ML20137E129
Text
El5.h0ADDIb h
8r urg
(
(
4 UNITED STATES
'
- Y'
^'j NUCLEAR REGULATORY COMMISSION M
r ADVISORY COMMITTEE ON REACTOR SAFEGUARDS g
W ASHINGTON, D. C. 20555 g....+
June 5,1982 l
Dr. M.hailo D. Trifunac 1488 Old House Road Pasadena, CA 91107
Dear Mike:
Two questions have arisen in connection with the response you gave to i.
l my question concerning recurrence frequencies of increasingly severe l
earthquakes at the Midland site..
1.
On Page 336 when you use the term return period, is this per year?
2.
Richard Holt mentioned that when they tried to go from your assumptions, as stated in the transcript, they had difficulty getting the results you presented.
(They appeared to expect smaller exceedance probabilities).
Do you have any comment?
Sincerely.
s David Okrent
- s.x.., c w.,
.s L(hkg.sg/
") w(< w C6+
1
~
8508230147 850722 PDR FOIA DELL 85-363 PDR fatt 'l
l
?
I I
i e
23 MR. OKRENT:
I thought some levity might help.
l 24 I am going to go on to the next part of this seismic 25 discussion in a moment; but let me forewarn the seismic l
Co'?*'maq O P D F O P T tir.
e r use r c e
,r n
,*. c.
- t C,3yp % %fd, 2'^'"---
.(
i 234 experts here, meaning from the Applicant and the Staff and the ACRS consultants.
At the end of the discussion on the seismic safety 6
issue, I will be interested in your considered opinions on what you would give as your expected value for the l
e earthquake acceleration tied to REG GUIDE 1.60, so we have a certain spectrum in mind, which has a probability of nonexceedance of 99 or a probability of exceedance of one in a thousand, let's say, a probability of exceedance of one in 10,000 and one in 100,000 per year; okay?
Is'the question clear, so you have time to reflect on it?
- j (Laughter.)
- 4 MR. OKRENT:
By the way, I will also be interested in what you would say if people asked you to
't give a number for which you could be 90 percent confident; and I will accept as an answer you have no basis for giving a high confidence level; but, I mean, that tells
'i one something about our problem.
U However, I would prefer that you try to play that game, also.
It's not a game, really, because it's fairly fundamental to the question that' we are trying to get at.
- ?
Why don't we go on, then, to the engineering part of l;
E' the discussion?
~
i 6
+..,,
f
(
6 I do want before we recess for tonight to come back 7
to the question I posed earlier to the seismologists.
8 Namely, what is their estimate of that earthquake 9
having a return frequency of a thousand years, 10,000 10 years, a hundred thousand years?
11 Will the spokesman for the Applicant start off?
12 MR. KNIGHT:
We had hoped, the seismologists and 13 myself and the others who had thought that, perhaps, the 14 meeting, before it would end tonight -- had hoped to be 15.
absent tomorrow.
We have to drive to Detroit, which would 16 mean that we would have to leave fairly early.
17 Is there any way that we could satisfy your 18 requirement, either in writing or to have a member of 19 Staff, which, certainly, would be represented here, to 20 pass their judgments onto the other members of the Staff?
21 MR. OKRENT:
I am not sure what it is you are 22 requesting or proposing.
23 MR. KNIGHT:
I understand you would like us to
(
24 be here for the opening part of at least the meeting 25 tomorrow.
O
4
-g j
315
( _
1
'MR. OKRENT:
I was hoping to get the answers --
2 THE VOICE:
Tonight.
3 MR. OKRENT:
-- this evening.
4 MR. KNIGHT:
I am sorry.
5 l
MR. OKRENT:
I said before we recess.
If we 6
can, we will.
l
.7 So, again, I will give the Applicant ~ first 8
opportunity.
9 MR. KLIMKIEWICZ:
I just happen to have brought 10 some slides.
11 (Laughter.)
j I
12 MR. KLIMKIEWICZ:
I have a response to your 13 question.
I am not so sure that it's an answer.
'('
14 (Slide.)
15 Shown up on the screen now is results that we had 16 given in a draft report.
These represent a family of 17 seismic hazard curves to be used in a seismic PRA.
18 Now, I believe what needs to be discussed is how 19 this band of curves is developed.
20 Several sets of input are used in order to develop 21 these types of curves.
22 Initially, the seismic hazard curves for modified 23 intensity that I showed earlier I used as a basis for
(
24 developing these sets of curves.
25-Also, we had used subjective weighting of models,
(
C o ll * ** 1 r.
C C C A D*f*T V O C FDTTTFr,
f"M
(
(.
316
~(
1 the three tectonic models that I had described early.
We 2
had favored model one, weighted it as
.5, model two at.3 3
a'nd model three at
.2.
(
4 The intensity model curves were then transferred 5
into sustained accelerations using the data published by I
6 Nutley, 1979 I
7 I believe part of the problem with defining-8 confidence levels is in a type of data that is available --
9 if I can take a minute to change the slide --
10 (Slide.)
11 12 MR. OKRENT:
It's not every subcommittee that 13 has this type of a meeting.
(
14 MR. COOK:
This one could go down in history.
15 (Whereupon, there occurred loud volumes 16 of music, generated by the band Radiation.)
17 MR. OKRENT:
I hope the reporter is getting 18 everything.
19 (Laughter.)
20 MR. KLIMKIEWICZ:
Shown on the slide now are the 21 data of Nutley, 1979 22 As you can see, the data are very limited at 23 intensity 8 and higher earthquakes; and for determining
(
24 probabilities at very low annual frequencies you really 25 need to examine the types of ground motions at intensities k
C f %? "# $C E I D O D T T IIf,
- EDUTf7,
{ P f,
(
l l
t l
317 l
(.
1 8, 9 and larger.-
l 2
In order to develop the curves that were shown in t
3 the previous slide what we did was fit a model to this 4
type of data, assuming a value of -- while fitting linear 5
models through the median values and the standard 6
deviation values, the standard deviation values are shown 7-as dotted lines.
i 8
Also, having fit that type of model through there, 9
it illustrated that maxima for this type of data fell out 10 about two-and-a-half standard deviations.
That was used i
11 to define sort of an upper range considered in the family 12 of curves shown earlier.
13 If you can put the other slide back up, pl e as e'.
(
14 (Slide.)
15 Using the distribution that was fit to the Nutley 16 data, the weights for the three models and also the annual 17 frequencies which were determined in the original hazard 18 assessment, we had.to assign weights of probabilities to 1
19 all of these curves, and their weights were adjusted such 20 that the cumulative distribution was one at the outside 21 curve.
22 We interpret these-as probabilities rather than 23 absolute confidence levels that we are modeling to a
(
24 reality situation.
25 From the standpoint of this kind of curve, the 90th
(
g O $
$h O
I N
(
'(
318
(
1 percent confidence is pretty much defined by this solid 1
2 line, and that intersects the SSE shown as a blue line at 1
3 about -- at close to ten to the minus four.
4 So using this type of an approach, it sort of 5
verifies that at fairly high confidence the SSE. level is 6
close to ten to the minus four; but, again, we are dealing 7
with limited data in making this type of assessment.
8 One other aspect of doing probabilistic assessments 9
is that normal assumptions made in calculating 10 probabilities result in curves that generally 11 over-estimate the historical observation, even at low 12 levels of motion that have been observed, well, 13 frequently.
(
14 You can see here that the probabilistic results 15 shown on two correlations of peak acceleration to 16 intensity, those of McGuire, 1977, and Trifunac and Brady,
'17 1975, show that the probabilistic results are conservative 18 when compared to the historical estimated ground motion 19 levels at the site.
20
_(Slide.)
21 This is seen here, plotted against the OBE.
Ground 22 motion levels are the estimated ground motions resulting 23 from an intermediate earthquake shown down below and a k
24 Parkfield earthquake, which pretty much represents an 25 estimate of ground motion resulting from the 1812 New k
e n " " ?.a q o r o o r t I t'C S e c u I c g, L 7 9,
(
/
l i
l 319 i
(
1 Madrid earthquake at the Midland site.
2 So I guess the answer is that you can extend 3
probabilistic assessment to this degree, but whether it
(
4 represents absolute confidence has yet to be resolved.
5 Clearly, you can make many more assumptions about l
6 tectonic models and ground motion models.
7 MR. OKRENT:
But this is what you are proposing 8
as an answer, if you have to give one; can I phrase it 9
that way?
10 MR. KLIMKIEWICZ:
Yes.
11 (Laughter.)
12 MR. OKRENT:
Okay.
Thank you.
13 (Laughter.)
~
14 MR. POMEROY:
Mr. Chairman, might I just ask, 15 what is the answer,7 I am sorry.
16 (Laughter.)
17 MR. HOLT:
What is the question?
18 (Laughter.)
19 MR. POMEROY:
Can you speak into the mike?
20 MR. KLINKIEWICZ:
Sure.
I believe the answer is 21 that given the assumptions that we used in this analysis, 22 namely, the Nutley data which correlates sustained motions
~
23 versus intensity, the Cornell McGuire approach for k
24 calculating probabilities out of site, also our subjective 25 weighting of tectonic models which is based in part on
(
.~1.,,,.
. e o o m r.,c.
o r o,,1 c =
1 -,
_., _ _. _, _ _ _ _. _. _ - _ _._,.__,. ~___ _ ___ ___
f t
320 l
1 eastern seismology as well as seismology in the central U.
i 2
S. region, also based on geology.
I I
(
3 Given that, and assuming all of that information suchasthatdefinitionofannualfrequencyofsomegroundl 4
I motion is weighted, this is a representation of that, of a !
5 i
l 6
synthesis of that seismic hazard analysis.
7 It represents the-distribution in ground motion 8
versus intensity and also the uncertainty in defining 9
seismic recurrence rates for tectonic model and also the 10 effect of variation of upper magnitude, because we assumed 11 the larger upper magnitude for one of the source models.
12 MR. OKRENT:
Again, just so we distinguish 13 between what it is when I talk to my graduate student and
(
14 I say, "Oh, just do a. perimetric study because we are only 15 interested in methodology,'" here this is your best 16 judgment that you are giving us as well?
17 MR. KLIMKIEWICZ:
That's right.
18 MR'. OKRENT:
Thank you.
Were you answered, Dr.
19 Pomeroy?
20 MR. POMER0Y:
I am not sure.
I think your 2 'i question was what is the earthquake in a thousand years, 22 in 10,000 years, in 100,000 years?
23 J4R. OKRENT:
What he did say is you can read k
24 that curve and from it have 90 percent confidence the safe 25 shutdown earthquake, which looks to me like.12 G, is a
(
e m....
y.
, c o r o,1,ic.
< r o,, r e r
,-m
(
321
(
1 ten to the minus four earthquake.
i 2
Did I read that correctly?
3 MR. KLIMKIEWICZ:
That's right, that's right.
(
4 MR. POMEROY:
Fine.
5 l-MR. OKRENT:
Then, of co,urse, you can read it 6
ten to the minus five, ten to the minus three.
Okay?
7 MR. POMER0Y:
Fine.
8 MR. OKRENT:
Ten to the minus five at 90 percent 9
confidence.
As I can eyeball it from here, it's of the 10 l
order of.17 G.
11 Is that a fair reading?
12 MR. KLIHKIEWICZ:
1 can't see from here but it's l
13 probably so.
(
14 (Laughter.)
15 MR. OKRENT:
Thank you.
16 MR. OKRENT:
Dr. Trifunac.
17 MR. TRIFUNAC:
First of all, this is sustained 18 acceleration.
This is generally not acceptable type of 19 acceleration.
Only some people use that; most people 20 don't.
21 The.other thing is that if you look at these curves, 22 they all bend and they all reach a saturation level, which 23 I cannot explain with confidence that is very high, what k
24 it means; but I suspect that it is coming from the fact 25 that model attenuation function has a flat cutoff region
(
e -,, --, e o r o n e - 1, n o r o,,
(
l 322 I
(
1 for the source; and there is no data for that, either.
2 So you are really reading numbers from parts of the 3
curves that are highly extrapolated and you don't have the 4
data.
5 MR. KLIMKIEWICZ:
I believe I said earlier that 6
we assumed a truncation at 2.5 standard deviations.
I 7
HR. OKRENT:
Dr. Trifunac, I am going to suggest 8
that for this part of the meeting, it's fair to ask what 9
is on the viewgraph if someone uses a viewgraph but not 10 why.
~
11 (Laughter.)
12 MR. OKRENT:
What I asked was that the that 13 people give their opinions and I didn't say we were going
('
14 to ask them to justify their subjective judgments.
15 It would be very interesting but that would go 16 beyond the call of duty, sort 6 of.
Thank you.
17 Let's go to the Staff.
18 MR. KIMBALL:
I also have a viewgraph.
19 (Slide.)
20 In the spirit of Dr. Okrent's request, Dr. Reiter 21 and I wi1~1 both offer our individual professional 22 opinions; and this is not to be taken as a Staff position.
23 (Laughter.)
(
24.
MR. KIMBALL:
I must say in the review of the 25 Applicant's seismic hazards analysis, we did not make a
(
er.:,3 r. o ron omr. e een ce 1 -.,
(
(
l 323 f
1 definitive enough review to come up with a specific return 2
period for the SSE.
3 However, we did review it.in quite a detail, so we
(
l 4
can make some judgment on that alone.
5 l
In addition, in my opinion, the most definitive 6
study that has been done on the seismic hazard is that 7
done for the systematic evaluation program.
8 Quickly, in this viewgraph the slide shows the two 9
closest sites and the sites in Michigan, the Palisades 10 site, which is the closest, and the Big Rock Point.
11 Even though it looks like an open triangle, that is 12 characterized as the one th,o'usand year uniform hazard 13 spectrum.
If I were to approximate that, it's probably
(.
14 somewhere between 5,000 and 10,000, and the Staff says 15 it's on the order of a thousand to 10,000.
16 Based upon looking at the Applicant's hazard 17 results, looking at relationships such as Murphy and 18 0'Brien and Trifunac and Brady, not 1 coking at sustained, 19 looking at the terra results, I would say I have moderate 20 confidence that it's in the order of a thousand to 10,000, 21 probably closer to 10,000 or ten to the minus four, 22 essentially concurring with what George says there.
23 Now, to extrapolate into the realm of ten to the k
24 minus four to ten to the minus five, I would have low 25 confidence in what I am about to say.
No numbers.
(
l l
---==-... ---..--
f f
324 s
l
(
1 i
Probably somewhere in the order of intensity 7 dash 8, 2
which depending upon which correlation between 3
acceleration and intensity you were going to use, and
[
4 would probably range somewhere between.16 and.2 G.
l 5
MR. OKRENT:
Okay.
Thank you.
l l
6 l
MR. REITER:
My name is Leon Reiter from the l
l 7
Staff.
8 Just a brief comment mentioned before with respect 9
to Parkfield.
We can, of course, go into great detail as 10 to why.our decision about Parkfield as to exclusion, 11 non-exclusion was taken, but I just want to assure you 12 that we did give this a great deal of consideration.
13 Our conclusion was not that Parkfield represents the
(
14 typical record for this particular area but that given the 15 uncertainty and the source conditions for eastern 16 earthquakes and the uncertainty in Parkfield, we think it 17 appropriate to include it within the range of records used 18 to assign a site-specific spectrum.
It is not 'to say we 19 are rejecting the argument.
20 Right now we are in the middle of funding a research 21 program to examine this item specifically, the effect of 22 surface or near-surface rupture upon records.
We' hope to 23 have some results within a few months.
k 24 With respect to the probabilistic esyimates, 25 although we did our results entirely separately, I agree
(
S c'"*
t e; 9EPOPTIS'O SEPVTre. tvn.
l
(
1 I
l l
325
(
1 with Jeff that if I had to make a very good guess, this l
2 would be somewhere in the order of ten to the minus three 3
or ten to the minus four, probably a lot closer to ten to 4
the minus four.
i 5
With respect to ten to the minus fifth, Iwouldlikel i
'6 to paraphrase the response of Jim Devine of the USGA.
7 The chance of me giving a probability at'this 8
probability'is very low.
9 (Laughter.)
10 l
MR. OKRENT:
Dr. Pomeroy, can we get your 11 opinion?
12 MR. POMEROY:
I don't want to belabor the point.
13
.I think that simply I would want to state first that
(
14 there are great uncertainties in what we are dealing with 15 here, as everybody is now reasonably well aware.
16
.There is no need to go into that much further, but I 17 think there are uncertainties in almost every parameter 18 that goes into a probabilistic analysis.
19 Those uncertainties are generally orders of 20
-magnitude; and as long as they are orders of magnitude --
21 for example, the slide that was shown earlier that showed 22 you an. order of magnitude in standard accelerations.
23 As long as you have that kind of uncertainty, it's a
!(
24 concept of predicting an exact acceleration or an exact 25 intensity and I think is nothing more than a sort of
(
i CfHHT!C DIDODTYUn CEDif*F7 TPD
{
326 l
1 scientific gut feeling when you come down to it.
2 I am really very reluctant to do that; but as long 3
as the Chairman has asked to do that, I will do that.
4 I would say that if you are talking about ten to the 5
minus third, you are probably talking about an intensity l
f 8, plus or minus one unit.
6 7
When you talk about ten to the minus fourth, you are 8
talking about intensity 9, plus or minus two units.
9 When you are talking with ten to the minus five, you 10 are probably talking about intensity 10 plus or minus 11 three units.
12 That says that there is no real estimate there.
13 That's all I wanted to say.
\\
14 MR. OKRENT:
Dr. Trifunac, do you want to 15 comment today or tomorrow morning?
16 HR. TRIFUNAC:
I can comment tomorrow.
17 (Laughter.)
18 MR. OKRENT:
Dr. Trifunac, walked to me during 19 the break and said if I gave him to tomorrow morning, he 20 would work for an hour and come up with an estimate based 21 on some arithmetic.
22 So I am going to give him that privilege.
He is 23 going to tell us before he leaves, k
24 So that you don't miss too much more of the 25 festivities, let me suggest what we might plan for
(
en,m.n eeenetron sreurce i,
(
i l
327 f
1 tomorrow.
~
2 Clearly, as I indicated earlier, we cannot complete 3
the original agenda.
However, I think it was useful to 4
devote this time on seismic, because I hope we don't 5
really have to do any more on it in detail at the full 6
committee meeting, either, just a summary.
I 7
I would suggest that we pick up the item on AC, DC 8
first thing tomorrow morning and then pick up the items G 9
and H.
10 I wasn't sure whether that was going to be short or 11 long, but possibly Item I and then as much of K and L as 12
.we can in the morning.
That would be my intent.
13 After that we would see where we were.
You know,
(.
14 that might take us through the agenda time; or if our 15 experience is any indication, it might take more than 16 that.
17 Are there any comments on that proposed revision of 18 the agenda?
19 (No response.)
20 MR. COOK:'
Would you care to start at 8: 00 21 o' clock perhaps?-
22 We are.at your disposal.
23 MR. OKRENT:
Let's see.
Are there any comments
(
24 on starting at 8: 00 o' clock?
25 MR. FISCHER:
Just a minute.
(
SONNTAG REPORTING SERVICF. iTD
(
(
j c
328
{
1 MR. OKRENT:
I am sorry.
It seems we are bound 2
by the Federal Advisory Committee Act both to have open l
3 meetings and to begin them at the time published in the l
~
(
l 4
Federal Register.
5 We don't have to end at the time published.
I 6
l (Laughter.)
7 MR. OKRENT:
So we would be willing; but, 8
unfortunately, in the same way as'this evening, for 9
example, were we ready at 5:00 o' clock, we couldn't begin I
10 until 6:00, because that was the published time.
11 Does the proposed revision of the agenda 12 inconvenience anyone, that they desperately wants a topic g
13 covered tomorrow morning that is not currently proposed?
14 (No response.)
15 MR. OKRENT:
Any comments?
Going --
16 MR. SULLIVAN:
Dr. Okrent, we may have to ask l
17 for a slight re-arrangement of the management, because we 18 are flying some folks in from Big Rock tomorrow morning 19 early.
20 When they get here, we will discuss it with them.
21 MR. OKRENT:
All right, fine.
Well, you can 22 think it over.
We are flexible.
23 Let me know first thing in the morning if you want
(
24 to do something different than what I proposed.
25 With that I will recess the meeting until tomorrow
(
SONNTAG REPORTING SERVICE. LTD.
l l
329
(
1 morning at 8:30.
2 (Whereupon the meeting was recessed 3
until the hour of 8:30 o' clock A.
M.,
4 May 21, 1982.)
5
-(The meeting was recessed at 6
10:30 o' clock P. M.)
7 8
9 10 11 12 13
(
14 15 16 17 18 19 20 21 22 23
(
24 25
(
1 1
c ott o+? a d D P D o D
- T t't"*
C t* Dif T r a t==
(
i i
6 Berore guing into the agenda for today's meeting, I 7
am going to call on Dr. Trifunac.
He owed us something 8
from last night.
9 Why don't we deal with that matter first?
10 Dr. Trifunac?
11 MR. TRIFUNAC:
Yes.
I have done some 12 calculations that are preliminary, but here is the 13 summary.
14 I am going to give a list or its, and then if you
(]}
1 15 want to accept that, I will give you the conclusions.
16 (Laughter.)
17 MR. TRIFUNAC:
Okay.
The first if is if we 10 assume in quotation marks Michigan Basin seismicity 19 applies.
20 If you recall, there was a figure, which I believe 21 was figure No. 5 in one of the presentations yesterday, 22 which showed the recurrence relationships for the area, 23 and Michigan basin is the case which corresponds to a very 24 large area subject with a uniform seismicity rate, and I 25 am assuming that this is a meaningful way to come up with i
50NNTAG REPORTING SERVICE, LTD.
(
(
334 1
a preliminary estimate of probabilities.
(
2 So this is assumption No.
1.
3 Assumption No. 2 is that seismicity is uniform in
(
4 the area around the site, in large areas necessary; and I 5
am going to assume, also,.that the largest earthquake can 6
happen in the area a's modified by Cal intensity of 8, with 7
a cutoff level recurrence relationship, is going to be 8.
8 The second assumption is that we want to incur the 9
PSV, the psudo relative velocity spectrum for 5 percent 10 damping.
I picked 5 percent.
It's sort of a nice number 11 in the middle.
12 This could be done for any damping.
13 The third assumption is, then, that the logarhythm
(
14 of base ten'or sub V is equal to 1.1 minus 1.0 M sub B.
15 This is, again, essentially from figure No. 5 whica was 16 presented yesterday, per year, per ten to the power 17 kilometers squared.
10 All of this data is really based on the intensities, 19 and I have used the Nutley-Herman 1978 relationship 20 between MB and I class, which states that M sub V is equal 21 to 0.5 times the max number intensity plus 1 75.
With 22 this going back into N sub C relationship, I find that 23 logarhythm base ten, where N is the number of earthquakes, 24 is equal to approximately 0.18 minus 0.5 I, where zero 25 stands for intensity.
(
SONNTAG REPORTING SERVICE, LTD.
l
(
335 1
This gives me the relationship for the area I want
(
2 to consider.
o o,
3 The next if is that I am using intensity based risk
(
4 calculations.
I donft believe that we have enough 5
statistics and data to use M sub V in this area, even 6
though this is the way calculations were done for the 7
other estimates, a
a 8
The other if is my assumption that the life of the 9
plant is going to be of the order of 50 years, and so this 10 is an input that I needed.
11 So if you are willing to accept some of these its, 12 or at least consider them, the conclusions are tne 13 following.
14
{
When the probabilities of exceeding the following 15 figure accelerations are approximately according to the 16 following numbers -- and I will give three numbers in i
17 order.
10 First I will say what is the acceleration I am 19 talking about.
20 Then I will give approximate probability of 21 exceedance, given the above conditions, and then I will 22 give an estimate of the term period.
23 I picked up numbers which are in the range that 24 might be or interest.
(
25 So the first acceleration is.05, 5 percent G, l
l l
SONNTAG REPORTING SERVICE, LTD.
I
I 336 o
1 probability of exceedance is about 80 percent.
Return 2
period is about 30 years.
3 10 percent G,
.10, probability of exceedance is 4
about 40 percent.
The return period is close to 100 5
years.
6 15 percent G,
.15, the probability of exceedance is 7
somewhere in the vicinity of 15 percent.
Return period is 8
about 300 years.
9 The last entry is 20 percent G,
.20.
Probability of 10 exceedance is on the order of 8, certainly less than 10 11 percent, and the return period is 600 years.
12 This concludes my summary.
13 HR. OKRENT:
If I can just ask two questions to
(
14 understand what you were saying.
15 When you said you were assuming that there was a 16 certain region in which you were accepting the stated 1'I seismicity, is that what was called the -- well, 10 categorized as the smaller region on the map or designated 19 Michigan basin or the central stable region?
Which of 20 those two was it?
21 MR. TRIFUNAC:
I believe it is the smaller 22 region; but if you look at the map, you will find that 23 this is really a very large region; and the other things, 24 the others are hot pockets further away, are quite a few 25 kilometers away and, therefore, the effect they produce on
(
SDRNTIC 7tET6ETING SERVICE, LTD.
(
f o
e i
337 1
the final result is very small, yes.
,-(
2 MR. OKRENT:
Okay.
And then, just to make it 3
clear -- I think it is -- that when you are talking about
,(
4 probability of exceedance, that is over the 50-year 5
period?
6 MR. TRIFUNAC:
Yes, that is right.
7 MR. OKRENT:
Thank you.
All right.
t 1
_ _ _ _ _ _ _. _