supplement 1, there-are-some
)7
( y.y
)
25 specific answers given to specific questions that a utility
?
w.
-. -... + -.
'1 318 1
asks and the tendency is, we're trying to make broader use jj -e 2
of those in the original intent.
So, I think this time we l
3 will try and focus more on the general issue.
l 4
Supplement 2; I'm not sure how many of you are I-5 aware of that, but that looked at the first year of 6
operation and evaluated the reporting and the reporting 7
content and then gave some recommendations on how things 8
might be improved.
Again, Mark mentioned that, and it 9
showed how the quality had improved over the succeeding 10 years.
g.
1 11 The first area of real past staff initiatives to 12-try and implement soma changes occurred in the 1987/1988
. /T 13 time frame.
The staff found that some of the information
)
A
)
'~'
14 that was not coming in under the new rule, but had come in j
15 under the previous rule was being used in PRA and in some 16 other. studies,-and some of the information was not coming in 17.
'through NPRDS.
So what.we did was try and propose the 1
l l
J 18
.following; we had proposed a section under Part 50 to bring-L E
'19 '
in train level. reporting -- train level unavailability 20 reporting at the safety -- for the safety systems. Now, this 21-
--- what we-envisioned was a tabular monthly report format.
22-We would have been requesting information on the system and
-23 the subsystem, the component, the cause, any corrective 24 actions and the durations sf the unavailability.
Now, this f.
/s
(,
25 would'have brought in the information for all causes,
319 L
l 1_
preventative maintenance, testing, corrective maintenance,
-s p
s._f 2
equipment failure, personnel error, and the intent of this I.
3 was to. kind of fill the gap and we think provide useful l
4 information for everyone. Particularly it would have 5
provided information for identifying risk significant 6
equipment in some of the safety systems.
7 Now, as part of the rule making, we also had under 1
8 consideration some adjustments to 50.73, particularly in the 9
ESF area.
What I have identified here is we were 10
- considering exempting the reporting of failures when the --
11 not failures, excuse me, ESF actuations when the systems 12 were not required to be operable.
We also considered
/~'}
13 exempting some of the unneedeu actuations of ventilation in
- x/
i 14 certain ventilation and isolation systems.
Now, there were 15 a number of reasons -- these proposals didn't get formally 16 promulgated at-that particular time for a number of reasons.
17 This was one of the questions that we had posed to us before
-18
- the meeting as to what we had' considered in-the past. So 19 that's why we are presenting it.
-20 Raji, if we can go to the next slide.
21 okay, now as far as-current staff initiatives are 22-concerned, as I mentioned, thy big difficulty is where do we 23 draw the lines, or how do we narrow the bands so that we can 24 meet the overall needs of safety as well as the other needs f%
( )
25 that are involved in the reporting arena. Now, we've had
l l'
320 1
-discussions within AEOD.
We've had discussions with other f~s
/
Y
[ i,j/
2
-NRC staff in the regions and with some of the owner's groups s
3 in trying to get these things off of the ground, our 4
original -- our focus right now in the short term has been L
5 trying to find a mechanism where we can eliminate the I
i 6
unneeded reactor water clean-up system isolations and some i
7 of the control room emergency isolation actuation reporting.
.8 A number of these just don't seem to be fitting into any 9
particular area of study, although you can't say that again i
10
'for all. So what we are trying to is draw the line.
11 We are looking at pe asible administrative rule 1
12-change or possible use of exemptions, and I think that was
[N 13 mentioned previously in an-NUS seminar in June.
In the
(
).
'~'
' 14.
interim, I might mention though, as:the. region mentioned, i
15 why'we.all need to comply with what's existing now.
We i
16 lalso have underway a supplement.3 to NUREG 10.22, and as has 17 been mentioned a couple of times, we plan on bringing in the
'18
' input from you people in this series of workshops, as well' t
19 as factoring in the experience that we've gained in
,i 20.
intoracting'with you over the last five years on individual 21 cases.
22
.Now, the goals for this document are primarily to 23 improvo the consistency of interpretation for what the i
--f _
24 requirements are.
They also would be to clarify the f s.
.( ).
25-
' report.ing to bring in the types of things that we feel are e
~
321 J
1 meant to be. reported and are covered by the current rules
,_q
/
\\
)
i
.r
_f that there may not be general agreement on.
Some of them --
2 3
particularly things like Eric Weiss was mentioning.
4 Long-term, what we have under consideration is 5
more of a systematic look, standing back and saying, "Okay, 6
enough time has past.
We have seen tra change in 7
performance within the industry, and we just need to start 8
taking a look to see, you know, what changes should we'be 9
. making in LER reporting," and this would more than likely 10 involve a rule change.
So that's a longer term type of l
11 effort that is at least under consideration.
12 That concludes what I have to say, Tom.
- (~y 13 i
s
/-
^ ' ' ' '
14 NRC PANEL DISCUSSION - INDUSTRY FEEDBACK ON 50.73 REPORTING 15 MR.LNOVAK:
Thank you, Jack.
16 What we would like to do.now is just go into our 17 general panel discussion on 50.73 reporting and when we've 18
'had'enough time, then we'll take a break.and then we'll get 19 back to the safe guards issue.
-20 So, I think this is a good time to bring up 21-anything that was left over from the 50.72 discussions, as 22 well as anything you've heard on LER reporting.
And-again, 23 just' find the microphone and introduce yourself and ask your 24 question or provide any other information you think is
'in #
'5 relevant.
2
322 1-MR. TUNSTILL:
Jack Tunstill with Florida Power s L
. \\s,,f 2
Corporation.-Did I understand you to say that the results of 3'
this workshop and the other three that will be held will be 4
incorporated into supplement 3 to 10.22?
l 5
MR. WILLIAMS:.It was our intent to try and 6
capture some of the critical questions or at least some of 7
the ones that could lend themselves to a generic answer from-8 this workshop.
We've also got a lot of other things we've 9
accumulated over the past two years, and there's some on ESF l
10 actuations.
There's some on other issues, and we may even l
11 refer' to other documents, depending on the status of those 12' documents.
F iN 13 Let me also take the opportunity to mention _that 4
Y 14-there were_ cards given out, I think. So if anyone doesn't 15-want to get to a microphone and ask a particular question, 16 you can give it to us or send it to us, and we'll entertain I
17
.it that way, s
18 Were there cards given out?
19 MR. CROOKS:
Yeah.
120 MR. BETHAY:
I'm Steve Bethay with Georgia Power 21 again and I.have several questions, and.I'll talk until it 22 looks like somebody else wants to get them.
23 The first one ties back to my earlier statement g..
about tell us why you want to know, and it relates to your 24
( %.)
)
25 comment, Mr. Williams, to FP&L on ESF component actuations
323 4
t' 1
from non-ESF signals.
I infer from your answer that l
^
2 equipment wear and tear must be of concern to you since 3
safety of the plants are obviously not impacted by a non-ESF 4
closure of a valve, or in his case, start of a pump, I S
think. If that's the concern, then maybe the industry should
!6 adopt a philosophy of reporting all ESF operations, which e
7 would include surveillance tests -- which I don't think you 8
want, pre-planned actuations, which I don't think you want.
9
.It.seemslmore emphasis should be put on failures than 10-actuations, with the possible exception of RPS actuations.
11' We're interested in those.
You.are interested in those.
"12
.Can_you give us examples where_ESF actuations,'not failures, Ej'"%c 13
'have lead to.some generic communication, and if so, how
\\ ']
14 often has that occurred?
l 15:
MR. WTLLIAMS:
I do have things that come to mind.
16 There are some HVAC actuations, for_ example, that come to 17 mind, and'we've just-done a small study, a technical review-
'18 on HVAC, control. room HVACs, in fact.
So, that's.one
( 19 '
example that comes to mind.- There a)e examples even in
-L 20L there that1come to mind where there was an actuation that y
- i 21-didn't have any immediate impact, but~there was a cutoff of i
.22 the supply of air to the control room and there was an 4
'23 affect on the operators some time ~after that that we found 2 41 interesting But there are some examples.
[c
(
25 (To Mr. Rossi.)
Did you have any in mind?
a f,
m y--+
W 9
324 1
MR. ROSSI:
I don't have any generic N_ /
2 communications that I can think of that we put out.
I do 3
know that there have been cases where we've seen a lot of i
l 4
spurious actuations of things like ventilation systems and 5
RAD monitors at a particular plant over, you know, a couple 6
of months where it is conveyed to us the feeling that there 7
were so many of them that it might lull operators into 8
. ignoring a real one, should a real one occur, and we have 9
Worked through the regions with particular plants to address 10 problems like that._ So even though an individual one may 11 not concern us, a whole series of them may draw into L
12-question how the operators would view an actual real one if Q
13
.one should occur.
N.,)
14 MR. WILLIAMS:
On the other hand, there's another l
151 point brought out by your question which is very l
h 16 meritorious, and it goes to this concept'of key components i
17 or certain -- operation of certain major components in the 18-plant that we're-interested in and trying'to capture that l
19 with an ESF rule or with a rule. It's different than trying=
20-to capture the component failures.
One seems easier than (3
21 the other.
But again,_the way that we organize the rule in 22.
the.NPRDS puts us in the situation we're in.
So, we may be.
23 able to-straighten that out in the future, but that would'be 24
.a'long term rule-making kind of effort.
cm.
25' MR.-BETHAY:
Just a suggestion on those.
For the l
r
- 325 i
1 long. te rm, I like the idea.of a monthly report on that kind 7-~
. \\_,<):
l
^
2 of thing, including the monthly operating report or 3
something like that.
4 MR. JORDAN:
I don't know if it's helpful, but I 5
do recall an information notice that I wrote some years ago 6
on an ESP actuation. It had to do with a safety injection 7
signal that was caused by a walkie-talkie, and there was a -
8
- obviously any ESF actuation we didn't want from a cause we 9
didn't want, and we wound up putting an information notice 10 out on that.
Also, I think we've seen some correlations-11-between loss of off-site power in use of walkie-talkies and
. l 12 I think it was about four or five years later that we wound
. [y 13' up putting a piece of-correspondence out on that.
Those are
\\/
14' just two examples of things that I have personal experience 15-with.
-16 lHR. ROSSI:
We=did find a problem once where we 17
-were getting inadvertent swap-overs of ECCS systems from the
-18 refueling water. storage tank to dry' sumps, and I think'in 19'
' looking at that particular set ~of events, we found some 20 actual problems-with plants that needed to be addressed.
21 You know, some times these inadvertent actuation things, ik' 22 when theyfare looked at closely, can drop out' other problems 4
23
. that might not have been found otherwise.
And it comes back 24 to one of the things I said earlier, it's very difficult for
/
,I i]/ -
-25 us to write a level that will definitely drop out all the o
g h
j me-
-Y*
l-326 L-1 stuff that we don't need to know about and still capture
> -} '
N-2 those that are useful to know about.
So, we're going to 3
probably keep the level such that we capture everything we L
4 think we need to know about and it will probably still have 1
5 some things in it that we don't need to know about.
6 KR. NOVAK Let me just add one thing. Being the 7
. moderator, I have to limit myself to how much discussion we 8
can entertain on my own,part as well, but your question L
9
.wasn't -- you know, it didn't -- we had a little bit of 10 advanced warning of it, and I asked the staff in DSP, in the 11 Division of Safety Programs that's responsible for things 12 like the air system study, the service water case study,
![
i 13 whether or not engineered safety feature actuations provided L
s~,/
14 any information.
The thing is, it may be the-trigger-for p
l 15 the report, and if you go back to air systems, for example,.
16 and some people say, well, they weren't safety related, so
'17 maybe they weren't reported, this whole database that we 18-
'have takes the whole event.
So while you don't necessarily.
[
19?
use the ESF actuation per se, it does give you a trigger for
- l. m l
20 the report.
And I think in talking to the author of our 21 case studies on air systems and so forth, he was very clear 22'
~ that information relevant-to that kind of effort may in fact have come out of some unintended engineered safety feature 23-L
[L p;
24-actuations, but when you proceeded to look at the event more
~(_,
25 carefully, you began to understand the air system a little
r 327
?
1 more, So there is'some of that that comes off, and I just i
- ,.s\\
j I
2' wanted to let the record show that we look at every part of
%/
3 the LER.
So, you just have to recognize that.
I'm sorry.
4 Now you have another question.
5 MR. BETHAY:
Yes, sir.
Let me change subjects.
6 Different paragraph from 50.72 or 3.
This is regarding 1.
7 unanalyzed conditions and outside design basis.
50.59 8
defines an unreviewed safety question, at least in part, as 9
an increase-in the probability or consequences of an i
10 accident and malfunction.
NSAC 125,-that we are all moving L
11 towards, defines consequences in terms of part 100 limits.
4 12 Would you agree that an unanalyzed condition or outside
- i,
[~'N 13 design basis is roughly equivalent to an unreviewed safety
.'~ question, and if not, how would you define the design basis t
15 for;the purposes of reporting, and how do you see the 3
L 16 distinction between a design basis and a design input or a i
~17 design detail?
l1.
l, 18 MR. WILLIAMS:
I think we struggled with a lot of 19' that in the design basis guidelines, and that's probably 20
-about.the.best forum for that kind of a question.
That's 21 not so much a reporting question as an engineering judgment E
22 question.that lends itself to a process to resolve -- that's 23 described in those guidelines.
So I think that that would 24 be the best place to turn to right now to answer that 0,
( )
25 question as opposed to the reporting criteria or the i
.~
g w
7-328 h
1 reporting scope.
MR. BETHAY:
It I ', j-~(h
\\~ /
2 is addressed in the NUMARC guidelines, and it's not very 3
clear there.
It answers it two different ways.
Okay, I'll 4
leave that as a question on the record.
+
y 5
The last one I have, and then I'll hush, is 6
regarding the operation of a condition prohibited by tech 7
specs.
Do you consider a deficiency identified in a tech 1
8 spec surveillance procedure, notably they occur in LSFTs, in 9
which there may be a gap in the tested logic to constitute a 10 missurveillance, and that's reportable under paragraph L
11 (a) (2) (1) (b), or would you consider that a procedure l-12 deficiency that needs to be fixed?
P h
13 MR. WILLIAMS:
I just say-it depends on whether
[d 14 there was a tech spec violation or not and pass it over.
15 MR. ROSSI:
I'll give you my personal opinion.
I 1
l 16
'always_have a lot of difficulty with these kinds of very 17 specific questions because, you know, we run the risk of
-18 setting a precedent, but let me give you my personal opinion 19 on that one.
If you have a surveillance procedure that's L
20 missed a key part of the surveillance of-a piece of safety l
21 equipment for a number of' years and now you find.it, I think 22 that's reportable as a tech spec violation because I don't-f 23
-think you've met the surveillance requirement of the tech
~24 spec.
Now, let me give you the reason why I think we ought jn.
L 25 to know that.
That, to me, sounds like it's something that
f' i
329 1-could be -- if you have that problem, it's likely that other
}. "~s
> A~ /
2 plants are going to have the same problem.
So by reporting 3
it, we're going to make others aware of the problem and get
- 4 it fixed across the industry.
So that's my personal answer 5
on that.
I don't know whether anybody wants to --
]
6 MR. WILLIAMS:
I can give you a specific example.
]
7 There's a plant with an auxiliary feedwater system that's 8
had-that system refurbished a couple of-times, has Woodward 9
governors.
The governors have been refurbished by Woodward.
1 0,
10 There were springs put in there that were not customized 11 springs as they should have been to perform the safety i
12 function.
There was surveillance testing that didn't ' pick
- /~:; -
1 t
13 up the operation-of those particular components the way it I
,v
'14 should have and didn't really test in the mode that the i
l 15
-equipment would be demanded.
There is some-test
.1 16, improvements in an. industry initiative and an INPO y
1 17 initiative.
There are some testing things.that they look-l 1
i 118
'for that will help improve that, and right now we're trying.
19
'to figure out how~to deal with that particular problem.
It L[
20 was never reported.
So-really it's -- it's case by case.
21' There'is safety significance in this.one. How much or how.
o<
b 22 l'lttle ils the subject of study, but it's not that simple.
1 L
23
'Ift it is defined as a tech spec violation, it is reportable-l 24 now, and if it's not -- and that's pretty -- that's the l
i,Y
(,,/
25 closest line. As Jack said, there's a band.
And once we get l
h l:
L
i
\\
1 330 I
J 1
to a situation like this aux. feed, the one I started j ' lj-'s.
l - } -l-2.
discussing, a lot of judgment comes into play.
3 MR. ROSSI:
I think we do see reports from various 1
l 4
licensees on gaps in the surveillance.
I think they are l'
5 reporting that, a lot of utilities are.
6 MR. BETHAY:
There are lots of LERs.
In procedure 7
update programs you find lots of those things, and our 8
position has been to report them. Most of them are not that 9
significant.in and of themselves.
Have you seen where L
10
-you've taken, for instance, a Hatch LER, where those things 1;
l 11 have been reported time and time again over the last three 12 or four years and had gone to another plant and said, "Do i
[ k' 13 you have the same problem because you have not done a L
i' L,i 14:
procedure update program?"
I'm not picking on anybody.
15
.It's just a question.
16 MR. ROSSI:
Well, we usually would collect
'17 important ones like that togeth ar anc probably put out a I
18 notice or.something where we see them, but I don't know
'19 whether we've done'that on that particular one or not.
20 MR. REYES:
Ernie, can I add to that?
This'is 21 Luis Reyes.
Typically in the Region, we are aware of those 22 examples, and we'quickly try to disseminate that information 23 to similar designs. So, I think you'll find out that on an 24-informal review that goes in the region, that quickly -- the A
i
\\
(,,j/
25' question is, "Okay, how many are there?"
Let's call the i
f 1
331 l+
?'24 -
1 residents at those sites and make them aware of that and-o l i
+s:,
\\--
2 pass the information to the utilities in an effort to try to
~
3 help disseminate the information, and that's a common l>
4 occurrence.
MR. CROOKS:
-5 This question came up at the NUS seminar in June and it was l-6-
-- it involved the in-service testing program under Section 7
.4.05 of the tech specs.
At that time what was posed was, 8'
there were some valves that were being tested in the forward 9
-flow condition for their safety function and the -- but they L
10 were not being tested in the reverse flow condition, again, L
11 for the safety function of isolation.
And the question was 12
. posed to us at the time by the utility, and we did a fair
,N
)
13 amount of looking at it because they tested the valves and
- V 14 said, gee,-the valves in this case were operable.
But we 15 took.the overall experience with check valves, and we also
.16 ~
' looked at the particular valves in question and talked with 17 the people in NRR and the project manager'regarding the' 18-seriousness of what was missing, and I guess we came to the
-19
-- we came to the. conclusion that they should report it 20
'because that would at least bring the information forward,-
21 sf again, we could provide some generic feedback.on those --
22 that particular instance.
23 MR. WILLIAMS:
That was kind of a -- that was
,ss 24 another one.
Sometimes there are controversies within the
~I h
Ns/
25 staff on these issues, and it's healthy to mention it
332 y- $ _
1 because that was one where if people did a good review with
\\
L
's 2
their IST and tried to add things to their IST program, they p
3 might'have to submit LERs every time they found something 4
that.wasn't in the program, which was a disincentive to
.5 improve the program and NRR wanted the program improved, 6
yet, we had to report. So these kinds of issues are best
'7 dealt with through discussions with the staff.
That's the 8
only thing we could suggest right now, i
9 MR. NOVAK:
We've got a question over there.
10 MR. ROBERTS:
My name is Dave Roberts.
I workffor 11 Virginia Power at the North Anna Power Station.
I'm the 12'
' supervisor of the Station Nuclear Safety there.
My question l
/~s D(
deals with one of the other controversial issues brought up
)
13 14 this. morning and that's serious degradation of the nuclear 15-power plant.
We have determined, for example, durir.g an 16 outage situation that one train of the safety system,-for 3
.f l'7 -
example, a containment isolation valve, is definitely
' noperable, did not pass its local leak rate test. In order i 19 to determine.if that is a serious degradation of a principal 20 safety barrier and'therefore to report'it, do we have to 21 then apply a design basis event to that condition, assuming 22 a-single failurelof the~other containment isolation valve?
i 23, Would we consider that one valve failure when determining if i
24 we report or not?
25 (No response.)
333 1
MR. NOVAK:
Let's not all jump to the microphone b A;/~wiYl U
2 at once.-
3 MR. WILLIAMS:
I would like to cnswer that as a 4
single component failure, but I'd have to go back and look o
1 V
5.
at it to tell you the truth.
So we could take that as a i
1' 6
. question.
7 MR. REYES:
Let me add to that.
There are some i
8 cases where a single valve failure causes you to leak so 9-much that you can't even pressurize it, and clearly your 10 release limits, you can't calculate them. So there are 11 circumstances where the failure is sua severe that you have 12 to raise a question, but that's not the case every time.
1: si 13 MR.-ROBERTS:
But I understand the issue is the --
ij-14 the redundant component is definitely fully operable, and i
15 the only condition where you would actually be seriously L
~16 degraded would be under a single failure. situation.
So do 17 you consider that a serious degradation?
18-MR. ROSSI:: I guess my' feeling ~is, if you've got 19 redundant components and you have reasonable belief that one 20 of those components wouldn't have worked for a lengthy period of. time well beyond what you would tua allowed by the t
tech specs, that that's outside the design basis at 3 I
minimum and is reportable.
I think we see -- I think we see E
a lot of reports on valves that leak during tests of m,
k I
s,j 25-containment isolation.
So, I think people generally are l
4
L' 334 yN 1
' reporting those. But if there's reasonable belief that the I
2 redundant component was out of service for a lengthy period 3
of time.beyond what you are allowed in the tech specs, that 4
to me would be probably outside the design basis.
5 MR. NOVAK We can entertain some other questions.
6 MS. TULLEY:
A trend that I'm kind of seeing in L
7 some of the answers here to questions on the SFs and outside 8
design basis seems to be a lot from the perspective that L
L 9
we've got to look at things on a case-by-case basis.
And it L
10 seems'to go against what you are trying to do in the rule, L
l 11
. hichcis give some general guidelines for the type of events w
12 that you'want to report. So that's something that concerns l;. p '
13' me.
I think what the owners' group has tried to do in some H
~
14 of the proposals that they've presented to you, is try to 1
~15 come up with some further guidance that will clarify l
16' situations that aren't currently clarified, not narrow the j
L l
17 band down so tight on a threshold, but give some better l
18 guidance out there..And I'm just curious, one of the things 19 that we've presented to you-is the concept of reportability 20 on outside design-basis, and-the contention that we're l
21 making is that it would only be reportable if you actually L
22-
. severely impact some of the major parameters under which the L
l 23 plan is licensed.
Like, you actually exceed 10-CFR 100 j
l 24 limits, or a problem is-shown to occur with fuel clouding
,~
(sg 25 temperatures as a result-of some change that you've found or L
335 1
degradation-that you've found out in the plant.
or there's i
2 some severe challenge to primary containment or reactor t
3 coolant pressure boundary.
We came up with that approach j
4 and we really believe that it's consistent with the way 2 is 5
written right now and that it refers to any event or 6
condition that resulted in a condition of the nuclear power 7
plant being seriously degraded.
And some of the comments 8
that I hear today seem to be indicating a much lower 9
threshold of reportability to that issue. I'd just be 10 interested in the comments on that.
i 11 MR. WILLIAMS:
Let me make one comment because we 12 discussed this outside for a second, Cindy.
One of the 13 things we've tried to say is how we use the data, and the j
14 original rule is really clearly a product of initiatives-
]
15 from the' feedback of operating ~ experience.
It was put in a
16
. place to get full sequences, engineering evaluations, all i
17 ESF's were requested because it was thought that there would i
18-be something in the sequence' of events involving. an ESF f
-19 actuation that would be interesting and worth reviewing for 20 operating' experience.
The rule was put in place to support 21 the team function of the submittal and analysis of 22 operational data.
The validation of the design basis of a 23 plant, a licensing issue, if we wanted to continue to monitor the validity of a safety analysis or the licensing 24 25 basis would be a different rule.
We would write a different
e 336 i
rule for that kind of a function.
To support the feedback i
2 of operational experience, the analysis of all the events in 3
the operating industry to feedback lessons to licansees as i
4 information notices. To maka sure we don't have a serious 5
action is a different function, and the rule was set for a 6
very low threshold.
Now, when we try to atch 50.59 or any 7
licensing criteria against that rule, there's a lot of 8
commen areas, but there'n some areas that are not common and 9
have a much lower threshold, and that's why.
10 MS. TULLEY:
Just to follow up on that.
11 Understanding why the threshold or the rule is set the way 12 it is, how do you intend to respond to utility concerns that 13 were expressed in NUREG 13.95 with the low threshold and 14 what I would centend is potentially a threshold that's i
15 moving downward, not moving upward in the application of the 16 rule, and how do you justify that against the fact that 17 unfortunately LERs are a very highly visible document, 18 whether we like it or not, and viewed and potentially 19 misinterpreted by a great many people like PUCs, pSCs, i
20 intervenors and whatnot. I guess I'm curious how -- given 21 some of the comments you folks have made -- you are going to 22-be responsive to the concerns raised in the NUREG.
23 KR. JORDAN:
Okay, perhaps I would try to respond 24 to that With the existing rule, we can make some 25 adjustments, we thinx, in a relatively minor rule change 1
-n.
1 337 1
that will take away so.ae of the enerous reports that 2
industry and NRC would agree have not been beneficial to us 3
in learning safety lessons.
The philosophy that we have had 4
on reporting is not in my view, one that we expect to 5
change.
That we are, in fact, trying to find relatively low l
6 level individual safety significance eventc that contain 7
elements that will help teach us a lesson and allow it to be 8
extracted and conveyed.
And so the idea of only seriously 9
degraded or severely exceeding those kinds of terms are not 10 where I expect to lead the rule.
I hope that we can educate 11 the public and we can educate the PUCs in terms of what the 12 reporting is really trying to do.
It wasn't aimed at 13 evaluating the licensees for rate making.
A by-product of 14 somebody doing a careful review might -- and some PUCs have 15 done it -- might be something they choose to do, but it's 16 not the purpose of that reporting information.
And I would
)
17 not plan to revise the reporting in order to make the 18 industry look good.
So, I would like to make that very 19 clecr.
I don't plan to do that.
Now, in order to gain the 20 information the. agency needs to do its job and feed back 21 experience to you, I'm very interested in adjusting it so 22 that you're not wasting resources reporting things that are 23 not important in terms of having a role in feedback and that 24 we're not wasting resources collecting and screening that 25 kind of information.
So that's the way I would like to go, I
/
b.
338
'~N 1
and I would certainly encourage industry guidance in
\\
2 conveying in a uniform fashion within the industry how to
)
3 meet those kinds of goals.
So,'if getting back to that kind l
4 of philosophy helps, that's what I would offer.
1 5
MS. TULLEY:
One more follow-up question and then 6
I'll hush up.
Do you anticipate -- you mentioned this 7
supplement 3 that you're working on.
Number one, I would be 8
curious of the time frame that you are working on. When do l
l 9
you expect to have it out?
And number two, do you 10 anticipate that the interaction with the BWR owners' group 11 and further discussion on the owners' group proposals will 12 be of assistance and possibly impact in some way or form rn 13 that supplement?
l -(G) l 14 MR. JORDANt Okay, I won't commit the staff to a 15 date, but that guidance would be in months, not years, and 16 it would be a fairly direci. outcome of these meetings that 17 ve are holding now.
I would hope to use all of the industry 18 feedback that we have received, and work that NUMARC is 19 doing and the owners' group are conducting to try and 20 provide guidance.
But, I have to make it very clear that 21 the guidance has to be within the scope of the existing rulo 22 or a modification that we make to the rule.
So, if va make 23 a modification to the rule in the next few months, then there will be a revision for 10.22, unless we do it at the 24
(
25 same time, m)d I think -- I do expect both of those things
339 J ^x 1
to occur.
Minor revision to the rule and changes to the
(\\ ')
2 guidance.
So there may be some jockeying as to whether we 3
have to do a two-step guidance or a one-step guidance.
4 KR. NOVAK Question?
5 KR. WEBB Yes, sir.
My name is Tom Webb. I'm an 6
'ngineer at the Vogtle site, and I would like to say that I 7
agree with Cindy.
She was asking earlier for some guidance 8
on what comes outside of design basis or what's an 9
unanalytod condition.
I know you are reluctant to come up 10 with apecific guishlines, but I was hoping that maybe in the 11 future we might see some -- at least some general types of 12 guidelines to f. ell us how to evaluate those kinds of l,,)
13 conditions.
I know in the past, if plants are reluctant to
(
/
14 Itake a report or if they don't know for sure whether they 15 are outside design basis or in an unanalyzed condition, they 16 sometimes are forced to go back to their original designer 17 and the engineer / architect has nothing to gain by saying 18 that he screwed up and it puts him in a position where he 19 may want to use his pencil to write backwards, forwards and 20 sideways to say that there's no problem here.
And we're 21 hoping for some kind of relief in that area.
Can you 22 comment?
23 MR. JORDAN:
I guess I would comment and say that 24 the design basis in terms of calculations that support your
\\_,/
25 design basis in some plants is very poor, and you know, it's
t 340
'~N 1
a historic problem in terms of that not having been N'
2 translated from the architect / engineering firm to the 3
utility.
And there is a design reconstitution that some 4
utilities are doing to try to have that basis fresh and more 5
easily understood by the operating and engineering staff.
6 If there is a question, we can't give you guidance.
You 7
know, if you have a question about whether or not it's 8
within your design basis, I can not provide you guidance 9
that would help you decide whether it is.
You would have to 10 go back to your analysis or do another analysis to establish 11 it.
So if that's the question, I can't help you, 12 MR. WEBBt I don't think we are looking for t
! n)
(
13 specific guidance in any particular type problem but a Ns/
l l
14 general type guidance would be appreciated.
15 MR. JORDAN:
Okay, I think the question has been 16 raised sufficiently that we need to clarify in our own 17 guidance to you what we understand being outside the design 18 basis means, and perhaps we can express that more clearly.
19 But once you have a question, you still have to go back and 20 do a review against your early documentation or a new 21 analysis to determine that.
22 MR. WEBB Obviously. Thank you.
23 MR. JORDAN:
Okay.
24 MR. WEBB:
Another question I had.
Mr. Williams,
-s 25 before lunch, you made a statement that LERs would be looked x_,
341
' 'y 1
at when considering plant extension, and when I heard that,
(
)
\\ s' 2
immediately the term bean count came to my head again, and I
~
3 didn't know what you meant, and I thought I would ask you if 4
you could tell us exactly what you meant.
5 MR. WILLIAMS Now some of the relevant issues for 6
life extension and aging and plant aging ore coming from 7
LERs and that's what I was refer' ring to.
8 MR. WEBBt Relevant issues.
9 MR. WILLIAMS:
Relevant issues.
10 MR. ROSSI:
I think what he's saying is that the 11 LERs may very well have problems that are in there that are l
12 relevant to aging kinds of problems.
It's the problems that l
[w)'
13 are being looked at, not a -- certainly not a bean count.
\\v l
14 MR. WILLIAMS:
No, not a bean count.
15 MR. WEBBt Thank you, sir.
16 One last question.
You also mentioned, sir, that 17.
there is a significance factor you assign to LERs between 18 one and four and I hadn't heard this before. What is the 19 difference between a one and four?
20 MR. WILLIAMS Well, AEOD has a significance 21 categorization, and over the years it's really been to 22 identify generic issues.
A category four is not a very 23 significant LER in terms of further-on study and it's 24 essentially a grading system on an engineering scale for the
,1s Y
(,,/
25 initiation of engineering studies, category one event,
342 1
there is very few of.
They are very important. category 2
,s I
)
\\ _,/
2 events, we have a couple of hundred a year, in that range, 3
to 250.
Those kinds of events are the kinds of events that 4
might initiate a study or further action by the staff. They 5
are of generic interest in AEOD.
If you would like more 6
information on that, you can certainly just call us up.
The 7
question of significance grading or rating of LERs, one time 8
-- we also administer the abnormal occurrence program to 9
report to Congress in AEOD, and we, in concert with the 10 staff, select and write up the abnormal occurrences that are 11 reported to the congress every quarter.
At one time, I 12 tried to pull together all the significance scales that we
/N 13 have, and I had twelve, and I think we have a new one now.
N-14 It tries to sweep across all the past history.
You'll find 15-one other significance rating mechanism in the performance 16 indicator report. In volume 2 of that report, there is a 17 definition of a significant event that we use and we look at 18 LERs with that definition. So if you have any questions on 19 how we grade LERs or what our significance criteria are, 20 give us a call and we can send you another document.
21 MR. WEBBt Is it possible to find out what the 22 ratings are for the various pisnts?
23 MR. JORDAN:
Let me maybe stop us from being in a 24 direction we don't need to be.
The four characteristics p
)
25 that Mark referred to are really an internal screening tool, v
+
343 e'
1 They really don't have an association with risks or anything
\\s '
2 else. It's a way of the staff deciding how much more werk in 3
real time they would do on a particular licensee event 4
report.
So it's not a lasting grade or characteristic in 5
risk.
The two things that are, in my view, are the 6
characterization as a significant event that Mark referred 7
to that is in the quarterly performance indicator report.
8 We do bean count, if you will, significant events for a 9
specific plant, and those are complex events, multiple 10 failures, and it is a criteria that we have developed and 11 put into that performance indicator report, and they are 12 one-zies and two-zies for plants on a quarterly basis.
[,,)
13 Now, the other thing in terms of significance is V
14 the accident sequence pre-cursor analysis of an event in 15 which one establishes on a relative risk scale how risky 16 evento were and that's saying that given the failure that 17 occurred, the probability of core damage was ten to the 18 minus four, ten to the minus three, et cetera.
That's the 19 one that I place the greatest credence in in terms of seeing 20 that there was an event that had potential risk to the 21 public and then there needs to be more emphasis or less 22 emphasis on that kind of event across a group of plants.
23 MR. WEBBt So this information is available then 24 to the plants then is what you're saying?
O(,,)
25 KR. JORDAN:
Absolutely.
The NRC does nothing
344 1
that is not publicly available and available to the plants.
r_s I
)
x,/
2 MR. WEBB Thank you, sir.
3 KR. NOVAK A question on that side.
4 MR. REPKA:
Thank you.
David Repka with Winston 5
and Strawn. I just wanted to raise for the record a viestion 6
-- a somewhat subtle question that I know some licensees and 7
myself have grappled with over the years, and that question 8
is, does 50.72 and 50.73 create any obligation on the part 9
of the licensee to re-evaluate long past or historical 10 conditions at the plant.
An example of that might be a 11 situation where several years ago you removed a piece of l
12 equipment and you replaced it with a different piece of 13 equipment that is undisputably in compliance with 14 regulations, but nevertheless you've gone on and you've been 15 doing an ongoing analysis of the piece of equipment you i
1 16 removed, and low and behold, five years later, you discover L
17 there is a question with this piece of equip'.nent. Maybe it 18 didn't perform the way we expected.
Maybe it wasn't in L,
(
19 compliance back when it was in the plant.
Is that the sort 20 of thing that 50.72 and 50.73 encompasses because clearly 21 there is no longer a presently reportable condition per se?
22 The only comment I would make, I suppose, is, I know that is l
23 a dif ficult question, and I know that may raise information 24 that's relevant to the NRC and relevant in safety
,()
25 significance, but I just wonder whether 50.72 and 73 are the l
l
.,r
345 1
appropriate reporting mechanisms; particular 72, since you
,s3 l
)
\\- /
2 think of it as going through the red phone and the 3
notification center and this is regarding a condition that 4
maybe hasn't existed for years and years.
5 MR. JORDAN I think I'll take a crack at that.
6 There are not very many cases in my view that part 21 refers 7
to operating plants, but I think I could say that that's one 8
in which it might.
If you have information on equipment you 9
formerly used that other utilities may still be using that 10 you have -- that's come to your attention, then I think 11 there's a moral obligation and even a legal one under part 12 21 to notify us, so that we can in fact convey that on to l
/ )
13 other utilities.
I think that's a case where it really
\\- /
1 14 wouldn't fit 50.72 -- 50.73 perhaps.
It's curious because 15 in a previous discussion, I couldn t think of anything that e
16 an operating plant would have to report under Part 21 that 17 wouldn't be already covered under 50.72 or 73, but that is 18 one.
19 MR. NOVAK:
We have some time for some'more 20 questions.
21 KR. SMITH:
My name is Willie Smith.
I'm a 22 compliance engineer for Plant Vogtle.
50.73 (a) (2) (5) talks 23 about any event or condition that alone could have prevented 24 the fulfillment of the safety function.
I was wondering if
(
'r
(,/
25 I ceuld have some comments about why the word "alone" is
346 1
within that criteria.
's-2 KR. WEISS:
Yeah, I think I could help you there.
3 The word "alone" did not appear in the proposed rule but i
4 were in the final rule as a result of public comment. This 5
was -- I actually used to cover this in the presentation I 6
gave because it does come up frequently.
And the nature of 7
the public comment was that the rule as proposed was too 8
vague.
Any condition could prevent the fulfillment of a 9
safety function.
All one needed to do was to hypothesize 10 that we happen to be in another state or we happen to have 11 another failure, and we said, "No, that wasn't what we had 12 in mind.
What we had in mind was that that condition alone lo
\\
13 could have prevented the fulfillment of a safety function.
14 You didn't need to image a different set of failures."
So 15 we put the word "alone" in right af ter the proposed rule l
16 making, and that is a frequently asked question.
l l
17 MR. SMITH:
A similar question is the criteria for 18 an unanalyzed condition. It talks about a condition that --
l 19 unanalyzed condition that significantly compromises plant 20 safety.
Maybe a comment on why the word "significantly" is 21 in that criteria.
22 MR. WEISS:
Well, again, when the rule was at the 23 proposed stage, there were those people that trivialized the 24 reporting requirement. They said, "Well, anything
/~,N
\\
\\s_,/
25 compromises plant safety.
Gee, the clock falls off the
347 e"N 1
wall, the guy's not going to know what time it is."
I mean, I
)
\\'
2 they came up with the most absurd suggestions as to what the 3
rule was encompassing.
So in discussion was the attorney 4
who was working with us. We agreed upon using the word S
"significant".
What a reasonable man would consider as i
6 significant.
Safety significance is what we had in mind,
[
7 not something utterly trivial.
In a sense, we fell into the 8
trap that we've been trying to avoid for years, and that was i
9 to make the rule overly legalistic.
We wanted to have a 10 realistic threshold where one engineer talking to another 11 could say, "Yes, the failure of this pump would have an l
l 12 impact on safety," and it is not 6-it would prevent the
/
h.
(
]
13 fulfillment of a safety function under a certain design 14 basis accident, and it's not a trivial thing.
One could 15 find components in almost any safety system, which if they 16 failed, compromised safety to some extent, but it's so l
17 trivial, you know.
The indicating light might fail two I
l 18 hours2.083333e-4 days <br />0.005 hours <br />2.97619e-5 weeks <br />6.849e-6 months <br /> less than the estimated meantime -- previously 19 estimated meantime before failure, but gee, that's trivial.
20 That's not quite what we had.in mind.
We didn't mean to l
21 sharpen the pencil to the point of reductio ad absurdum So 22 words like "significant" got stuck in the final rule that i
l 23 weren't there in the proposed rule and that "alone" could a
24 have prevented.
I remember that very clearly, got stuck in 7 (V l
l l (s /
25 the final rule.
I remember almost the day it happened.
1 l
348 1
MR. JORDAN:
What day was that, Eric?
2 (Laughter) 3 KR. CROOKS:
I had one thing that I wanted to 4
mention, too.
In what we are looking at in the reactor 5
water clean up system, ESP area, and the HVAC isolation 6
area, in reducing reporting, we are trying to come up with 7
some provision that would still bring in information where 8
there are high frequency of these types of events.
We 9
recognize some of those terms are going to need to be at 10 least couched or some guidance given. So that's something 11 that, you know, you might also consider in which you -- in 12 your efforts.
What would be -- you know, I guess, what dc
/ 'T
~
i
/
13 you think in that regard.
And there are a number of 14 examples where some plants were having, you know, I guess, a 15
-- well, what was considered a very high frequency of, let's 16 say, reactor water clean-up system isolations that, you 17
- know, were coming from a variety of things.
But over a 18 short period of time, they weren't really being focused on 19 and weren't being resolved until finally some issue was 20 raised and some of it was because of the, you know, we'd 21 have to keep reporting these things, so we need to do 22 something.
That's not -- I mean, the focus, it just helped 23 to bring focus on the concerns.
So we feel that's still an 24 element that -- you know, on an individual plant basis.
We
,y
)
\\~ /
25 would need some information on that.
349 g
1 MR. NOVAK I'm going to entertain one more
(
/
N' 2
question and then we'll take a break.
Does someone have the 3
last question?
4 MR. HILL:
Richard Hill, Assistant General 5
Manager at Farley.
On the supplement 3 yau are talking 6
about issuing, along the lines of what we just talked 7
about, about the tse of the word "alone" being added, would 8
you entertain an*/ ideas of working with any industry groups 9
or giving that supplement 3 out for comment before you issue 10 it?
11 MR. NOVAK:
I think the answer would be, sure.
l l
12 Yes.
It would be a peer review or something like that,
[m 13 which would be fine.
'u 14 MR. JORDAN:
Is our attorney still here?
Within 15 the bounds of our advisory committee act, we will -- when we 16 provide anything for comment to any industry group or any 17 public citizen group, we do it to all.
So it would have to 18 be a publishing in some Federal Register notice or something 19 that -- here's the guidance, here's when we plan to issue i
20 it, and if you have any comments, provide them by date 21 certain.
The only problem with that on our part is that we 22 then have a longer time period to get the guidance out.
23 I kind of hope from these workshops that we're j
1 24 going to-be able to narrow things down pretty well, and will u s 25 have had a pretty good cross section of industries input, so s.
l l
350 f'^
l that the industry feedback would not change or, you know,
\\/
2 blow the next step, so that we could do it in a reasonably 3
timely fashion.
So I'm sort of counting on this kind of a 4
workshop done four times to give us a 95 percentile and then 5
further interactions clearly in a public format still to get 6
it to the 100 percent level.
L 7
MR. WILLIAMS:
I wanted to make one question there 8
-- one comment there.
's we develop our guidance and we get 9
the questions from you it would be very helpful if we keep 10 the questions focused on the feedback of operating 11 experience.
When we try to focus on single words like 12 "significant" or "alone", where we are using engineering
/
13 judgment, and that was the intent of the rule, we're trying i
14 to play the ropes on the reporting rule.
The staff is very 15 good at playing the ropes on the reporting rule.
We don't 16 want to spend a lot of time doing that.
We'd rather 17 eliminate the reports that are not of interest to us or 18 safety significant in our view and your view and get the 19 ones that we are not getting.
And it's important to keep 20 our eye on the goal because we could get very distracted and 21 spend a lot of time interpreting where the boundaries are 22 and there wouldn't be any reduction in reporting LERs.
We 23 could define things to the lowest level in every case and 24 the engineering judgment would be legislated out of the 7-x_/
25 whole thing.
So it's important to keep our eye on the goal
351 r3 1
of feeding back important information, as opposed to trying f
\\
,1 8
2 to draw too fine a line.
I think the use of engineering 3
judgment in our industry is so important and this rule has 4
gotten us so far in improving performance 1.9 the industry 5
and the information notices are now up to 80 or 100 a year.
6 There's a lot of information going out in industry networks 7
and NRC's networks, and we've done a lot of good with this 8
rule.
So, if we continue with the intent, we'll continue to 9
avoid, hopefully, serious events, rather than focusing where 10 the ropes are.
11 MR. NOVAK We're going to take about a 15 minute 12 break.
Let's get back up here around 2:30 or so.
'm) 13 (A short recess was taken.)
L/
14 MR. NOVAK:
We're prepared to begin the talks 15 regarding safeguards.
I wonder if we could more or less get 16 in our places.
Thank you.
17 What we would like to do now is turn the meeting 18 in the direction of reporting of safeguards events.
- Again, 19 we'did see that this was an issue of concern based on the 20 impact study, and we have with us today three people who 21 will give brief talks on subjects related to this.
I would 22 like to start off first then with Bob Dube.
He's from the 23 office of NRR, and he'll then be followed by other members.
i 24 Go ahead, Bob.
[
y
\\,/
25
352 1
SAFEGUARLC EVENTS (10 CFR 73.71)
/
i
(_,)
2 KR. DUBE Thank you.
I'm Chief of the 3
Performance Assessment Section of the Safeguards Branch in 4
NRR, and we have two other safeguards staff with us on the 5
panel this afternoon.
We have Nancy Ervin who is also in 6
our branch. Those of you who work in the safeguards area who 7
have an interest in event reporting I'm sure know Nancy.
8.
She is the co-author of a regulatory guide in the NUREG that 9
deals with event reporting. She has been the primary staf f 10 contact for questions involving new interpretations of the 11 existing regulations since their publication.
12 We also have with us today Joan Higdon.
Joan
[ '$
^
13 works in our Office of Nuclear Materials, Safety and
~')
a 14 Safeguards.
Joan has been doing the data analysis, and 15 again, any of you that have been working this area know 16 Joan. She's been around and met with licensee groups for the 17 last couple of years.
18 We'll do our very best to stay on schedule.
I'm 19 going to make a few introductory remarks.
Nancy will then 20 spend 10 minutes talking about one hour reporting, and Joan 21 will spend 10 minutes talking about logs.
22 The safeguard requirements in 73.71 are very 23 similar to the safety regulations on event reporting that 24 we've been discussing already today.
We do have a fx..
()
25 requirement for one hour reporting for those events that are
353 1
significant.
Less significant events have to be logged s.
[
)'
\\
2 within 24 hours2.777778e-4 days <br />0.00667 hours <br />3.968254e-5 weeks <br />9.132e-6 months <br /> and that log has to be submitted quarterly m-3 to NRC.
Most of the -- 73.71 defines what the Commission 4
considers to be a significant event.
Most of those 5
definitions involve actual occurrences, such as an actual or 6
attempted threat to commit significant physical damage to a 7
power reactor or unauthorized entry into a protected area or 8
a vital area or actual or attempted introduction of 9
contraband into a protected area or vital area.
10 There is one other group of events that are 11 considered significant.
Those events that involve the 12 failure or degradation of the physical security system that
T 13 could allow unauthorized or undetected access into either a l (h 14 protected area or a vital area. The important consideration 15 '
there is that if those -- if those failures are properly 1
16 compensated, then the event can be logged.
If they are not 17 compensated, then they involve one hour reporting.
18 The existing requirements were published in June 19 of 1987 along with a regulatory guide and a NU7EG that 20 provided additional guidance.
We now have three years of
-21 experience with that guidance.
We've been in the process 22 for some time now evaluating that experience and we are in l
23 the process of changing both the reg guide and the NUREG.
F 24 We recognize that it will be several months at the minimum l
/%.
(,)
25 before we finish that effort.
We have also recognized that
354
~s 1
one of the problems that we have is that some unimportant
)
(_/
2 stuff has been reported on a one-hour basis.
We feel that 3
it is important to deal with that particular issue as 4
quickly as possible.
So rather than wait for completion of 5
the NUREG and the regulatory guide, we have prepared a 6
generic letter which would not impose any new requirements 7
and would be strictly voluntary, but would provide licensees 8
with the flexibility of logging some events, which in the 9
past have been reported under the one-hour reporting 10 provisions.
That particular generic letter just last week 11-was sent to our CRGR.
They have not had an opportunity to 12 review it yet, but we hope to have that out in the riear
(
13 future and Nancy will spend some time telling you about that 14 effort.
15 Nancy.
16 MS. ERVIN:
As Mr. Dube mentioned, we developed 17 the generic letter in order to provide an immediate revision 18 to some of our guidance that's published in order to reduce 19 the number of unnecessary events that are coming into the 20 NRC operation center, and also to reduce the reporting 21 burden on the licensees.
22 In developing the generic letter, we evaluated the 23 safeguards events that had been coming into the operations 24 center for the past several years, and we evaluated them e
f))/
(
25 based on their safety significance and also based on the J
e m -
355 1
immediate actions taken by NRC and the licensees.
In doing
_s,
[
)
\\_,/
2 the evaluation, we determined that some of the events were 3
being reported in accordance with our guidance, which based 4
on our safety evaluation, wo determined we needed to revise.
5 Also, some of the events were being reported because our 6
published guidance didn't provide enough clarity in some 7
areas.
So what I'm going to do is go over what we have in 8
the generic letter as far as the reports that have come into 9
the center in the past several years as one-hour reports 10 that can be logged instead of being reported within one 11 hour1.273148e-4 days <br />0.00306 hours <br />1.818783e-5 weeks <br />4.1855e-6 months <br />.
Before I talk 12 about the specific events, I'd like to go over a couple of
/'~~ h 13 areas where our policy has changed that are generic.
The 14 first one deals with proper compensation, and it is a l
15 revision to our current policy. Our current policy says that 16 licensees should report events to the center within one hour 17 that have the potential to allow unescorted or undetected 18 access into a protected vital material controlled access 19 area, or to a transport.
We are changing that guidance so 20 that: you can go ahead and log the events if extenuating 21 circumstances cause you not to meet that ten minute time 22 frame of compensating the event.
23 When we allow this reduction in your reporting, it l
24 was providing there was no malevolent intent with the rh.
/
i
( )
25 degradation. Also that nothing adverse resulted from the
.n...
356 1
delay and that you take appropriate measures to insure a 7_(v 2
more timely response in the future, one example of what 3
could cause this is, perhaps a contractor or a vendor 4
becomes aware of a safeguards problem and he doesn't tell 5
security in a timely manner, so as a result you can't comp 6
it within the ten minutes, but you do compensate it.
You 7
check into the event and you find that nothing has happened 8
adverse from it, and you institute the proper compensatory 9
measures.
Well, that's not the kind of event that we want 10 coming into the center because it is not significant.
It's 11 true that in order to compensate it properly, it should have 12 been done right away within the ten minute time frame.
But g-k(
13 that doesn't make it an event that should come in and it
.'s 14 would take up the time of the licensee and the NRC in a 15 reactive measure.
It is one that should go in the log and 16 come in quarterly.
17 When you do have to put it in the log and it's 18_
because it took more than ten minutes, we would like you to 19 note the cause of the delay in the log.
20 The other policy issue is in the area of fitness 21 for duty events.
Significant fitness for duty events are 22 now reportable under part 10 CFR 26 and not under 73.71 23 anymore. And the' fitness for duty performance data is to be 24 submitted under the provision of 10 CFR 26.71(d).
There are
.fN lV) 25 going to be occasions -- they'll be very rare -- but there r
337 1
may be occasions where a significa..* event in the safeguards 7-,s 2
area is caused by a fitness for duty evea,*..
If that 3
happens, the fitness for duty aspect should be submitted to 4
NRC in accordance with Part 26 and the safeguards aspect 5
should come in under 73.71.
If you have to make a 6
telephonic report for both events, in the generic letter, it 7
gives you the option of making one telephone call instead of 8
having to make two.
If you choose to make one call, it 9
would be -- if it's a significant event under safeguards, it 10 would have to be within one hour of the occurrence of the 11 event and then it would have to be followed up by the 30 day 12 report as is already required by the rule.
/)
13 Now, what I'm going to do is just go over the
,)
\\
14 examples that are in the generic letter, that as Bob pointed 15 out, is over in CRGR for review right now, and these are 16 examples of safeguard events that don't need to be recorded 17 within one hour.
You can log them instead.
When I say you 18 can log them, I do mean if they are properly compensated in 19 accordance with the guidance that's already published and 20 already been approved through CRGR and out for public 21 comment and everything.
This is reg guide 5.62 and NUREG 22 13.04.
With the exception of the couple of policy changes 23-that we just discussed, the allowance of the extension of 24 the ten minutes and then the fitness for duty events.
/
t
[
25 The first one is a design flaw or vulnerability in
\\_/
,a w
e,, - -e,
338 7-~
1 a protected controlled material or vital area of a 1
i(
/
2 safeguards barrier.
This is a change to our currently
./
3 published guidance.
Previously, if you had a design flaw or 4
degraded barrier thal had existed for more than ten minutes, 5
it was a one hour report, and e. gain, we found when we did 6
our evaluations, that even though these may have been 7
problem areas, they have may have even involved violations, 8
they still were not the type of significant event that NRC 9
needed to react to and that you needed to react to for all 10 the events that we've had.
Now, if something does come up 11 when you are investigating the degraded barrier, that's l
12 already in the current guidance and if something adverse l
[s 13 happens, then, of course, it would be a one-hour report.
14 The second one is a failed compensatory measure, 15 such as an inattentive or sleeping security -- such as t
l 16 inattentive or sleeping security personnel or equipment that 17 fails after being successfully established as an effective 18 compensatory measure for a degraded security system.
If i
l l
19 security personnel are ineffective because of alcohol or 20 drugs,-the security degradation is reportable under 73.71 21 and the licensee should include the positive results of the 22 for cause test in the data submitted to NRC under 26.71(d).
23 The next one is discussing contraband.
Discovery 24 of contraband inside the protected area that is not a
,A-i
/
25 significant threat can be logged.
For example, such a
339
,rS 1
condition could be the discovery of a few bullets or a
)
's 2
weapon that was inadvertently left unattended or unsecured i
3 by the security force.
Another example could be contraband 4
found in a vehicle located in a park r; lot outside the PA, 5
contraband that's turned in by the ( iver of a vehicle 6
before the search and before the vehicle enters the 7
protected area, or contraband that is discovered during a 8
routine search and can be explained satisfactorily as not 9
constituting a threat.
10 The reason that there is no report necessary to 11 the ops center is because these events can be explained 12 satisfactorily and they don't have any adverse consequences n
(b' to the facility, 13 14 Compromise, including loss or theft of safeguards 15 information is the next one.
That can be logged if it could 16 not significantly assist an individual in gaining 17 unauthorized or undetected access to a facility or in an act 18 of radiological sabotage or theft of S and M.
19 I have to go back to the other one for just a 20 second.
That was about a paragraph that was about 11 21 sentences long and it just hit me -- the last example that I 22 gave you about discovering contraband during a routine 23 search that can be explained satisfactorily as not 24 constituting any threat.
On that particular one, because
,I,\\
25 it's found during the routing search -- and this went for
360 1
all of them that were outside of the PA, so I'll back up 7 ~3
\\_ f
(
2 just a little bit.
If it's a vehicle located in a parking 3
lot outside the PA'or if you've got a driver that turns it 4
in before the vehicle comes in the PA or you discover it 5
during the routine search, again before it goes into the PA, j
6 and they can explain it satisfactorily, on that one, you 7
would not have to make the one error'and you also would not 8
have to log on that particular one.
The reason is because 9
it didn't come into the protected area.
That's why it would l
10 not require either report.
l 11 The next one deals with loss of all the AC power 12 supply to the security systems or loss of all of their
(N 13 computer systems.
These can be logged, provided adequate j k, m
14 compensatory measures can be maintained until the systems 15 are restored.
If the power loss or the computer failure 1
16 would not enable unauthorized or undetected access, this l
l 17 would be another where not only would ycu not have to report 1
i 18 it within one hour, but you would not have to log it either, l
19 and the reason for it is because it would not have allowed 20 the undetected or the unauthorized access. For example, a l
21 computer failure would not require reporting if it's negated 1
22 by an automatic switch over to a functioning back up system 23 without a time delay. Also, a momentary loss of lighting 24 caused by a power interruption would not require reporting 1
,/ m
(_)
25 if the loss could not have allowed unauthorized or
361 r
1 undetected access.
And that's really one of the key factors
(,b 2
in determining some of these events that we are talking to 3
you about that when we did our evaluation, we found that e
4 these were not only events that should not be coming into 5
the center, but they also didn't have to be logged.
The 6
rule defines the events that need to be logged as those that l
7 could allow undetected or unauthorized access into one of 8
these areas.
9 Then the other part of that rule for logging is 10 the degradation of the security effectiveness of your plan, L
E 11 or the actual condition of a degradation, but again with I
l 12 that the same logic applies.
The way you need to look at it
, -'s 13 is if you could have allowed undetected or unauthorized i
i
\\
'14 access or if it'was really a degradation. If it wasn't, then l
15 it wouldn't even have to be~ logged.
16 The next area is one that's got a lot of examples l:
17 and this is one where we had a lot of: events coming into the y
l 18 center initially and we seemed'to work most of them out, but 1
19 occasionally we still get one or.two in.
Most of these came-20 from lack of clarity in our guidance.
They deal with the I
21 access authorization program and they are partial failures.
12 2 Whenever I give you one of these events, please bear in mind l'
23 that it's a partial failure of an otherwise ef fective total 24.
access authorization program and that's what we considered b
l T
25' when we decided, do they need to come in the center within O
u l
\\
362
'r x 1
one hour or not.
(
)
\\~ /
2 The first one is a vendor who has been cleared and 3
authorized to receive a badge permitting unauthorized access 4
to the PA's and BA's, who inadvertently enters the protected 5
area through a-vehicle gate before being searched and before 6
being issued a badge.
The licensee discovers the event, 7
searches the individual, issues the badge and takes 8
corrective action to prevent reoccurrence.
There was no 9
malevolent intent, nothing adverse happened from it.
The 10 individual had been completely screened.
He was cleared, 11 and he didn't deliberately try to circumvent the system.
12 This event can be logged instead of being called in within 1
s i
).
13 one hour.
d
~-)
l
'\\
L 14.
The next one is search equipment that fails where l'
15 the licensee does not detect the failure immediately and 16 therefore people go into the protected area unsearched.
If
(
17 these individuals -- if they did not know the egr.ipment went 18 down and they know that they are subject to tb.e search and l
R 19 the licensee takes the appropriate measures to correct the 20 problem once he discovers that he had the failure, that l
l 21 means that he switches to other equipment available.
If 22 possible, he brings the people back. Whatever -- whatever.is' 23 effective for that type event'and nothing adverse happens, 24 then again, you don't have to call the report in in an hour.
.O,)
25 You can log it.
,1
363 1
Now, if the licensee discovers this search 2
equipment failure before anyone goes through, if he'c got 3
'five search trains up and one of them goes down and he r
4 closes it off before anybody goes through and sends them all 5
into the other four, then you don't have to log that either.
6 It's not a one hour.
It's not a log because you haven't had 7
an event happen that could have allowed the undetected or 8
the unauthorized access. You caught it and you sent them to 9
another train, so that one would not require a report.
10 If an individual who is required to have an escort 11 for a particular area inadvertently becomes separated from 12 his or her escort, but the escort or another person who is 13 authorized unescorted access recognizes the situation and 14 corrects it and nothing adverse happens from it, then that 15 can be logged.
If an individual has to separate from his or
-16 her escort to use the restroom that has a limited means of 17 ogress and the escort remains nearby with full view of the 18-egress area, no report is required.
Again, you haven't had 19 an event here.
You've got the escort right outside the 20 door. The' individual goes in and uses the restroom and comes 21.
right back out. No report, no log entry.
22 The next one deals with an employee of a
-23 contractor or a licensee who enters the VA improperly 24 without realizing -- a licensee contractor, without 25 realizing that the card reader is processing the preceding 1
I l
364 1
1 employee's card, or the employee walks in behind another 2
employee without using a key card.
This event can be 3
logged even if the employee was not authorized access to any 4
VA if there was no malevolent intent and the entry was 5
inadvertent.
And a lot of licensee judgment comes in on t>
these.
Each case is specific when it happens at your f
~I 7
facility.
If there was no malevolent intent, then you can
)
8 log the event.
You don't have to call the center.
9 The next one is about an individual who enters a 10 VA that he is authorized access to by inadvertently using an Il access control medium, either key _ card badge that was i
12
. intended for another individual who is authorized unescorted 13' accessito that area.
They are both authorized, it's a 14
' logged. event.
It is a system failure that you want to i
15 correct before something more serious happens but it's 1
16L certainly not a one-hour report..
17 The next one is about an individual authorized 18 only protected area access, who incorrectly is issued a 19
. badge granting vital area access but he doesn't enter, he or 20 she does not enter any vital areas or they. don't enter any-21 vital areas with malevolent intent.
It the individual is 22-incorrectly issued a badge but they can't reasonably use it 23-because it requires, say a PIN, a personal identification 24 number and they don't know that number and the' number is 25 required.
You have to hit'it on the keypad before you can l
p,
365 1
even enter the protected area.
The event doesn't have to be 2
reported or logged if it's promptly discovered and the 3
reason for it is, because it's not reasonable that that 4
individual could have figured out the combination of numbers 5
required, and if they had, if malevolent intent enters into 6
it or the individual is deliberately trying to defeat the L
7.
security system, then that becomes something different.
But 8
when you have a PIN just to get into the protected area and 9
they can't use it, then you haven't had an event that could 10 have allowed the access.
11 The next one involves improper control of access control media, including picture-badges, keys, key cards or 13_
access control computer cards.
And this included loss or 14 off-site removal.that could be used to gain unauthorized or 15-undetected access.
Proper _ compensation in order to log the 16 event instead'of calling it in within one hour would include 17 preventing l successful usa of the medium and initiation of 18-
_ measures to determine if the medium-was used during-the 19:
_ period that it was lost or off site.
If you do determine 20
.that it'was used during this period, then it would be a one-1 21 hour2.430556e-4 days <br />0.00583 hours <br />3.472222e-5 weeks <br />7.9905e-6 months <br /> report and the time frame would start from when you
- 22 discovered that it was used.
23; If the licensee determines that the medium could 24 not have been used to gain access, then you don't have to 25 report it and you don't have to log it. I'll give you some k'f.
366 1
' situations that would be ones that you would not have to yyI,)
2 report.
If.the authorized individual only momentarily takes o
1 3'
the badge outside of the PA and the event is immediately i
-4~
discovered and corrected by return of the badge before a
-5 compromise could have reasonably occurred -- and this is a 6
very quick turn-around, it doesn't mean that would have timo j
7 to take the badge off-site and compromise the badge.
This
{
8 would be where they walk out and they suddenly realize 9
they've still got it in.their hand or on their coat and they 10 turn right around and they come back in.
11 If the badge or key card is only momentarily L
12 misplaced and the. event is discovered and corrected before jO 13 anyone could reasonably use the device for entry, or if a 14 badge is automatically deleted from the system when taken 15' Loff-site, a new badge with a different access code is issued 16 to the individual involved upon re-entry and the previous
' 17.
code is not used again in your security system.
That's the-i 18 only time so far that we have determined that you wouldn't t
19 have to make-aireport where a badge is taken off-site.
It 20
.would'have to be where the code is deleted immediately, you 21 don't ever key that code in again and the individual gets a 22 new badge when they come back in.
1 231 MR. DUBE:
Nancy, I think we've just about used up
.24' our time.
g 25 MS. ERVIN:
Okay.
Well, we've only got two more jj e
w
367
- j s; 1
to go, so I might as well go ahead and give them to you.
(
)
. (._/ "
2 The next one is a card reader failure that causes 3
vital area doors to unlock in the open position or to lock 4'
in the closed position but with no functioning alarm, this 5
can be logged now, again, as long as the proper comp 6
measures have been taken and nothing adverse happens.
t 7
Further, if the card reader failure causes the VA doors to 8
lock in the closed position and the door alarms function 9
properly in the closed position, you don't have to report it 10 or log it.
i 11 The last one is involving pre-employment screening t
P 12 records and this is to include falsification of a minor
/
L13 nature-or inadequate administration, control or evaluation
(
14 of psychological tests.
Unescorted access of the individual 15 should be cancelled or suspended ~until the identified 16 anomaly is resolved.
If the licensee discovers that 17 unescortedLaccess would have been denied, based on 18 information that's developed later from these portions of
-19 the program that weren'tLdone completely, then a one-hour 20 report would be required, and that's as currently stated in 21 the reg guide in the NUREG.
22' I was going to go into the definition of one hour
'23 a little bit but Bob didn't -- Bob Dube had hit on that, so 24 I thought the time would be better spent to let you know d
Y L
(
)
25 that these are the things in the generic letter that will
.~
368 l
1
. allow you to reduce that one-hour reporting to logged
>h
'(
l 2
events.
l a.
L 3-
.Now, Joan is going to talk about the analysis 4
system.
L 5
MS. HIGDON - Good afternoon.
I'm Joan Higdon, L
6 Manager of the Safeguards Event Logs Analysis Program.
What 7
I would like to do is take a few minutes to give you some 8
background information on our program and its goal for those L
9 that may not be familiar with it.
It's a fairly new L
10-program, been in existence for about two years.
l 11~
The~ Division of Safeguards and Transportation has 12-responsibility for conducting and implementing the Logs 4q 13 Program ~.
Activities associated with this effort are the
[
Y i/
14 review and analysis of report events in a quarterly log and s
15 providing feedback-to NRC and industry cf analysis findings L16 and statistical data.
The goal of:this program is to serve L
L 17' both audiences.
The logs are a tool, one tool, to be used 1
18 forLimproving safeguard system performance.
LW 19 Emerging from this program thus far are a number i
of cases where event logs and feedback data resulted in 20
'21:
improved equipment operation or reduced human error.
'22' Additional staff resources have been dedicated to this 3
?
23 program which will enable us to perform a technical analysis 24 of the ovent data and further refinement of the feedback m.
- ' j 25 data, and of course the results of this will be given back L
1 1
.-~
-we
=
s n
369 1
directly to licensees and appropriate NRC staff.
t
. f'N.
[d f
I 2
The quarterly log events are categorized based on v
l-L 3
=the root cause of each event.
We are focusing on the l=
4 specific component that failed, type of human error or L-5
~ environmental factors.
Statistical data is provided to NRC 6
and the licensee with regional industry average in a L
7 quarterly feedback report.
The averages are not standards 8
of performance or the norm but are to be used only as a i
9 point cf reference to be evaluated in light of site-specific l,
10 characteristics, circumstances, facility design, equipment, 1
L 11 environmental factors, et cetera.
L 12 Although the numbers are useful in doing trend
.y-'s 13 analysis, we have noticed that they do vary significantly 1
7 14 between facilities because of site-specific characteristics
\\ --
15 and other factors.
Therefore, emphasis is being placed on 16 identifying'and evaluating the root cause of unusual trends f
171
-and reported events.
18 The event logs and'NRC subsequent analysis of the I
19
' event data is geared at being a positive' approach to 20 maintaining effective safeguard: system performance.
We want-21; to identify-possible. indicators of system weaknesses, focus E
22 inspection resources, but we want to do it as a positive 23 approach in a positive manner and to set NRC and the 24:
licenseeT2p to succeed, not to fail.
,N.(d The licensees-are using the data to perform a F
25 v
+,
l' 1
i' 370-1 self-assessment of their facilities, equipment and l
fs N_,
2' procedures.
This program will give the licensee an 3
opportunity to identify and take self-corrective action as a 4
result-of the. logs and feedback data.
I 5-There are a number of cases where the event log g
6 and feedback data highlighted a need for certain changes at l
7 a facility.
Many licensees are performing a root cause 8
analysis based on the event logs and quarterly feedback 9
reports.
The results of these analysis have resulted in i
10 modifications to certain equipment or security procedures i
l' L
11 which improved equipment reliability or reduced human error.-
l 12 The results are provided to' industry since these findings
'~'
D}
-13
-may have application at other facilities.
And some of the 14 specific areas that have been improved or changes have been
.15 made are-identified and listed in your handout.
-16 There are a number of new initiatives to this 17 program that are taking place.
Additional staff resources.
18-have been dedicated to the log analysis.
The data is now 19' undergoing a technical ~ review and issued as a companion to 20 the; quarterly feedback report will be a technical report.
21 And we will be focusing co specific areas, such as certain 22 equipment performance, environmental influences or security 23 procedures.
In addition, work has begun to normalize the r~%-
24 data, grouping like facilities together based on size, j( )y 25' population, environmental conditions.
Over time, the
a 371 1
quarterly report will be revamped to present the data -- to j-i
^
2 present a more realistic picture of what is going-on at a O
3 facility in their practices.
4 our staff is sensitive to licensee concerns and 5
needs with regard to this program.
For those of you that
/
L 6
attended the Orlando meeting, we appreciated the opportunity j
7 to meet with you, the individual security associations
,l 8
separately, and the feedback that we.got.
It was very 9
.useful, and we have looked at it and evaluated it and we are l
10 making changes accordingly so this logs analysis program can l
.11 be useful to both-NRC.and the licensees,
- 12.
I think we can entertain questions now at the g
ll /^}
13 panel discussion.
Thank you.
l Q) 14
'15 NRC PANEL DISCUSSION - INDUSTRY FEEDBACK ON 73.71. REPORTING 16 MR. BETRAY:
I told our' resident I was going to i
17-hush, but I'm'not.
L-18 First of all, I want to comr.end you -- I'm Steve-
,19 Bethay again, Georgia power - :on your efforts.
I think it 20' is going to be a big help to all of us.
j L 21; I'm not familiar with the quarterly report that 1-L',
22L you mentioned to-licensees.
.Is that something that you send l
23-out regularly?
l 24~
MS. HIGDON:
Yes, sir.
Where are you from, do you l f~)
!? (j 25 mind my asking?
l.
~
372 1
MR. BETHAY:
Georgia Power.
N) 2:
MS. HIGDON:. Okay, the quarterly report is a
3
-distributed'to all reporting licensees, NRC regional staff 4
and appropriate NRC headquarters staff, and at the corporate 5
level on an as-requested basis.
If you are not on the 6
mailing list, I'll be more than glad to put you on -- and 7
'this applies to anyone here.
You are not limited to a 8
certain number of copies at a particular facility or 9
-headquarters staff.
10
. MR. BETHAY:
Is that a safeguards document?
11 MS. HIGDON:
No, sir, it is not.
The quarterly 12
. feedback report identifies the facilities by facility code P-e 13 numbers, this is a random code number that was assigned so V) l 14
-the report, number one, would not become safeguards 15 information, and secondly is we did not feel -- meaning us 16 at NRC -- that:we wanted to identify the particular facility L17 by name, just as respect for your privacy.
You may consider i
18 it proprietary information or, you know, whatever.
Of 19 course if you want to share your number, of course.you're 20 responsible for that, but we didn't' feel comfortable doing 21 that. 'So we did the code number for those two reasons.
22-MR. BETHAY:
Okay, I'll~get with you afterwards to 23
.get on the ist.
24 MS. HIGDON:. Sure.
'25 MR. BETHAY:
The other question is an v
373 1
administrative one that we wrestle with a lot.
The
~s
,)
2 correspondence that we send in.to the NRC dealing with 3
security issues in general, lots of times we will very 4
conservatively classify the whole thing as safeguards 5.
information.
Does that present you a problem?
Would you 6
rather we didn't do that or would you just as soon we did?
7 MS. HIGDON:
Is this on the event logs?
1 v
8 MR. BETHAY:
On anything; event logs, LERs, plan 9
changes, TQP changes.
Would you rather see us a-10 conservatively call the whole~ thing safeguards or pick and 4
11 choose?
12 MR. DUBE:
We would leave that up to you, we've i
/j 13 spent a fair amount of time over the last six to eight weeks 14 trying to deal with that issue.
Some of our regional R15 offices have been concerned, there has been some management 16 concern within NRC that maybe some utility management have L17 not'been seeing information that they should be seeing l
18
'because it's identified as safeguards information and 19 therefore, the distribution of that'information is being 20 restricted because of that handling problem.
-i 21-We would leave that up to you.
We have no problem v.
22 within NRC handling the information, but there may be some-23 information that you would like your management to see that p,
would be more readily available to them if you limited the 24
.r9 I
i i
25-information a little bit better.
Q i
l>
D
l 374 a
1-MR. ELLER:
Ron Eller with Duke Power. Company.
-,__T i
2 I'd-like to encourage the prompt issuance of the
-3 generic guidance being prepared for the 73.71'reportability 4
and I'd like to ask if we can get a date that we can expect 5-that by.
i F
6 I would also like to ask that a separate workshop 7
be prepared for safeguards reportability coincident with the 8
release of this generic. guidance.
9' KR. DUBE:
I'm not sure we can give you a date.
V 10' There are a lot of initiatives right now that we're working E
11 on and-like everything else, we have to juggle priorities.
y 12 In addition, as you know, a NUREG has to go s
c rN g 13 through an internal review process so that even after we A
W
/
14' complete the document to our satisfaction at the staff 15 level,:there-will be an additional period where we'll go 16 through internal review.
We're also going to be publishing 17 it for public comment obviously.-
18 MR. NOVAK:
Any other~ questions?
19 MR.-TUNSTILL:
Yes, sir,' Jack Tunstill.with
!20 Florida Power Corporation again.
'i
, 21' It has to'do with the issuance of this ge.teric-
'22 letter.
I assume that it -- it sounds like, and I i
23 compliment you for at least going through it in detail -- it
- gA
-24 sounds as though it's reduced commitments for us more than 7.4
- i 25
-anything else.
1
,,~
. ~
375
/~w 1-Yesterday in the backfitting discussion, of course v
.I
)'
'\\^
2 we. talked about how generic letters would probably be made 3
available to-the industry while they're still in the draft 4
phase.
Has this thing -- as I said, it's quite extensive, I
5 from just sitting.here listening to it, and I'm not an 6
events reporting type person.
7 For our folks back home to do a little bit of 8
planning, because.obviously we're going to have to turn 9
around and change procedures, go through a little effort
~10-here, and if we had some of this information -- has it 11 alreadyLbeen sent out?
i 12 MR. DUBE:
No, no, it is not effective yet, it is p
L _,s~(.
(
13 still undergoing internal review. We hope to have that
'N_/
14 completed sometime in the near future.
't 50 15 Let me make two other points.
The intent is to
\\-
16 try to decrease-unnecessary'one-hour reporting to the ops
.17 center.. That's the reason we give this particular portion 18:
of the effort priority-treatment.
To make sure that:we're 19 not inadvertently creating problems for.you, our intent is-20 to make it voluntary so if there are any elements of it-that 21' you choose not1to take advantage of, that's an option that 22 you'll have.
23 KRe TUNSTILL:. Well I would ask, in the spirit of
. tu.
24 our backfitting program that we have where the utilitias are s
k' )N. '25 at least afforded an opportunity to look at generic letters m,
p g
376 VN 1
when they go up in requirements for us, that we also be
- l l '
.1 p
~2 afforded the same. opportunity when they come down in 3
requirements -- just a request.
4 MR. NOVAK: Any other questions?
l 5
MR. DUBE:
On that last comment, understand that 6
this is-an interim effort and that if you have any comments 7
on it, those comments would be addressed when we complete 8
the NUREG and the Regulatory Guide.
That same information 9
will eventually be' incorporated into those.
10 Looks like we may be finishing early here.
11 MR. NOVAK:
I don't see any more questions.
If
= 12 -
one does come up,lwe'll bring it up again, it doesn't have 13
-to bellimited to just this time, u,
114 What we were going to-do now, as the agenda l15 suggests, was open the' meeting up summary ~ discussion.
That 16
.gives you an opportunity to ask'any questions or bring up 17 any points that you've heard any time today, and why don't l18 we just see'if there's anything that people are'looking for.
I 19 And then I'll ask the panel members if they have 20 any what I'l1 call closing points that they would-like to i
~
21 make based on what they've heard today.
22-So'we'll start first with the audience.
I'm sure 23 our closing-remarks will spark some questions, but let's 24 again give you some opportunity.
Yes, sir?
l f
\\_,/
25' l
l l
377
- ,q 1
SUMMARY
DISCUSSION x/
2 KR. MOORE:
Brad Moore with Alabama Power.
3 On ESF reporting, let me just ask, is a partial 4
ESF' actuation reportable?
An example is this might be where 5'
maintenance is doing some work on a termination cabinet, 6
electrical junction, a short occurs and one valve is 7
actuated, a containment isolation valve or one valve in a 8
series, such that the real ESF piece of equipment doesn't 9
. actuate.. Would '; hat be reportable under the rule?
I 10 MR. WILLIAMS:
Again recognizing that, you know, 11 it's not.a final interpretation by the staff or anything 12 like that, there are some -- ESF actuations generically are j
< X..
j p
13.
reportable.
There are some. situations which may not meet N~-l
. 14' the definition of ESF actuation, I think you're pretty'close 1
,i
,15 to one.
So that may not be reportable under guidance that 16.
we issued, but I really think we've got to write down these 17' responses and1obtain some_ generic understanding of their 18 meaning or what the' exact facts surrounding that particular 19:
situation you've'got are.
'I think the' latest information on l
20.
ESF actuation'is that there is a staff position on it-and it b
21'
.would say_in general a partial ESF acs-ition is probably L~
-22 reportable.
If it's, you know, a half SCRAM, it's not-23-reportable.
So there is some -- due to the nature of our 1
24 business there are some situations which seem to be in the i
l
_j l
's 25 middle ground and we have issued guidance to take care of 1
r
,j.-
378 1
those in the past and I think we can answer your question f-i
\\s -
2 with guidance in the future here.
3 MR. MOORE:. I bring this up in light too of the 4
fact that an LER -- I'm sure you probably recognize this,
'5'
.but it takes on the other of two to three man-days to 6
prepare, it goes through, in our case, a full plant review
- 3 7
committee where you have top level managers, plant manager, 8
involved in the review, and going through the corporate l
9 office through a vice president for signature.
So there is 10 a fairly significant amount of time that an LER takes 11 through the plant's resources and we like to think that
.i 12 we're1using our criteria to help make sure that we're
'j5.
13 applying our resources.in the right place while still 14 supplying you the information you'need on the generic 15 matters.
16-MR. WILLIAMS:
Well I think:again, the first place 17 to do;it would be the resident, the region and then j
18-headquarters and if you don't get an answer, you.can just 19 give us a call directly and we'll certainly get you an 20 answer.
21 MS. TULLEY:
This'is the time for wrap-up 22 comments, right?
23 I guess I'd just like to start off'and I guess my 24' comments are on behalf of the BWR owners group pretty much b).
(
25 at this point -- I do think -- I'd like to thank you for 1
m
379 i
1 what we did here today because I think you have given us t
jest)..
1i_-h 2
some perspectives, a little bit more perspective en where s
3 you were coming from with respect to the rural.
I 4
acknowledge that we may not have liked everything that we 5
heard in that respect.
6 You noted at the beginning of this that the type 7
of things you wanted to discuss in this workshop were, you 8
know, why.do we have to report things that we have to 9
report, what should we really report and how can we improve 10 the process.
And you've acknowledged that you're probably 11 going to be putting out some new guidance.
12 In wrapping up, I guess the thing that I would
/
13 like to note is I applaud the idea of trying to get the A
14
. threshold-band a little bit better' defined.
I recognize 15 that there is never going to be a' fine ~line for determining 16
.reportability on any of these. issues,.they're very, very 17 difficult.
But I am ailittle bit concerned'with some of the.
.18 things that I've. heard, that-the way the. threshold band is 19
- going to be further defined is very much to the low end.
g 20 I would recommend -- or let me say one other 21 thing, the owners group has never expressed a. desire, per 22 se, to raise the threshold, we just want to better define 23 it.-
But what I would urge you is to continue your-talks 24-with us.-
I think we could provide you with a lot more
,s-k,
> valuable input on the struggles that we go through in making 25.
a,.
l 380 1
those reportability determinations and have further j_
}
b.
_,f[
~
2 exchanges in future meetings or something so that we
?
3 understand better exactly what it is you folks are really 4
looking for and then we can explain to you some of the 1
5 problems or many of the problems that we deal with in trying j
6 to identify that information for you.
I just want to urge 7
that there should be a real interchange between the two of 8
us before any.further guidance comes out.
9 It may even be the beginning'of what you 10 acknowledged is really a longer term effort, which is
'11 sitting down and taking a systematic look at what the rule 1'
I 12 does now and.I believe things have been brought up about 1
J(
13' looking at where we-want to be four or five years from now.
i)
14
'I-hope in.doing this guidance, that we aren't unduly 15 constrained by where we were six years ago.
We've-learned a 16 lot and I' hope we.take advantage of.that.
L 17 Thank you.
18 MR. NOVAK:
Thank you.
19 I'm goin7 to ask the panel if they have some, what t
'I'll call at least closing observations.
Lou, do you want l-
!?
21.
to start?
22 MR. REYES:
This is Luis Reyes with Region II.
23 There's a couple of. points I'd like to make.
I don't want i
24 to get-lost in the all day conversation.
g
.(
25 I personally have had the opportunity to review
(
i i
1<
em,e
381 l
every 50.72 report in.this region for the last four years, 1
91 2
so every one of them comes to my desk and I review them.
I 3
wanted to make a point that overall I feel the utilities in 4
the southeast are complying with the rule and are reporting 5
at the threshold that we want.
There are some examples 6
where we have differences of opinion and I think you have 7
found that we talked to you, we brought it to the table and 8
at least for the time being we resolve it, realizing that 9
there are some differences of opinion.
10 Another issue that was brought up today was the 11 one-hour versus four-hour reporting, and I want'ed to e
12 emphasize there are occasions where you in retrospect report
. g-,
j 13_
four hours and upon your review and ours'it should have been
'14 a one-hour report, and there are some cases where you
,l 15 attempt.a one-hour report and'you don't quite make the 60 16
- minutes, j
17' I hope you will agree that in terms of enforcement 18' we use as much discretion as we can te make sure we don't 19-count minutes-and look into your attempts to have an honest 20-effort to report on' time.
You'll find.out my staff comes to 21
.my' office with issues such as that, you' reported in four 22' hours, upon further review.it should have been a one-hour 23 report and we use as much discretion as we can with
.24 enforcement if your efforts were such that you attempted to f_s 25
~do a good job.
u
- a
1 382 l'
The last point I wanted to make was you talked 3p_
Y
'q_,/j 2
about other agencies using the LER accounts and issues like 3
that.
I think Ed Jordan mentioned that our rules are
.t 4
intended to satisfy our needs and we're not planning to l
5 change our rules for public utilities commissions, but we 6
are trying to do a better job in the SALP report to project 7
if you have a low threshold reporting, if you submit 8
voluntary LERs and give you perhaps a viable mechanism to 9
show other agencies that we feel you not only are complying 10 with the rule but perhaps for the bean counters it's wrong 11 to count the beans because you're meeting a lower threshold 12 than the 1 rule requires.
We didn't do that in the past,
/"'9 13 We're trying to do that now.
J 14
- 10R. NOVAK:-
Pierce.
t 15-MR. SKINNER:
I only have a couple of comments 16 that'I would like to make.
' 17 '
.After listening to most of the conversation today,
'18 I think a good majority of the questions that come up as far 19 as reportability can-be handled by-the licensee's L.
20' communications directly with the resident inspector staff.
L
=
I L
21 The staff, when they'have a problem, know where to go to, to i
122
-help you and assist in making a determination as far as
-23 whether an item is reportable under the one-hour, four-hour 24' or not reportable at all.
So based on that, I guess the i,Vj 25 thing that I would like to stress is that you establish a i
w o
+y-
383 1
good communication with_the resident inspector staff.
.,esq s/.
2 MR. NOVAK:. Thank you, Pierce.
3 Why don't I go on the other side of the table.
4 Ernie, did you have any comments you wanted to make?
[
5 MR. ROSSI:
Yes, I think I do.
From my 6
perspective, I think the 50.72 reporting is working quite 7
well.
I don't think there are a lot of things that need to j
8 be reported that are not being reported, and furthermore,'I
]
9 think that all the issues that we need to know about, we're 10.
finding out'about, either through the reports or through the 11-information that comes from the region.
12' And certainly my intent in coming down here wasn't
[]
13' to try to_ lower the reporting threshold and I think you need LJ 14
_to use the NUREG guidance and letters that you may have 15'-
received as the guidance on that.
I think my main intent 16:
was-to try to tell you how we use the reports and why we 17 wanted to hear about the problems, to sort of alleviate some 18
. frustrations that you may have in having to do the reports.
j 19 But from my standpoint, I think the system is 20.
working well and I-think the raports get reviewed well.
21-I don't think I have anything~else -- oh, one 22 other thing -- I do think it's important that we look at 23 ways to try to eliminate some of the things that we clearly 24.
don't need.
I think the RWCU actuations and reactor trip b['.
i, 25
-system actuations when the plant is shut down and the rods
+#
e-
- +, -
4
384
~s 1
are already in, that we ought to try to eliminate those and-j p
s/
2 I would hope that we can find a large majority of those that 3
we can eliminate relatively easily, and I think we do need 4
to do that.
I'm also convinced now that we probably need 5-some more guidance on what does and doesn't constitute an 6
ESF actuation in terms of partial actuations.
And I think 7
we need to do that in a very systematic way and go through I
J 8
our CRGR process so that we have an agency-wide position on 9
it.
10 MR. NOVAK:
Jack, do-you have any comments?
11 1 MR. CROOKS:
Ernie picked up the few things that I
. 12-was going to say regarding the rule changes and the guidance s.
'v
(
J 213 that -- you know, I think we've covered that we plan to do V
. 14.
it in a systematic and open fashion and try at least to 15 accomplish I:think the goals that we're all trying to get 16 to.
I 17' MR. NOVAK:
Mark.
18 MR. WILLIAMS:
I just had a couple of comments.
19 First'of all, I hope that we've shown today that
~ 20~
'there is a system set up for the feedback of experience.
21-
-The reports that are sent in to us, we use.
Ernie described l
22 a daily and a, weekly and a systematic review of all the 23 reports, and we.have engineers looking at all that 24.
.information and you're seeing feedback and studies and
,_s 25-noticos and a lot of-information over the past few years.
385 1
That system is working for you, we're part of it, we work k.
2 hard every day to try to make it work to pull the things out 3
of.those LERs.
We spend a lot of time and res'.ources on it 4
at the NRC.
5 One of the interesting areas that has come up is 6
'the outside design basis and the unanalyzed condition.
It 7
turns out that that's a batch of LERs that we just got done 8
looking at in some detail.
We had Idaho pull a whole slew 9
of those.
Those are some of'the most interesting conditions
-10 that we find.
Many of them are not counted in any 11 performance indicators or anything else, but that kind of 12 information really should be the focus.
If you remember the A{.
13.
pie chart,-there are only a few percent that were in that
-!v.
14 category and we expect to see more and the design basis 15-
. reconstitution programs will give us probably.more of that 16' kind of information.
17 But the areas where engineering judgment is really
- 18..
required are some of the sticky areas on reporting.
And' 19 balancing reporting with corporate pressures and goals is 20 something that we have to live with in our business.
21 I hope we've at least shown you tha'. we.are
-22 reviewing information.that you send in and will try to use 23 it to the best of the industry.
24-MR. JORDAN:
Okay, I just have a few comments and V
25 I'd like to support Luis' comment that generally we are
1 386 1
quite satisfied with the level of reporting.
That is not a 7
l'wr}
2-big issue.
I think in the discussions we identified some 3
areas where there have been failures, but they're the kind 4.
you count on your figures as opposed to big areas that need 5
big changes.
And those are handled on a case-by-case basis.
6 I would urge, within your own organizations that 7
you communicate the philosophy for reporting so that your
'8 own staff' understand that the NRC-is really not in a bean-9L counting mode with regards to event information.
We really 10 are looking for the important stuff, to use that term one 11 more time.
That information that may have safety relevance 12-
.at your facility, for your own reviews, that may have safety
/O 13 relevance for your neighbors' review and'that may, through
q,l
\\
- 14 INPO and NRC review, provide some perhaps obscure lessons
.l M
- 15 that can'be then fed back to all utilities and improve-the 16 overall safety at the plants.
And.so if your folks 17 understand that we're not really asking for reports because 18 we count beans, that would help a lot. -Ar.d I'm not sure 19 that's' apparent to everyone.
20 We. certainly do feel frora this and our previous 21 meetings there is a need for us to adjust both requirements 22 and' guidance in order to use your scarce resources and ours 23 more effectively, and to-eliminate-the non-useful reports 24 and we'give-you a pledge that we're going to work on that.
25.
And that is in months, not years.
1 I
387 1
We feel that industry in the meantime must i
s-i y/:
21 exercise care so as not to attempt to change reporting 3
requirements through guidance.
We have to work those 4
together.
We can't change the requirements through 5
guidance, we can interpret, as long as we're within the 1
6 bounds of the requirements.
I 7
For the NRC, I think at the next workshop, I want 8
to make sure that we make a little stronger ties between the 9
lessons that we've learned from reported information and 10 submitted reports'so th'at you can see the relevance of the 11 reports to our lessons.
I think that's an important 12 element.
We're telling you how we go about doing it, but we i
r~N 13 haven't given you enough examples of the cases, and every f
s V 4
}-
'~'
14 piece of generic correspondence that comes to you; 15' information notice, bulletin, generic letter, has behind it 16-a series of events that have been extracted through the very i
17-process that we talked about.
We didn't_give you enough 18 examples,.I think, and-maybe those of you that attended the 19 backfit workshop yesterday, we talked about a number of 20 those pieces of generic correspondence in detail.
And if I 21 had-it to do-over.again, I think I would probably have the j
~
122 sequence of the two workshops reversed.so that we would talk 23 about-the event reporting _first and go through some of those 24-instances and then talk about the backfitting that fy i
' t 25 frequently in fact comes out of the information we learn e
e
- e --mee f
=
388
~
1 through the events.
- y s.
s/
2 I think a lesson to industry that I hope maybe the 3
next workshops -- we won't be able to do it next Monday and l
4 Tuesday in Region I, but perhaps in subsequent, is to have a 5
little more participation on the part of the industry about 6
' how you use event-related failure to report inforr ation.
I 7-participated in a workshop a week ago -- it was actually a 8
meeting, a topical meeting, ANAS SME, in which there was a 9-discussion about the industry's use of operating experience.
10 And I found that very beneficial.
I was impressed by the i
11 utilities that were represented and the programs that you 12-have~and how you're using operating experience information.
E13 Duke-Power was one of the utilities represented in this 14 particular panel that I was on.
So I think this sort of a L
15-group might benefit from a couple of those discussions and 16 I'll_try to, for subsequent workshops, bring one of those s
17-people into giving a discussion.
18 I think. lastly I would say it has been a very
' 19
. candid and open workshop.- I appreciate the opportunity to L
1 20 meet with you,'and clearly the thing we have to keep doing p
21 and do better is communicate.
If you realize that our 22 purposes are in fact'to feedback information to you, to 23 receive the kind of information that we can extract lessons 24 from so we can feedback to industry and so that you can j'^]
~
that's the goal and that we can revise i
(_,)
25 extract it yourself, 3
~
. ~... _...
e
389 h
1 the-reporting requirements and the guidance so that wc're 2-all more effective in doing that.
'3 We didn't develop and don't plan to develop 4
requirements so that'we can, for instance, assess licensee 5
performance.
That's always a byproduct of any of our tools 6-in.that regard.
i' 7.
That's the end of my comments.
~8-MR. NOVAK:
Let me ask Bob, Nancy or Joan if you 9.
people have anything you'd like to add.
10-(No response.).
11-MR. NOVAK:
Well-based on what you just heard, we 12 certuinly have time to entertain any statements that you E]
-13 would like to leave for the record.
The direction of these
(
g L14 workshops are flexible.
I think we're a little bit locked
-15 into'.what we can do next Monday and Tuesday, but-I think we 16 want to provide the information that is most meaningful.to 17 you, so we would look at that and if y'ou have some ideas on l
18 what.you really thought you might want to hear differently, 19 certainly we can take the time and explain some of our
~
20 reviews that we do and the results that come out.
This is 21.-
what Ed was referring to.
In fact, it was interesting, just
.22-to take a minute, that even in an independent review of a 23 research activity, there was mention that it's important 24 that operating experience be disseminated as broadly as it 7
1 i
25 can.
We try to'do that.
We -- actually we provide this
.. =. -
390
-1 information internationally.
We try to help and look for
/
\\
(s,,)U 2
help from other countries.
We are learning more and more 3
from other operating reactors outside of the U.S.
4 I don't think there's a need to drag it on.
I 5
certainly appreciate your efforts here, I appreciate the 6
questions and -- there was a question about how one could 7
obtain a copy of the transcript.
I'd like to suggest that 8
if you want a copy, certainly you can call me back at the 9
-office next week and I'll know how it can be obtained.
I'm i
10 sure there's a nominal cost.
There will be copies of it 11~
placed ilt the PDR, so if that's the avenue, we'll tell you 12 when it will arrive thare and then you can obtain it from
(N 13 our public document r..
in Washington.
We.may make
\\)-
l 14 arrangements where it can come through in some other
-15 approach, but if you do want it, you can call me.
Or if
.i 16 there are.any other questions regarding this, the area code 17 is 301-492-4484.
We'll be in the office certainly on Monday 18 and then most of us are. going up to King of Prussia Monday.
19 afternoon, so we'll be out of the office Tuesday, but we'll 20
.be right back in.
21 We do appreciate it.
Cindy?
L-22 MS. TULLEY:
When is the transcript going to be 1
l l
23 available, how: soon, do you know?
24 MR. NOVAK:
If I can get a hand signal from the
[ '}
! (
j gentleman who is listening.
25-l
391 4
1 THE REPORTER:
Five days.
[
2-MR. NOVAK:
Five working days, that's not too bad, 3
that's very good.
4 Again, thank you very much and we'll see you soon.
5 (Whereupon, the workshop was concluded at 3:51 6
p.m.)
7 8
s' 9
10 11
~
12 13 14
.. 15-16'
' 17 18 19 20 21 22' 23
.24
- 25
l REPORTER'S CERTIFICATE This is to certify that the attached proceed-c ings before the United States Nuclear Regulatory Commission in the matter oft NAME OF PR0CEED1NO D,'ent Peporting Workshop DOCKET NUMBER:
PLACE OF PROCEEDING Atlan+a, Georgia were held as herein appears, and that this is the original transcript thereof for the file of the United States Nuclear Regulatory Commission taken by me and thereafter reduced to typewriting by me or under the direction of the court report-ing company, and that the transcript is a true and accurate record of the foregoing proceedings.
TU--
Official Reporter Ann Riley & Associates. Ltd.
0 4
5 d
f0 A$ fO h
seem ammeer.
enants Date BCS P/ 37 PDR g
I L
Pee mese ans neewm b
Per Dessense Per Ceavoresten
[
h-
^
PerOsnesten poseese topsy Per Year lasormemen Seesee l
h8*
Juegh REAAMS
$dfff /*fAA'S fNd SA/[LddO MAMA /dl /W54/CL Y AVJ//44/[ /d THE PM Y$44/$CA//f MS MAC Ct/dyr 4tAeg7,4sg gypgy,g,ynp 1
4 74 4 W 4, d'A 1
'A#7.2. 9, /99o
/],
PAddS 3/9-3 y/
~
CC (pters eMLt')
,)EA4/ AA7HJf pp/t 4 C7M/7 DO $107 eee this kem es e RECORD of aessents. eeneemmeos, sesonata, emeronees, and seminer setens M Wepens, org, symeo!. Agency / Pest) l Mesm he.--Sieg.
DENU/S jf LL /Sog/
- enene nu i
2 INVf eas4.am P0ftM 41 (nev. 7-76) rPu een CbE-11.806 h 5. coo.u.a..>....
sp.
I i
r
.......