ML22230A188
ML22230A188 | |
Person / Time | |
---|---|
Issue date: | 05/17/1978 |
From: | NRC/OCM |
To: | |
References | |
Tran-M780517 | |
Download: ML22230A188 (1) | |
Text
1GJM r uR./ ro ECRETARIAT RECORDS
- NUCLEAR REGULATORY COMMISSION IN THE MATTER OF:
SECY-78-137 BRIEFING Place - Washington, D. C.
Date - Wednesday, 17 May 1978 Pcges 1-60 Telephone :
(202 ) 3A7-3700 ACE - FEDERAL REPORTERS, INC.
Official Reporten *
.d..<14 North Cap itol Street Washington , D.C. 20001 NATJONWIDE COVERAGE* DAILY
DISCLAIMER This is an unofficial transcript of a meeting of the United States Nuclear Regulatory Commission held on May 17, 1978 in the Commission 1 s offices at 1717 H Street, N. W., Washington, D. C. The meeting was open to public attendance and observation. This transcript has not been revie~*1ed, corrected, or edited, and it may contain inaccuracies.
The transcript is intended solely for general informational purposes.
As provided by 1.0 CFR 9.103, it is not part of the formal or informal record of decision of the matters discussed. Expressions of opinion in thi~ transcript do not necessarily reflect final determinations or beliefs. No pleading or other paper may be filed with the Commission in any proceeding as the result of or addressed to any statement or arg~ment contained herein, except as the Commission may authorize.
(
1 WHITLOCK UNITED STATES OF AMERICA CR 7525 barb 2 NUCLEAR REGULATORY COMMISSION 3 PUBLIC MEETING 4 BRIEFING ON SECY-78-137 5
6 Room 1130 1717 H Street, N.W.
7 Wasnington, D.C.
8 Wednesday, 17 May 1978 9
10 The Commission met, pursuant to notice, at 1:45 11 p.m.
12 BEFORE:
13 DR. JOSEPH M. HENDRIE, Chairman 14 PETER A BRADFORD, Commissioner 15 VICTOR GILINSKY, Commissioner 16 ALSO PRESENT:
- s. CHILK 17 D. BUNCH
- s. LEVINE 18 M. MALSCH L. GOSSICK 19 E. CASE H. DENTON 20 R. BLOND 21 22 23 24 Ace-Federal Reporters, Inc.
25
I 7525.01.1 2 P ROCE E D I NGS
- bw 2 CHAIRMAN HENDRIE1 If Me can come to order.
3 The Commission meets_this afternoon for a 4 briefing by the Staff on assessment of relative differences 5 in Class 9 accident risks and evaluations of alternatives 6 to sites with high population densities.
7 Lee, you have Harold Denton with you and Marty 8 and Darrell Bunch.
9 MR. GOSSICK: Mr. Denton would like to introduce 10 the subject and he will give the briefing.
JJ MR. DENTON: The paper concerns how the Staff 12 tr.eats coremelt accidents in doing ass.essment of sites with 13 high population densities.
14 Just by way of background, the StaffJs first 15 exclusive consideration of coremelt accidents with regard 16 to high population sites was in the Newbold Island case in 17 1973. As a part of our NEPA review, Mr. Munsing, who was 18 Director of Regulation,. informed if he would move to a site 19 of lower pop~lation density -- he informerl-the applicant that 20 .he should move to a site of lower populatin density, the 21 difference in population was about a factor of 10.
22 About a month later, the Staff developed guidance 23 on population density around sites and has be.en .following 24 this guidance every sinc.e. So, since 1973, ,we hav.e not had 25 a site higher than this guideline for doing detailed
7525.01.2 3 assess~ents until last year, when th~ site came in that
- b.w 2 tripp,ed the Staff-'s value, which required us to do a detailed 3 look at alternative .sites that might be available at lower 4 population densities.
5 The Staff-'s paper described the approach that 6 the Sta.ff had taken ln looking at the higher population 7 - density sites. What we want to do today is to briefly go 8 through what is in the Staff-'s paper and the relatiohship 9 between NEPA reviews and coremelt accidents in relationship 10 between what the Staff is presently doing in this area Jl -and other Staff papers that are being prepared.
12 The one point I did want to make is that these 13 500 .people *per square mile value that has been used by the 14 Staff to give a detailed consideration of population 15 density has been tripped only once in four y.ears, and it was 16 ,the tripping of that value implication. which has now been 17 withdrawn that had led to the development of the material 18 in the Staff paper case.
19 MR. CASE: Let me make i t crystal clear. That does 20 not mean that a SJi.he::: is unacceptabLe if it trips. It only 21 ,means that you should do a more detailed evaluation of 22 al terna ti ve sites.
23 CHAIRMAN HENDRIE: As best I can remember, 24 Harold, that policy which mentioned the population, the 25 trip-level population density, but it s.eems to me that that
7525.0J.3 4
-
- bw 2
3 was our way of dealing with the reluctance of the Commission in.those days to discuss the m~tter in any more detail.
managed to get, as I recall, the number thereby into the We 4 MR. DENTON: *-- into the Standard Revie.w Plan and 5 various other places.
6 COMMISSIONER GILINSKY: And that didn-'t stop you.
7 CHAIRMAN HENDRIE: Even then at other times, 8 who knows, even now the Staff has its business to conduct 9 and manages to get it conducted. Please go on.
JO MR. BUNCH: You previously noted the title of the JI paper. It is rather long and complicated. It certainly 12 it touches on several complex and controversial issues.
13 (Slide.)
14 In s~bstancs, though, as Harold was indicated, 15 the-heatt of the paper is fairly straightforward. Basically, 16 is there a way that we can provide improved guidance in our 17 NEPA reviews on alternative sites, particularly alternative 18 sites with high-population densities. During the 19 development of the paper we received a number of comme.nts on 20 the subject particularly dealing .with possibls policy 21 implications in other areas, rather than the strict review 22 of alternative .sites, policy implications relating to the 23 treatment of Cla..ss 9 accidents in the licensing process, 24 possible implications in our siting reviews and_ths like.
25 COMMISSIONER GILINSKY: Why are you only dealing
7525.01.4 5 with Class 9?
- bw 2 MR. BUNCH,: I wi.11 be discussing that in just 3 a moment. The subject is really residual risks, ~11 of 4 the spectrum of .risks are associated with reactor located 5 in a giv.en site.
6 MR. DENTON: *We routinely include accidents 3 through 8, 7 8 b.eing design basis accident. 9 is .shorthand for a.11 8 accidents .beyond design basis accidents, including those 9 that lead to coremelt.
JO MR. BUNCH: As I will discuss in a moment, the
.JI reason .we have highlighted -insth'eititcl.e::.:~J;c:1.l;l,s 9 accidents is 12 that the great bulk of the reviews *that we have done in the 13 past in our e.nvironmental assessments have tneateded Cla.ss 9 J4 in only a very qualitative way. And recently because of the 15 insights that have been developed in several cases, we 16 thought perhaps more detailed evaluation .was in order, and 17 that is what we will be discuss~ng in just a moment.
18 COMMISSIONER GILINSKY-: I thought that we wer.e 19 considering Class 9 accidents in environmental reviews, as 20 oppo~ed to sa~ety reviews.
21 MR. BUNCH: Only in .a qualitative sense. There 22 is an acknowledgement of the poss ibl.e risks associated with 23 Class 9 accidents.
24 MR. DENTON: Reference is made to the results of 25 WASH-14DO and a general discussion of that, but not any
7525.01 .5 6 detailed quantification of potential consequences.
- bw 2 MR. CASE: The boilsrplate is the same, regardless 3 of_the site location, except whers now the pbpulation 4 density is about 5.00 persons per square mile.
5 COMMISSION GILINSKY: It is an acknowledgement of 6 _the possibility of a Class 9 accident and not much more.
7 MR. DENTON: Cla.ss 3 through 8 are tailored to 8 11t the site and the msteorology of the plant and Class 9 9 is described as --
JO COMMISSIONER GILINSKY: Is that because how our
.J I regulat~on is structured?
12 MR. DENTON: Yes.
13 MR. KELLEY.: That has been cha.llenged since the 14 study.
15 MR. BUNCH: You may recall that in the memorandum 16 of April 24 we were aavised that you Mere reserving the 17 judgment on the specific r~commendations in the paper, pending 18 a revi.ew of siting policy or a status report on the related 19 papers on siting policy, and that is basically what we will 20 be talking about today, if I might have the next viewgraph.
21 (Slide.)
22 As .was indicated by Harold, the Staff and its 23 safety reviews have always considered population density as 2*4 a re1atiYe1y impot'tant part of the review. Population 25 den51ty is basically Shorthand for a way of characterizing
7525.01 .6 7
.risks associated with accidents. The perception being that
- bw 2 all othe~ things being equal, the consequences of accident 3 release are going to be higher at a loc~tion with a greater 4 ~ population than with a lowe~ population.
5 COMMISSIONER GILINSKY: Let me understand.
6 Class 9 accidents or evaluations of Class 9 accidents play 7 no role in the choice of sites for rsactors at the present 8 time; is that correct?
9 MR. BUNCH: No, sir. Not totally. Ln a recent JO Commission paper, I believe it is 111, we tried to
,11 articulate the current accident analysis practice. The 12 nume.rical calculations that we do, do not include the 13 kind of Class 9 accidents that are discussed in reports such 14 as WASH-14DO. There is a very large source te.rm that involves 15 .basica.lly a full release o.f ..:fue-:fission gases and a very 16 substantial portion of the iodine. You wouldn 1 t expect to 17 get that kind of source term without that, but it is assumed 18 that the containment is intact, and that the engin.eered 19 . safety f.eatures are effective in controlling and mitigating 20 the consequences of those accidents. Because we do rely 21 and give credit on engineered safety features to provide 22 some containment of even very large source terms, we have 23 had to use other devices, such as controls on. population 24 densities _or special reviews in densely populated sites to 25 assure that reasonable protection is afforded even for these
7525.01 ~7 8 more severe events.
- bw 2 So that is a long answer to say that when we 3 look at the site, we look at how far that site is from 4 centers of population. One of the things *that we keep in 5 mind is the possible implications that might be associated 6 with a very large accident.
7 COMM I SST ONER GI LINSKY: Is that written :do.wn 8 anywhere? Are there specific guidelines on it?
9 MR. BUNCH.: There are no specific guidelines as 10 such. The statements of consideration to JO CFR Part IDO JI do point out that the definition of a population center 12 distance and their requirement to keep some distance between 13 a reactor and a census of population is in part based on 14 consideration of Class 9 accidents. Beyond that there are 15 no specific guidelines.
16 MR. DENTON-: We hav.e not Looked at the 17 consequences of Class 9 for plants which fell within the 18 envelope of the Standard Review Plan population density 19 numbers. It was only this one case that feil without, that 20 prompted us to take a hard look at how we should treat the 21 consequences of coremelt in Class 9 accidents, when it was 22 outside the envelope.
23 MR. BUNCH: In any event, prior to the 24 enactmsnt of NEPA, We have relied on the Safety Review as 25 the principal vehicle to consider the risks associated with
7525.0).8 9 sites that are located in fairly densely populated areas.
- bw 2 Even then, there has been an acknowledgement that population 3 density is just one of the very many factors that are 4 important in the site selection process. With the 5 enactment of NEPA the Staff has been able to devel-0p and 6 us.e a more structured approach toward the balancing of 7 various environmental, economic and safe*ty-related 8 consideration in their siting reviews.
9 And as the top &.u1.JJ1i:t:. indicates, there is on 10 NEPA reviews, an attempt to perform such a balancing.
JI Now, as Harold just mentioned, when we have a 12 ,proposed site that is in a relatively isolated area, .we do 13 not give particular weight to small differences in 14 population bet.ween Site A and Site B, the altern~te site 15 and the proposed site.
16 Other facto.rs, economic and environmental, have 17 more weight than the NEPA review process. When we have such 18 a situation such as arose in Newbold Island, where the size 19 of the population was an important consideration, there are 20 guidelines ,that are used to .institute a special, more 21 indepth review of all of the factors. Those guidelines are 22 included in two o.f the enclosures to the S:ECY. paper.
23 24 25 l_ ----- - ------ -
10 7525 I do~'t plan to go into them, except to note that
~HIT:sp
- 2 2 there is a threshold value. If the population is in that 3 vicinity or hi~her_at the proposed site, then we do undertake 4 a more detailed review.
5 Again, most of the sites that we have existing today 6 are relatively isolated. More~than 90 percent of the existing 7 sites are in areas below the guideline values we use for a 8 special review. As Harold indicated, there has only been one 9 instance since these Newbold Island-related guidelines were 10 developed that we have had an occasion to institute the special 11 review processes *. because _of,-,popuJ_ation density considerations.
12 They are not limits; they_ are just triggers .or thresholds.
13 When we get done with the review and the NEPA balancing, there 14 has yet been no determination as to the acceptability of the 15 site one way or the other.
16 COMi.1\1:ESSIONER G*I LINSKY:
1 How do you compute that 17 density? Is i t simply the average density across that circle?
18 MR. BUNCH: Yes, sir. You compute it at various 19 distances -- one mile, five miles, l0~miles, 20 miles, 30 miles 20 and look at what the average density is in each circle out 21 to a distance of 30 miles.
22 MR. CASE: If i t trips any one, then you do the 23 special review.
24 COMMISSIONER GILINSKY: You could have a city of a Ace-Federal Reporters, Inc.
25 million 30 miles from there?
. 11 sp2 MR. BUNCH: Yes, that is correct.
2 MR. DENTON: That is only one way we look. The other 3 was weighing the density by meteorology, which gives a_;;_differ-4 ent way of looking at pop~latioQ. And then in this review of 5 this detailed look, we used the same consequence modeling code 6 that WASH-1400 used that went out to hundreds of miles, so we 7 wouldn't stop if we saw a city that was on the fringe 8 of 30 miles; we would use one of these other techniques to 9 bring it il.h. If you do _a_**screeri, you do the 500 and it is iwell 10 below, you can be sure there is not a site much better than 11 that available in the populated parts of the country.
12 MR. BUNCH: Generally, with the 500-people-per-13 square-mile number, you can't stay within that value , if. you 14 haueJa city of 100,000 or more, say perhaps as close as 10 15 miles, to the facility, a large city with those gµidelines are 16 going to have to be somewhat removed from reactors.
17 With the application that we did receive that caused 18 the trip levels to Le exceeded, there was a need again to look 19 at the pros and the cons of the various alternative sites, 20 and accordingly, to try and figure out how much weight to give 21 to population relative to the other factors that might be 22 both advantageous and disadvantage6us2at the various sites.
23 That brings us specifically to the subject that is 24 discussed in the paper, and our attempts to find l;:>ettercways to Ace-Federal Reporters, Inc.
25 assist in our determination of what kind,)6f weight, what kind
12 sp3 of balancing process to undertake.
2 If I might have the next slide, please.
3 (Slide.)
4 Joe asked earlier about class 9 accidents and whether 5 or not they are the only thing that we look at, and I answered 6
no.
7 COMMISSIONER GILINSKY: Or whether you look at them 8 at all.
9 MR. BUNCH: As was indicated in our NEPA reviews, we 10 do discuss ail classes of accidents, from the most likely to the 11 leastllikely, in general terms. We discuss the reviews and 12 assessments of the design basis accidents that appear in the 13 qualitative safety evaluation reports, and we discuss, in 14 terms the class 9 accidents,-arid make reference to the more 15 detailed quantitative assessments that appear in WASH-1400.
16 They are not tailored on a site-specific basis to any 17 degree, but there have been many instances where assessments 18 have been made to try to determine the relative importance of 19 class 9 accidents.
20 I think, as you have heard in previous discussions on 21 WASH-14 00, there is a general percep;tion::.*1that after the plans 22 have been designed to mitigate and safely accommodate all the 23 more credible accidents, what residual risks are left are 24 largely those associated with the very, very unlikely events Ace-Federal Reporters, Inc.
25 that class 9 events --
13 sp4 COMMISSIONER GIBLINSKY: How does the overall risks cif 2 class 9 accidents compare with the lower *classes?
3 MR:.:' BUNCH: We have performed some analyses. There 4 are more extensive analyses in the WASH-14..00, andithey indicate 5
that the risks associated with the class 9 accidents might vary, 6
say, in the range of 50 to a few hundred manrem a year, using:
7 the methodology of WASH-1400. These are from class 9 accidents.
8 They are relatively small impacts.
9 MR.- DENT.ON: They appear to be in the same range as 10 the occupational exposure per year basis.
11 COMMISSIONER GILINSKY: How do they compare with 12 the one through eight?
13 MR. DENTON: I think my view is the one through eight 14 are somewhat less than the residual risk of class 9.
15 CHAIRMAN HENDRIE: That's the way I recall.
16 MR. DENTON: They are down an order of magnitude.
17 COMMISSIONER GILINSKY: We are screening i t out on 18 the basis of,p,irobability, even though the risk is computed to 19 be greater from that source.
20 MR. BUNCH: I think~it would be fairer to say that we 21 looked at both probability and consequences, and we have a 22 fair perception of the range of consequences that might be 23 associated with class 9~,accidents, and were convinced when we 24 considered their probability that the risks are very low.
Ace-Federal Reporters, Inc.
25 Those numerical analyses that we have done support that view.
14 sp5 COMMISSIONER GILINSKY: We are saying, the risks are 2 still greater than from the one through eight?
3 MR. BUNCH: They are both small numbers.
4 COMMISSIONER GILINSKY: That's fair enough, but I
- 5 guess it is not immediately clear why you would use one for 6 citing purposes and you wouldn't take into account the other 7 which is larger.
8 MR. BUNCH: I am not sure I understand your question, 9 sir.
10 MR. CASE: What he is saying is, you use it for 11 citing purposes, the trip level, but once you get to do your 12 NEPA evaluation, you only debit the class 1 through 8 risks; 13 you don't quantify the class 9 risk for a particular site after 14 you have gone through the site selection process.
15 MR. DENTON: The way I see it working is along this 16 line. The part 100 provides a specific way to assure that 17 individual risks are low, given design basis accidents, and it 18 prescribes the calculational method. It also has admonitions 19 that where very large cities are involved, special consideratio 20 should be given. But i t is easier to calculate design basis 21 accidents because of the historical precedent for looking at 22 themr and accidents that are beyond design basis, how systems 23 performed wasn't known until 1400, or those types of techniques 24 became available.
Ace-Federal Reporters, Inc.
25 1400 is still being looked at by the Lewis Committee
15 sp6 *in certain facets. 1 So we have not done detailed calculations, 2 site-specific, except in these individual cases where they trip 3 one of the staff citing guidelines to try to get a relative 4 view of the difference between sites.
5 I think we still see an inability to be absolutely 6 right in terms of calculating class 9 accidents. We think that 7 the techniques that are available are good at looking at 8 relative differences and choosing between sites.
9 MR. MALSCH: One,:,p:roblem has. been that under Part 100 10 which are the Commission citing guidelines from the safety 11 standpoint, the reviews have always focused on the dose guide-12 line values in the part a~ the 25 rem whole body and 100rem to 13 the thyroid ... Those are only associated with what fart 100 14 refers to as credible accidents, which by definition are eight, 15 not nine accidents.
16 In applying Part 100 to individual sites, the focus 17 has been not on class 9 but on class 8.
18 COMMISSIONER GILINSKY: But credible means certain 19 categories of accidents which we choose to consider.
20 MR. MALSCH: Because of probability reasons. The 21 line between credible and incredible is based upon probability 22 considerations, although the regulation doesn't specify a 23 credibility number.
24 COMMISSIONER BRADFORD: Wherei:'is the line?
Ace-Federal Reporters, Inc.
25 MR. MALSCH: It is not specified in the regulations.
16 sp7 Various staff guidelines have been developed to use in reviews.
2 MR. CASE: That is really the issue in the ATWS that 3 we discussed earlier this morning: where do you draw the line.
4 MR. DENTON: The staff review has generally been
\
5 deterministic. We say -- we .determine many of the parameters 6 such as SSEs in a deterministic mode and not a probablistic 7 one. So anything that falls within the design basis accident 8 envelop5 is looked at in a deterministic mode.
9 The issue here that we are raising is, how do we 10 treat in NEPA environmental statements of alternative sites 11 these residual risks associated with ones beyond design basis 12 accidents?
13 I thiBk the staff practice has been that for sites 14 that had population densities within existing ranges, we relied 15 on reference to 1400 as a generic treatment of what the 16 residual risks were, but where the proposed sites exceeded,lthe 17 values, we have taken a hard look at alternative sites to 18 those and have included in part of that evaluation an analysis 19 of the potential consequences of class 9 accidents at those 20 sites.
21 MR. BUNCH: If I might have backup viewgraph B-4, 22 I think I might be able to finish the response to your 23 question, Commissioner.
24 (Slide.)
Ace-Federal Reporters, Inc.
25 As discussed in the paper, we did attempt to find
sp8 ways to better rank all sorts of residual accident risks in the 2 NEPA review process. This indicates one. such display where 3 we just attempted to monetize everything and to get a rough 4 figure of merit, if you would, of the relativ.e differences 5 between these sites.
6 The value of the numbers is not particularly impor~
7 tant. They are small relative to other costs that were 8 generally included in the NEPA review of alternative sites and 9 associated environmental impacts.
10 What we are trying to do h~re, again, is recognize 11 that with sites that were relatively densely populated areas, 12 we wanted to make sure that the standard practice still had 13 meaning and value. This particular evaluation was used to 14 indicate to us, do we still think we are iri:~the~.satne. ball park, 15 and I think the conclusion from this type of analysis was, yes, 16 we were for the particular site in hand. The accident risks 17 were still very low in terms of differences from site to site.
18 And in all cases, they seemed to be low.
19 So, again, all classe~ of accidents are taken into 20 account in the NEPA reviews, and what we are talking about here 21 is, there is some threshold where we have to look to see if 22 our standard practice of treating things in a relatively 23 qualitative way.1:::and the cost-benefit evaluations deserves a 24 little more rigor.
Ace-Federal Reporters, Inc.
25 This particular example would appear to me to
19 sp9 indicate that at around the guideline values;,perhaps it is not 2 yet required, but nonetheless it did provide some indication.
3 MR. DENTON: This assumes that we have the same plant 4 at each of those five sites, and using the techniques of 1400 5 we calculated late cancers, manrem, early fatalities, monetized 6 them all to give the common dollars, looked at cost of 7 decontamination of land. So this is in terms, then, of the 8 per-year residual risks from class 9 accidents at the sites, 9 using WASH-1400 type methodology, which does show that the 10 values monetized are small.
11 MR. SEGE: It also assumes WASH-1400 pro~abilities 12 constant from site to site.
13 For example, the proximity of Edwards Arsenal and 14 the possibility that that might affect probability of an 15 initiating event or something else that might occur at some 16 other site that would not be factored in.
17 MR. DENTON: This was not intended -- they would be 18 covered in the other parts of the review. You are right, 19 though. This is the same plant, the same probabilities at each 20 site.
21 CHAIRMAN HENDRIE: This is the consequence model.
22 MR. DENTON: Yes.
I 23 MR. BUNCH: Yes, sir.
24 MR. CASE: The same model that the ACRS talked to Ace-Federal Reporters, Inc.
25 you a week or so about.
20 spl0 COMMISSIONER GILINSKY: This is a small point but 2 you generally take these things out to three places.
3 MR. BUNCH: They ought not to be taken out of more 4 than one place.
e-2 5 I . might *-note: If you look at the left-hand begin #3 6 c_o'l:umn; =you .. will see whether the proposed. site is below or above 7 the trips i:oints. You will see that there are substantial 8 differences in these five alternative sites from the standpoint 9 of population. About a factor of 10 of the highest and lowest 10 one might have expected because of this range of population 11 densities to have a significant difference in the residual 12 risks of accidents.
13 These kind of analyses take the other factors that 14 are important in looking at the consequences of accidents, 15 meteorology, the distribution of the population around the 16 plant at various degrees, all the other things that can be take 17 into account, arid wraps them altogether and comes out with an 18 integrated assessment.
19 And these indicate that competing factors, such as 20 the winds tend to blow away from the population centers, '_
21 squeeze down the*apparent differences between sites, and taken 22 altogether, the population figures may be a little misleading.
23 So the answer, again, is, there.\is not a significant 24 difference.
Ace-Federal Reporters, Inc.
25 ;CHAIRMAN HENDRIE: That is one sort of consideration,
'I
21 spll but the error of those -- on that consequent modeling is pretty 2 large, isn't it?
3 MR. LEVINE: If I may, there are large errors, 4 uncertainties, in the calculation indicated in WASH-1400.
5 Furthermore, the model as designed is designed for calculating 6 100 reactors as an ensemble and not per reactor. It is not 7 designed for individual site calculations. It is probably good 8 enough to 10 or 15 miles, because the Gaussian meteorological model i 9
pretty accurate out there, so the early kind of results, the 10 early health effects that you would see from large doses in 11 close are probably better than the far-out doses.
12 On the other hand, the far-o~t~dcises~are sort of 13 manrem-dependent, and i t doesn't matter very much. It is not 14 clear how good the model is for far-out calculation, because 15 one doesn't have downwind data, and the effects would have to 16 be included in the model.
17 MR. BUNCH: The point here is not that the results 18 are very accurate, are very precise. One of the good aspects 19 of the crack model is that it can take the population, distri-20 bution, it can take meteorology to the extent that we are able 21 to characterize it with all its limitations, and take them 22 together,.a_iritegrate them, and give you something that is more 23 complet:e, daT ,more complete picture. It still has limitations.
24 COMMISSIONER GillLINSKY: Suppose we took Indian Point Ace-Federal Reporters, Inc.
25 or Zion? What was the significance of these sites?
22 spl2 MR. DENTON: These were the candidate sites that were 2 originally propo~ed, as the most suitable site.
3 MR. CASE: Perryman was the one that tripped the 4 SQQ.persons per squarm.~ile. These were alternative to 5 Perryman .. This is a live case.
6 COMMISSIONER GILINSKY: I see.
7 MR. DENTON: Now, we have done this sort of thing 8 for higher population~sites. I think with regard to Newbold 9 Island, about a factor of 10 for Newbold Island was 10 site that prompted the 11 COM-'.lISSIONER GILINSKY: Isn't the risk, :the-WASH-1400 12 risk, concentrated in a very small number?
13 MR. DENTON: In the higher population sites.
14 Newbold Island 15 COMMISSIONER GILINSKY: Like a handfull of plants out 16 of a hundred.
17 MR. DENTON: Yes, sir.
18 COMMISSIONER GILINSKY: Do we have numbers on that?
19 Have we 20 MR. BUNCH: On the relative --
21 COMMISSIONER GILINSKY: Yes. Have we done this for 22 existing sites, this kind of analysis?
23 MR. DENTON: For Newbold Island the numbers turned 24 out in dollar values, it wasapproximately a factor of 10 high, Ace-Federal Reporters, Inc.
25 hi,~her than these.
CHAIRMAN HENDRIE: The difference is significant, as
, I I
I 23 spl3 you move from site to site. The calculational method has a 2 nui:nber of uncertainties in it. If you look at the absolute value 3 of the numbers, I expect the uncertainties quoted in 4 WASH-140O<app).y., the factor of 5. How much error do I associate 5 with the relative differences down the column? It seems to me 6 that there is still an error of considerably involved 7 because the model was set up to treat the ensemble, and I don't 8 know quite what it does when you put in one and crank one site 9 and
- then crank .on through to an end number and then go back and 10 put in another site.
11 I think there are still some errors, but what are 12 they factors of?
13 MR. CASE: I thought they told me, within a factor of 14 2, it is not significant. A relative factor of 2 is not 15 significant.
16 MR. DENTON: Yes, sir.
17 MR. BUNCH: I believe, if you look at the model just 18 as a way of computing consequences from a g:rlJyen accidental 19 release, you can see systematic differences between the sites.
20 I think, if you are careful about what -- the purpose 21 and the application of the model, you ought to be able to dis-22 tinguish bettv.een sites , if. you start seeing factors 3, 5, 10 23 differences.. There are still uncertainties, as -Mr. Levine was*
24 pointing out. The assumptions* as to the meteorological trans-Ace-Federal Reporters, Inc.
25 port are fair.ly:c:si:(llple. If you find that there are differences
24 spl4 in the topography from site A to site B, that is not normally 2 taken into account in these calculations, and you have to be 3 very carefu~ -- and if you see a difference between 2 and 3 --
4 to see whether or not that is due to the assumptions you have 5 made or due to the differences in the site.
6 So it is no better and probably no worse than a lot 7 of other techniques that are used to discriminate between 8 sites.
9 MR. DENillON: Then you have the other various indica-10 tors --
11 MR. CASE.: It is better than raw population density.
12 MR. BUNCH: What this viewgraph shows is the same 13 type of information comparing now a site that has a substantiall 14 higher population than the ones indicated in the previous view-15 graph.
16 The rightfhand column is based on the Newbold Island 17 population and just looks at six different kinds of consequence 18 that are computed in the WASH-1400 consequence model.
19 Again, don't pay too much attention to the specific 20 numbers, other than to note that of the five values for the 21 left-hand side, there are some differences, although they are 22 relatively small. The Newbold Island population was rather 23 higher, considerably higher than any of these five, and now you 24 begin to see differ~nces that are more appreciable, and they Ace-Federal Reporters, Inc.
25 generally run a factor 6f~~baut 10 higher than the less densely
25 splS 2
I personally would take these to indicate that the crack 3
code can show some signifieant .:* differences based on population 4
density.
5 MR. DENTON: In all of these tables they have beennQ~-
6 malized to 1 £ram Perryman. They are all in the basis of 7
each parameter, compared to Perryman.
8 CHAIRMAN HENDRIE: Did I not read the figure right?
9 Why is Newbold relatively low to Perryman in infant mortality?
10 Is it close in population?
11 MR. BUNCH: Frankly, I don':.t know the answer to that.
12 The population at Newbold Island is somewhat higher 13 than in Perryman, than we would have expected.
14 CHAIRMAN HENDRIE: It is not in, close, or is it?
15 MR. BUNCH: I b~li~ve it is about 100,000 people at 16 Newbold Island, and these particular values.
17 CHAIRMAN HENDRIE: Within what?
18 MR. BUNCH: Within 10 miles.
19 CHAIRMAN HENDRIE: But once you come down to a mile or 20 two, that goes to zero effectively,uor,:under five miles. I 21 think that was one of what the applicant felt_.was a redeeming 22 grace of that site, that even though the numbers were large, as 23 you went out, that right around the plant within the several-mile 24 Ace-Federal Reporters, Inc. range i t was low. I wonder if that is th~ effect we are seeings 25 here.
26 spl6 MR. BUNCH: Thabis possible. The difficult thing-sjin 2 the code is when you get unusual :_results, such as this, results 3 that you may not have expected. You ha,.ze to go back and look 4 at precisely what was the population distribution to confirm 5 that in fact that is what would be predictable.
6 CHAIRMAN HENDRIE: Now, these are all cut off at 30 7 miles.
8 MR. BUNCH: The c6nsequence assessments are taken out 9 t-o-* 500 miles. As Mr. Levine was pointing out, the manrems 10 generally occur in a range of 30 to perhaps 200 miles or so.
11 The principal. source of that. And acute fatalities are in an 12 area of perhaps a few miles out to perhaps 25 miles, as 13 calculated.
14 So the estimates take into account the population out 15 to very large distances.
16 CHAIRMAN HENDRIE: I am surprised at the manrem and 17 the differences between that and Perryman.
18 Well, no matter. Onwapd.
19 MR. DENTON: I think, in doing this, we had some 20 questions also about why we were getting these sorts of 21 results, and we didn't use these results in our final comparison 22 of sites.
23 CHAIRMAN HENDRIE: Perryman ts southwest of the 24 Newbold site, by what, maybe?
Ace-Federal Reporters, Inc.
25 MR. BUNCH: I think about 12 miles or so.
27 spl7 CHAIRMAN HENDRIE: It's more than that, isn't it?
2 Perrymanris -- southwest of Perryman. But it is more like 30 or 3 40i something like that.
4 MR. DENTON: You mean, upcreek?
5 CHAIRMAN HENDRIE: No, Newbold.
6 MR. DENTON: Yes, you're right.
7 MR. BUNCH: The,:-,p9pu1ations that we are using 8 MR. DENTON: North of Baltimore.
9 CHAIRMAN HENDRIE: Which is about ho.w_;ifar away from 10 Newbold? 50 miles?
11 MR. DENTON: At least that.
12 CHAIRMAN HENDRIE: On the other hand, as you start 13 sweeping out the circles, the code runs out to 500 miles, at 14 which place you have got the whole Mid Atlantic population for 15 both sites. I am just surprised at a factor of 10 in manrem.
16 MR. LEVINE: That is an ensemble. Perryman is near 17 the ocean, and we stop calculating over the ocean. The~e is no 18 population density. Perryman., for instance, is a lot closer to 19 the ocean than is Newbold.
20 MR. DENTON: It's about the same.
21 CHAIRMAN HENDRIE: But you< sweep.; 0ut,-'*the .same .pepula-22 tion. You have got the center of a 500-mile circle that is 23 placed 50 miles south then there is great population mass, and -
24 ~R. BLOND: You are still dominated by the population Ace-Federal Reporters, Inc.
25 in certain areas and by the probability of the wind blowing in
)
28 splB that direction. It is the combination of the events that you 2 ~fg indicating what is happening on that chart.
3 MR. LEVINE: Wind direction probability.--
4 MR. BUNCH: I thin~ this discussion should serve to 5 indicate, when you start using results such as the crack code, 6 you have to be very cautioµs about drawing too much meaning 7 into the things, and in fact that was the comments offered by 8 research, and they are::_.ii.mcluded in our paper, and I believe 9 both OPE and OGC noted that there were difficulties in direct 10 application of these kinds of analyses, which we had also 11 noted.
12 If I might have the fifth viewgraph, we can now turn 13 to the related discussion on how these ki~d~_:of analyses might 14 impact on whether activities related to siting policy.
15 (Slide.)
16 This was discussed in your letter of the 24th of 17 April.
18 The first one is PRM 50-10. As .,theL$ubj.e6tf.indicates, 19 the rule-making proposed would be to require explicit considera 20 tion of cwas.s 9 accident risks whenever a new or novel siting 21 or design approach was suggested.
22 This one is related to the SECY paper in the sense 23 that they are talking about an unusual situation in the SECY 24 paper, and we have indicated that we thought that we at least Ace-Federal Reporters, Inc.
25 had to go back and see whether the standard practice was still
29 spl9 applicable.
r 2 It is a discretionary sort of thing, a trigger an~
3 approach that says, let's make sure that things are still as 4 they appear to be.
5 The proposed rule-making goes rather b~yond that and 6 would require a mandatory and explicit analysis of cla~s 9 7 accidents under the defined conditions.
8 Now, the status of that action is that we have advise 9 the petitioner that our response to the petition would await 10 completion of the staff*' s liquid pathway generic study, and the 11 petitioner may withdraw the proposed rule-making, since the 12 liquid pathway study satisfies the intent of this petition, at 13 least the remedy that he was seeking.
14 As I mentioned, SECY 137 is related to this in the 15 sense that i t accepts the thrust of the propoied rule-making as e-2&3 16 i t relates to our NEPA site reviews.
17 18 19 20 21 22 23 24 Ace-Federal Reporters, Inc.
25
7525.04. l 30 The next one is a proposed rulemaking to modify 10
'te 2 CFR Part 1,00 so as to provide _explicJ t numerical population 3 density criteria. A paper h~s been prepared for concurrence
-4 by Standards. There are basically four options to this paper.
5 .One of th.em is the obvious, to accept. One of them is to 6 deny. One of them is to accept in part and say, y.es, we think 7 there may be value in pu~ting numerical population density 8 cr-1teria into the regulations, but that. such criteria ought 9 to comport more closely with what the staff~s current practics 10 is. Dur final option is to deny the petition, but to note J1 that we do have ongaing activities related to our .reassessment 12 of Part 100, and that this particular action would be 13 considered under that envelope of activities.
14 The next categ.ory of papers are those on general 15 siting Policy. Standards has prepared a copy, a paper for 16 concurrencs. At_ this time, there are discussions taking place 17 between Standards and NRR as to the appropriate scope and 18 ~ontent o~_this paper.
19 There was a second formal paper, to consider 20 alternatives to current policy, general policy statement, and
.21 that pap.er is being deferred pending resolution to th.e
-22 statement of curre.nt policy.
23 The SECY paper before y_au today .would articulate one 2-4 narrow aspect of our general siting review, that aspect 25 dealing with our NEPA review of alternative ~ites.
31 Papers on accident analysis for Part LOO, siting 2 assessments are also in preparation. I mentioned that SECY 3 ~8-lJl was completed. That is the paper on current practice.
4 Standards has solicited comments from various staff members 5 and:.- officers on possible alternatives to current practices.
6 It is currently scheduled for completion in July of 1978, 7 although there may be some delay in that because of the 8 activities in the other papers I mentioned.
9 Emergency planning. In earliar months, the staff 10 has discussed the possible amendment to Part 50 to provide for JI emergency planning beyond the LPZ _in certain instances. This 12 paper, SECY 137, is not really related to the emergency 1-3 planning, except to the extent that the analyses that we did 14 in support of the paper provide some input as to the risk that 15 might .be considered in developing an e,mergency plan.
16 Finally, I have noted that the sta.ff has under .way a 17 study activity, generic study activity to determine whether or 18 not to update or other.wise dispose of the proposed annex that 19 provides guidance on treating accidents in .environmental 20 statements. The NRR has developed a category, a task plan, to
.21 scope and schedule this effort. We expect it to be complete, 22 hopefully, some time next year.
23 Now, if you have any questions on any of these, I 24 can try to answer them. We have several members from 25 Standards who can provide more detail should you request it.
7525.04.3 32 Standards .who can provide more detail should you request it.
2 CHAIRMAN HENDRIE~ Slide back up to that one which 3 assigned some .dollars to the five sites.
4 (Slide.)
5 CHAIRMAN HENDRIE: I wQnder if_this is going to be 6 typical. That is, the proposition here is to say, lnok where 7 population density for a proposed plant hits the trip line, we 8 .want to look more carefully at alternative sites than would 9 0th.er.wise be the cas.e. And part of . this more careful Looking 10 is a consideration of relative Class 9 risks for the original JI and alternativ.e sites. If the results 12 Now, in order to use that Class 9 element in the 13 comparison, you have to find some way to reduce its effects to 14 units .which are compatible with the un-1ts in. which other 15 ,* effects will be measured between the alternative si_tes. And 16 what you have done here is to go to doilars, as is normal in 17 these sorts of analyses, and used a consequence model from 18 WASH-14.00 as the tool tog.et there.
19 And it turns out that the .difference between sites, 20 even between over .factors of ten in population out to 30 21 miles, turn out to be within what people believe to be sort 22 of the slop in the calculation ** The right-hand column says 23 that these five sites are indistinguishable for Class 9 24 purposes with regard to the di1ferences between them due to 25 the whole residuum of accidents with regard to Class 9. So I I
I I
I
7525.04.,4 33 for these five sites the Class 9 exercise is a .wash. That is, 2 it doesnJt distinguish between these sites in any meaningful 3 waY.
I I 4 MR. BUNCH: That is correct.
I 5 CHAIRMAN HENDRIE: Now, is there something special I
I 6 about. this array of alternative sites, these five sites, that 7 makes them c.ome out that way? or is this likely to be the 1*
8 general outcome for alternative sites?
9 MR. DENTON: I*'d Look at it in this light, that 10 Perryman ,was just barely over the 500, just barely tr;ip,ped,.
JI So it is not too far over the criterion that we were using as 12 a *tr.:i;p that it says that at that level you are not apt to find 1.3 sites that vary much with regard to residual Class 9. Of 14 course, I took it as something of a confirmation, at least in 15 this part of the country, of the 5DO as the trip point.
16 Now, it doesn 1 t say that if you found sites that 17 were very isolated, as you can find in the Southeast or 18 perhaps som.e sect ion of the country, you might find more of a 19 differsnce. But certainly in the Northeast I think this is 20 representativ.e of wb.at it wiJl be between sites, especially in 21 a site which comes close to the 5.00 number. It doesn*'t get 22 too +/-ar out of line from it.
23 COMMISSIONER GILINSKY: The 5.00 only applies from 24 zero to 30 miles?
25 CHAIRMAN HENDRIEz Yes.
r I
7525.04.5 34 MR. DENTON: It was picked based on Looking a_ll the 2 way out to IDO and trying to find a number; and it was based 3 on looking at different regions of the country and so forth.
4 It is really a number derived for the Northeast U.S., ,which 5 was the most densely populated.
6 CHAIRMAN HENDRIE: If you take the sites and plot 7 the population within the circle of Radius R.varsus Radius I, 8 you get a series of curves that go up like th is. Some are l o,w 9 and rise; soms rise fast and flatten. You get an array IO betw.een the sites *
.11 If you go to a large enough R, they a 11 converge.
12 And why is that? At R equals 3,000 miles, you have got the 13 population of the United States, ,wherever the site is.
14 <Laughter.)
15 CHAIRMAN HENDRIE: And on a more reasonable scale 16 for reactor.s in the eastern part of the country, by the time 17 you get out around 40 or 50 miles for any site, almost any 18 s.i.te in the eastern part of the country, you have begun to hit 19 .major metropolitan centers. And as you go on out, you just 20 pick up more, and-the population number just converges on sort 21 of. the population of the Northeast and.the population of the 22 Southeast. So that you don*'t get any distinction, rea.lly, 23 between sites if you go way out, if you take large 24 If you come in to sma 11 R and say .we will only 25 consider i.t within two miles, then you say, gee, wait a
11525. 04 *. 6 35 minute, that is gexting too close to the plant. The effects 2 clearly stretch further than that. And there are differences.
3 30 was sort of a guess-and-by-God cutoff point taken about as 4 far out as you could go and still see reasonable differences 5 between the existing array --
6 COMMISSIONER GILINSKY: What is the relation of the 7 30 miles to the radius in which there would still be effects 8 in a Class 9 accident that would be expected to o~cur?
9 MR. DENTON: It is bey,ond the radius at which you JO would expect fatalities to occur. It is nowhere near as far J1 out as you would expect man-rem doses to occur.
12 COMMISSIONER GILINSKY: Some of the man-rem in 13 other words, .would the 30 encompass the bulk of the event?
14 MR. DENTON: We have some studies ongoing to look 15 at that. Perhaps someone in the audience knows the answer.
16 MR. LEVINE: What was the question?
17 MR. DENTON-: What p.ercentag.e of man-rem is picked up 18 in certa.in distances with the era ck .code?
19 MR. BLOND: Between about 30 miles and 200 miles 20 you pick up probably more than 90 percent of the man-rem.
21 COMMISSIONER GILINSKY: So 90 percent of the man-rem 22 would be beyond the 30 miles.
23 MR. LEVINE.: Yes.
24 MR. BLOND: On a risk basis, yes.
25 MR. CASE.: But it is a.11 constant any.way, because
r 7525.04.7 36 it is all essentially the same beyond 30 miles or maybe 50 2 miles.
3 MR. LEVINE~ May I say somsthing about that chart?
4 I think it is very germane. You ar_e accepting that model as 5 the correct model. !~think there are reasons for questioning 6 it. For instance, you made an observation about the totals 7 being nearly the same. If you look at the constituents of the 8 totals, you f_ind it all .comes from the man-rem. And as you 9 point out, the man-rem are .almost all the same, because in JO general the population is almost all the same in large
,1 I distances.
12 However, you now have to examine the dollar values 13 assigned, and they counted $1,000 per man-rem and $1 million 14 per fatality. In fact, I think the $1,0DO' per man-rem is 15 probably high by two orders of magnitude. And if you change 16 it, then you get a markedly different result. And in fact, if 17 you £orget about man-rem and just lDok at early fatalities, 18 than you see significant differences.
19 So there are different ways to parcel this. And if 20 you take into account life-shortening effects, as opposed to 21 fatalities, you can then compare, make a comparison betwsen 22 the relative dollars in an early fatality and the relative 23 dollars in a late fatality.
24 So one Dught to examine this model more carefully, 25 and~ one might be able to make more sense about. how to judge
7525.04.8 37 di.fferences .in sites.
'te 2 COMMISSIONER GILINSKY: The point I .was going to 3 make is a different one, is that we are using a criterion for 4 whether or not to lDok at Class 9 accidents whose effects are 5 beyond 30 miles. We are using a criterion which counts the 6 population within the 30 miles. My question is~ Why wouldn 1 t 7 we do it for every site? Do you follow me? In other words, 8 why bring in the question of Class 9 accidents only in those 9 cases where the population exceeds some limit in the zero to 10 30 mi.le rang.e when th.at isn't where the bulk of the effect of
.11 the Class 9 accident is .anyway?
12 In other words, if we wanted to take into account 1.3 Class 9 accidents, it would seem to me we would want to take 14 them into account in all places.
15 MR. DENTON: I think you are asking two di.fferent 16 questions. One is, in the NEPA analysis, why not take into 17 account Class 9 accident consequences in every case?
18 C0MMISSIDNER GILINSKY.: Yes.
19 MR. DENTON: L think that is -- the revision of Annex A 20 is the category.~ct~sl>'.__ .action plan, and it requires having a 21 tool. that is competent to get an accurate calculation on it.
22 That is the category A that Del mentioned that we are redoing 23 the ways that we cal.culate three through eight, to make sure 24 that .we are doing them consistently with nine. And we would 25 hope that within a year we would have three through nine all
17525.04.9 38 done on a ~onsistent manner on every case.
-te 2 COMMISSI-DNER GILINSKY: You are proposing to do the 3 Class 9 in certain cases?
I 4 MR. DENTON: We-'re only proposing to do it where it I 5 is different than the usual plan.
6 COMMISSIONER GILINSKY: Why limit it to those cases?
7 MR. BUNCH: If we look at monetizing or any of the 8 other tests and we take into account both the calculated 9 consequences and the calculated risks of abilities, the 10 judgment .we generally have is that the numbers are very low.
Jl .Ths calculation is not particularly sasy to perform. And 12 simple sconomy says if we can conclude that the outcome is not 13 going to influence the balancing between alternative sites in 14 any significant way, it doesn 1 t seem to me, at least, to be 15 particularly beneficial to go through the exercise of a 16 numerical calculation.
17 Even in this particular instance, it is not clear 18 that if a future case came up, that there would be benefit in 19 doing a sp.eci.fic crack calculation. I think we might well be 20 able to say, yes, there are differences in Class 9 risks here 21 versus one in a very isolated area. The sites are relatively 22 homogeneous, the candidate sites here, and based on work such 23 as has been done here, we have every expectation that if we 24 did a calculation, it would only show, as in all the other 25 cases, the risks are very low.
I I .,
I I
I I
I I
.h 17525.05. 1 2
CHAIRMAN HENDRIE:
39 But you seem to_think that they are significant enough to take into account.
I 3 MR. DENTON: Because there are choices between I 4 alternative sides; whereas, it might af~ect your choice I
I 5 between alternative sides; .whereas, if it didn-'t affect the I 6 basic decision, if population density is low to begin with 7 at the .preferred site, then we would not learn information 8 that would, say, go to other _sites.
9 I think that we have looked on it as a decision --
10 COMMISSIONER GILINSKY: There may be other reasons Jl . 1or sites, for looking at one or another site, even though 12 the, p.opulation density is lower than the 5_00 per square mile.
13 MR. DENTON: We do that anyway. We look at other 14 alternative sites from an snvironmental standpoint, but ~e 15 don*'t give population particular .weight, unless it gets 16 high enough to warrant this detailed look at th.e different 17 consequences.
18 COMMISSIONER GILINSKY: I don't see ,why you would 19 attach _more significance to that in the av.er 5-00 square mile 20 persons.per square mile cass than in the under ~00 persons 21 per mile case when the f a.cts 22 MR. CASE: This is just the minimum effects. The 23 accute effects are much closer in than 30 miles.
24 COMMISSLONER GILINSKY.: But you are lumping them 25 altogether.
7525.05.2 40 MR. CASE: That may not be the way to do it. Maybe 2 you ought to look at sach o+/- the individual ones.
3 MR. DENTON: We have done it also separately, but 4 we have_tried to put them together to get a total fix. But 5 ws also look at ea£h one of these parameters separately.
6 MR. BUNCH: As Chairman Hendrie pointed out,_this 7 particular graph is where we try to convert everything into 8 numerical values so that w.e hav.e a figure of merit to compare 9 against other numerical Nalues.
JO In the paper, ws tabulated acute fatalities, other
.11 kinds of impacts that might be associated with the Cla.ss 9 12 accident and looked at those separately.
13 As Mr. Case has pointed out, it is not completely 14 nbvious that this kind of lumping together provides you with 15 all the right insights into comparison between the alternative 16 sites.
17 MR. DENTON: Th.e 30 miles was really*, a way to --
18 that the staff thought would help differentiate when you 19 could .distinguish between sites with som.e -- more than a 20 factor of 2.
21 In this case, it is barely distinguishable between 22 the ones. And I think fo.r the standard .case, we have just 23 referred to 14.00 as being the best representation of Cla_ss 9 24 residual risk, applicable to sites in general and applicable 25 to .any particular site that was not an exceptional one.
7525.05.3 41 MR. KELLEY: Aren't most of the .fatalities going 2 to be in ths 30 mile?
3 MR. DENTON-: Yes.
-4 MR. KELLEY: Then that..,s a good reason.
5 CHAIRMAN HENDRIEJ You are talking about something 6 differ.ent. Harold and company here are talking about an 7 additional element that they think it would be appropriate 8 to use in comparing alternative ~ites.
9 What you are saying is, look, when you go ahead 10 and try to total up the environmental costs of this plant,
.11 why don.., t you run through and take th.e n} 11 range of Cla_ss 12 9s?
13 The answer is they are trying to work out some 14 reasonable computing machinery to be abls to do that on a 15 consistent basis ~hrough the whole list.
16 COMMISSLONER GILINSKY: The additional part is because 17 the bulk of the Class 9 effects ssem to be beyond the 30-mile 18 limit.
19 MR. DENTON: For man-rem, that is true.
20 COMMISSI.ONER GILINSKY: If they weren't, then I 21 would say what they are suggesting makes a lot of sense. In 22 other words, they left out the Class 9. If you are ge~ting 23 over the 500,person per square mile limit, then you may have 2'4 some population problems, so we are going to l.ook a lLttle 25 more c.ar.efully.
17525.05.4 42 MR. CASE.: If I had weighted acute deaths, a 2 factor o_f 10 higher in cancers, a factor of 10 lower, it would 3 admit that point and you would say fine. It depends on how 4 much w.eight you want t.o assign to each one of these stands.
5 And Sol makes .the point that w.e have weighted 6 man-rem by a factor of LOO t.oo high.
7 COMMISSIONER GILINSKY: You may think you have 8 ., w.eight.ed fatalities t.oo high, too.
9 MR. DENTON: I think the 5-00 served its purpose.
JO CHAIRMAN HENDRIE.: DonJt a+/-tack a policy which has JI been anormously ~seful in an almost totally informal way.
12 (Laughter.)
13 MR. BLOND.: The.re i.s one other point that might be 14 made and that ,would be that the population that you have got, 15 or the -- excusB me the figure -- I just lost my point.
16 (Laughter.)
17 MR. CASE: ~00 per person is gnod, maybe you were 18 going to say.
19 MR. BLOND: No. A design basis accide.nt in 20 Class 3 through 8 accidents don't really given you any other
- 21 merit that you can measure population in close at all because 22 there is no iacts that will essentially be generated by this.
23 You need.the Class 9 accident in .order to .make any .evaluation 24 of consequence.
25 CHAIRMAN HENDRIE: Population dependence.
7525.05.5 43 MR. BLOND: That"s right. So it is the Class 9, 2 the .use of the Cla.ss 9 accident that is r.equired to see 3 what is going to be the effect on a population.
4 MR. DENTON: I .would think that we would be 5 revising this 5.00 number as a result of the actions that hav.e 6 been menti-oned. It was developed before 1400 was available.
7 And so it was based strictly on lnoking at population density, 8 variances in the country and choosing a number based on 9 what was available with regard to sites, rather than this 10 . detailed knowledged of where fatalities occur, man-rem occur, J l or decontamination costs occur.
12 L think that ,with the knowledge we have now got, 13 we can go back and develop a ,much more sophisticated indicator 14 of relative consequen~es. But in terms of its usefulnass, 15 it has resulted in proposed sites which have been in the low 16 range of alternatives available in the applicant service 17 region.
18 MR. MALSCH.: .. One of the problems has been that the 19 part JOO reviews have tended to focus on the doses at 20 selected .distances from e.ssentia-1ly Class 8 accidents. As 21 technology improved, that particular calculation didn 1 t turn 22 out to be .useful in contro-1ling other densly populated 23 areas.
24 When part 1~0 was put out way back, 15 or so years 25 ago, it ~as stated that it was intended that the Part continue
~--
17525.05.6 .44 Commission policy of keeping reactor sites away from dense populations .
- h 2 So once the dose guideline value calculations didn~t 3 servB that purpose, the starf had to look to other kinds of 4 methodologies to achieve .what was thought to bs the 5 underlying purpose of the part; namely, to have some way of 6 avoiding selling near-density populate.d_ar.eas. And this is 7 just the latsst in those.
8 MR. DENTON.: Back in th-.e days that we are discussing 9 in J73, there was a continuing trend in applications to go 10 to higher and higher populated, or more populated £ites, Jl with Nswbold being the highest.
12 I think that step -- and when the line .was drawn 1.3 there and when the starf adopted criteria to determine that 14 practice,. no.w I think you are co.rrect to question the 15 relationship between that and .what we now know about 1-4.00.
16 It is not the_right way to present it.
17 CHAIRMAN HENDRIE: The simplicity o.f the trip point 18 compensates for a lot of the objections that you might have 19 because of .. this simp.licity. It do.esn"t .require much of a 20 struggle at all to know whether you are in the ballpark or 21 ab.oVe .or below.
22 The.r.e is a lot of advantage in that.
23 COMMl.SSIONER GILINSKY: If you are belo.w in that 24 limit, do you take population into account at all in 25 assessing the advantagss of alternative sites?
7525.05.7 45 MR. DENTON: No.
2 COMMISSIONER GI LINSKY:. You don-'t.
3 MR. DENTON: If it is belo.w 5DO --
4 CHAIRMAN HENDRIE: Is there a special case of that 5 where one site or another has some peculaiar evacuation 6 prope.rties?
7 MR. CASE: Yes, but I _.understood his question to 8 be on the NEPA side. On the NEPA side, no.
9 MR. DENTON: Comparing most of our rackups of 10 alternative sites, we include population, but we don-'t J1 sxplicitly take it into account if it is below, if the 12 proposed site is below the trip point.
1.3 COMMISSIONER GILINSKY-: Suppose you were Looking at 14 a site which is below that trip point and you had these sorts 15 of answers, calculations available. Would that as a practical 16 matter affect your decision in any way? Would this affect 17 these residual accident risks, weigh in the balance 18 significantly?
19 MR. DENTON: I think at these levels and recognizing 20 the uncBrtainty that is in them, we would tend to focus more 21 on the environmental differences between the sites and 22 differences like this.
23 COMMISSIONER GILINSKY: You are saying it wouldn*'t 24 really amount to anything?
25 MR. DENTON: Not at this level. That is correct.
7525.05.8 46 COMMISSIONER GILINSKY: Why would it in the 2 over 5D0 category?
3 MR. CASE: It might not either. It just says 4 look and see. That is what the policy says.
5 CHAIRMAN HENDRIE~ It didnJt here as calculated 6 for this array.
7 MR. DENTON: We had expected a bigg.er difference 8 to turn out, based on the difference in the zero to 30 mile.
9 CHAIRMAN HENDRIE-: On the other hand, I expect 10 Raven's-- .. would st i 11 be perceptively different?
.11 (Pause.)
12 Richard, you have been silent through all of these 13 questions to this point.
14 COMM ISSI ONER BRADFORD: SQ.~l th.:i,,~gsi §i~:_: _4-r_:J:c:sJ-pwly
- 15 MR. BUNCH: That basically conciudes the 16 presentation we had planned today. The only final remark I 17 .would add is, again, there is a need in the NEPA reviews to 18 consider and balance the various economic environmental 19 safety issues associated with alternative sites. And we do 20 so, as Harold .was just pointing out.
21 What the substance of_ this discussion is, when w.e 22 run into an unusual situation and we feel the n.eed, and it is 23 so suggested in our paper to take a look and* make sure that 24 we are still satisfied, at least from the population 25 standpoint that there is no special need to take accident . risks
7525.05.9 47
.or give risks great .weight in the overall balancing.
2 We have found this particular device to be helpful 3 to us in that a.sse.ssmen t. I am not sure that unl e.ss we have 4 a different kind of ensemble of sites in the next alternative 5 site review, that ~e would necessarily invoke this 6 particular msthod. It does seem to have some aspects. As 7 has been indicated several times in the discussions today, 8 there are complications in it that have to be approached 9 cautiously.
10 There are several activities underway to better Jl understand the values and.the limitations on the crack 12 model in site specific evaluations and in evaluations of 13 alternative sites.
14 COMMISSIDNER GILINSKY: Is i.t your view that the 15 vuse. of*~ this model is inappropriate in some ,way:-?.=,?
16 MR. LEVINE-: This is a first effort. I don-'t -.want 17 to be critical of it. But if one wants to use it on a more 18 regular basis, .one ought to do it more c,arefuJly.
19 You can get right out of the ~E_IR ::E-~PJ:\l:ltt,, :i:=tG1 20 example, that a latent cancer fatality is worth somewhere 21 between $10 and $LOO per man-rem. And if you apply life 22 shortening considerations, it is worth about a third of that.
23 So it comes down to $3 to $30 a man-rem, or about $10 a 24 man-rem, because you only lose about a third of a life, and 25 things like that.
7525.05.JO 48
.sh And then I think that you want to weight those 2 things, weigh the various factors, as opposed to just 3 adding and se.e what you think about it. Then you might get 4 some index 5 COMMISSIONER GILINSKY: But even if you reduce the 6 ,man-rem by LOO - -
7 MR. LEVINE: Weil, then, but the question is if they 8 are all the same, still, or roughly all the same, you want 9 to use thos.e .as pa.rt of your index or .not. And then one has 10 to think that through.
JJ How would you weight, again, the importance of
- 12 13 14 early fatalities versus late fatalities on scales other than just life shortening, for instance?
COMMISSIONER GILINSKY: These are vague indications.
15 MR. LEVINE: You might want to look, for instance, 16 at a large number of sites to s,ee what kinds of variances 17 you .do g.et and convince yourself they are sma.11 or if they 18 are large, you might f~nd it a better way of combining the 19 things together in a model to give you more useful guidance.
20 I am not sure how important it is. I. think there 21 was a mtsstateme.nt made earlier that the risks from accidents 22 are not the same as those from routine effluent releases.
- 23 24 25 L think that they are in terms of man-rem, at least, three orders of magnitude smaller from accidents than from --
per year, than from a routine effluent release.
7525.05 *. 11 49
.h 2 MR. DENTON: I meant to say -- I thought they were comparable occupational exposure.
3 MR. LEVINE: Occupational exposure is about the
- 4 same as routine effluent releases. I misunderstood.
5 COMMISSIDNER GILINSKY: Comparing to the other 6 accident categories.
7 MR. LEVINE: If you want to lnok at accidents and 8 forget that, then the large accident end of the spectrum 9 gives you 10 MR. DENTON.: There is .one other area w-h'e~e 1.we t\fs'e Jl these numbers. That is in comparing the health effects of
- 12 13 14 alternative energies. There we don~t monitor them.
leave them in terms of health effects in comparing to-coal.
MR. BUNCH: That concludes our remarks.
We 15 CHAIRMAN HENDRIE: Peter, have you developed any 16 questions? I se.e that you have de-veloped a pu.zzled Look, 17 if it is not quite Yerbalized yet.
18 There were assorted commentaries on the paper.
19 Jim?
20 MR. KELLEY: I-'.11 pa"5s.
21 CHAIRMAN HE NORI E.: Yo u'.11 pass. Does OPE .want a quarter's worth of the discu.ssion here?
22 23 MR. SEGE: Yes, Mr. Chairman.
24 CHAIRMAN HENDRIE: Go ahead.
25 MR. SEGE: We have seen the great unc.ertainties
7525.05.12 50 su.rroun.ding these relationships which everybody .in different 2 terms has indLcated a keen awareness of, the uncertainties 3 about .the di1ferencs in probabilities in accidents, the 4 uncer.tainties -- the nature and magnitude of an accident 5 that .might occur, ths different uncertainties regarding I
I 6 *hether topographer emergsncy actions, population shifts I 7 over time, uncertainties and possible.controvarsies about I
I 8 the gross simplif.ication as applLed to the results in terms 9 of translatiDn to dollar terms.
JO So the question, the basic question really arises, JI *how.much calculation such as this really a.dds to knowledge 12 about a specif.ic site?
1.3 Is that addition al know 1 edge rea 11 y real? Do .we 14 know precisely what the risks are in preparing these sites?
15 Is ,that know 1 edge stemming from this cas.e analyses pr irnar ily J6 i 11 usory.?
17 If the knowledg.e is not that real, .then the question JS occurs whether some sort of simple population guidelines that 19 .might sxclude siting above certain population densities or 20 generally permits siting below certain population densities 21 and perhaps create a zone in between where there may be a 22 presumption that could be overcome.
23 Whether the simple guidelines .wouldn*'t do just as 24 well as case calculations, on the basis that is now possible 25 and actually proposed, whether it does not introduce po.ssibly
I 175;:5.13 51 damaging illusions of .kno.wledg.e, whether we don't possibly 2 t.end to gloss ov.er. things that dow't happen to appear in 3 the "model of calculation.
4 I
5 6
7 8
9 10 e-5 .I 1 12
- 13 14 15 16 17 18 19 20 21 22
- 23 2A 25
7
~5 .06. l 52 gsh I am not suggesting an answer; I am suggesting 2 a question regarding two ways one could go in terms of 3 following some sort of guideline based on generic evaluation 4 of Class 9 accident possibilities versus (-inaudible) case.
5 Tlie:te; is an intuition which says that low population 6 densities around reactor sites are better than high population 7 densities. The consequence model calculations donJ.t seem 8 to correlate very strongly ~1th that in.tuition, and it is 9 not clear why, and it makes me somewhat uncomfortable that IO there is no such s.trong correlation *bt:1tthere is no clear
.II showing that the intuition is likely to be wrong.
12 I would feel more comfortable if there were more 13 of a correlation between results and the intuition that it is 14 better to site in less densely populated sites.
l5 Lastly, I have -- the question has been raised 16 before in the course of the discussion as to whether it is 17 really appropriate to confine Class 9 considerations to just 18 the 500 population, but that there are other situations where 19 it could be notably a situation in which the compensating 20 containment design features which make it possible to, in a 21 very small population zone, such as --- well, under a mile 22 and whether in those cases not considering the Class 9 23 possibilites might not result in an incomplete and perhaps 24 misleading evaluation of the proposed site.
25 With this, I would like to leave the question whether
25.06.2 53 gsh the power of this analysis applied to individual cases is 2 sufficiently strong for the intent and purposes and whether 3 consideration of Class 9 accidents should properly be 4 confined to just those cases of 500 and more population, or whether 5 there aren't certain pt:lj!_eJ".;..S.1,:1:;i:t:t_?-\-tj;a:r:is--, where consideration should 6 be given to Class 9 accidents also in some sort of fashion.
7 COMMISSIONER GILINSKY: I would like _to pursue this 8 a little fur_ther because I don't think I was clear in what 9 I said before.
lO We used the population density in a 30-mile circle
.11 to decide whether or not a site needs population standards 12 of some sort.
13 MR. CASEi At varying distance.
14 COMMISSIONER GI LINSKY.: It is a little more 15 fine grained but it goes out to 30 miles.
16 MR. CASE: Yes.
17 COMMISSIONER GILINSKY-: So what we seem to be sayin9 18 is that we care about what ha ppen,s out to 30 miles. Now if 19 a site goes over that limit, it seems to me we ought to use 20 some method that applies to that 30-mile circle.
21 Now it is a separate question as to whether or not 22 you ought to be looking beyond that circle. If you think you 23 ought to be looking beyond that circle, I think you ought to I -
24 be looking beyond it in every case.
25 But it seems to me that looking beyond it is not
25.06.3 54 gsh the way to solve the problem within the circle.
2 MR. DENTON~ We 11, but 3 COMMISSIONER GILINSKY: I don-'t know if that is 4 any clearer.
5 MR. CASE: You are saying that ~here is a logical 6 inconsistency in what we do.
7 COMM! SS !ONER GI LINSKY,: Not really, because what 8 we are doing --
9 MR. BUNCH: We are doing an analysis using the 10 Wash 1400 model. It turns out that that model turns its 11 calculations out to 500 miles. We do perform assessments of 12 the e.ffects within that 30-mile area.
L3 Typically, what you are going to be interested in 14 are the early fatalities as you calculate them or -- what-'s 15 the other figure of merit? There are about 6 or 7 principal 16 ones that we use, some of which are more specific to the 17 close in population. Clearly, using the criterion that 18 refers to the close in population, one of the things you 19 want to make sure of is those tests that can come out of the 20 Wash 1400 type calculation will give you insight on the 21 within 30-mile differences. You are going to look at it very 22 carefully.
23 MR. DENTilN: Let me go at it another way. I 24 understand what you are saying. But basically, there are 25 three decisions or recommendations the staff makes in the
25.06.4 55 gsh environmental area. Do you need more power? How should you 2 generate it? And when we compare how you generate it, we 3 do include in that comparison of the health effects of 4 generation all the health effects out to all .the numbers 5 of miles. And three, then, is where you build it.
6 So in looking at where you build it, we don-'t 7 presently give detailed attention to population if we know 8 by inspection that the population density is near the minimum 9 that can be achieved in that service area.
10 That is what our standard review plan does, simply
-11 it assures us that we don't have .to look hard because we are 12 not likely to find sites that are a lot better than the one 13 pro posed if the Qne p.roposed is within .this very simple 14 criterion.
15 But when we look at alternatives of energy sources, 16 we are including all of those .heal th effects beyond 30 miles.
17 COMMISSIONEr, GILINSKY.: But nothing ever comes of 18 that analysis.
19 (Laughter.)
20 MR. DENTON: It is an important decision point. I 21 think if it came out the same way, something would come of it.
22 COMMISSIONER GILINSKY.: But when we get down to 23 dealing with a reactor and deciding where to put it, we look 24 - out to 30 miles. That may or may not be the right thing to 25 do, but in any case, that is what we do.
I I
25.06.5 56 gsh Now if it goes over our limit ---
2 CHAIRMAN HENDRIE: Wait. I don*'.t _think that-'s quite 3 the right thing to say.
4 MR. DENTON: If it is within 500, we look at the 5 normal part 100 indicators and we don~t give detailed 6 attention to trying to find better sites between 275 and 7 .31 O people per square mile.
8 COMMISSIONER GILINSKY.: Right. But if it goes over 9 the 500, I would think we would want some kind of measure 10 that applies to that 30-mile circle, some different test.
I I MR. DENTON: The first test that .we used .to have, 12 tests that we used .to have were nothing other than drawing 13 on draft paper relative numbers of people, plotting where they 14 lived, close and far out. Then we came up with the SPI 15 index, which was a way of .weighting people by meteorology 16 but not by specific meteorology because it gets complicated 17 and you get direction dependence. And we looked one time 18 at weighting people in the 30-miles by the weather in each 19 sector, but that is getting very complicated. And then when 20 the crack code came along with all of its defects, it was a 21 way of .weighting all the population and the people and the 22 weather.
23 CHAIRMAN HENDRIE: Even.
24 MR. DENTON: Even. It was the best .way available 25 at the time.
- 25. 06. 6 57 gsh MR. CASE: That is what we have been doing up e 2 until now. It is not perfect.
3 MR. BUNCH: I think your point is well taken, that 4 of all of the things that might come out of such as 5 assessment, we ought to make sure that the principal tests 6 that are looked at are those that relate to the zero ~o 30 7 miles.
8 It turns out that because this was a NEPA review 9 and because we do monitize a lot of impacts that are 10 associated with the various sites, we thought it would be 11 useful .to go the next step and try to monet.iz~ the residual 12 risks, just .to give people some additional perspective.
13 I think from the standpoint of making a comparison, 14 it would be perhaps to be more helpful to key in on those 15 that relate to the zero to 30 mile population differences.
16 CHAIRMAN HENDRIE: But if you don't monetize them, 17 how do you compare .the population effects with other aspects 18 of the alternative sites?
19 MR. CASE: I think that you could still :mom..et.:i;z e 1
20 them. But I think Vic is saying, he wonders whether you ought 21 to ime>'f.l.e~t:i;ze beyond 30 miles, because your basic assumption 22 going in is they are all the same beyond 30 miles, so you 23 don*'t worry about that. You only worry about things up to 24 30 miles. And therefore, that is your assumption, then stick 25 with it.
___ _J
25.06.7 58 gsh COMMISSIONER GI LINSKY.: There are two parts here.
2 One is how you deal with the zero to 30 miles. And if you 3 are prepared to go above that 500 limit, then I think that 4 you ought to look more carefully within that circle. And 5 there may be ways to distinguish sites *. If you think you 6 ought to be looking beyond that, I don~t understand why you 7 wouldn't look at those cases that exceed the limit.
8 It strikes me as being a separate question whether 9 or not you go beyond the 30-mile circle. And if one decides 10 that one ought to look beyond, I think that you ought to look 11 beyond. I think that you ought to look in all cases.
12 I don/t see why you are putting some and not others.
13 MR. CASE: I think that we are saying the same 14 thing a little differently.
15 MR. MALSCH: You would have the same problem if you 16 are worried about coal mine accidents in the absolute balance 17 from a nuclear power plant. You would wonder if you could 18 draw distinctions of up and beyond 30 miles.
19 COMMISSIONER GILINSKY: Harold was saying that in a 20 lot of these other cases, in an absolute sense, it wouldn~t 21 amount to all that much. And if that is true, I suspect it 22 is also true in the cases where the 500 person per square 23 mile has been ex~eeded also.
24 MR. CASE: Yes.
25 MR. DENTON-: Except for very high populated sites.
- 25. 06. 8 59 gsh That's true. As we said at the beginning, the Wash 1400
- 2 ~ study shows that most of the risk averaged together is due 3 to those few sites at the high end.
4 CHAIRMAN HENDRIE: Any other q.uestions?
5 (Pause.)
6 MR. KELLEY: I would like an inf ormat iona l comment.
7 The party tells me that this issue of Class 9, et 8 cetera, it would be litigated before the appeal board. So you 9 may get a decision from the appeal board fairly soon. It may 10 be narrow focus. I don't know., But you may get something JI in your in-file before too long.
12 This has been a generic discussion today, but I 13 just wanted to flag the fact -- as I understand it, the 14 applicant claims that the staff can't look at Class 9 at 15 all, and that is the issue.
16 CHAIRMAN HENDRIE: When you say relatively soon, is 17 that days?
18 MR. MALSCH: That is, the oral argument has been 19 scheduled for late next week before the appeal board.
20 MR. KELLEY: It might be mid-summer.
21 MR. MALSCH: Briefs have bsen filed by the staff, 22 applicant and other parties.
23 CHAIRMAN HENDRIE: There are two recommendations in 24 the staff paper which was in the commissioner action paper.
25 Would you prefer to lull this a while and then deal
I I
I I
I 25.06.9 60 I I
gsh with the concurrent sheet basis rather than -- I feel a 2 little reluctant to se.t up .and take votes, .L must say, at 3 the moment, and, indeed, this afternoon 1 s exercise was 4 intended as a briefing .to help focus the Commission 1 s attention 5 on these matters rather than as a decision session.
6 So I would say thank you very much for the 7 briefing and discussion.
8 The Commissioners are admonished to k.eep it on their 9 desks for actions as they feel able.
10 <Whereupon, at J:20 p.m., the hearing was adjourned.)
11 12 13 14 15 16 17 18 19 20 21 22 23 24 25