ML20217M763
ML20217M763 | |
Person / Time | |
---|---|
Issue date: | 02/19/1998 |
From: | NRC COMMISSION (OCM) |
To: | |
References | |
RTR-REGGD-GENERA NUDOCS 9804080195 | |
Download: ML20217M763 (180) | |
Text
..
ggpo%i.
awM 4 cx .. c au$27@kk1 #+4emn, $kny y;fgs@p&m%
pyg e w q.N OFFICIADTRANSCRIPT y 9h%% y:p;t &
aaK %w;Ngsggggy#"%p&W p m p ,wy:m#1 : ~
- OF"PROGEED.INGS @mp*g h%w Mlh w% m %y ~
I-
+ .
., .- mr
% '. , y [ ,,j . .
~.
7r .o ' UNITED STATES OF AMERIC . ' ". . .:
d- s '$ y
-.~
$q@.mrng?g!iW@V@.
k /b
~
k hMfk h.,[f Mhdkd[h5Nh)Mk.%h)IhfMh
M@$$$
h b(
EM M
b vi .i .
h .
II y* .L w gvwggg ?< '
~.4, .
.e 1[ HOP. ONM"4 ' - "'
"W4
,.f?f. f*9 l>f 1
'W ,Q' yy ? "i '?
.. > s ,. > w . ,
i..*- #
....A
, , 4,+
6 l. . s A.
".'.'.*m..v.-
( $, -
j t
.lyl'.fj'
.s . . w. l 8 U 1( A_I . ('
. N.I I L 3 ie\
. ...,..;. 7. . . .- .
- t. .
d/[F #
hkhYSMii)iIi'l \ .
/
i
- a. ..
- Vi ,dhh.- i N" ..u. .'fthN dlPN P D
't T ?t l's 4 D Jjl,lL.( 0, -
- d. .M. . . . M_(M. _ .:. -- . . .
,, ..;t , . .
-
- 1.
, ,( -
4 -.,-a.,+-
pggj g,2 " ' '
'4 "A
-- .l
- .. . . .o. ..,
g ,
W 9.. b ( ) t i.I u * l o :NNk .
'.3 i..': . . a, .,
s y,4f(n).'6 g.lfg@ J' - ,
.. r
)
O 3- . ATiBN2t@ndh && , V'M*D W %
hT} ' , hf 8t .
lBs .
s 9804080195 980219 PDR COMMS NRCC .,
,..s. ..
CORRESPONDENCE PDR
.V<' '2. N. "'~ -
_____..____J
nn"-*' -
Ji:' - .
.s
. y . g7 , s. ,
" ,2' '
~ Nyg gh ' e, o s . s
- m. m n' '
,k.# MMim,, . , , L y
gOg-wwqakp Y!,$q da y? bY v n w5.... e ? $. m l
$??W5 b M.o if,3#@f'$pu$!Q.$4NOFFIGIA - m. - ,RA'N.SC. RI sg.
- m. w.,;m,DY
, n
<. m . ~ .
r .
S 4,
$srdh 3l ?$NUNITNDYSTATESLOFNNIEInh73$h,kkSIP h:i{lk_@&a$m$an$$$$f' m
$ o %$$W?$hWhQ&Wked~?Wb" IhTORY$GQBRh (l V y -
cn- .
j
- By,. ,
. , e s' e ,
. W.
.s{;~ -
., . m.t
'[TitigL'>'
l ' ' 'avDM:91En14
,.' ' a * %;
j n' crgtejbagrir JIL7A2 i_ J .- #3l[r ~
. ~ . . '
, .'( .
[
, ' 3 l
t , .
.a. s .o - -
,< .a l .i *
)
- p. (, "> .'.
is.
, e j
j ., 4
.O e -
v > . ,- .
i p ,. -
9 ., .
- n e
V
, ,~. r , .
p <ma f y 4 a <a.t .
- p. , ,
4, a .
- s. .J
- p
- p. , . ,
N f
-hy^ - q c '
A '
g ,..
~
j t t 3 l ,
w
, . . , .,e.
r n' '
( *
.4i 3
Qy +
C d'2' t :s - -
~
sr -
f j
- - 's , ,
I I
, 1 Y, p u
- r. .
p k '
- a. .~
u 3
9
, v1 y
g gr f
(.
1
-r t
,'W
' 9 ._ .
- y. , > c ,
l ,, -
y J, '4 g
'V(
l ,
J -
, lb'
- ' y pr. t , t ,
n .
- g. .
g .. t - ..
IV - -
a,' u *,
_ , . , c .c e
- d ./ .
r ;
,gj '+
g '
4
_ , r i e -'h qp ~<
q)_.
__ - 9 z.
i n
~
y, W
, s'.n '
a) ,
3 ,
- '- 1 d e i g -
~
,- 3 ana m-n---2~ - n<
gg
) -
, > gr 1 a .
x .'
- x g, n . - -v .w-mv vv was4}
5 s -
y, ,
- *s,- e , ' ' '
1 o
330 *
, t I . , ,- <
s P ', --
NW7]e' 150
' ' kG,et(/D:
~
. ngton i
td > , ' f ^ ,
, " 02)f ,
m Q -
r '
~
t 2
- 5 ;
+ c. V
\ - . - - h. s x A a u GL _ . _ -v '
P ':r 9004000195 980219 1: 4 PDR COMMS NRCC "
b.'
~
A o .;
g_ -. CORRESPONDENCE PDR t i d.J.,.,_A f (
gin ma JL 2 . . > __cL -
_ __._____--J
143 1 UNITED STATES NUCLEAR REGULATORY COMMISSION 2
3 ***
4 WORKSHOP ON 5
DEMONSTRATING COMPLIANCE WITH RADIOLOGICAL CRITERIA 6 FOR LICENSE TERMINATION - METHODS TO CONDUCT 7 A FINAL STATUS SURVEY AND DOSE MODELING 8 ***
d 10 U.S. Nuclear Regulatory Commission 11 Two White Flint North, Auditorium 12 11545 Rockville Pike 13 Rockville, Maryland 20852-2738 14 15 Thursday, February 19, 1998 4 16 -
17 The above-referenced workshop commenced, pursuant 18 to notice, at 9:03 a.m.
l 19 PARTICIPANTS:
20 CHERYL TROTTIER, Chief Radiation Protection &
21 Health Effects Branch, NRC -
~ '
22 CHRIS DAILY
, 23 THERESA BROWN {
24 RALPH ANDERSEN, Nuclear Energy Institute 25 MAN-SUNG YIM, North Carolina State University
\ l t
ANN RILEY & ASSOCIATES, LTD.
Cot.rt Reporters 1250 I Stteet, N.W., Suite 300 Washington, D.C. 20005 l (202) 842-0034 1
- ]
144 1 PARTICIPANTS: [ Continued]
2 RICK ROBERTS,-RFETS-SSOC
- 3 PETE LITTLEFIELD, Duke Engineering Service 4 CARL GOGOLAK, Department of Energy 5 COREY McDANIEL, EOP Group ,
6 HENRY MORTON, Potomac, Maryland 7 ERIC ABELQUIST, 8
ALEXANDER WILLIAMS, Department of Energy 9' BRUCE MANN, Commonwealth Edison 10 TOM POTTER, Radiation Protection Consultant 11 EARL SAITO, Combustion Engineering (
'12 A. JOSEPH NARDI, Westinghouse 13 GREG CHAPMAN, Nuclear Fuel Services 14- KEN.DUVALL,-Department of Energy 15' BILL COOPER,-GPU Nuclear, Oyster Creek 16 BEVERLY GOOD, GPU Nuclear 4 17 f
18 I 19 20 21 .,..
22 23 '
i 24-25-1 \
l l
-ANN RILEY & ASSOCIATES, LTD.
l' Court Reporters .
1250-I Street, N.W., Suite 300 Washington, D.C. 20005 i
(202)-842-0034 '
L. )
145 1 PROCEEDINGS 2
[9:03 a.m.]
L 3 MS. TROTTIER: Good morning, if.I could get your l 4 attention, please. We're going to try and get started on 5 time.
u 6 As you can see, we have a little bit of technical
!' 7 problems here, too many computers, I think.
8 Anyway, my name is Cheryl Trottier. I'm the chief 9
of the Radiation Protection and Health Effects Branch in s 1
10 research in NRC. Our branch has been charged with the task-11 of putting together this regulatory guide. What we're going 12 to try to accomplish today is to walk you through the 13 methodology that you will find in NUREG-1549.
14 After the meeting yesterday I realized'that a lot 15 of you, unfortunately, don't have all the time in the world .
16 to sit and review these voluminous documents that we have 17 been putting on the website. So I think it might be 18 helpful, and that's what the. staff is going to do this 19 morning,'to spend a little bit of time, maybe even up to a L 20 couple of hours walking you through the methodology.
21 Hopefully that will make it a little easier for you to come - -
22 up with issues.that you want to raise before us, before we 23 get this guide sent forward to the Commission.
l 24 I do want to reiterate what we have -- at least at 25 a-staff level -- decided to try with this guidance, and that I i
ANN RILEY & ASSOCIATES, LTD.
Court Reporters l 1250 I Street, N.W., Suite 300 Washington,.D.C. 20005 (202) 842-0034 1
146 1
is to put it out as a guidance document and including the 2 NUREGs that support the regulatory guide for interim use and 3 comment. What we're hoping to accomplish by that is to give 4
you an opportunity to provide more reasonable comments to us 5 because you're trying to use it. I think part of the 6
problem during the normal public comment period is you're 7
not really using the material, so it's hard to judge where B its flaws are. So what we're hoping is by putting it out 9 for a period of comment that we will get some more useful 10 feedback from you.
11 I realize that a real problem with this point that 12 we're at now is with does methodology is it's the piece you 13 have seen most recently. In fact, you know, it's barely 14 been on the web I think for maybe -- I don't know, maybe two 15 weeks if it's been there that long. So we realize you're at ,
16 a real disadvantage.. This is a fairly new methodology that 17 we're trying to incorporate here and we would like to get --
18 at least today what I'd like to get is some feedback on l 19 whether the approach is even reasonable. What we're 20 proposing here is an approach where it's basically a 21 decision framework to help you make decisions on cleanup. -
22 And it starts out with a screening approach. We believe I 23 there will be licensees -- granted, not licensees with major 24 contamination, but licensees with minor contamination which 25 would actually be able to use the screening values that will i
ANN RILEY & ASSOCIATES, LTD.
Court Reporters 1250 I Street, N.W., Suite 300 Washington, D.C. 20005 (202) 842-0034
i 147 1 be presented in the guidance documents.
2 Once a person goes through that step, then there 3 are other steps in there that help them make decisions on j 4 what's the most cost-effective steps to take. In most cases i
j 5 it's simply fine tuning the information that goes int _o the 6 model, making that information more site-specific. The I 1
7 screening values are based on very generic'models. They're 8 based on averages that are probably not realistic for the 9 majority of sites. So, therefore, you wouldn't expect the l 10 values to stay the same once you input your own !
)
11 site-specific data, i l
12 So, I think, you know, the approach is different, 4 13 we're not giving you a tabl.e of cleanup values to say, okay, i
j 14 these are the values you have to meet. It's an entirely l
\
15 different approach. And I' guess what I would really like to 16 find out by the end of the day, what you think-about an i 17 approach which includes a screening table and then 18 methodology for moving away from those values in the '
19 screening table.
20 With that I will turn this over to Chris Daily who 1' 21 is the project manager on the dose methodology NUREG. She's. -
22 being assisted today by Theresa Brown from Sandia who did a 23 lot of the work on the DandD model for us and hopefully 24 you'll be able to get some useful information out of this 25 today. And I want to thank you for coming.
ANN RILEY & ASSOCIATES, LTD.
Court Reporters 1250 I Street, N.W., Suite 300 Washington, D.C. 20005 (202) 842-0034
l I
J 148 1 MS. DAILY: Good morning. We' c still trying to 2 make some adjustments, so if it turns out that you can't 3 read what we have on the screen, just let us know and we'll {
4 try and do more adjustments so you can tell what we're I l
5 doing.
6 What I would like to do this morning is spend some 7 i' time going over what i's in 1549, how we're defining 8
screening, how the models are constructed, how the parameter 9 distributions were developed, just to give_you a general 10 overview.
11 Theresa is going to go over some example 12 applications and then we'll open it up for some more 13 discussion and for specific questions about different parts 14 of the guidance that we're talking about.
15 To start with, I'd like to talk about what.
16 " screening" is. So the next slide.
l 17 [ Slide.)
18 MS. DAILY: If somebody in the back wants to try )
19 and change the lights a little bit and see if we can get it i
I l
20 a little dimmer up front. Thanks Mark. Oops, that's a I l
21 little much. .
22 [ Laughter.)
23 MS. DAILY: I'll keep going while they work on the l 24 lights.
l l 25 As we said before, the purpose of this model is to l
ANN RILEY & ASSOCIATES, LTD.
Court Reporters 1250 I Street, N.W., Suite 300 Washington, D.C. 20005 (202) 842-0034 t
149
~1- do screening. And we're defining screening in a very narrow 2 sense. The purpose of screening is not the develop an exact 3 representation of dose at a specific site. What it's 4 designed to do is to cover all of the potential pathways,
- 5 all of the potential mechanisms for exposure to a rechptor.
6 That is not going to be precise. It's not going to be 7 ex'act. 'It's not a true representation of does, that's not 8 its purpose. Its purpose is to make sure that if you 9 demonstrate compliance with screening there's high assurance 10 that you are complying with the requirements to dose in the 11 rule.
12 So default screening is based on generic modeling 13 that should cover sites all across the United States. We 14 expect it to be cost effective in that if you demonstrate 15 compliance you don't have to do a lot more work. And the 9 16 point behind this approach is that you should only do as 17 much as you need to do. If your true does is say, two 18 millirem per year, there's no reason to spend thousands of 19 dollars trying to make sure that it's not two and a half i
20 millirem per year, or five millirem per year, you're well 21 below the dose criterion. -
22 As you get closer to the dose criterion in terms 23 of what the actual dose from your site is, you have to do 24 more site-specific modeling and do a more reasonable 25 estimate of your dose if that appears to be a cost-effective I
ANN RILEY & ASSOCIATES, LTD. !
j Court Reporters 1 1250 I Street, N.W., Suite 300 Washington, D.C. 20005 (202) 842-0034
150 l
-1 thing to do. In some cases it may be more cost effective to i
! 2 simply remediate a section of your site. And that's l
3 supported in this methodology also. It gives you the option 4 to use what's the most reasonable approach for getting your 5 site released and demonstrating compliance. ,
6 The idea behind this is that we'would have a very 7 simple to use approach at the beginning for screening and l 8 that there would be minimal requirements on the licensee ~or 9 the NRC to review these sites.
10 When you demonstrate compliance under screening, I 11 want to emphasize the whole point is that you have high 12 assurance that you have demonstrated compliance with the 13 criteria of the rule.
14 [ Slide.)
15 MS. DAILY: I'd like to move on and talk a little .
[
16 bit about exactly how the models themselves are constructed.
17 Some of you have already seen this information, but I think 18 it would help for us to get back to a baseline here.
19 We're primarily talking about two scenarios at the 20 moment. Originally we had four scenarios, now we're talking 21 about building occupancy which is reuse of a commercial .
22 facility that has surface contamination inside the structure 23 and the other scenario we're going to talk about is a 24 residential scenario which is reuse of a site that has some l
25 soil contamination.
ANN RILEY & ASSOCIATES, LTD.
Court Reporters 1250 I Street, N.W., Suite 300 l Washington, D.C. 20005 l
(202) 842-0034 L.
151 1
The residential scenario is based on the 2 assumption that there are thin surface layers of 3 contamination in the soil. The property is going to be used 4
for residential purposes and light -- what you might 5 consider to be light farming. We are not assuming that 6 someone gets 100 percent of their diet from the site.
7 Originally with the version that was released back in 1990, 8
1994 timeframe the default assumption was that you would 9 receive 25 percent of your total diet from the site.
10 The idea behind that assumption is that we would 11 be accounting for all of the transport mechanisms at a site 12 that has soil contamination. So pathways that we are 13 talking about or resuspension of soils, ingestion pathways 14 linked to agricultural products at the site, like growing 15 crops, food stuffs for animals, and then external exposure .
16 from the soils and from within a structure that you would 17 construct on the site after the site had been released.
18 [ Slide.]
19 MS. DAILY: The next slide is just a visual 20 representation of the pathways that we are trying to 21 incorporate. Remembering that uhat we're doing is, since 22 we're starting from screening, we're assuming that all of 23 these pathways are active. And the idea is that when you 24 get to a specific site you can designate certain pathways as 25 not being appropriate for your particular site, but when you ANN RILEY & ASSOCIATES, LTD.
... . ,p,ourt Reporters 1250 I Street, N.W., Suite 300 2 84 I b3
152 1 start all of these pathways were there. So from soil 2 contamination there's a path to air for resuspension, 3 there's a path from direct soil ingestion, there's a path 4 from direct soil ingestion, there's a path from external 5 exposure.
u 6 Soil contamination also is incorporated in the 7 uptake to plants, ingestion by animals and then we model the 8 transport of the radionJclides to the unsaturated zone and 9 then down into an aquifer. And that water is then taken up 10 by plants, it's ingested by animals and then it's ingested 11 directly by the human receptor at the site.
12 [ Slide.]
13 MS. DAILY: The water model that we're using is 14 very simple and we expect that the primary purpose of that 15 model is to indicate whether or not there are any .
16 significant pathways.from contaminated water.
17 Obviously when you get to a specific site and you 18 need to get a better representation of those pathways you 19 can substitute a more complex model, if that's appropriate.
20 [ Slide.)
21 MS. DAILY: The building occupancy scenario is, as. -
1 22 I said, based on thin surface sources within a structure. I 1
23 We're assuming that the primary exposure is a long-term, 24 chronic, low-level exposure from normal reuse of that 25 building. There's no assumptions that the sources are going ANN RILEY & ASSOCIATES, LTD.
Court Reporters 1250 I Street, N.W., Suite 300 Washington, D.C. 20005 (202) 842-0034
153 1 to be disturbed through processes'like renovation of the 2' structure. This is just normal occupancy of the building 3 and1that's how the' parameters are developed.
4 The pathways'that we're looking at are inhalation, 5 _ secondary ingestion, and external exposure. ,
6 [ Slide.)
7 MS. DAILY: And graphically we just show that
, 8 there's three ways for the contamination to affect the 9
receptor, either directly through external exposure through 10 secondary ingestion and through resuspension.
11 [ Slide.]
12 MS. DAILY: We've performed parameter analysis' !
13 over the last couple of years and the purpose of that '
14 analysis was to determine if the assumptions that are 15 imbedded in the design in the parameters in the construction .
16 of the model result in something that is consistent with 17" what we intended for this model and for screening.
18 The idea being that as you move from screening to 19 something that's more site-specific, as you make a more
_20 realistic estimate of your dose, the dose should decrease 21 which is kind of the definition -- the basic definition of -
22 starting from screening.
C .
23 The' approach for the parameter analysis was to l
-24 identify default parameter sets that -- this is hard to read 25 .too, I apologize.
-ANN RILEY & ASSOCIATES, LTD.
Court Reporters 1250 I Street, N.W., Suite 300 i Washington, D.C. 20005 (202) 842-0034
p 154 l
)
1 [ Slide.]
2 MS. DAILY: The fault parameter sets that limit 3 the probability that when you go to a specific site and L 4 calculate the total effective dose equivalent on a 5 site-specific basis that that estimate would be highar than 1
L 6 the default, total effective dose equivalent that you would.
7 calculate with screening. And linked to that is the 8
requirement that from a regulatory point of view, if you do 9 underestimate the total effective dose equivalent, you would 10 like that overestimate to be limited -- to be small. So 11 that's part of our analysis also.
12 (311de.]
i 13 MS. DAILY: The idea behind all of this is 14 supporting a finding that the final default parameters are
.15 prudently or reasonably conservative, not excessively 16 conservative. We're not trying to do something that has no 17 basis in reality, but we are trying to make sure that we're 18 accounting for all the potential pathways and that screening !
19 is definitely doing the screening. !
20 We want to assure that there's a low probability ;
21 that the criteria will be exceeded if you had more realistic. - "
22 information.
23- [ Slide.]
l 24 MS. DAILY: The parameter analysis itself for the !
25' residential and building scenarios, we went through and 1
ANN.RILEY & ASSOCIATES, LTD.
Court Reporters 1250 I Street, N.W., Suite 300 Washington, D.C. 20005 (202) 842-0034
1 155 r
1 defined the parameters as being behcvioral or metabolic or 2 physical and then we treated them slightly differently in l 3 the analysis.
l 4 We determined ranges in distributions if those 5 were appropriate for each of the parameters, and then we are 6 performing the analysis for all radionuclides and all 7 pathways combindd.
8 [ Slide.]
l 9 MS. DAILY: I'd like to spend a little bit of time 10 talking about how the parameter distributions were h
11 developed. The parameter distributions are not at all
)
12 intended to be bounding or worst-case distributions. The
! 13 physical parameters rey M.sent what we expect to be real 14 variability across the United States. So the variability, 15 for example, the variability in the depth to groundwater .
16 from one site to the next, the variability in precipitation 17 rates in different regions of the country, that sort of 18 thing.
19 They are based on measured values, so they're not 20 theoretical maxima or minima. They exclude physically
{
21 impossible situations, and we have accounted for some of the- -
22 correlations that we could account for. So there's no water 23 running up hill, there's no puddles of water constantly on 24 the surface of the soil and things like that.
i l 25 The distributions that we developed for these 1 i
ANN RILEY & ASSOCIATES, LTD.
Court Reporters 1250 I Street, N.W., Suite 300 Washington, D.C. 20005 l
(202) 842-0034 1
156 l- 1 physical parameters were input. We used a Monte Carlo shell 1
2 over the DandD software. So DandD itself is a deterministic 3 model, but we did a probabilistic type analysis using a i
4 shell so that we could select values out of these t
5 distributions when we developed the dose distributions for 6 the final output.
? [ Slide.]
l 8 MS. DAILY: The behavioral and metabolic l 9 parameters were. handled differently. Those parameters 10 directly represent what we're calling the average member of 11 the screening group. The rule specifies that the dose is to 12 the average member.of the critical group. For the purposes
{
t 13 of screening we've defined a screening group which is I 14 essentially.a generic representation of a critical group 15 across any of the sites'across the United States.
16 What we've done is define the screening group as 17 being composed of adult males. In the case of the building I
(
18 occupancy scenario, it's workers in light industry. I In the 19 case of the residential scenario, it is residents at that i 20 site. So people that would normally live and occupy a house 21 on a soil contaminated site. -
22 [ Slide.]
23 MSc DAILY: Once we had developed the 24 distributions that represent these parameters -- and I'd j 25 like to emphasize that the way that we developed these j
j ANN RILEY & ASSOCIATES, LTD.
Court Reporters 1250 I Street, N.W., Suite 300 Washington, D.C. 20005 (202) 842-0034
)
i 157 ,
1 (
distributions is to look at annual average values. We're 2 not looking for acute effects. This is an average over a 3 one year time period for the purposes of the rule. It's }
4 supposed to be reasonable for something that could occur for t
5 the duration of a year.
6 !'
So, for example, going back momentarily to the 7
physical parameters for something like resuspension, if you 8
have a situation where you have high flow rates and high 9 resuspension, obviously you would deplete the source in a 10 year.
So that is a situation that was excluded from the 11 construction of the parameter distribution.
12 For the behavioral parameters once we had 13 developed these distributions we then -- and we agreed that 14 these were reasonable distributions that represented the 15 average member of the screening group, we selected the mean 16 of those distributions as input into the analysis. So those !
17 '
parameters are handled as constants within the analysis.
18 And we assume that the licensee who is looking at the 19 construction of the screening group in looking at those 20 average values can use that information to determine if that 21 screening group is representative of the uses at their site. .
22 If you need to move to a site-specific critical 23 group which you may need to do when you move away from 24 screening, you can evaluate exactly how the screening group 25 was constructed and look specifically for places where it ANN RILEY & ASSOCIATES, LTD.
Court Reporters 1250 I Street, N.W., Suite 300 Washington, D.C. 20005 (202) 842-0034 l
158 1 doesn't match what would be appropriate for your site.
2 So in the case of a screening group that is the ;
3 resident farmer,.if for your site certain pathways are' 4 precluded, for physical reasons or whatever, that would be a 5 modification that would allow you to use a site-specific 6 critical group that would be different than the screening 7 group that we assumed.
8 [ Slide.]
9 MS. DAILY: So the concept of the average member 10 of the critical group is. incorporated within the 11 distributions for these parameters.
i 12 Okay. So once we have these parameters developed, 13 these distributions developed, we're doing our analysis, how 14 do we decide 1.ou to develop a default-set of parameters? At 15 the moment we don't have a default set of parameters. There .
16- are still some remaining technical issues that'are going to 17 have to be resolved by the Commission and then we can -- we 3
\
18 expect to be able to move forward and develop default 19 parameter values.
20 And the way that we would like'to be able to do {
21 that is -- when you do this analysis there's going to be -
'l 22 more than one set of parameters that meet your criterion at 23 a certain level of confidence. What we would like to do is 24 make sure that the parameter set that's selected is the 25 least extreme within the requirements for that level of l ANN RILEY & ASSOCIATES, LTD.
Court Reporters 1250 I Street, N.W., Suite 300 Washington, D.C. 20005 (202) 842-0034
I 159 i
l 1 confidence and that they're all about equally conservative l 2 with meeting that criterion.
{
3 So we are trying to stay away from a situation l
4 where some of the parameters would be set at their 95th 5 percentile and some of them would be set at the 99th s We're l 6 trying to control for that kind of situation, for example.
7 [ Slide.]'
8 MS. DAILY: Another option that we are looking at 9 as opposed to having a single set of parameters is to have i i
10 parameters that are based on a regional basis or perhaps on 11 licensee-type basis. That still has to be resolved, but 12 it's certainly something that we're looking at and I would 13 welcome any comments that you might have on that or 14 suggestions for what would be a reasonable approach for 15 default parameters.
16 [ Slide.] -
17 MS. DAILY: The technical basis documents that 18 we're going to be talking about today -- primarily we're 19 going to talk about NUREG-1549 which is the decision 20 methodology and how to perform dose assessments.
21 The scenarios and the calculations are covered in - - -
22 NUREG/CR-5512, Volume 1. Volume 2 is going to hopefully be 23 published this year and it is the user guide for the 24 software. ' Volume 2 is also going to have a listing of the l 25 changes that have been made to the models and the ANN RILEY & ASSOCIATES, LTD.
Court Reporters 1250 I Street, N.W., Suite 300 Washington, D.C. 20005
, (202) 842-0034
160 i
i 1- calculations since Volume 1 was published.
-2 And then' finally, Volume 3 contains the detailed 3 information about parameter distributions and default i
t 4' parameter selection. Since we haven't completed the default 5 parameter selection, that one is not published yet. But I 6 did include as an attachment to 1549, that's out there in l 7 the lobby, the description of all of the parameter' 8 distributions, all of the work that we've done on defining 9 exactly what those parameters mean and the results of that 10 description of parameters. That's an attachment at the back
? l 11 of 1549.
l 12 So right now it looks very, very thick. The front i 13 part is relatively thin, that's 1549. And if that's what 14 you're mostly interested in, that's fine. If you want to 15- get into the technical details, they're attached.at the 16 back. .
17 [ Slide.]
18 MS, DAILY: Just to give you an idea of what the 19 contents of 1549 are, for those of you who haven't had a 20' chance to read it yet. We start with a general introduction 21 and in that introduction is a short section that describes ...-
i 22 the assumptions used in the modeling and a short statement 23 of what he had hoped to accomplish in the interim period 24 while this is out for the coming year. And I'd like to 25 emphasize that what we would like to achieve in thic interim ANN RILEY & ASSOCIATES, LTD.
Court Reporters 1250 I Street, N.W., Suite 300 l_ , Washington, D.C. 20005 1
(202) 842-0034 l
)
I 1 161 l 1 period is some testing of the approach, more review of the
~
2 parameter distributions, how they were constructed. l 3
I can tell you that for the building occupancy 4 scenario, resuspension is a very important parameter for 5
those radionuclides that are sensitive to inhalation.u And 6 there is very little information out right now on indoor 1 7 resuspension, especially thingb such as the change in 8 resuspension over time.
9 There's lots of information about that for outdoor 10 sources, but none for indoor sources. And I think it would 11 be very worthwhile to have people do measurements at their 12 facility and especially I know some people have been doing 13 measurements but it hasn't been published. And it would be 14 very helpful to everybody, I think, to have more of that 15 kind of thing published in the open literature. !
16 (Slide.] .
i 17 MS. DAILY: After we do an overview of the 18 decisipn framework, then we've broken the document up into 19 three main sections. The first section which is Section 3 :
20 talks about the use of the framework for licensees that use 21 . generic screening. So it's a simple walkthrough for a . . -
22 simple situation. Exactly how we would assume that you 23 would use a decision methodology and the screening tables.
24 Then in Chapter 4 we talk about a situation where 25 someone needs to make minor modifications to the default l
ANN RILEY & ASSOCIATES, LTD.
Court Reporters 1250 I Street, N.W., Suite 300 Washington, D.C. 20005 (202) 842-0034
f 162 1 parameters. And then in Section 5 we go over everybody else f 2 up to site-specific modeling, developing your own conceptual l 3 model, using other models outside of 5512, and other issues.
l 4 And probably this afternoon we'll be likely to get some 5
comments on issues associated with site-specific mode, ling 6 that either are not incorporated into 1549 yet or that are 7 unclear or'that you think need to be added to 1549. )
8 [ Slide.]
9 MS. DAILY: The appendices to 1549 first in 10 Appendix A there's default concentration values equivalent 11 to 25 millirem per year. That really needs to be -- the 12 title needs to be changed. Because since we're talking 13 about screening, we're not saying that doses ~are exactly 14 equivalent to 25 millirem per year. We're saying, in those 15 situations where screening is appropriate, these values 16 could be used to demonstrate compliance with the criterion.
17 In Appendix B we have the total effective dose 18 equivalent factors, the DCFs, in terms of millirem per year 19 per picocurie per gram.
20 In Appendix C we describe the scenarios and the 21 pathways and talk a little bit about critical group. Then .
22 we describe dose modeling.
23 E has parameter descriptions and information for 24 changing parameters. The intent of that appendix is to give 25 you an idea of a shortened version of what is in the ANN RILEY & ASSOCIATES, LTD.
. Court Reporters i
1250 I Street, N.W., Suite 300 Washington, D.C. 20005 (202) 842-0034
I 163 l
1 attachment. Short descriptions of the parameters for the i
l 2 different scenarios, what information sources could be used i
3 to suggest alternatives to the default values, understanding
! 4 that default values will come later, and how those l 5 parameters are being used.
L -
6 We will be including some information from the l
- 7 sensitivity analysis in that section also. It's not there 8 at the moment. It was just completed.
9 F is a discussion of area factors. The way that 10 it is constructed right now, it is based on an earlier 11 version of the parameter analysis. That will be updated 12 before the final version of this document is released. But 13 the concept behind the development of those area factors is 14 expected to remain the same. The numbers themselves will 15 change.
16 And finally, in G at the end, there's a series of 17 examples that are similar to what Theresa is going to be 18 going over in a few minutes.
19 [ Slide.]
20 MS. DAILY: Now we'll go back to the framework, 21 the decision methodology. A lot of you have seen this - 4 22 ,
before, but I'd like to just step through it very simple 23 manner here and Theresa will go over it in more detail next.
24 The idea is that when you start this process you 25 will bring in whatever information you have. If you don't i
ANN RILEY & ASSOCIATES, LTD.
Court Reporters 1250 I Street, N.W., Suite 300 Washington, D.C. 20005 (202) 842-0034 i
i
164 1 -- under the screening assumptions you don't have a lot of 2 site-specific information so primarily what you are i
3 gathering at that point, if you're doing screening is a 4 site-specific source term.
5
! And then you step through the process of sqenario 6 definition, definition of the pathways, the conceptual )
7 models, if you're using -- if screening is appropriate for 8 you, that has been done in DandD. If it's not, that's where 9 'you would be developing your own site-specific approach.
10 At step five you can make the decision of whether 11 or not you comply with the criterion. If you do, you take 12 care of any remaining ALARA requirements if there are any 13 remaining requirements and do your paperwork requ'irements to 14 release the license.
15 If you don't demonstrate compliance at step five, , {
16 then you move into the meat of this decision analysis 17 approach. You evaluate all of the options that are 18 appropriate for your site for decommissioning, you select 1
19 among those options, do cost benefit analysis, if 20 appropriate, data worth analysis, if appropriate, it just 21 lays out a standardized methodology so that the idea behind -
22 this is that going in everybody will understand what the !
23 process is, what the requirements are, and a little bit 24 about what is going to be looked for in making these 25 decisions. Then you implement your decision. If it's l ANN RILEY & ASSOCIATES, LTD.
I Court Reporters 1250 I Street, N.W., Suite 300 Washington, D.C. 20005 (202) 842-0034
165 1 appropriate, you do your remediation, if it's appropriate, 2 you develop revised parameter values and put them back into 3 the model and cycle around again.
4 This is not expected to be an infinite process.
5 Theideaisthatthisallowsyoutodoamoreaccuratp 6 representation of the dose at your site as far as that is 7 necessary or reasonable for your site. You don't keep 8 iterating forever until you finally demonstrate compliance.
9 At some point you're going to have to actually take an ;
10 action.
11 If you've been through this cycle and you've 1 12 gotten a better representation of the dose from your site 13 and you're still exceeding the criterion, you're going to 14 have to look at things like remediation, or restricted use, 15 or some other option. But the general process here I think 16 lays out a pretty - .is laid out in a fairly straightforward 17 manner.
18 But we definitely want more comment back on this 19 kind of an approach, places where you think it's helpful, 20 where it's not helpful, places where you think additional 21 information would be useful. .
22 [ Slide.]
23 MS. DAILY: Okay. Just a summary of this 24 methodology is that it's an iterative process.
25 Decommissioning is an iterative process. It's not -- if you l AHN RILEY & ASSOCIATES, LTD.
l Court Reporters 1250 1 Street, N.W., Suite 300 Washington, D.C. 20005 (202) 842-0034
l 166 1
i look at those screening numbers, it's a pass / fail thing, 2 it's not it at all. As Cheryl has said, these are not 3
cleanup values, they are simply a starting point. For some 4 licensees they're going to be an ending point and that's 5 great.
Thentheydon'thavetospendalotofmoneyjdoing 6 unnecessary analysis. But for other people that need a l
7 better representation of their site, it's expected that you '
8 move on from there.
9 The idea behind this process is that there's a 10 structure where adding site-specific information will reduce 11 uncertainty and reduce the projected dose.
12 We are hoping that this approach will help you 13 optimize the process of decommissioning and that it will lay -
, 14 out a little better the process of interacting with i
, 15 stakeholders as appropriate and with the NRC through this' l
l 16 process so that you , don't go far down the line.and then 17 realize that you've got to do something totally different. i 18 [ Slide.) !
19 MS. DAILY: The hierarchy that we've been talking 20 about, just to reiterate a little bit, is that default 21 screening is a reasonable place for a lot of people to .-
22 start. It's not necessarily where a lot of. people will end !
l l
23 up, but it's a reasonable place to start. It's based on l 24 generic models and default parameter. I 25 When you move on to site-specific screening, still )
ANN RILEY & ASSOCIATES,~ LTD.
Court Reporters 1250'I Street, N.W., Suite 300 Washington, D.C. 20005 (202) 842-0034
167 1
as part of the iterations or a gradual approach and you 2
l don't have to do that, you can go straight to site-specific 3 modeling if that's what you want to do. But we're giving 4 you a gradual approach if you want to use it. Site-specific 5
type screening would still use the generic models, but you l
6 can modify parameters and pathways as appropriate based on 7 site-specific information. '
8 [ Slide.]
9 MS. DAILY: In terms of bringing site-specific
- 10 into the analysis, you can use a variety of information. In
'11 our examples we'll go over some of the sources that would be L 12 reasonable places to get this kind of information. You i 13 would logically want to use the most inexpensive information l
L 14 first, but you can also efficiently use more expensive 15 information as long as you know where it fits in the process
! 16 and where the most efficient place is to inject this 17 information.
i 18 At the moment we're proposing that if you use l 19 site-specific information that falls within the predefined 20 distributions for the parameters additional uncertainty l
l 21 analysis would not be necessary. -
22 [ Slide.]
23 MS. DAILY: The approach for this is discussed in 24 1549. It will be discussed in terms of the parameter 25 -distributions and selection of defaults in Volume 3. And I
)
ANN RILEY & ASSOCIATES,' LTD.
. Court Reporters 1250 I Street, N.W., Suite 300 Washington, D.C. 20005 (202) 842-0034 i
168 1 guess it would be interesting to see -- I'm interested to 2 hear your comments on this and see how it would work in 3 actual application in the field.
4 [ Slide.]
5 MS. DAILY: The idea behind site-specific modeling 6 which I'll just cover very briefly here is covered in 7 Chapter 5 of 1549. This can be anything from using some of 8 the default models and parameters with some pathways such as 9 the groundwater pathway using a more complex model or a 10 site-specific model to using completely site-specific 11 models. Alternate scenarios from what we've proposed, any 12 of the appropriate pathways and parameters for your site.
13 [ Slide.]
14 MS. DAILY: And I want to emphasize here that if i 15 the models and parameters that we have developed in 5512 are
. q 16 not appropriate for your site and you know that, don't use )
17 them. And if you have site-specific models and parameters 1
{
18 that's fine, justify those values for your particular site. l 19 And that's really a reasonable approach for any kind of 20 modeling, I think. I don't think we're saying anything that I 21 is unusual or unexpected here. ,.
22 You always are using a site-specific source term 23 for any of these approaches. And even for site-specific 24 modeling there may be some pathways that don't need a lot of 25 detail and you could continue to use the values that were ANN RILEY & ASSOCIATES, LTD.
Court Reporters 1250 I Street, N.W., Suite 300 Washington, D.C. 20005 (202) 842-0034
r 169 1 set up in 5512. '
2 Now, I want to turn it over.to Theresa and let her 3 go through some example applications. Then after she is 4 done, we'll take a break and then we'll come back and have 5- comments and questions. '
i 6 MS. BROWN: Well, I think Chris set that up fairly 7 well.' These are truly hypothetical situations. We're going 8 to go through two sites, two example sites.
L 9 [ Slide.]
10 MS. BROWN: One is just a very low contamination 11 of cobalt-60. It demonstrates the screening process. And 12 the other one is depleted uranium contaminated soils.
13 [ Slide.]
14 MS. BROWN: Now, you probably can't read the 15 framework anymore, but essentially the first example goes 16 through the first five steps through the screening process, 17 the color happens to be pink. The pink box is for the first 18 level of the screening in the decision methodology. And 19 then the second example goes to the area in blue, a very 20 simple level two or site-specific type analysis with the 21 DandD models. . 4 22 .[ Slide.]
23 MS. BROWN: In the cobalt-60 example they go 24 through the very first step which is collect existing 25 information. They happen to know through operational ANN RILEY & ASSOCIATES, LTD.
, Court Reporters 1250 I Street, N.W., Suite 300 l Washington, D.C. 20005 (202) 842-0034
170 1
surveys that they don't have any move than two picocuries 1 j
2 per gram of cobalt-60 in soil contamination.
3 So as I step through this screening process,'then 4 the second step is then they select the residential 5
scenarios with pathways which are predefined in 5512,, Volume 6 1.
7 [ Slide.]
8 MS. BROWN: Normally they would just run DandD 9 with the default parameters for that scenario.
l 10 [ Slide.]
! 11 MS. BROWN: In this case, since we have to do an 12 interim solution, they have to look up the default dose 13 conversion factor for cobalt-60 in the table in 1549 and 14 they would calculate the dose. Multiply that times the 15 concentration. In this case the licensee decides to use the , l 16 maximum concentration because it's easy to defend and they
! 17 get through screening-in the first steps it's the 13 18 millirems, a little bit less than 14 millirems.
19 [ Slide.]
[ 20 MS. BROWN: fk) if you look at this, it seems like i 21 it's a complicated process of stepping through this
~
22 framework, but you've already proceeded through the first I
23 five steps fairly quickly, however long it took them to go 24 out and do their historical site assessment and look up a i 125 number in a table. !
ANN RILEY & ASSOCIATES, LTD. 1 Court Reporters 1250 I Street, N.W., Suite 300 Washington, D.C. 20005 (202) 842-0034 l
171 1 [ Slide.]
2 MS. BROWN: In the second example we've picked a 3
location because in this case they actually go out and' 4
collect some site-specific information that's actually out 5
on the web and done a little bit of site-specific analysis 6 in terms of collecting soil samples.
7 So this one is a depleted uranium site from a 8 metal processing use.
9 [ Slide.]
10 MS. BROWN: In going through their historical site 11 assessment, they discover they have two areas to be 12 concerned about. One, they are physically separated and one 13 of them is a spill where they have fairly high 14 concentrations of depleted uranium -- ten -- 30 picocuries 15 per gram of depleted uranium, and then another spill --
16 an-ther location where they have a building that they've 17 taken down, and so it covers a much larger area than the I 18 fairly contained spill.
19 [ Slide.)
20 MS. BROWN: And that's just that situation 21 graphically. One is outside a storage facility where the - -
22 spill was and the other area is this very large area where 23 the building has been removed.
24 [ Slide.]
25 MS. BROWN: They also go to the look-up tables to ANN RILEY & ASSOCIATES, LTD.
Court Reporters 1250 I Street, N.W., Suite 300 Washington, D.C. 20005 (202) 842-0034 l
172 l 1 look up the dose conversion factors for U-238, U-234, U-25 1
2 and multiply that times their concentrations, put that in a l 3 little spreadsheet on the next slide.
4 [ Slide.)
, 5 MS. BROWN: And look up their dose. In this case 6 both of the areas exceed the screening criteria. One is 188 7 millirems per year and the other is 60 in this screening 8 analysis. So they decide to look at what their options are 9 in terms of collecting site-specific information, 10 remediating, looking at the costs of remediating, and what 11 their combinations of options are, or partial remediation in 12 terms of cleaning up one of the sites, but possible leaving 13 the other site to. gather site-specific information and 14 reduce the dose.
15 [ Slide.]
16 MS. BROWN:. When they start looking at what these 17 options are for collecting site-specific information, 18 they've looked at the sensitivity analysis results in Volume 19 3 and they see that there are three primary sets of 20 parameters that influence the dose more than any other. The 21 absorption coefficient, the Kd for uranium, the wet-to-dry .
22 conversion factor, and just the soil type which is 23 correlated to a lot of the parameter values in the model.
24 They also look at data collection costs so that 25 they can compare the cost and the value of collecting this
( ANN RILEY & ASSOCIATES, LTD.
Court Reporters ,
1250 I Street, N.W., Suite 300 1 Washington, D.C. 20005 I (202) 842-0034 )
I
- 173 l
{
1- information. /nd they know that to collect a soil i
.2 infctmation - 6 soil-type information, it's just a matter 3 of doing a litm;e una search and it will take a few hours to l 4 do that.
l And ths1 .0 go cut and get the Kd data that they i 5
need, they need to go and get uncontaminated soil samples of 6 the same type, simt.lar chemistries, and do a variety of 7 samples for that.
And so we've estimated the costs based o'n
'8 some of the data collection costs in the GEIS, again, just
- 9 for the example.
10 [ Slide.)
11 MS. BROWN: The one reason that they know that the j 12 Kds are one of the sensitive parameters is, again, based on 13 some of the information that_will be provided in Volume 3 of 14 5512. And we planned to present this information in the
! 15 forms of graphs or tables. This is the distribution of .
16 total effective dose, equivalent for your U-238 and the other.
17 uranium isotopes'look similar to this. And you can see that l 18 as you get up around the 90th percentile it's the tai] of .
l' 19 that distribution that really shows the wide range of dose l
20 so that you anticipate being able to reduce the dose if you 21 collect.some information on the sensitive parameters. - !
l 22, [ Slide.] !
!- 23 MS. BROWN: And in this case, and I'm sorry you i-24 can't see'this very well, this is the CDF, the cumulative !
25 probability distribution of the sampled Kd values. And if l ANN RILEY & ASSOCIATES, LTD.
. Court Reporters 1250 I Street, N.W., Suite 300 Washington, D.C. 20005 (202) 842-0034
i-174 1 you can see it -- let's see -- no, you can't, but the value 2 for the 95th percentile in this case what was used to do 3 this screening calculation is at the very low end of this 4 range, it's at the tail of the range, and the median value 5 in this distribution is up around 100 milliliters peq gram.
6 So then the licensee would then know that there's a good 7 opportunity or there's a high probability of going out and 8 collecting site-specific data that would move them closer to 9 this hundred towards the median, and so they feel like -- or 10 you can get the idea that there would be a very'large change 11 in the simulated dose.
12 [ Slide.]
13 MS. BROWN: Another one of their options is to do 14 a remediation. And just a quick way to see if that's even a 15 viable option'at this point, even though you wouldn't .
16 necessarily want to use these screening calculations to set 17 cleanup levels, it's just to look at how much it would cost 18 to clean that up right now.
l 19 For area, it's a small area and a small spill.
20 Those costs are actually fairly reasonable. They are a-
- 21 limited range, and again, this range of costs for disposal -
22 is from the GEIS and in this case the range in values is due 23 to different disposal facilities.
24 [ Slide.]
l 25 MS. BROWN: And then-for area B you see that that l
ANN RILEY & ASSOCIATES, LTD.
Court Reporters 1250 I Street, N.W., Suite 300 Washington, D.C. 20005 (202) 842-0034
175 1- cost gets up close to a million and much higher. That range
~ '
2 is much higher for the large area of contamination. So the 3 licensee goes through this analysis and there.isn't any 4 fixed way that they have to come to their decision, so 5' they've decided, they'll go out and they'll collect this 6 literature information on the soil type because it's fairly
.7 low cost even though they don't think it will have much of 8 an impact on the simulated dose using the DandD models.
9 They will also go_out and collect some Kd values 10 because they anticipate a fairly large reduction in -
11 simulated dose and it's a reasonable cost again. And then 12 for Area A they decide just to clean that-up because the 13 remedial costs are fairly small and it's expedient.
14 . -
[ Slide.]
15 MS. BROWN: This is just a simulated table. They ,
16 went out,.they collected 20 samples for the Kd; They ran 17 the analysis, they got a range in values over an order of 18 magnitude or so, and the mean value is close to what the 19 mean.value of that PDF was just by coincidence. And they 20 decide to use the smallest value of Kd just because it's 21- easy.to defend and when they run the analysis with DandD -- - *-
22 [ Slide.]
23 MS. BROWN: -- it turns out that'that is 24' ' sufficient to' reduce the dose.
~
So here is --- they would i 25' update all their parameter values'for the site-specific i
-ANN RILEY-&LASSOCIATES, LTD. !
. Court Reporters.
1250 I Street, N.W., Suite 300 Washington, D.C. 20005 -1 (202) 842-0034-1
176 ;
1 1 parameter values including the values they got from the l 2 literature and the minimum Kd value they measured and the 3 simulate value goes down to about 10.2 millirems per year.
4 And then they can come to the conclusion that Site B or Area 5 B where you had the building can be left for unrestricted 6 use -- can be released for unrestricted use just based on 7 , collecting one site-specific set of data.
8 [ Slide.] 1 l
9 MS. BROWN: In Area A they just remediated. This 10 puts us back to our decision methodology and obviously these l 11 are very elegant examples, they're simplified just for this 12 presentation. But even though the process of going through 13 this framework looks very detailed, and this is not 14 necessarily the most elegant way to present going through 15 the decision process for a very complicated site, you get 16 the flavor of how you would optimize your decision. What we 17 hope to do eventually is automate the sensitivity analysis 18 so that it doesn't take as much work to go through these 19 tables and look and see what the sensitive parameters are 20 and see how much it might change the dose. We would like to 21 automate that so that you see that sensitivity analysis by ,,
22 just sliding your cursor over and saying, if I can reduce 23 this parameter value or increase this parameter value by 24 this much, what would be the net effect on dose and use the 25 simulations we've already done to help provide that I
ANN RILEY & ASSOCIATES,' LTD.
Court Reporters 1250'I Street, N.W., Suite 300 Washington, D.C. 20005 (202) 842-0034 l
4 177 L .1 information graphically.
l 2
l-L But at this point you can dofit just based-on good 3 old fashioned looking it up in the figures.
4 c
And that's the :end of my examples. Did you want.
l 5 to' continue?
c u
.6 [ Slide.]
7- MS. DAILY:
! Okay. -Just to.let you know what we're 8- planning on doing for the rest of the' morning and'for the 9 rest of day, we've gone over some examples. We've talked i
u
.10 about some of the existing information. When we come back 11 from our break we'll start talking.about'the screening 12 approach, first, I think, and then we'll move on
'13 site-specific issues. So if you want to' break now. If 14-somebody.wants to bring the lights up back there/and be back
- 15 in like 15 minutes.
4 i 16 [ Recess.) ,
17 MS. DAILY: Okay. If we can get sta';ted again, 18 please, do you want to take your seats.
19 Hello, if everybody wants'to take their seats i 20 we'll go ahead and get started now.
21' ;
For those of you who were interested in getting 22
.:opies of the. presentation, we're going.to be making copies- !
23-
.over lunch and we should have them when we restart'this 24- afternoon. If you have to leave before then and'you want a. )
25 copy, just leave your card with us and.we'll get a copy off ANN RILEY & ASSOCIATES,fLTD.
. . Court Reporters
- l. 1250^I~ Street, N.W., Suite 300 l Washington, D.C. 20005-
'(202) 842-0034
l 178 1 to you.
2 What I would like to do now is start opening this
- 3. up for comments and questions and generally follow the 4
' outline of 154'9 in terms of structuring how we step through 5 this. But before we start that formal' process, if there are 6
some general comments or. general questions that people want 7 to cover -- or'I know some people were goin'g to have to l 8 leave a little'early, so if'they want to bring up some l
'9 issues right now, we may not be able to get a full response
! 10- .to you right now,.but for anybody that needs!to get 11 something done this morning, if you want to come up now.'
12' MR. ANDERSEN: Ralph Andersen with the Nuclear.
l 13 Energy Institute.
j I want to say first that that was'a very 14 helpful presentation this morning. I appreciate your all
,, 15 taking. time to do that. I think I have a better i
16 appreciation'of wher,e you are and where you're,trying to go 17: with the model. ;
I 18 We're utilizing a task force of about 15 companies l j
19 that are licensed to operate nuclear power plants to review i I . .
20 all of this guidance as it's been coming forth. >
21 Obviously due to the time constraints we have not
~
22 Lhad the opportunity to make an extensive review of 1549. We
- 23 i are beginning'to work our way through that. But on that i
124 . task force'there.are:about eight companies thattare i 25, ' operating plants: currently and there are seven who'either .i j
t ANN RILEY'& ASSOCIATES, LTD.
Court Reporters 1250'I Street, N . W . ., Suite 300 l Washington, D.C. 20005 (202) 842-0034
I 179 1 have plants that are shut down or may be confronting a shut 2 .down plant in the near future. So they are actually in the 3 space of being very near-term users of this interim 4 guidance. And I wanted to make a general' comment or two in l 5 that regard.
6 The plan is, of course, to hav's this interim 7 guidance issued in the near future for some period of time '
8 for use and comment. Looking at our task force, for 9 example, and I think this will be representative of your 10 broader population of licensees, you'll have many people who 11 will be reviewing it from a hypothetical point of view, even 12 though they may be using their own site-specific data and
- 13 then you will have others that will be using it in the real 14 world in real time. So at the same time they are evaluating 15 the methodology they are, in fact, interacting with the NRC .
16 in a licensing framework to. decommission their-facility and 17 to terminate their license.
18 The concern is not having a clear understanding of 19 where you all are with this particular model and methodology 20 with regards to either benchmarking models or BNB codes, or 21 what level'of peer review has occurred. It's very much an . -
22 at-risk venture for those people who are actually in the 23 decommissioning and licensing termination process. And I i
24 would suggest to you -- even though I certainly don't expect 25 an answer from you here today -- that as part of your ANN RILEY & ASSOCIATES, LTD.
Court Reporters 1250 I Street, N.W., Suite 300 Washington, D.C. 20005 (202) 842-0034 L
180 1 process of going to the Commission with the interim guidance 2 that you point out to them that somehow they need to j 3 establish policy on how much at-risk licensees may be, 4 particularly with this portion of the guidance,.because I .
5 think the other portions are more familiar to us just., merely 6 because they've been available for a longer review period.
7 l
i And I think it would be unfortunate to' pick up on 8 opening comments that you have in 1549 itself. You 9 recommend in that that certainly licensees should have NRC 10' review their DCGLs before they go forth making resource 11 investments and so forth. Well, use of this methodology 12 kind of poses that situation. If the licensee derives DCGLs
- 13 from this methodology and pr'oposes them to the NRC,'really 14 their basis for the DCGL itself is, well, that's the method i
15 you handed me. You don't really have sufficient info'rmation l
16 to justify this model back to you.
l 17 'And, therefore, if we go forth what that, commic 18 . resources, make plans, make submittals on the' docket, then a 19 year from now you make some degree of adjustments to the 20 model and the parameters and the underlying assumptions.
21 That could create some very awkward situations and possibly - ^
22 very costly difficult situations for licensees who have 23 proceeded down that path. So I would just raise that with 24 you as a policy issue to be addressed somehow concurrent
! -25 'with the issuances of the interim guidance, and I believe l i
I ANN RILEY & ASSOCIATES, LTD.
Court Reporters
-1250 I Street, N.W., Suite 300 2 84 I b3
y.
X 181 1
' il -that needsLto be~ addressed. ~(
L l '2 . .MS. DAILY: Thank you.
!; ~3 MR. ANDERSEN: Additionally,-I gue'ss'I'd like to 4 make one mment.about the screening model because I am one 5 of those people;that will need to leave soon.-
W.
6 It regards .specifically the AIARA step within the L 17 'decisi n. process..'Two comments. One-is, if you use the
['
o 8 screening mode?. and you're successful in that regard, de n 9 facto, have you also satisfied the ALARA criteria? ~And the l- .
10 reason I ask that is because how else would you derive a-
'11 ~DCGL to usie iv the ALARA process if not the'one that you 12 usedLfor the screening method.
'13' It wouldn't seem to me to be appropriate that
\;
.i 14- first you would screen yourself as needing no further action-15 and then you would go back in and use what would nominally 16 be the most restrictive DCGL'to do an ALARA analysis. And~
17 at the same' time it isn't clear to me it would warrant the
~
18 expense of going in and deriving.a. site-specific DOGL to do' 1
19 .that. . So I'd just leave that-with you as a. comment-more--
~
3 20- than anything else. '
21? Secondly,.as far as where.it exists in the process. -. -
22 ,for people how do not successfully pass-the screening model, 23 I1think you need -- if-this isnin'the guidance _already, I i
24- apologize, I haven't got there from here.yet -- but I-think !
- 25 you need to talk somehow to the issue of the appropriate
(, ANN RILEY'& ASSOCIATES,LTD. ,
l- , Court Reporters .
. 1250 I Street, N.W.,, Suite 300 Washington, D.C. 20005
~
1 l
(202). 842-0034
182 1
selection of DCGLs for use in other parts of the guidance.
2 l
I think I'm clear on what your decisions are you
- 3 need to make from a policy view on screening, but this~also 4 becomes the source of input to both MARSSIM and'to the ALARA 5 analysis.
And I'm most intrigued at this point by what you 6 would use for the ALARA analysis? Would it be a best 7 estimatc, a 95 percent-estimate? So, again, I'd just leave 8 that as a -- is that already addressed in the guidance? I 9 guess that's an easy question?
10 MS. DAILY: .It's partly addressed.
11 MR. ANDERSEN: Okay.
12 MS. DAILY: It's expected that the ALARA analysis-13 is embedded in the process of evaluating options and 14 determining the reasonable approach. So that from our 15 . diagram in steps 8, 9, and 10, that would be where you would 16 be performing your ALARA analysis. And the ALARA analysis 17 in that case would be the same as the dose assessment that 18 .you're doing would be a more site-specific analysis which is' -
19 what you're doing with your dose assessment at that point.
.20 MR. ANDERSEN:
. Okay.
21 MS. DAILY: The other ALARA that we talked about -
22- in the examples, when you are just doing screening, in that 23~ _ case, that's basicallyfa good housekeeping type as you're 24- closing out your site. It's not a formal calculational' 25 ALARA analysis.
ANN RILEY & ASSOCIATES, LTD.
Court Reporters 1250 I Street, N.W., Suite 300 Washington, D.C. 20005 (202) 842-0034 l
1
i l
183 1 MR. ANDERSEN: Okay. Well, thank you. And, 2 again, I appreciate the presentation.
3 MR. YIM: Man Sung Yim, North Carolina State 4 University. I think this is a new approach we discussed 5 this morning, is really a step forward and tremendous 6 progress in implementing some of the benefits of the recent 7 progress in performance assessment approach, the phased 8
approach using uncertainty issues in the decisionmaking 9 process, these are, I think, have tremendous benefits for 10 industry for the future as long as we can take care of the 11 initial confusion and so on.
12 This probablistic approach has been also proposed 13 by NRC branch technical division on low-level waste 14 assessment and even though that has not been published, it 15 is probably going to be implemented and now we have this 16 probabilistic approach even though we have deterministic and 17 probablistic combinations in this guideline there will be 18 major working holes for the future of the industry.
19 And my feeling is this is really a paradigm shift 20 even though this has been practices in high-level waste
- 21 performance assessment, not in low-level waste or . - -
22 decommissioning.
23 And as we know -- I'm sure you have discussed some L
24 of those issues in your appendix, but when.you use a 25 probabilistic approach your reserves are really driven by ANN RILEY & ASSOCIATES, LTD.
Court Reporters 1250 I Street, N.W., Suite 300 l Washington, D.C. 20005 l_ (202) 842-0034
i 184 1 your probability distributions. And you can play with your 2 probability distributions and pretty much determine your 3 results. And it seems to me at this point we don't have any 4 guidelines and really in terms of if you want to do your own 5 probabilistic data evaluation and do a site-specific ,
6 analysis. I think there has to be some guidelines --
7 supporting guidelines implementing this approach to benefit 8 -- to come up with the maxin.um benefit.
9 With the use of this approach and also at the same l 10 time you -- assuming that your NUREG-5512 model I understand 11 that your model is different from the original, but you're 12 assuming that model is the most conservative. And if you 13 want to change that model, that model changed as always to 14 produce the lower dose so then the question is, do we really 15 feel comfortable about the conservatism about your models? .
16 It seems to me at this point you are not addressing the i
17 uncertainty of your models.
18 For example, you have discussed an example of
! 19 using Kd approach for uranium, your example case.
! And as we l
20 know, for uranium there is an argument -- a controversy over 21 using Kd approaches, transport of uranium is one order of -
l 22 surface transport. It's not really Kd controlled, it's more 23 solubility, or maybe even colature transport. So if we 24 start talking about those model issues, I think you want to 25 make sure that the model that are you are going to suggest ANN RILEY & ASSOCIATES, LTD.
Court Reporters
' 1250 I Street, N.W., Suite 300 Washington, D.C. 20005 (202) 842-0034
185 1
is really the -- what you-think reasonably, but still very 2 conservative approach that people do not have any confusion.
3 Of if people want to come up with some different 4
approaches, maybe you should make sure what kind of degree 5 of conservativism you built into this 5512 model so that 6
they know what is the initial boundary conditions they are ,
7 playing with. I 8
But at the same time, I think this approach will 9
have a -- as I said earlier -- tremendous benefit. So I 10 think we need more guidance and detailed information from 11 you so that people or industry can be comfortable. i j
12 MS. BROWN: In terms of the conservatism of the 13 models, you are correct, that has not been fully tested yet.
14 There has been some testing, and obviously a lot of thought 15 put into what is a -- what would be a prudent like 16 conservative modeling assumption or model for doing this 17 screening analysis. Part of that, I believe still is in a 18 draft state and still needs to be published and should ;
19 probably be included as part of 1549 or Volume 3. I'm not 20 sure yet. But that guidance is forthcoming and there should 21 be some additional testing in terms of the conservatism of - -
22 the models and some guidance from moving to more complex or 23 less conservative models. That's in the plan, it just will i
L 24 be done during this interim stage, i
l 25 MR. ROBERTS: I'm Rick Roberts from the Rocky i
ANN RILEY & ASSOCIATES, LTD.
Court Reporters t
1250 I Street, N.W., Suite 300 l Nashington, D.C. 20005 (202) 842-0034
I l
186 1 Flats Environmental Technology Site. In NUREG-5512 there is 2 another exposure scenario. There's the renovation worker 3
exposure scenario which is not covered in your methodology.
4 1
And will there be instances where the renovation worker 5
cleanup levels in picocuries per gram will need to be,used 6 at a site that would be over and above the requirements in 7 the methodology? '
8 MS. DAILY: The original intent of that renovation 9
scenario was to handle volume sources inside structures and 10 primarily to look at limiting short-term acute effects like 11 when somebody removes a wall from a facility and the wall 12 was contaminated all the way through, would that result in a 13 situation that was more restrictive than we would have 14 controlled with the long-term, chronic, surface source 15 approach.
16 We had to only look at two scenarios in this 17 current analysis because of source requirements. In our 18 earlier analysis the renovation scenario did not dominate, 19 but we hope to go back and pick up that scenario and find 20 out for sure that there aren't any situations where that 21 would be more restrictive than the building occupancy -
22 scenario. It's just up till now we haven't seen that it was 23 more restrictive.
24 MR. ROBERTS: Is there any way then within your 25 methodology you could state the criteria that would actually
(
I ANN RILEY & ASSOCIATES, LTD.
Court Reporters 1250 I Street, N.W., Suite 300 Washington, D.C. 20005 I
(202) 842-0034 l
187 1 be -- you would need to have in order to have to calculate 2 the cleanup levels based on the renovation worker so that 3
people will know that it actually does not apply to their 4 site?
5 MS. DAILY: We'll have to -- I would take t_ hat as 6 something that needs to be added into the guidance.
p 7 MR. ROBERTS: Thanks.
8 MR. LITTLEFIELD: This is Pete Littlefield, Duke 9 Engineering. You said this morning that the default 10 parameters were still being worked on. So, I guess, what is 11 in the version of the DandD program that we have right now?
12 Are those default parameters that are in there going"to be 13 changed?
14 MS. DAILY: Yeah, the version of DandD that's out 15 right now, we've been calling an interim release.
16 MR. LITTLEFIELD: Right.
17 MS. DAILY: Specifically because the defaults that 18 are in there are basically place holders. They are not 19 considered something that woald be sufficient to use for 20 demonstrating compliance with the regulation. We will be 21 replacing those parameters as soon as we determine what -
22 level of conservatism is going to be used for screening, 23 then we can go back in and do the calculations to replace 24 those values.
25 MR. LITTLEFIELD: And, likewise, the numbers that ANN RILEY & ASSOCIATES, LTD.
Court Reporters 1250 I Street, N.W., Suite 300 Washington, D.C. 20005 (202) 842-0034 i
f l
188 1 been published in the draft NUREG-1549 those will also be 2 changing as you change the default parameters?
3 MS. DAILY: They will change based on what the l
4 i
Commission decides in terms of level of conservatism since 5
those are default concentration values from the dose ,
6 distributions. So they will not be impacted by a change to 7 the default parameters. It's not quite the same thing. The j 8
default parameters would allow you to calculate -- well, it 9 depends. If we come up with a single set of default
! 10 }
parameters, then we -- let me start again. l 11 The tables that are in 1549 right now are based on 12 the full distributions on a radionuclide-specific basis.
13 And they give a range of values for different levels of 14 confidence. Obviously when we settle on one level we'll 15 have default concentrations. The other things that we're 16 looking at that would make those tables a little more 17 complicated is if we have regional versions. So we would 18 have multiple tables. Or if we have other types of 19 groupings, those tables would look different in the final 20 release. Does that --
21 MR. LITTLEFIELD: When I think of default - -
22 parameters, I think of modifying things like occupancy times 23 for homes and things, and those would all impact on those 24 values that are published.
j 25 MS. BROWN: The PDFs that were used to generate i
I i l ANN RILEY & ASSOCIATES, LTD.
Court Reporters 1250 I Street, N.W., Suite 300 Washington, D.C. 20005 l (202) 842-0034 !
189 1 the dose distribution, the output, for this analysis unless 2 those change as part of this review process, we haven't 3 represented one of the parameters appropriately. We haven't 4 set a probability distribution that really represents the 5 uncertainty in that particular parameter bank. If thpse are i
6 left in tact, then the tables for what the screening value 7 wbuld be, either for the dose conversion factor, or for the 1 8 concentration, as soon as they pick a confidence interval 9 that would be set and that would not change. 1 10 Then we would go through this process of doing the 11 parameter analysis to aslect the default parameters.
f 12 Selecting a default parameter does not change the values in 13 the tables because the tables are based on that confidence 14 ' int'erval .
15 MR. LITTLEFIELD: Would you have any suggestions .
16 for somebody that needs to calculate their dose factors at 17 the present time? That is, they're in decommissioning right 18 now and should we be using DandD in its present version.
19 Should we be using the tables that are in the draft NUREG, 20 should be using RESRAD? Any suggestions along those?
21 MS. TROTTIER: I would suggest that in the interim- -
l 22 you contact the licensing staff, either the region or 23 headquarters whoever you have been working with. I wouldn't
! 24 guess at it. You know, we're going to try to get this 25 guidance out in order for it to be usable by the required i
ANN RILEY & ASSOCIATES, LTD.
Court Reporters 1250 I Street, N.W., Suite 300 {
Washington, D.C. 20005 (202) 842-0034 -
I l !
l l1 190 '
1 implementation date. But I realize, you know, it doesn't i
2 leave very many months. So for some people it may not be 3 enough time. So I think to get the right answer the best 4 thing to do is to talk to whoever is the appropriate one for l
l 5 your site and not take a chance.
i j 6 ME LITTLEFIELD: Okay. Thanks.
7 MS. DAILY: Okay. If thee aren't any other -- oh, 8 sorry, go ahead.
9 MR. GOGOLAK: Hi, Chris.
10 MS. DAILY- Hi.
11 MR. GOGOLAK: Carl Gogolak, Environmental l 12 Measurements Lab,. DOE. I'd just like to make a couple of 13 general comments and one of the -- I agree very much with 14 the first comment that was said in terms of this being a 15 paradigm shift, and being what I might phrase "a graded !
16 approach to modeling" which is very much similar to the l 17 kinds of graded approach that the MARSSIM took in terms of l
18 effort of survey level. !
t 19 l l
I And that I think is to be applauded, is it a :
]
l 20 paradigm shift. The problem that I see, I think, is that in i 21' its current form the 1549 really kind of -- it has the - -
- 22 appearance of being very complex. And one of the things I 23 that we learned in going through the development of the l 24 MARSSIM process is the simpler and more understandable we 25 could make things, the better off we were. And in some ways ANN RILEY & ASSOCIATES, LTD.
Court Reporters 1250 I Street, N.W., Suite 300 Washington, D.C. 20005 (202) 842-0034 L
i I
l i
191 I i
1 l
I found the' initial discussion of the three-step process in 2
the original NUREG-1500 to be somewhat more straightforward 3 than the process that's described here.
4 It's called a " decision method for dose 5
assessment" and I'm not really sure at a certain point l
6 exactly what decision is being made. In the example that 7 was given for cobalt, an example was given that there was a l 8 two picocurie per gram maximum found and this was the source l 9 term in the model. But it doesn't given any indication as 10 to this is the maximum of how many measurements, how was )
1 11 that determined, what is the plus and minus on that maximum, 12 and these are the kinds of things that coming off the survey 13 side, you know, I feel are important at the interface 14 between the modeling and the survey.
15 Th6 cante to the ratio of the uranium. There was a 16 statement made of what the survey had done, and this was the 17 mean, but there was no plus or minus on thdt mean and no I l
18 consideration again of how that would interface with the 19 modeling.
20 .
So, I guess, in summary, I think that the graded 21 approach is really, really good. And I think that that . .
22 approach was I think originally expressed in the NUREG-1500 ;
23 very well. But I would like to see a more straightforward 24 kind of implementation of this that would make it possible !
25 for people who are maybe not experts in dose mode?.ing to be I
ANN RILEY & ASSOCIATES, LTD.
s L .
d5Uit Reporters * ' "?
1250 I Street, N.W., Suite 30b Washington, D.C. 20005 (202) 842-0034 l
L l 192 1
l able to make these kinds of decisions and also have those 2 decisions accepted by people who may be in a position to 3 review these things, but they themselves are also not 4 experts in dose modeling.
5 Now, in some cases where, for example, we see that 6 there is a parameter like Kd that's driving things. Then I 7 wonder if it isn't almost possible to derive, you know, 8 default maps of areas in the country or soil types or 9 something that would not require somebody to do a million \
l 10 dollars worth of soil sampling and analysis to move these 11 parameters because we know it's so important. And that 12 would make it, I think, a lot easier for people to move off j 13 of the screening level which, you know, will be sufficient 14 for a larger number of people who don't have very much 15 contamination and that's okay. But I think we have to make i 16 it a lot easier for people to move off of that default. And l
17 right now it still looks like an awfully intimidating 18 process to get off of that -- those default values.
i 19 MS, DAILY: I have to say that for the most part I 20 agree with your comments. We've been working on getting l 21 1549 to be a clearer representation of what we're trying to . .-
22 do and we're still working on that. So, you're right, it 23 needs work to be easier to use.
24 We are continuing to work on the parameters in 25 terms of developing more support for licensees who are INN RILEY & ASSOCIATES, LTD.
Court Reporters 1250 I Street, N.W., Suite 300 Washington, D.C. 20005 (202) 842-0034
193 1 developing revised values. And that is something that is i
2 going to take some time.
3 The information that's in Appendix E right now is .
4 the first approach to that, but it is still too complicated 5 for licensees and things like Kds, we need more information u
6 that we can give to licensees.
7 MR. GOGOLAK: I would also suggest that perhaps 8 you would look at broad scenarios too. I mean, some of the 9 uses of. buildings would lead to some pathways being just 10 ridiculous to even consider. I mean, there are some sites 11 where even 25 percent consumption resident farmer are not 12 ever going to be a likely scenario.
13 MS. DAILY: And that's incorporated in 1549. It 14 may not be stated as clearly as it needs to be, but 15 definitely -- what we are trying to is make it clear enough 16 when you start, exactly what the assumptions were that you 17 then have the information you need to step away to say that 18 you made this assumption, it's not appropriate for my site 19 for.this reason, and therefore, that pathway is not l 20 appropriate, or that pathway can be modified based on this 21 additional information. That's the intent. ,
22 MR. GOGOLAK: Good, I hope you achieve it.
.23 MR. McDANIEL: Corey McDaniel with EOP Group. The l 24 comment I had was kind of along the same lines from the 25 discussion on the critical group in the appendix, page C-4.
ANN RILEY & ASSOCIATES, LTD.
Court Reporters 1250 I Street, N.W., Suite.300 L Washington, D.C. 20005 l (202) 842-0034
194 1 There's a discussion of adequate justification where i 2
examples are provided of particular situations where you may l
3 not have to use the resident farmer scenario or, you know, 4 as stated as the default. The examples given are an urban 5
industrial -- an urban situation, an industrial situation, 6 or a particularly rocky environment. And then after that, 7 you know, kind of towards the end i't says that these are --
1 8
these could be appropriate for the situations noted here 9
based on their characteristics which kind of indicates, you 10 i know, here, these are three examples that we're putting l 11 l forth that you can use, and these are the three examples. l 12 And my recommendation would be twofold. One, that 13 you may want to state that these aren't the only examples, l 14 which has been my indication from speaking with Chris, and l 15 with Dave Fauver that, you know, this is meant to apply to a 16 number of different scenarios. And also, any further 17 discussion on again the intention. I know the intention of 18 1549 is to provide all this flexibility, but sometimes if 19 you just read it, like in this case, it looks like you're 20 saying that these are three scenarios where you don't have :
21 to use the resident farmer and maybe other sites -- other -
22 situations don't apply.
23 For example, I know of at least two sites that are 24 wetland sites where it would seem to make sense to say that l 25 if your site is a wetlands, first of all it's not going to l
- ANN RILEY & ASSOCIATES, LTD.
Court Reporters 1250 I Street, N.W., Suite 300 Washington, D.C. 20005 l
(202) 842-0034 l
t
1 195 1 make too much sense to say somebody is going to be living on 2 the site. And also it may be difficult -- or from an 3 institutional control standpoint if the site is a designated 4 wetlands and you've got a governmental, you know, 5
institutional control, per se, in place saying that y,,ou 6 can't live in it, you know, is that going to be a -- could r
7 l that site then be released for unrestricted release because \
j 8 you have, you know, both the person can't live on the site 9 and the law says that the person can't live on the site.
10 And I k now we discussed that briefly, you said it's a gray l
11 area. I think that's going to be one of the areas that for 12 a lot of people is going to be important to be clarified. I 13 So for this particular section I would like to 14 see, maybe either more examples included or a discussion 15 that there's a broad number of examples and it's going to be l 16 up to you to prove your case in each situation.
17 And if you have a site that has a bunch of large 18 rocks on it you cay say automatically that we've got rocks 19 and nobody can live on our site. I'd just like to see a 20 little bit more clarification on that. And possibly the 21 inclusion of at least two other examples, a site -- a desert. -
22 site that is like, you know, Death Valley, or something like '
23 that, you're not going to have resident farmers, and also a I
l 24 wetlands site where the person can't live on the site, you i
l 25 know, full time. A little more discussion on that I think ANN RILEY & ASSOCIATES, LTD.
Court Reporters 1250 I Street, N.W., Suite 300 l Washington, D.C. 20005 (202) 842-0034
r-1 196 1- would be helpful.
2 MS. DAILY: Okay. I'm going to put that under the 3
heading of more information needs to be included in 1549, 4- some clarification.
5 MR. McDANIEL: Okay. And that is something that 6 will go to the -- when the report goes to the Commission, 7 comments from this meeting will be included, or will there 8 be any more changes made to 1549 before it goes to the l
9 Commission that maybe this could be included?
10 MS. TROTTIER: I notice how she keeps looking over
! 11 here. We will attempt to incorporate as many of the 12 comments as we can. As you might guess, timing is of the 13 essence since I have basically one day left. But there are 14 some like that -- there are some changes that need to be l
15 made to 1549, minor things. But there could be some minor 16 changes made before At goes to the Commission. Otherwise we 17 will tell the Commission what comments came from the meeting 18 and that we will make these changes. We will either make 19 the changes before the goes out, or we'll make the changes
~20 during the interim review period. Well tell the Commission, 21 in any case, what we plan to do. . .-
22 MR. McDANIEL: Okay. Thank you. ,
23 MR..MANN: Bruce Mann, Commonwealth Edison. A 1
24 general comment on the guidance materials. Yesterday we i t
L 25 went through this draft regulatory guide -- I'm sorry, draft I ANN RILEY_& ASSOCIATES, LTD.
Court Reporters 1250 I Street, N.W., Suite 300 Washington, D.C. 20005
.(202) 842-0034 !
l l )
197 1 -- yeah, draft regulatory guide, it's called DG-4006.
2 In looking through it with respect to the topic {
3 for today's meeting the discussion is very limited and I l
4 would like to see sort of a beefing up in this document. If 1
5 this is the guide -- the guidance versus the NUREG-15,49, I'm l 6 a little confused on how this is going to be handled in 7 terms of a customer for this. But the philosophy that was 8 discussed at length yesterday, the role of this document is 9 not consistent in my mind with respect to the dose modeling 10 portion as regards the -- it's a rather lengthy discussion 11 in this document in comparison as far as survey 12 implementation. It would be nice to have sort of "one-stop 13 shopping". I think at this level the recommendation I would 14 make would beef up that discussion in this draft guide.
15 Also I'm a little confused -- again, I apologize 16 because I haven't waded through all the details of 1549, I'm c ,17 a little confused in the overall decision framework that's 18 needed to support compliance with the new Subpart E rule.
19 We have attended workshops with respect to ALARA, and 20 obviously attended workshops with respect to surveys and now 21 we're discussing dose modeling. And in all these meetings -
22 the terms, " decision model", " decision methodology" is used 23 and I'm a little confused in regard to this. I would like 24 to see -- I'm concerned that it's not an integrated approach 25 to decision modeling, whatever you want to call it in these ANN RILEY & ASSOCIATES, LTD.
Court Reporters 1250 I Street, N.W., Suite 300 Washington, D.C. 20005 (202) 842-0034 l
198 1 different aspects that you need that a licensee would 2 conceivably need to demonstrate compliance.
3 The ALARA part had a decision methodology and I 4 would just like to see more explanation and integration 5
because right now I'm concerned that we have three d{fferent 6 portions of this guidance that deal with this broad topic 7 and they're not coherent. And I'm concerned that we end up 8 traveling down more or less three different paths to
- 9 demonstrate compliance and to implement surveys and make 10 decisions about when remediation -- additional remediation 11 is required to achieve different ends.
12 I have a question about what -- is it too early 13 for questions on this draft guide? Or do you want to defer 14 those until later?
15 MS. DAILY: Why don't we let Cheryl respond to the 4 16 first part of your comment and then we'll get to your 17 question. -
18 MR. MANN: Okay.
{
l 19 MS. TROTTIER: Let me tell you a little bit about 20 our philosophy of why we prepared the guide the way we did.
21 As you know, regulatory guides represent staff positions. -.
22 Unfortunately, the negative of that is regulatory guides are 23 not cookbooks. It creates a less-than-perfect situation.
24 But it is just the Agency policy that that's what regulatory 25 guides are for, to represent staff positions.
l l
ANN RILEY & ASSOCIATES', LTD.
Court Reporters 1250 I Street, N.W., Suite 300 Washington, D.C. 20005 (202) 842-0034
! l
l l
i 199 1 MARSSIM is not an NRC document.
i MARSSIM is a 2 document that was jointly developed by many agencies. NRC ;
3 .cannot change MARSSIM. If NRC wants something done a little 4 differently, the regulatory guide is the mechanism for 5 indicating that.
6 That's not true with 1549. It's a staff-generated j 7 document. We can make that document say whatever we feel it 8 needs to say.
9 I did not want -- in fact, the regulatory guide 10 started out having language in dose modeling, and then we i
11 had 1549. What happens over time in the NRC is that l
12 documents get changed, they reference other documents, the !
l 13 other document doesn't get changed, you now have two 14 documents that don't say the same thing, people become 15 totally confused. So I felt that it would be a lot clearer l
16 that for dose modeling the regulatory guide would just say, 17 use 1549, and that's really where the decision came inI.
~
i 18 I realize that it is hard to follow. I think this l
19 is something we're going to work on over the course of this i
L 20 year. But, again, I think part of this is the problem that 21 just simply comes in with the format of regulatory guides. -
l \
-22 NUREG documents have the ability to be in whatever 23 form you want them to be in. And they can also be a lot 24 easier to follow than regulatory guides. So we will try to i 25' find a mechanism for-making the whole process a lot clearer.
l l
ANN RILEY & ASSOCIATES, LTD.
Court Reporters 1250 I Street, N.W., Suite 300 Washington, D.C. 20005 i (202) 842-0034
I I
200 l
1 Another point I want to make at this time is, you 2
have to realize that this regulatory guide addresses only 3 one part of decommissioning, the final release criteria.
4 That is it's only purpose. It does not walk people through 5 the decommissioning process. ,
6 We were asked to write this guidance to implement 7 the final rule and so we are limited in what topic's we can 8 discuss. And, in fact, I believe there will be a standard 9 format and content guide prepared for license termination 10 plants. And I think there is one currently.for 11 decommissioning plans, but I can't remember now.
12 But, anyway, nonetheless it's something we 13 recognize needs to be clarified a little what the process 14 is. And I'm hoping that when the standard format and l 15 content guide is prepared that will also provide additional .
16 guidance on the various steps and, you know, when licensees 17 need to make various decisions. So hopefully that will 18 solve some of those problems.
19 MR. MANN: A question on regulatory position i
l 20 Section C, draft guide and it talks about the obtaining NRC !
21 approval of it's DCGL,, that doesn't mean to imply that -
i
{
1 22- you'll have a single derived concentration guideline value j i
23 for a given site, does it?
l
,24 MS. DAILY: No.
I 25 MR. MANN: In response to your response, I don't ANN RILEY & ASSOCIATES, LTD.
Court Reporters l
1250 I Street, N.W., Suite 300 Washington, D.C. 20005 (202) 842-0034
201 ;
1 want to dominate the discussion here, but I'm a little bit 2 concerned about your statement that this guidance is not 3 intended to be applicable to the overall decommissioning 4 process, if I understood what you said. There were I 5 statements made in the presentations this morning and in 6 earlier presentations asking the question, is this guidance 7 practical for these models an'd decision methods; do they l 8 appear to be useful for licensees to make real-world
{
9 decisions about decommissioning? And I think it's a paradox 10 that's posed then if we have to answer that question based 11 on our own objectives.
12 I would have to take issue with that because the 13 endpoint decommissioning is license termination and we have 14 -- and particularly in the utility industry, we have to come 15 up with a decommissioning -- or factor that into the way the ,
16 utilities do decommissioning. And maybe we're-talking at I 17 !
the wrong level here, both at my end and at your end about ;
18 this. But I think the implications of this rule are 19 potentially very significant if you take literally what's 20 being offered here as guidance and try to factor it into the 21 decisionmaking process in terms of extent of removal of -
22 facilities and equipment and so forth, there's a lot at 23 stake here. And the big constraint that's not addressed is 24 the economic costs and the time costs of money and all those 25 things.
1 I
ANN RILEY & ASSOCIATES, LTD.
Court Reporters '
1250 I Street, N.W., Suite 300 l
Washington, D.C. 20005 (202) 842-0034 L
F l'
202 1 So I think we should be careful, I think both 2 sides. I think we need more dialogue between the industry l 3 and the Commission perhaps. It's not at this level that I 4 perhaps that should be addressed, but I think the 5
implications and the financial exposure potential is ,
6 significant.
7 ' MR. SAITO:
l Earl Saito, Combustion Engineering. I 8 had a question about your last comment that tLis is not -- !
9 this is just for decommissioning -- this is just for final 10 status survey. What we talked about today or the flow 11 diagram here is not just final status survey, the flow 12 diagram here is complete decommissioning. So I'm very 13 confused as to your statement and the discussion. Can you 14 help me clarify that?
i 15 MS. TROTTIER: I don't think I said " final status '
16 survey." This only addresses the radiological criteria for l 17 license termination. That is the only purpose of this 18 regulatory guide. Meeting that rule, meeting Subpart E to l 19 Part 20, that is it's only purpose. It does not do anything
, 20 to meet any of the requirements in Parts 30, 40, 50, 70, 72, i
21 it just meets the requirements in Subpart E to Part 20. ..
22 MR. SAITO: So it's just to give us the limits 23 that we're working for, is that -- !
24 MS. TROTTIER: How to meet the limits.
25 MR. SAITO: How to meet the limits.
ANN RILEY & ASSOCIATES, LTD.
Court Reporters 1250 I Street, N.W., Suite 300
{ Washington, D.C. 20005 l (202) 842-0034 l-l
r-
- 203 1 MS. TROTTIER
- The limit is in Part 20. This is 2 how to meet the limit.
3 MR. SAITO: How to demonstrate compliance?
, 4 MS. TROTTIER: Right.
5 MR. SAITO: Okay. I'm sorry, I was a little 6 confused there.
- 7 While I have the microphone, one other question.
8 The ALARA requirement here seems to be inconsistent with 9 what the ALARA requirement was in the previous meeting where 10 ALARA was a planning tool rather than an end compliance 11 demonstration. That you would use ALARA to make decisions 12 on what work should be done, not that at the end of the 13 process you said, am I ALARA. And there seems to be an 14 inconsistency between the guide and what has been presented l 15 today. And I was wondering how that could be rectified? d 16 MS. DAILY:, Well, I think we intend the same I
17- thing. ALARA is-incorporated in the process, in the i I q l 18 decisionmaking process. So it's not intended to be j 19 something that you tag on at the end. If I'm understanding 20 what your statement was.
l 21 MR. SAITO: Yeah. ,,
22 MS DAILY: The diagram can be misinterpreted, I 23 think,.in that there's that box there at the end that says 24 ALARA and really that's there for the licensees that are 25 doing screening. They're not going to go through the whole 1
a ANN RILEY & ASSOCIATES' LTD.
Court Reporters 1250 I Street, N.W., Suite 300 Washington, D.C. 20005 (202) 842-0034 l
j
l 204 1 process. They're not integrating ALARA in their 2 decisionmaking about remediation or more site-specific 3 modeling or whatever, they're just going through, they're 4 demonstrating that they're meeting the screening criteria 1
5 and then there's any remaining ALARA issues like good, j l 6 housekeeping type issues, but it's specifically not a 7 calculational type approach. It's the same thing that 8 you've done traditionally in the past as part of your 9 operating program, your ALARA activities. So it's not meant 10 to be a whole separate analysis at all.
11 MR. SAITO: Thank you. -
12 MR. WILLIAMS: Alexander Williams, I'm with the i
13 Environment Management Office of the Department of Energy. !
14 First of all, I would like to follow up on a 15 couple of items from the November meeting. I'm one of the .
, 16 DOE sponsors of the RESRAD code and I was very: surprised at 17 some of the things in November about how RESRAD and for that l 18 matter the MEPUS and PRESTO codes were treated there.
l 19 Through the years I've spent, oh, perhaps a third or a half 20 million dollars of the taxpayers' money doing various things 21 with NRC involving RESRAD. We run about 20 diff'e rent -
22 training workshops at NRC either headquarters of field 23 offices training NRC people RESRAD. code. We've also added l
[ 24 some radionuclides to RESRAD at the request of NRC. We also 25 have made certain code changes on a recommendation of NRC.
ANN RILEY & ASSOCIATES, LTD.
Court Reporters 1250 I Street, N.W., Suite 300 Washington, D.C. 20005 (202) 842-0034 L
205 i So imagine my surprise to go to this meeting in l 2 November and discover that there was some question as to 4 l
3 whether the previous NRC approval of use the RESRAD by 4 Commission staff and licensees was still valid especially 5 since I wasn't even aware that that was an issue. Impgine 6 my surprise that there were all sorts of requirements for 7 RESRAD and other computer codes concerning quality control 8 and verification and validation, and in fact, RESRAD, at l
9 least, had been previously approved, and I presume that the )
10 approval was based on a -- I think a stellar quality control 11 record that RESRAD has. So this is all very puzzling to me 12 and I thought that I would inquire about this right now. ;
13 You do, of course, have a letter from DOE on this 1 i
14 same subject. But it left me puzzled since I've spent a lot 15 of my time personally on all this training, and, you know, a I 16 fair amount of the taxpayers' money, I'm just very puzzled i
17 about the whole thing. And I thought I would inquire.
18 MS. DAILY: I think you're --
19 MR. WILLIAMS: Is this a fair question?
20 MS. DAILY: I think you may be misinterpreting 21 what we were trying to do in November. If you remember, -
22 what we were trying to do was lay out for all of the codes 23 asking the developers of those codes to let people know what 24 those codes did, what QA they had gone through. We were not l
[ 25 saying that RESRAD had not gone through QA, we were giving l
l t
ANN RILEY & ASSOCIATES, LTD.
Court Reporters 1250 I Street, N.W., Suite 300 Washington, D.C. 20005 (202) 842-0034
206 1 you an opportunity to share with the public what that QA t
2 happened to be. We have not called in and questioned the 3 previous use of RESRAd, so I think you may be 4 misinterpreting what we're talking about here.
5 The point of what we're talking about is that when 6 you use a code at a site, and the means any code, you need l 7 to put some thought into whether or not that code is '
8 appropriate for your particular situation, and that's what i 9 we have stated.
l 10 When you do a site-specific analysis, we're simply 11 saying, make sure that that code covers the things that you 12 think it needs to cover for your site, it incorporates the 13 pathways that are important for your site in a way that you i
i 14 think is reasonable for your site and that the parameters a- 15 that you use, the site-specific. parameter values make sense 16 in that context. .
17 We apply that same criterion to our own code. We 18 have to look at our code and make sure that it is 19 appropriate for the intended use. If you make any changes 20 to the parameters, you must justify those parameters as 21 being appropriate for your site. .
22 MR. WILLIAMS: Well, forgive me for going back 23 into the record of the last hearing, but there were a number 24 of statements made about how licensees would have to come in 25 and justify'other codes, other parameters, how they would ANN RILEY & ASSOCIATES', LTD.
Court Reporters 1,250 I;S.treet, N.W., Spite, 3,00 Washington, D.C. 20005 (202) 842-0034
l' 207 1 have to come in perhaps with a QA package 'cn1 any other code, i 2 )
and there are a number of codes out there that have had '
3 extensive use, have had extensive QA, I really think that 4 based on the presentations back in November and some might i 5
say that I'm a biased party, but based on the presenqation 6 back on November, there was more QA for any of the other 7 three codes than there was for DandD. And'that's easy to 8
understand conceptually because DandD has been more recently 9 developed in either RESRAD, or MEPUS or PRESTO. And I think 10 that each of the other three codes has advantages and 11 disadvantages to it, but this isn't the place to get into 12 that.
13 I agree with what you're saying, at the same time, 14 this whole treatment of the other codes, not just RESRAd, 15 but the other two as well, seems very puzzling to me and it 16 appears to me.to perhaps be what I refer to as the "NIH 17 syndrome" and I'm not referring to the institution down the 18 road here either. Because it seemed based on what was in 19 the November meeting that all sorts of hoops and hurdles and 20 stuff were being established by the NRC staff which were to i
21 be shown or demonstrated or proven or whatever by licensees . - -
22 who use another code while these same requirements or hoops l
23 and hurdles were not being applied to DandD. And it left me 4
24 very puzzled because, you know, through the years I've done 1
25 an awful lot of stuff with the NRC staff in terms of RESRAD.
i i
ANN RILEY &' ASSOCIATES', LTD.
Court Reporters 1250 I Street, N.W., Suite 300 l
. Washington, D.C. 20005 (202) 842-0034
l 208 ;
1 And it's just left me very surprised and sort of puzzled and 2 sort of wondering what was going on.
3' MS. DAILY: I think it would be helpful if you l
4 would look carefully at, especially, Section 5 of 1549 and !
i 5 point out in additional comments, we have received your \
{
6 first set of comments --
7 MR. WILLIAMSt Sure.
8 MS. DAILY: -- and we will address them. Point 9 out specifically where you think RESRAD or another code is 10 being treated inappropriately, or if there's information in 11 there that you think is not reasonable, information we !
12 I should include, it would help to have specific comments on 13 that section.
i 14 MR. WILLIAMS: Okay. Well, the likelihood of !
15 having significant and meaningful comments between now and )
16 tomorrow, if in fact tomorrow is the date, is about zero as 17 I think we all recognize.
18 MS. DAILY: But as we have stated, we will be :
19 trying to incorporate any changes as appropriate before this 20 is published and we do have an interim use period where we 21 _will be making additional changes. I think it would be very. -
22 worthwhile for you to make your comments.
23 MR. WILLIAMS: Okay. Well, you know, this was all 24 very surprising when the matter first came up back in 25 November with what I felt was no notice whatsoever, and l
l l
ANN RILEY & ASSOCIATES, LTD.
Court Reporters 1250 I Street, N.W., Suite 300 Washington, D.C. 20005 (202) 842-0034
209 1 given that I had worked with any number of people both at I 2 NRC headquarters and at field locations on the RESRAD 3 training in particular, this was all very surprising and 4 very puzzling to me as I explained earlier.
5 Second of all, I would like to address a te,chnical 6 issue which is the whole issue of Monte Carlo analysis. l 7 Thi's is based on my experience with RESRAD. I think that 8 with DandD you'll probably find the same thing. We have 9 experienced some difficulty getting DandD to work properly, 10 so I can't say that there was experience with DandD. And 11 this is one of the problems with DandD as it was pointed out 12 back in Novenber that there were come bugs in it, and there 13 hadn't been benchmarking.
14 MS. DAILY: Have you submitted those comments to 15 us?
4 16 MR. WILLIAMS: It's right in your record, ma'am.
17 It's in the record of the hearing. If you go to -- let me 18 see here, if you go to page 43 in Theresa Brown's 19 presentation, here's one quote. "We're in the process of 20 working out all the software bugs. In the same general area j 21 to date the code itself has not been included in any .
22 benchmarking studies." I would think that the -- I think l 23 that this issue would appear self-evident since it was your 24 transcript and your meeting --
25 MS. DAILY: But we're talking about bugs that you i
1 ANN RILEY & ASSOCIATES, LTD.
Court Reporters 1250 I Street, N.W., Suite 300 Washington, D.C. 20005 (202) 842-0034 l
210 I said that you ran into when you were attempting to run I l 2 DandD.
3' MR. WILLIAMS: Uh-huh. Yes.
4 MS. DAILY: So what I'm interested in is hearing i 5 exactly what those bugs were. , !
6 MR. WILLIAMS: Okay.
7 MS. DAILY: Especially if it's in writing.
8 MR. WILLIAMS: And it also might be the probl'em 9 that sometimes there are problems with computer 10 compatibility or operating systems on a computer.
11 MS. DAILY: Yes. I 12 MR. WILLIAMS: I'm, frankly, not the swiftest i 13 person around when it comes to computer activities.
I 14 MS. DAILY: A lot of people do run Windows 95, for l 15 example, and DandD was developed under Windows 3.1. So we 16 do have that 800 number and we do have an e-mail site that i
17 you can call and get support for any issues that you run 18 into when you're trying to run the code. And we track all 19 of those comments and will be incorporating any changes that 20 need to be made into the mode. But we certainly can't fix 3
21 something if we don't hear about. -
22 MR. WILLIAMS: Uh-huh. We have had some problems s
23 with RESRAD being compatible with various releases of 24 Windows and DOS for that matter. In one case RESRAD did not !
25 work with a particular-DOS release, and we discovered that !
i ANN RILEY & ASSOCIATES,' LTD.
Court Reporters 1250 I Street, N.W., Suite 300 Washington, D.C. 20005
-(202) 842-0034
l 211 1 it was a problem with DOS and not a problem with RESRAD.
)
2 Let me go into the technical issue, though, of i 3 Monte Carlo analysis. All of this Monte Carlo analysis or 4 uncertainty analysis is all well and good and scientifically 5 it's very sweet. Unfortunately, though, when it comes to {
u 6 cleaning up sites or decommissioning sites,.I'm not sure 7 that it'.s especially useful.
8 Basically for different classes of radionuclides 9 the assumptions you make on site use make a lot more 10 difference in terms of the dose than the uncertainty in a 11 lot of the physical parameters. For example, if you look at 12 those radionuclides that generate most of the dose from i
13 direct gamma exposure, such as cesium-137, cobalt-60, and 14 rhorium with its decay protogyny in equilibrium, most of
'E chese radionuclides -- these radionuclides get most.of their 16 dose by direct gamma, exposure.
17 The parameters, at least using RESRAD, that are 18 most sensitive for these radio.uclides are site occupancy.
l 19 So what makes the difference here are the assumptions that 1
20 you make on land use. If you assuc'e an industrial use, you 21 have a 40-hour per week occupancy, oz. the other hand, if you ,.
22 assume residential use, the occupancy t' the site will 23 certainly not exceed 100 percent. But th is .is an assumption 24 and any amount of uncertainty analysis of this particular 25 parameter isn't going to make an awful lot of difference l
ANN RILEY & ASSOCIATES, LTD.
i Court Reporters 1250'I Street, N.W., Suite 300 Washington, D.C. 20005 (202) 842-0034
I 212 1 because the assumption is basically a number that is a 2 surmise or best-guess based on sort of projected land use.
l i
L 3 MS. DAILY: Which is an excellent point, and 4 that's exactly why.when we looked at the behavioral 5
parameters and the metabolic parameters, since they are tied 6 directly into our defined screening group, those are not 7
included in the Monte Carlo analysis in terms of being 8 varied over at the distribution.
9 MR. WILLIAMS: But the gamma exposure half way is, 10 for the most part, dominated by site occupancy; to a lesser 11 extent, whatever. shielding you get from being inside 12 structures. I don't know how DandD comes out with this, but 13 in RESRAD, those are the most sensitive parameters.
14 Similarly, for those radionuclides that give most 15 of their exposure by inhalation, what's makes the biggest 16 difference here in terms of RESRAD are the assumptions made 17 -concerning the amount of dust in the air and the assumptions 18' made concerning particle size as to whether or not the 19 particles are of a size that they would or would not be 20 measurable. The difficulty here is that, yes, you can do a 21 Monte Carlo analysis, but you're dust loading in a . .- l 22 particular site is something that.is quite literally 23 measurable.
24 And while-doing something that looks -- doing an 25 analytical nothod that looks at lots of details, if what ANN RILEY & ASSOCIATES, LTD.
, Court Reporters 1250 I Street, H.W., Suite 300 Washington, D.C. 20005 (202) 842-0034 !
l 213 l
1 makes the difference is dust in air, I think that in terms !
2 of protecting the health it is in everyone's best interest 3
to spend your time and money looking at what's important 4 rather than a sophisticated analysic.
5 Finally, for your example with uranium in s, oil, 6
yes, indeed the distribution coefficient for uranium is very 7 important. That was shown as a lid on your slide.
One of' 8 the few things in RESRAD that I'm an expert in is the 9 geochemistry so I -- and I've done lots of analyses for 1' 10 sites with uranium.
11 And, yes, the Kd is a very important parameter, I
12 probably one of the most important. It directly affects how 13 fast uranium gets into ground water, there is no doubt about
{
14 that. At the same time, you have a presumption here that 15 groundwater is, in fact, being used for drinking water 16 purposes or for irrigation or for water and crops. And the 17 presumption that you have this infamous on-site well that's 18 going to give people radionuclides out of the groundwater is 19 something that is at least not true for the vast majority of 20 the people in the country. And, again, you have a 21 behavioral assumption or parameter that is quite subjective -
22 and is based on a worst-case analysis and to take a lot of 23 these assumptions when the important assumptions are really 24 not subject to a Monte Carlo-type analysis and then turn 25 around and on the remaining parameters try and do some sort ANN RILEY & ASSOCIATES, LTD.
Court Reporters 1250 I Street, N.W., Suite 300 Washington, D.C. 20005 (202) 842-0034 m
214 1 of Monte Carlo analysis seems to me to be an approach that 2 is not necessarily the best.
3 And, yes, the Department of Energy is looking at 4 uncertainty analysis for RESRAD. We have had an uncertainty 5
shell RESRAD for the last year and a half or two year,s on a 6 test and evaluation basis. Yes, we're going to continue to 7 the development of that, but at the same time I have 8 professional misgivings about that approach simply because 9 so many of the assumptions are really just that, they're 10 assumptions. And the things that make a difference are 11 controlled by people at least insofar as soil is concerned.
12 When you get inside of buildings this reliance upon i
13 parameters that are controlled by people is almost 100 14 percent.
15 What assumptions you make over occupancy, over any 16 kind of renovation, this is all controlled by people. And 17 to turn around and try and look at this with Monte Carlo 18 analysis would seem to me to be something that is simply 19 analysis for an aesthetic art form rather than for something 20 that would provide useful information.
21 I realize that there are probably some .
22 radionuclides and some pathways for which Monte Carlo 23 analysis might be reasonable. But certainly what I've run 24 into which are the uraniums and the thoriums and the ~esiums c
25 and the cobalts and to a lesser extent the americiums and 1
ANN RILEY & ASSOCIATES, LTD.
Court Reporters 1250 I Street, N.W., Suite 300 Washington, D.C. 20G05 (202) 842-0034 L
l l
215 1- plutoniums don't tend to be govern'ed by pathways for which 2 Monte Carlo analysis'would really seem to me, at least, to 3 make a lot of difference.
4 You might want to look into this a little bit more 5 and see exactly where and what circumstances it does ym}aa a 6 difference because the work that I have done on this is by 1
7 no means final wofd, but at the same time I have looked at I 8 parameter sensitivities for many of these radionuclides at i
]
9 many different sites and I'm fascinated by how most 10 radionuclides are dependent on a single pathway and single I
11 pathways in RESRAD are dependent upon, for the most part, a I 12 relative handful of parameters.
13 I think that you might find the same to be true, 14 and, yes, it's good to look at all pathways and all things 15 so that you make sure that you've covered everything and 16 that there isn't something undiscovere,d. But I think one of 17 the uses of the computer codes is for everyone whether it's 18 people in DOE, or in NRC, or in licensees, or anywhere else 19 to spend most of their time and money looking at the things 20 that are important and the things that really do protect 21 people. -
22 Thank you.
23 MS, DAILY: Thank you. I'd like to move on now to l 24 looking specifically at any questions that deal with 25 screening. Does anybody have any comments, questions, {
i ANN RILEY & ASSOCIATES, LTD.
. Court Reporters !
1250 I Street, N.W., Suite 300 Washington, D.C. 20005 !
(202) 842-0034 t
216 r-1 -statements they war t- to make related 'to the first -- it's 2~
actually Section 3 of 1549 where we talk about the screening 3- approach. I'll g.ve you a minute here to rustle through
- 4 your papers.
5 Whilu you're looking through things, I think 6 Theresa could go over a couple of facts about what we did
(
l- 7 with Monte Carlo analysis and what we didn't do with that-8 analysis.
9 MS. BROWN: I just thought it might be helpful.
10 NRC isn't saying that licensees must do' Monte Carlo analyses 11 for their sites. And maybe there's a misunderstanding about 12 how the Monte Carlo analysis was performed and used in'this i
- 13 ' case.
14 It was.the Monte Carlo analysis done for the
- 15 parameter analysis. That is, to try to come up with these
' f L
I 16 default screening values, again, is to describe the f 17 variability and uncertainty in'all of the parameter; values 18 that are used in the DandD models. And so we've already.
, 19 .done those Monte Carlo analyses and the idea was.to provide i
20 -the licensees with the results of those analyses to guide 21 their choice or'their determination of site-specific . -
22 parameter values. Those would be default or single 23 parameter values for either -- for any one of the parameter 24- values. I'm sorry, let'me try this again.
25 It would-be a deterministic analyses that the ANN RILEY & ASSOCIATES', LTD.
Court Reporters 1250 I Street, N.W., Suite 300
. Washington, D.C. 20005 l
~
(202) 842-0034 1
r 217 1 licensees would be performing. And our thought was that you t
2 i could use the Monte Carlo analyses that we had already done l
3 and the results from those analyses to guide determination 4 of what the most sensitive parameter values are and where 5 site-specific information might be useful. .
6 And so your comments in terms of how best to 7 present that information, how would it be most useful and l
l 8 understandable to you was the type of comment we were 9 looking for. I think,that one thing I would also like to 10 clear up is that DandD was designed just for doing this 11 s'creening, simple models, prudently conservative that you 12 could define parameter values or tables from the results of 13 this model that could be used for screening.
14 And the question that was asked in the last 15 workshop was, now, what are the other models that are out 16 there that have good QA that people are used'to using that 17 are user friendly that NRC is used to seeing come in house?
18 How would you move from DandD and what's the logical step to 19 move.to these other models? It was not to say why you would 20 use DandD over these other models, that wasn't the intent.
21 And I'm sorry that that was misunderstood. I hope that that. -
22 wasn't anything on my part.
23 The other thing that I wanted to say is that we've 24 always been required in-developing DandD to meet the same l 25 types of QA standt ds that's in our work process, to meet ANN RILEY & ASSOCIATES, LTD.
Court Reporters 1250 I Street, N.W., Suite 300 Washington, D..C. 20005 (202) 842-0034 l
l i
218 1
the same kind of QA standards that RESRAD and MEPUS have 2 been held to. It's to develop a comparable set of QA codes 3 and benchmark codes.
4 And as Andrew pointed out, we're just still in the 5
processofdevelopingthatcode,sowe'retryingtompetthe 6 same acandards that they've been held to.
7 MR. WILLIAMS: No, it wasn't anything that you 8 said back in November and it wasn't anything, Christine, 9 that you said back in November, but let me quote, and I will 10 leave the names of the guilty off of the quotes, but let me 11 quote from the record back in November. "I think the 12 guidance will be clear that the licensee needs to come in 13 and explain their pathways and lay those out and then 14 justify the deviations." Page 214.
15 Here is from page 220, "Say you did the same 16 parameter analysis and you picked the same confidence level 17 using some other cods, then you would still put in a source 18 g-termandlikelygetdifferentansNtreevenwiththesame 19 confidence level."
20 Continuing from another place, "The question is, 21 how does one interpret that in a regulatory perspective in . ,.-
22 making a decision?" You would have to pull it apart.
Any 23 way you -look at it, you are going to have to pull it apart 24~
and see why you are getting that different answer and then 25 odo the comparison to the NRC baseline. You're going to have ANN RILEY & ASSOCIATES; LTD.
, Court Reporters 1250 I Street, N.W., Suite 300 Washington, D.C. 20005 (202) 842-0034
1 219
. 1.
to get back to that at some point.
2- Here's-another quote, "Then the licensee would 3
'come in and make the argument to us about the parameters 4 that might need to be changed." That's from page 202. '
5- Here's;from-page 198, "Our. default set of ,
6 parameters doesn't' feed neatly into RESRAD or MEPUS or other 7 models because the default is intertwined with the 8 mathematical model in.DandD in 'the way the model is 9 . statement up." It's not-quite as simple as just feeding 10 .into the process that we are contemplating right now.
11 Here's another quote from page 208 --
12 MS. DAILY: Alexander, I think we have the idea.
13 If we could move on.
14- MR. WILLIAMS: Perhaps you understand my 15 puzzlement that there was a very clear message back in 16 . November that anyone.using any of these other codes was
- 17' going to have to come in with a QA package or show this and 11 8 ' demonstrate that and prove that, and yet we were told that-
-19 there were still bugs in DandD. We'were told that there'was 20 no benchmarking, there was no. attention-to do any kind of 21 field validation. And;perhaps you' understand my. puzzlement . ...--
22 --
23 MS. DAILY: Coming from your perspective I 24 understand.your puzzlement. I've tried to. respond by 25 telling you -~
ANN RILEY & ASSOCIATES, LTD.
Court Reporters 1250 I Street, N.W., Suite 300 Washington, D.C. 20005 (202)'842-0034
220 1 MR. WILLIAMS: Okay. Okay. Well, you know, I am 2 -- and perhaps you understand why I am making this an issue 3 now --
4 MS. DAILY: I understand.
5 MR. WILLIAMS: -- because there was some ,
6 unfinished business. And, you know, I certainly have some 7
more quotes from back in November that I will be happy to 8 share with everyone. But nonetheless, this is out of your 9 record and this is in every case of the last quotes it was 10 NRC employees who were doing the talking. So forgive my 11 misunderstanding, but the words were not mine, they were 12 from some of your colleagues.
- 13. MS. DAILY: Okay. And I'll restate, the intent is 14 that for any model that is used to demonstrate compliance, 15 there is a statement that that model is appropriate for any 16 model. .
17 MR. WILLIAMS: Okay. Well, that's fair enough 18 because, you know, my advice for years to both licenses and 19 NRC people who asked about using RESRAD for particular
- 20 applications was to take a good look at it, and to make sure 21 that it was appropriate, and in the case of the NRC folks, -
22 hey, if you're going to approve this for use, then make sure 23- your licensee'has a reasonably up-to-date version and that 24 they're giving you a diskette with the dataset so that you 25 can look at the data and spend your time looking at what's ANN RILEY & ASSOCIATES, LTD.
j Court Reporters 1250 I Street, N.W., Suite 300 Washington, D.C. 20005 (202) 842-0034
_ _ _ _ _ _ _ _ . _a
I- l 221 1 important rather than trying to input data into the code.
2 There are ways to do this that everybody wins, but I had the 3 suspicion or perhaps the misperception back in November that 4
-- and with reason, I think you would agree, that perhaps 5
there was a double standard that there was one standard for l 6 DandD and there was another standard for everyone else and 7 that there wasn't one standard fbr everybody. And I 8
appreciate your straightening out that misconception on my 9 part.
10 There was a misconception, though, as I pointed 11 out, and I'm glad to understand that this was all a 12 misunderstanding on my part. Or that some of your 13 colleagues misspoke and I misunderstood what they were 14 trying to say or whatever.
15 MS. DAILY: I think what we're all agreeing on is )
l 16 best practice. I don't think there's a question that best '
17 practice in doing dose assessments is too -- as a minimum, 18 have a code that is appropriate for your site. !
And I know 19 there has been some misunderstanding about DandD being ;
20 required to be used, or.especially to be required to be used 21 in a situation where it is known not to be applicable. And. ...
22 .that is definitely not our intention. l 23 MR. WILLIAMS: Sure, and I know some places where 24 RESRAD is an applicable leader and I would be the first to l 25 say so. But nonetheless, I think from these quotes you I l
JdH1 RILEY & ASSOCIATES', LTD. i Court Reporters 125J,I, Street, N.W., S.uit.e 300 Washington, D.C. 20005 (202) 842-0034
t 222 1 understand perhaps my puzzlement and misunderstanding or 2 being mislead by your colleagues or whatever. I would like 3 to point out that none of these quotes were things that you 4 said.
5 MS. DAILY: Thank you.
L 6 MR. ROBERTS: Rick Roberts. Has there been any 7 thought on the screening levels that they may be a little 8 bit too conservative? The site that I work on is a 9 transverse in the site dealing with plutonium and americium.
10 And when you look at Reg Guide 1.86 values for prerelease of 11 areas we have a value of 100 dpm per square centimeters.
12 But when you look on the screening level tables for 13 americium-241 and plutonium-239, you have a screening level 14 of 23 to 35 dpb per 100 square centimeters for the 95th and 15 75th percentiles. And we've been releasing areas for years j 16 based on the Reg Gui,de 1.86 values and this seems to show to 17 someone that maybe isn't as knowledgeable about the process I 18 that maybe we have prereleased materials at too high a level 19 in the past'and has any thought been put into it may be the 20 overconservativism of these screening levels? l 1
21 MS. DAILY: I think it's important to remember -
22 that first of all 1.86 was not developed as a dose-based l 23 standard. So, going to a dose-based approach ~is going to 24 change the numbers.
25 The second thing to keep in mind is that screening ANN RILEY & ASSOCIATES,~ LTD.
Court Reporters 1250 I Street, N.W., Suite 300 Washington, D.C. 20005 (202) 842-0034 =
I 223 1 is not for cleanup and it's not to represent a realistic 2 dose for a specific site. It's for screening. And when you 3 go to a specific -- I mean, it's possible that the numbers 4
that you're getting out of 1.86 are perfectly appropriate 5 for your site when evaluated on a site-specific basis. And 6 I don't think it's appropriate to go back and look at value 7
that were used under different criteria and developed for 8
different purposes and compare them directly to things that 9 we're doing now for a different rule.
10 So, we need to make that clear. And for building 11 surfaces, the difficulty really is that as Alexander has 12 said, most of those parameter are behavioral type parameters 13 and the one that is driving radionuclides that are sensitive 14 to inhalation is resuspension. And once again, it is very 15 dependent on the information that we have, developing that 16 distribution is totally dependent on our understanding of 17 the process of resuspension within a building and what has 18 been published in the literature on those processes.
19 So for a small site it doesn't make sense that you 20 would do a whole research project to look at resuspension 21 for your specific site, and the screening is hopefully going. .-
22 to be of use. For a larger facility it may be very useful 23 to do site-specific evaluation of resuspension, taking into 24 account the specifics of your facility including chemical 25 form and surface effects and things like that.
l l
ANN RILEY & ASSOCIATES, LTD.
Court Reporters 1250 I Street, N.W., Suite 300 Washington, D.C. 20005 (202) 842-0034
r 224 1 MR. ROBERTS: I guess I'm getting more towards in 2
using your screening level -- in using the upper end of your 3
distributions over whatever parameter set you've used 4
multiplying the 95th percentile by the 95th percentile or 5
however you did it to get your output distribution, maybe 6
that, what you're using is overly conservative compared to a 7
realistic -- any realistic site that's out there.
8 MS. BROWN: I think for the building occupancy in 9
particular it's not a conservative given the models that we 10 have and the dose limit that we have, the 25 millirem, but 11 they aren't conservative thdy're very -- actually fairly 12 realistic, the most realistic models that we've got within 13 DandD from that perspective. Because all it is, really, in 14 terms of physical parameter is this resuspension factor and 15 the rest are behavioral parameter and they're driven by the 16 scenario that's used,to make the decision. And that 17 scenario itself if not overly conservative because you're 18 looking at just occupancy and it's standard occupancy time 19 for someone in a building. And so none of that is 20 conservative. And that's not the 95th percentile times the 21 95th percentile, it's a random sample from a single PDF and .-
22 that's the resuspension factor because that's the only 23 physical parameter. All the rest are the mean values for 1 24 the distributions for the behavioral parameters.
25 So there's only one distribution in this case for l
} ANN RILEY & ASSOCIATES, LTD.
I Court Reporters l 1250 I Street, N.W., Suite 300 Washington, D.C. 20005 (202) 842-0034 1
225 1
that particular scenario that it's sampled from to get the 2
dose distribution and it is the resuspension part.
3 Now, is the resuspension factor, is it 4 representative? i It's based on measured values. The 5
question that I think we need help with is, you know, how_
6 can we make that more realistic, or are there additional 7
data sources that we didn't have, that we didn't collect, or 8
we didn't have at that time that NRC would find supportable 9 and that you could use to defend a different distribution 10 for the resuspension factor. A different bias to that 11 distribution for example.
12 And so that's the only thing that we can do in 13 terms of realism to make it more realistic and still have a 14 screening value for all possible sites. I 15 MR. ROBERTS: Okay. So --
l 1
16 MS. BROWN: It's not -- there's not -- it's not l l
17 based on, say,, physical limits for resuspension factor, it's ;
18 not just a random sample based on physical limits, it's a i 19 sample from measured values and the range of measured values 20 out there. So in my mind that's realistic.
21 MR. ROBERTS: So, I guess what I'm hear'ing is that. -
22 we've been releasing things at 100 dbm per 100 square 23 centimeters for years and that really that was -- that type 24 of release would give us higher than a 25 millirem dose to 25 an individual? '
l f ANN RILEY & ASSOCIATES, LTD.
Court Reporters 1250 I Street, N.W., Suite 300 Washington, D.C. 20005 (202) 842-0034
226 1 MS. BROWN: Not necessarily. No. Remember 2 screening is not for your specific site, it's looking across i 3 all of the sites that occur in the country now or new l
4 licensees also. And it's trying to encompass the possible 5 resuspension rates that have been measured. And when* we '
i 6 looked at those resuspension rates, we didn't just say I 7
'somebody has measured up to 10 to the minus 2 so we're going' 8 to incorporate that into our distributions, we said, how 9 were those things -- how were those values generated, how 10 were they measured, what did they represent, were they j 11 looking at long-term processes or short-term acute 12 processes. We're not interested in short-term acute 13 resuspension rates, we're looking at longer term rates.
14 So what we're trying to say is the distribution is 15 not -- the possible distribution that could ever occur under 16 any kind of circumstances, we tried to orient it toward the i 17 specific scenario in our screening group. But there's a lot 18 of discussion about resuspension parameter in the attachment
{
19 and it would probably help if you could take the time to
. 20 read for that one parameter, look over what we used to 21 develop that. Because there are other things that impact .
22- resuspension rates that we were not able to incorporate 23 because we didn't have enough information. But we know that :
24 there are other things that impact how easily something 25 resuspends, for example. And those are the kinds of things ANN RILEY & ASSOCIATES, LTD.
Court Reporters 1250 I Street, N.W., Suite 300 Washington, D.C. 20005 (202) 842-0034
227 1 that you may be able to incorporate on a site-specific basis l 2 to modify the dose that you're calculating. And 100 dpm at 3
your specific site the real dose from that 100 dpm is going 4 to be different than just the screening value. The intent 5
of the screening is actually to be at the higher end,of that 6 estimated range.
7 MR. ROBERTS: Okay.. I guess I 411 just end by 8 expressing a concern then that what it looks like in that we 9 have been releasing areas for a long time to a less 10 stringent standard than we're going to start releasing to 11 them in the future. And that might be a perception on my 12 part, but that's the way it looks and that's a concern.
13 MR. WILLIAMS: Let me help you out for just a 14 minute, Chris. Rick, first of all there are a couple of 15 things. First of all you've been releasing stuff to a given .
16 limit, all that means is that you're below that limit. I 17 expect that some of the things that you've released 18 basically has nothing there about background. So I wouldn't i 19 draw a conclusion that there is a problem simply based on 20 the fact that NRC has a screening level and you may or may 21 not be below their screening level. Because what you have -
i l
22 _been releasing you know it's under 100, but you, for the 23 most part may not know how far under.100 it is. So it i 24 doesn't necessary follow.
25 Next of all, I think the NRC folks did a very good ANN RILEY & ASSOCIATES, LTD.
Court Reporters 1250 I Street, N.W.,-Suite 300 Washington, D.C. 20005 (202) 842-0034 l
r 228 1
job of explaining because this was a screening level, not a 2 dose calculation. One thing you could do would be to go 3
back and do a calculation either using DandD or using RESRAD 4-build or something else and see exactly what sort of doses 5
you wind up with and see what parameters make a diff.erence 6 and what parameters don't. I think for plutonium at the l-7 l
i s inhalation pathway would probably govern and I think the L 8 would probably.be the most important thing, but I could be 9 wrong on that, ,
10-Anyway, I wanted to help you out some on that.
- 11 MS. DAILY: Carl?
12 MR. GOGOLAK: Yeah, I'm getting the impression as 13 I listen on and on that maybe we're asking one code'to be 14 doing too much because I heard you mention earlier that the 15 purpose in developing DandD was specifically to come up with ,
16 l
servitive screening. values that we would be able to.
17
. quantitate as being conservative, know that they would be 18 l conservative, have some idea of what 'the parameter' '
19 distributions and control on those were, and that the model 20 was designed-this way and that's why some of.the parameters 21 when you go in and change them they get to be, you know, . - -
22
'related in a complicated way to other parameters _and so it's
'23' -not always straightforward changing one parameter in the 24 model.
i 25 So this leads me to believe that perhaps it's more l
ANN RILEY & ASSOCIATES, LTD.
Court' Reporters
'1250'I Street, N.W., Suite 300 l Washington, D.C. 20005 ,
(202) 842-0034 )
i
[
229 1
difficult than it needs to be now to go to Step 2 using 2 -DandD. Because from the standpoint of a user I look at it 3 and I would say, okay, what I would like to be sure that I 4
know about a model is that if I do go in and start moving 5
parameters around to a different value that there isn't some 6
other parameter in -- that there isn't some' linkage to some 7
other parameter in the model that I don't know about that's 8
going to cause something strange to happen like waterflow up 9 hill or something like that. And as you well know 10 parameters sometimes are linked together, but there's not 11 always a cross-check to make sure that when you change one 12 it's effect on another one is appropriately changed.
13 And so, you know, when those kinds of linkages 14 occur, it makes the analysis necessary to get to a Step 2 15 much more complicated. So I'm wondering if we are not ask 16 asking a code that was designed for screening level now 17 trying to use it for Step 2 and Step 3, perhaps, is maybe 18 counterproductive.
19 I can only think of one example and this is when 20 I've tried to use area factors. And somebody else brought 21 up the idea about using screening for ALARA. And I've had a. -
22 hell of a time trying to figure out area factors using DandD 23 and I've read the Appendix F. It's complicated. And I 24 guess what I have trouble with is that for example the way 25 the external ~ exposure is calculated it is assuming an ANN RILEY & ASSOCIATES; LTD.
. Court Reporters l 1250 I Street, N.W., Suite 300 Washington, D.C. 20005 (202) 842-0034 i
I 230 1 infinite source even indoors. Now, that for gamma emitters, 2 the area factor is going to be affected primarily by the 3 external exposure.
i And there is -- but because of the~way 4
the modeling was constructed and I can understand, if you're 5
, doing it for screening, that.might be a perfectly rea,sonable 6 way to do, but the model wasn't designed to incorporate 7 different areas, so now you have the situation where it 8 becomes very difficult to make an adjustment for area and it L
1 9 has to be done by time of exposure rather than by actual 10 area. And that's just the example that stand out to me.
11 And so I'm wondering if we're not asking too much 12 from one model, and that perhaps other models that were not 13 developed specifically for screening might be more
. 14 appropriate. Perhaps what would be useful. What I would 15 personally find useful is if someone gave me guidance on 16 what range of parameters that I might use in another model.
17 For example, like RESRAD where I can put an area in and does 18 have the direct capability of computing for a specific area 19 external dose, what set of parameters that I should be using 20 at Step 2'in a model like that that was designed for this 21 kind of thing. -
22; I personally think that that might be more 23 p hel'ful. I just throw that out as a suggestion, but maybe 24 we're just trying to do too many things with a model that r
25- was designed for something entirely different at the
+
~
ANN RILEY & ASSOCIATES, LTD.
Court Reporters i
' 1250'I Street, N.W., Suite 300 Washington, D.C. 20005 (202) 842-0034 l
l l
231 1 beginning of the process.
2 MS. BROWN:
1 Well, I'm glad you don't want to put 3
modelers out of business. i 4 MR. YIM: l One comment. Man-Sung Yim.
5 MS. BROWN:
Can I respond to his question (irst?
6 And my comment is, first of all, that the greater detail !
7 that you look at these things -- this complex process of 8
having a contaminant that's transported through multiple 9
pathways to an individual the more detail you look- at it, 10 the harder it is to describe each and every parameter and 11 possible' parameterization. That's when you start needing 12 someone who is an expert to come in and help you do this 13 modeling and to put.these things out there for you. .
14 For NRC, I think the approach was, can we come up 15 with a very simple model, as few parameter values as 16 possible, and provid.e enough information to step through at k 17 least the first part of it, screening, and let there 18 licensees do it without a modeler. And so that's been the 19 goal for DandD and for this. And you were right, as you 20 move to a more site-specific analysis and you need more 21 complex modeling to help you understand what the potential . .-
22 pathways are and the intricacies and the physics within that 23
' pathway, then.you are going to have to use different models.
24' The question'is, should NRC provide guidance for all the 25 potential models out there and should they even provide I- ANN RILEY & ASSOCIATES, LTD.
I' .- . . Court Reporters. -
1250 I Street, N.W., Suite 300 Wa'shington, D.C. 20005 (202) 842-0034
l 232 I
guidance for which models to step forward to.
2
! I don't think in dealing with them that they feel i
3 comfortable in doing that. I think they feel more 4
comfortable in providing general guidance. If you come to 5
the table with another model, and that's certainly --
6 certainly valid, come to the table with one that's been QAed 7 like RESRAD, come to the table with defense for all of your 8
parameter values, why they're appropriate. And, granted, if 9
you have some information from NRC on a parameter value for 10 DandD that the NRC has already looked and said, here are the 11 ranges of values that we found in the literature that we 12 find valid, then that gives you some guidance for moving 13 forward. If there are parameters within your model that 14 they haven't looked at yet, they can't provide that 15 guidance, but you know from the documentation that they've 16 provided what sort o,f information they find to.be valid for i 17 documenting parameter values.
18 MR. GOGOLAK: Well, let me just say, first of all, 19 there's been a lot of talk about QA and QAing models here.
20 I have absolutely no idea what you mean when you talk about 21 QA. Does that mean that you're calculating what you think . -
22 you're calculating, that somebody checked to make sure that 23 you didn't misspell code, there aren't any syntax errors 24 because the QA -- it should be clear in this case, the QA l 25 doesn't mean that you have any idea whether your model is ANN RILEY & ASSOCIATES, LTD.
Court Reporters 1250'I Street, N.W., Suite 300 Washington, D.C. 20005 (202) 842-0034 1
I 233 i 1
1 any closer to some true dose than somebody else's model 2 because that's basically not determinable.
l 3
I think we end up in a situation where we have 1 I
4 dose particularly through internal pathway. External we can 5 go and measure absorbed dose and air, and that's at least L
6 close, ballpark. But when we start doing things like 1
7 ingestion, you know, even inhalation is tough because ask (
8 the people who modeled dosimetry for radon how easy i 9 inhalation was. And ingestion the further we get along in 10 this chain the more complicated it is and the more !
11 impossible.it is to really compare it to anything out in 12 reality. I mean, I've done modeling and I've done 13 measurements and I've done validations of models for 14 atmospheric exposure, external e).posure and even that's the I l
l 15 easiest one, and you're all over the map.
16 So I don't,really know, I mean, at some point 17 later today I'd like to know what really is involved with !
18 this QA. It's becoming a big issue and I don't know what l
l 19 kind of bugaboo this is, but as a user and as somebody who 20 is struggling with the idea of, okay, what is it that we 21 need to do in order to satisfy these requirements? You have. -
22 to have some idea of how to translate the dose rule into 23 something that you can measure in the field so that you can 24 demonstrate. And to my mind that's -- in this context, this 25 is my narrow view of what a dose model should do.
ANN RILEY & ASSOCIATES, LTD.
Court Reporters 1250'I Street, N.W., Suite 300 Washington, D.C. 20005 (202) 842-0034
234 1
Now, if you don't want to spend a lot of time and )
2 effort, if you have a table of values which the regulator 3
says, use these, these are acceptable for doing dose 4
conversions out there like you have in Table B, as you say 5 what DandD was designed for, that's fine. Plus for a lot of i 6
people who have very little contamination they don't need to 7
become dose modelers, they just need to look it up in the 8 table.
But here's what happens in reality, and I don't know 9
how many times I've had this discussion within and outside 10 the NRC.
Because I get people who come to me and say, well, 11 you know, you see these values here, 25 millirem is too low 12 a limit because we can't measure 25 millirem and I'm saying 13 that's not true that you can't measure 25 millirem, what it 14 is, is you can't take the 25 millirem, go to a default 15 screening table, take the equivalent concentration and then
=
16 go on the site and try to measure it for certain 17 radionuclides. Uraniums and thoriums we know tend to be a 18 problem. I We know that Ks of B tends to be the culprit.
19 The problem is that even today after we know this, 20 it's still problematical for somebody who has uranium or 21 thorium at a site.to go and easily convince somebody that he- -
22 can use something on a site that's different than the 23 screening model.
People go and they say, you've got to jump 24 through this hoop, that hoop, the other hoop, show this 25 data, that data, the other data. Otherwise what happens in I ANN RILEY & ASSOCIATES, LTD.
Court Reporters 1250 I Street, N.W., Suite 300 1 Washington, D.C. 20005 )
(202) 842-0034 I i
235 j 1
effect is that the minute you start throwing these hoops up, 2
you make it difficult for people to even get the data and 3
you make it even more difficult for somebody on the other 4 side to accept the data. And then what happens is nobody 5 uses anything but the default values. And this what,I was 6
encouraged in NUREG-1500, I was encouraged in 1549, all of 7 4 the discussions we had their idea was to try now that the 1 8 dose limits were getting lower and lower, the idea was to 9
try and make the dose modeling more realistic.
10 )
Because coming from the survey side I'm tired of 11 people saying, we don't have the equipment to show that this 4 12 rule can be met. And that's not true. We can measure 25 13 millirem easily from most radionuclides even in background.
14 The problem is, we can't do it when the model is so 15 conservative that we're actually using concentration' values 16 that correspond to one or two millirem. And I think that's
- 17
! what we need to get off of, and we've got to think about 18 things of how are we going to make it easy for people to do 19 this.
20 So if we know there's a case of KSDs that are 21 tough, why don't we have some default tables to make it .' '
)
22 easier for people, point it out for them, okay.
23 -
If we know that there is already a model where 1 l
24 it's easier to calculate area factors, why don't we look and !
J 25 see how that model might be used more effectively. If we l l
4 ANN RILEY & ASSOCIATES, LTD.
~
Court Reporters 1250 I Street, N.W., Suite 300 Washington, D.C. 20005 (202) 842-0034 1 4
l
' 236 1
have a large user community out there, it might be better 2
than trying to reinvent the wheel and, you know, you've 3
already got a situation where as far as I'm concerned 4
I infinite half-space exposure while it's fine for screening, 5
it's just plain wrong for any other kind of realistic dose 6 assessment. It's wrong. It's too much of an approximation 7,
and it's not hard to fix, it's not hard to model right. So l 8
.for me I have a hard time with that, I really do.
9 Particularly when there is in the discussion of area factors
'~
10 how to make an adjustment for the pathway that involves i 11 tracking in dirt on the bottom of your shoes that gets 12 resuspended inside a house. :
13 Now, this goes back to what William Alexander was i 14 saying about let's spend some of our effort in trying to 15 figure out which pathways are important. I mean, here you 16 can say, gee, we've looked at these pathways and this one is 17 never going to contribute more than XY, maybe there's one or I 18 two radionuclides, or one or two scenarios that we might'
~
19 have to look at, but why have this just incredible laundry 20 list of pathweys that get down to, you know, people who are l 21 standing on the side of a bank that's got sediment and, you .
22 know, take a dip in it and it's got some radioactivity -- I 23 mean, some of.these really minuscule pathways. As we said 24 before, it's good to brainstorm and have all of these, but 25 the value in this kind of analysis that can be done is to i i
l ANN RILEY & ASSOCIATES, LTD.
Court Reporters l 1250!I Street, N.W., Suite 300 Washington, D.C. 20005 (202) 842-0034
237 l'
say, look, there are hundreds of pathways,-but there's 2
really only two or three,that are important for these 3 radionuclides, and let's focus on those. Otherwise it gets 4 totally lost.
5.
So'here we have a situation where we've gop~the-6.
" Princess and the Pea" with the tracking in dirt on the 7
bottom of your shoes versus a way of calculating external 8
exposure that I personally would have a hard time living 9
with in a model that I would use that had any pretense to be 10 calculating a realistic dose as opposed to a screening. I'm 11 sorry, I just have that -- it's ---
12 MS. DAILY: I think your comments are very 13 helpful, actually, and I appreciate you making them.
14 MR. GOGOLAK: I wasn't making them to try, you 15 know, to be other than helpful. .
16 [ Laughter.]
17 MS. DAILY: Thank you. I realize that.
18 I think in terms of area factors especially, that 19 'needs to be simplified a lot and when it was first 20 developed, as you recall', it was maybe two years'ago and it 21 is not incorporating-the information that we have now. The . - !
i 22-sensitivity analysis that we have now, which was just !
3 23 completed, can be. incorporated into that'to simplify'it -
- 24 considerably.
25: Originally without that information I had to look w-
' ANN RILEY-E ASSOCIATES,' LTD. 4 Court Reporters j
1250'I Street, N.W., Suite.300 Washington, D.C. 20005 (202) 842-0034-
l 238 1 at all of the pathways. And as you're pointing out, it 2' makes.a whole lot more sense to just concentrating on the 3 pathways that are of importance.
I 4
In terms of your comment on external exposure, I 5 _ agree, if external exposure is of importance for your.
6: radionuclide, then you may need to do a more realistic 7 evaluation of external exposure. . At the moment what we have 8 is DandD using an infinite plane. We've.tried to make that
) 9 explicit as possible so that a licensee going in says, I've 10 got external . exposure aus my primary pathway, it's not 11 handled in a way that helps me in DandD, so I should move to 12 another model.
13 Now, obviously it would be much more helpful if we
{ 14- said it's handled in this way in DandD if external'exposureu l
l 15 is your primary pathway and you;need a better representation L 16- .of that. pathway go t,o model X and use this kind of i
17 -parameter. We just have not had the time, we haven't had 18- the resources to do that kind of work. I agree that it
- 19,' would be very helpful to be able to do that.
To the extent
.20 that.weLhave looked at that parameter, described it and said 21 how we are handling it, as.Theresa_said,'at the moment that . '
22- is'the level of support that we-have for licensees. But
? ~23- yours comment that it would be very helpful to have more L y 24 =information to give better guidance, I think is a useful' i l
- 25. ~ comment. '
i ANN RILEY'& ASSOCIATES, LTD.
CourtLReporters 1250'I Street,:N.W., Suite 300 l
Washington, D.C. 20005 (202) 842-0034
239 1
Well, at the risk of appearing in
(
MR. GOGOLAK:
i 2
the pay of Charlie Ewe, which I am not, and I have no vested 3
interest in RESRAD except for the fact that while we were 4
doing work in the MARSSIM and I was working on 1505 and we 5
were trying to develop the survey scheme, I needed to have 6
something where I could play with it and see how this would 7 work out in practice.
And I used RESRAD because it was 8
available, it was there to be used, it was version at the 9 time, 5.6, okay, it had been accepted in a lot part of the 10 industry.
11 I'm not suggesting that the NRC should go ahead 12 and, you know, start supporting every little Ma and Pa 13 model. Honestly, I believe modelers should be doing certain 14 things to help us out. But I don't think -- honestly the 15 world doesn't need another model. They are all basically 16 the same except for ,the parameters that you put in them.
17 And you may have a specific purpose for DandD and that's why 18 it was built that way, and I'd say, great, fine, use it for 19 that purpose. But, now, if there is another l
20 widely-accepted, easily-available model, and I don't mean i i
21 !
one that costs $20,000 necessarily, but one that might be . .
l 22 l available mostly for free, and that's had a lot of scrutiny, !
23 maybe those resources that you have at your disposal might 24 be better served by leaving DandD for the screening and for l 25 some of these other things, perhaps, looking to see how t
I ANN RILEY & ASSOCIATES, LTD.
Court Reporters 1250'I Street, N.W., Suite 300 Washington, D.C. 20005 ]
(202) 842-0034
I ,
240 1
another widely-accepted model that has been used also by the 2
NRC in the past might be more effectively used for other 3
purposes, other than the screening. It's just a suggestion.
4 Because what I see is that there's a lot of time 5
and effort going into trying to make DandD into something 6
that it didn't start out to be. That's what I keep hearing.
7 And it's just not making sense to me. '
8 Now, we're going to need to use something and I'm 9
wondering how we can get from Point A to Point B in an 10 efficient way. !
11 MR. YIM: A short comment, Man-Sung Yim from NC l 12 State.
Regarding the comment about the use of Monte Carlo 13 probabilistic analysis, I think that one of the concerns --
14 maybe not a concern, but comments we hear is the '
15 relationship between MARSSIM or any sampling survey between l 1
16
-- relationship between the sampling and the dose modeling.
17 People consider there is a gap. And I think use of 18 probabilistic Monte Carlo approach could provide a way to 19 breech that gap and help the sampling rate achieve more 20 integrated dose-based decisionmaking.
21 MR. MORTON: Henry Morton. One brief follow up on. .
22 what has just been said and particularly on Carl's comment.
23 And it has to.do basically with trying to refine the 24 conversation to deal with terminology of computer codes or 25 computer programs, model or multiple models within a ANN RILEY & ASSOCIATES, LTD.
Court Reporters 1250'I Street, N.W., Suite 300 Washington, D.C. 20005 (202) 842-0034
T 241 1
computer program, scenarios, pathways and values of 2
parameters, and I think perhaps the best way to try to focus 1
3 that is to try to ask a leading question. And that is 4
1 whether the DandD program will be made available in
! 5 electronic forum that can be -- in which a model within that 6
set of models might be changed in the event a licensee has 7
one pathway that needs to be changed?
8 MS. BROWN: Henry, can you clarify something for 9
me a little bit in terms of, are you saying substitute say a 10 different groundwater model or --
13 MR. MORTON: That would be an example. Any given 12 -- any given model. For instance, a lot of conversation --
13 we've had a lot of conversation this morning about the 14 resuspension model.
One might find when you go to do the 15 field work or experimental work tt-' you've suggested that '
16 you get the best information for the dust loading model, for 17 example, coupled with measurements and then that model then 18 might need to be changed from a resuspension model to a dust 19 loading model and perhaps a dust loading model and a surface 20 source term. '
21 So it's -- the question goes to the point of -
22 )
whether thic computer program will be available to change a 23 one model. The need there is that when you look at these 24 computer programs for complex sets of radionuclides, you may 25 have growth and decay over time from different nuclides. i ANN RILEY & ASSOCIATES,' LTD.
i Court Reporters 1250'I Straet, N.W., Suite 300 Washington, D.C. 20005 (202) 842-0034 l
l l
L
242 1 Then you have the section of the program that does the l
2 accounting of the various organ doses to choose the ones
! 3 that are fed into the calculation to get the total effective '
i 4 dose equivalent.
5 So, if I only need to deal with changing one '
i '
6 model, if that's a key model, and I don't have these other 7
characteristics of the program, then, in effect, I may have :
8 to go get another computer program, not just a model that's 9 as we've -- as the conversation has kind of loosely gone l 10 this morning.
11 So then if that's the case, the consequence it 12 seems to me basically is a choice of getting a computer j 13 program, in effect that's already been verified, because the 14 obstacles that Carl has mentioned, the obstacles that i
15 Alexander has mentioned, the costs that he's mentioned and j 16 undoubtedly the similar costs for you to develop DandD is 17 too much of a hurdle in practice for licensees for one 18 decommissioning job to go develop an entire computer program 19 that has the full complement of pathways in it. And then to 20 have to go through the hurdles of verifying that for you.
21 So that the practical result then is to say the -.
22 choice -- the choice is to either not, as a natter of 23 practice change that model in some cases. There are cases 24 in which we can integrate it in and get things in to DandD, 25 I think, but I think there will be plenty of cases in which
)
l '
ANN RILEY & ASSOCIATES' LTD.
, l i
Court Reporters I 1250 I Street, N.W., Suite 300 Washington, D.C. 20005 (202) 842-0034
243 1 that's not the case. !
l 2 MS. BROWN: I think that you have a lot of options 3
when you start looking at what your possibilities are for 4
-moving away from just this screening calculations. It i 5 I depends, again, on your source term which path ~ays w dominate, 6
which areas that we've done a lot of research in'just in !
, 7, general, in terms of modeling, and how many codes are out 8
there and you start looking at the range of possibilities, i 9 and they're tremendous, so it's hard to deal in I 10 hypotheticals with this huge list of radionuclides and all !
11 the potential things that could happen at any given site, 12 some unknown site. But you do have some insights that you 13 can gather and some techniques that you can use.
14 One thing you can do is go to another model that 15 just deals with a specific pathway, if you have a code to 16 run that model, or i,f you do a hand calculation and can '
i 17 i demonstrate that you've calculated properly. If you use a 18 more complicated model to say justify or hsip defend a 19 Parameter value in whatever code that you are then using to 20 do your dose calculation, that that's one option. You can 21 go in nd model some process in a more complex fashion just .
22 to try to help you defend a parameter value or a more 23 limited range.in parameter values. And so it doesn't 2<
necessarily have to all be integrated and run at one time, 25 that there are things that you can go out there and do that i
ANN RILEY & ASSOCIATES, LTD.
Court Reporters 1250'I Street, N.W., Suite 300 l Washington, D.C. 20005 (202) 842-0034 L- _
l 244 j 1 you don't have to have a complete package that goes from i
2 source term to dose. There are things that you can do in
, 3 terms of modeling and modeling in a more complex fashion to
! {
4 reduce your uncertainty even, to provide you with some l
5 insight into how things may behave and come up with a l 6 parameter value.
7 It's just a matter of being very clear in that 8 process of saying, here are all my assumptions, here's how I 9 can defend each one of the parameter values again in this 10 submodel and just p;oviding, you know, just clear, good 11 documentation for what you've done for NRC. And then how 12 you've used that information then to reduce your uncertainty 13 or to limit the parameter value that you want to use whether 14 it's in DandD or any other code. i 15 So there are ways to use more complex models that 16 aren't integrated as,part of a whole transport,. source term, 17 and dose modeling system. I think that in some ways we've l 18 tried to give some guidance on how you might go about that, t 19 but, again, it's been very general. It hasn't been terribly 20 specific. It's hard to do that for each and every parameter 21 value, or each and every aspect of the model. ..
22 MR. MORTON: Yeah, I understand that's true. In 23 general terms.one presumably can de those things, what I 24 think remains is an apprehension among a number of us is the 25 issue of the -- basically the reality of the amount of AMi FILEY & ASSOCIATES,' LTD.
Court Reporters
' 1250'I Otreet, N.W., Suite 300 Washington, D.C. 20005
'202) 842-0034 l
l
o 245 1 defense to be able to use_any other program and the I
2_ inability to do anything to make any changes other than 3' choose pathways and input parameter -- certain parameters 4 Tinto DandD.
5 So I don't want to beat that one anymore. ,
6 I think what I.would like to do with respect to 7
.trying to'get back to the track of this decision methodology 8
b p
and the processes that are described in the section that you 9 wanted to deal with next.
, 10 MS. DAILY: ~
Henry, we're running a little late.
I
! 11 901. MORTON: Sure.
12 MS. DAILY:
i _
Would you like to start us off right ;
l' 13 'after lunch?
l 14 MR. MORTON: That would be fine.
15 MS. DAILY: Is it acceptable to take a break now?
16 MR. MORTON; That will be fine.
17 .MS. DAILY: I know people are kind of getting a
'18 little antsy.
r-19 MR. MORTON: Sure.
20 MS. DAILY: So why don't we take an hour break and 21 come back here and we'll start back with Henry's comment.
22 Thank'you, Henry.
23 MR. MORTON: Sure. !
- 24 [Whereupon, at 12:03 p.m., the workshop was l
25 recessed, to reconvene at 1:03 p.m., this same day.]
i y~ ANN RILEY & ASSOCIATES, LTD.
Court Reporters 1250 I Street, N.W., Suite 300 b Washington, D.C. 20005 (202) 842-0034
,i 246 1
AFTERNOON SESSION 2
[1:14 p.m.]
3 MS. DAILY: Okay. If everybody would take their 4
seat please, we're going to get started again.
5 And like-I mentioned, before we broke for lunch, 6
Henry Morton is going to start off and we really appreciate 7
him allowing us to take a break for lunch and remembering '
8 his question. Henry.
9 MR. MORTON: Thanks, Chris. I think I'll try to 10 perhaps divide them into issues of process and product. In 11 the issue of process, what I look at is basically the 12 documentation and the guidance drafts before -- other than 13 1549 -- I interpret one process, and then when I look at 14 1549 I interpret another process. And from my point of 15 view, I think each of those should be acceptable and each of 16 them as presented lo.oks like they should be acceptable so 17 that the licensee basically should have the flexibility to 18 choose whichever is more appropriate.
19 With respect to the product in terms of trying to 20 figure or understand whether this -- the totality of this 21 rule works for us. We can look at each piece as a piece of . 4 22 the puzzle and I think presented in these various workshopsi 23'
. presented in the various reports we have pieces that on the 24 whole are reasonable approaches. Then I*get a concern when 25 I try to fit these pieces together to see'if this works for ANN RILEY & ASSOCIATED, LTD.
, Court Reporters 1250 I Street, N.W., Suite 300 Washington, D.C. 20005 (202) 842-0034
I l l
247 1
the cases that I have to work with. And in that regard I am 2 having-some difficulty.
3 First with respect to the beta version of DandD 4
and trying to get it to work, after -- I don't know, three 5
or four phone sessions with the computer people at Sandia we 6 gave up. This_was on a Windows 95 Pentium computer. So then 7 I need to rely on what I can get from the tables of 8 guidance, and I discussed yesterday somewhat what I could 9
extract from the-Appendices A and B from NUREG-1549 and what 10 the consequence of that appears to be for the uranium series )
11 and thorium series nuclides that I have to deal with.
12 MS. DAILY: Could I just interject here?
13 MR. MORTON: Yes.
14 MS. DAILY: We've been asked to remove uranium,
)
15 thorium, and radium from the screening tables because it's 16 felt that those numb.ers are not especially useful for most 17 licensees. So in the version that was handed out this 18 morning, those values have been removed from the tables.
19 MR. MORTON: Yes. So when I looked at that 20 version and see that the uranium series, the thorium series, r
21 and radium has been removed, now I'm left without anything . ...--
22 to gain any. perspective on whether this rule will work other 23 'than to go back to the best I can do then which is RESRAD 24 w
. ith either RESRAD defaults and/or PG-808 defaults and/or 25 what defaults I may have been able to get out of NUREG-5512.
I.
l-
-ANN RILEY & ASSOCIATES, LTD.
Court Reporters 1250 I Street, N.W., Suite 300 Washington, D.C. 20005 (202) 842-0034
248 1
But those are -- then I understand they're not necessarily 2
the defaults or the recommended values that I'm going to see 3 in DandD. So at this point basically I'm left without an 4
ability to judge now the perspective of the workings of this 5 entire rule. I can see the pieces, but in totality to see 6
the impact as to what the derived concentrations are or what 7
the allowable residual concentrations and whether I can go 8
measure those and have a measurement process, a statistical
- 9. process, and a linkage with dose-to-concentrLtion. Now, I'm 10- without a way to do that. So while I might be able to 11 endorse in principal or in view pieces of this puzzle, I 12 can't endorse the totality of it because I don't know the 13 -impact of it at this point.
14 MS. BROWN: In terms of running DandD I guess I 15 wasn't aware that there was anyone that hadn't ever been 16 able to get it to load and run. Probably since we didn't ,
17 hear back from you, we. thought the last fix might have 18 worked, I don't know. And I'm sorry about that.
19 MR. MORTON: Yeah, I had three or four phone H20 calls, so there should have been -- there was a record on it 21 'and we basically, jointly, decided to give'up. . ...
22 MS. BROWN: Right. And do you remember why? I'm 23 sorry, just out of curiosity?
24- MR. MORTON: I don't remember why.
25 MS. BROWN: Okay. I haven't had that problem with ANN RILEY & ASSOCIATES, LTD.
~
, Court Reporters l 1250 I Street, N.W., Suite 300 Washington, D.C. 20005 (202) 842-0034
249 1 any of my machines, and I've just -- I'm kind of an idiot
- 2 when it comes to installing these things, I just follow the 3 directions and it works. So, I can't give you any insight l 4 here on why it wouldn't load and why'it wouldn't run.
5* And if we didn't get that fixed, we need to find 6 out why, that's one thing. Another thing that I hear that I 7
want to be very cautious about is using the screeningotables 8 to set DCGLs, that really the screening levels and the 9
concentrations in those tables should be used with great 10 caution and only for screening to see if you can get through 11 this process without using anything other than your source 12 term $nformation. They should only be used for screening, 13 that you shouldn't try to make measurements and establish 14 cleanup levels based on the tables in 1549.
15 MR. MORTON: I understand that.
- 16 MS. BROWN:. But also -- but I do understand your 17 frustration with uranium and thorium. And I think that NRC 18 has thought about this to some degree, and maybe they should 19 answer this, but I'll put in my two cents.- If you have 20 uranium and thorium levels that are somewhere near
- 21 background you probably would not pass the screening at this. -
1 22 point. But you do have another aven~ue, and that's the fact 23- that they are near background and can't be measured above 24 background.
25 MR. MORTON: So, in effect, are you saying that I
ANN RILEY & ASSOCIATES,' LTD.
Court Reporters 1250 I Street, N.W., Suite 300 Washington, D.C. 20005 (202) 842-0034
250 1
the regulation is indistinguished from background --
2 outcome?
3 MS. TROTTIER: The 25 millirem is above 4 background. And I think you'll fine in -- Carl maybe needs 5 ta) help me out here.-- 1505 and in the Reg Guide guid,ance on 6 indistinguishable from background.
7 MR. MORTON: Yes, I've gone through that and I've 8 exercised it with sites.
9 So this basically leaves me with the question of 10 where to go from here with respect to these cases. Do I 11 basically have to get into a -- get into a site and then 12 discover the outcome?
13 MS. DAILY: I'm not sure-that I understand. If 14 you could restate that?
15 MR. MORTON: Okay. Well, let me shift a little 16 bit. I noticed and acknowledged that NUREG-1549 does not, 17 that I see, mention DCGL anywhere. It poses an approach and 18 a process to me in which I interpret, then I go through the 19 decision process and decide the -- decide the options. The 20 options might be to gather more information to help the 21 site-specific calculation, or it might be to go excavate or .
22 decontaminate. Then in following the process through, once 23 I have done that, then it leads me to take the results of my 24 final status-survey which in the end is a concentration or 25 an incremental concentration of the background presumably 1
I l -
ANN RILEY & ASSOCIATES, LTD.
Court Reporters i
1250 I Street, N.W., Suite 300 Washington, D.C. 20005 (202) 842-0034
251 1 putting in the average concentration of each nuclide into 2 DandD or whatever program is approved. Calculating the 3 dose, testing it against 25 millirem, if I'm less than 25-4 . millirem, I should be done. Is that what you intended?
5 MS. DAILY: Yes.
! =
6 MR. MORTON: Okay.
l 7 MS. DAILY: Yes.
8 MR. ABELQUIST: Hi, my name is Eric Abelquist.
9 I'm with the Oak Ridge Institute for Science and Education.
10 I just wanted to share a couple of my comments on 11 1549. And, again, I did have much of a chance to look at it i 12 Lin great detail. I should state up front that most of my
- 13. experience with the NRC's rulemaking effort has been with 14 the MARSSIM approach and participated in that effort for the 15 last four or so years. I guess my first comment'and I'll go l 16 ahead and state the , obvious, but when I looked at the 17 overall NRC's approach we're really adding a lot of 18 complexity to how we're going to do these dose-based 19 assessments in the future. And what I mean by that is {
20 currently, even though it's a hodgepodge of where we get our l l
L 21 criteria, at least someone could go to some of several ,. -
22 different documents and get a cleanup out. And that was-i.
l 23 where we started. And then from there the real effort was !
i 24 going out to the site in the following REG-5849, or now the i 25 MARSSIM manual demonstrating that you-have met that t l l
ANN RILEY & ASSOCIATES; LTD.
Court Reporters i 1250 I Street, N.W., Suite 300 l Washington, D.C. 20005 (202). 842-0034
L 1
252 1 particular release criterion.
2 Now, we're going to a dose-based rule and so the 3 nice part it is it's a dose basis. There's one number 4 that's going to unify all these cleanups. The difficulty is 5 it seems to me just initially there's just as much work 6
coming up with that initial cleanup value as it is to get 7 your site information to demonstrate compliance. So we 8 really added another step to the whole process. And I guess 9
it's probably pretty obvious to most people in this room, 10 but I just wanted to state that up front.
11 The next comment I have is when I look at Section 12 3 in 1549, the framework that licensees may use for the 13 generic screening and I got through the steps and my 14 approach to this, I was looking at.it last night, is to try 15 to see how this interface with the MARSSIM approach. Let me 16 state, for the recor,d, as we develop MARSSIM whenever we got 17 to discussing a DCGL we thought of that as a black box.
18 That somewhere in this development of the guidance to 19 support this rulemaking, there would be DCGL that would be 20 fed into the MARSSIM approach. And so as I was looking 21 l through.here the first thing I noticed was DCGL wasn't -
22 mentioned, some of the language did not see like they had 23 really considered MARSSIM to a sufficient degree. And I 24 think my uneasiness comes from the fact that I'm almost 25 thinking this is a parallel approach to demonstrating 1
ANN RILEY & ASSOCIATES, LTD.
Court Reporters 1250'I Street, N.W., Suite 300 Washington, D.C. 20005 (202) 842-0034
253 1 compliance where someone could choose either the MARSSIM f
2 approach or collect enough data and do a dose assessment to l 3 demonstrate compliance, and that'a really my overall 4 question.
5 Is this intended to be an alternate approach to 6 the MARSSIM or is it eventually going to be a more clear --
- 7 at least for me for a more clear interface of taking a DCdL I I
8 from this approach and then you're still within the overall 9 parameters of a MARSSIM approach? 9 10 MS. DAILY: I think it's clear that there is not 11 sufficient interface between MARSSIM and the dose modeling 12 and that's something that we intend to be working on over 13 the coming year. It really needs to be integrated a lot 14 better and explained a lot better. So it's helpful for you 15 to make comments on that, especially in specific places 1
16 where you think that it would be helpful to add references 17 to MARSSIM or we need to clarify what the interface is with 1 18 MARSSIM.
19 MR. ABELQUIST: One idea that I would have and 20 this is in the spirit of the MARSSIM. Basically ~MARSSIM 21 sells itself on the flexibility that's offered. And so when .
22 I was going through here one of the ideas I had would be 23 have an appendix with your screening values. And I know you 24 have that already, and you've pulled the uraniums and 25 thoriums off the shelf, so to speak, and I wish they would ANN RILEY & ASSOCIATES, LTD.
Court Reporters 1250 I Street, N.W., Suite 300 Washington, D.C. 20005 L
t (202) 842-0034
l 254 1
1 be back in there because we really need a starting point ~for 2 a lot of the approach to MARSSIM.
3 MARSSIM, for those of you who haven't gone through 4
it yet, basically involves doing a lot of desktop studies 5
first to see what your sample size is going to be and,then 6
you have different decision errors you could play with to 7 overall come up with a design that's e'fficient. And so I 8
think what would be beneficial to come out of this NUREG 9 would be screening values for all radionuclides. Then in 10 the spirit of MARSSIM someone could take the screening 11 value, use it as the DCGL, the default DCGL, run through 12 what decision errors need to be made, or what are acceptable 13 decision errors, see what the sample size is.
14 If the sample size looks okay, that given 15 screening value, then that's how you're going to design your 16 survey. If the screening value is so low as to require many 17 measurements or samples, then that would be the trigger to 18 the licensee to go back and do the leg work to feed into the 19 model to come up with a more site-specific DCGL. But you 20 should at first have a starting point where they could go 21 through and see what the sample size may be for the .
22 screening value. And so therefore you've minimized the work 23 that the licensee is going to have to do in coming up with 24 the value initially. It may be fine, but until I see what 25 the cost is and the sample is, you're sort of in a Catch 22.
l ANN RILEY & ASSOCIATES, LTD.
, Court Reporters 1250 I Street, N.W., Suite 300 Washington, D.C. 20005 (202) 842-0034 I
255 1
How much work do I do up front to come up with what I feed 2
into my model to come up with my site-specific DCGL.
{
3 That step might not need to be taken if the l
(
4 licensee could live with the resulting number of samples 5
that MARSSIM would require for this screening.
6 MS. BROWN: Would you sample before you knew the 7 you were likely te meet the 25 millirem?
8 MR. ABELQUIST: Well, in the -- and this is 9 another concern. When you look at the big picture of 10 decommissioning, you need to know what you're cleaning up to 11 begin with because you can't have your decay technicians 12 just haphazardly trying to decide where to clean up, where 13 to have the backhoe dig the ditch. So you do need a number i i
14 .somewhere and so you need to go out and do a i 15 characterization and that is a whole nother issue. How do 16 you design the chara,cterization survey that's going to 17 support the entire decommissioning effort? So, yes, you 18 would have to have some data initially to determination I
.19 where you're going to do your cleanup, but also to determine 20 what factors you're going to put into your code to come up 21 with a site-specific guideline. ,
22 MS. DAILY: This is actually -- I'm very glad you 23 brought this up because it's a fundamental issue. The way ;
24 that the methodology is designed for dose modeling, your 25 first thought is not what am I going to clean up, your first ANN RILEY & ASSOCIATES,' LTD.
Court Reporters 1250 I Street, N.W., Suite 300 Washington, D.C. 20005 l
(202) 842-0034 l
l 256 i
1 thought is, am I meeting the dose criterion? And in the 2 case where you have information about your source term, for 3 example, then that's what you would put in as your first cut 4 to evaluate the dose.
5 In the case where you don't have information c. bout 6 your source term and what you're trying to do is do what we 7 are used to doing which is look for a concentration, then 8 what you're talking about using the screening values to 9 decide whether or not you want to go and use those as a 10 basis for your final survey. There's a little bit of a 11 disconnect here. For licensees that can meet the screening ,
12 i numbers and they want to do something very simple and i 13 !
straightforward, then it makes sense to take those screening I i
14 numbers and evaluate what it would take to do a survey, do 15 your survey to confirm that that's valid for your site and 16 you'd be done. ,
i i
\
17 There are other ways of looking at this process 18 outside of orienting everything toward a survey. But it's {
19 also important that people not use these values as cleanup !
20 levels. They're absolutely not valid as cleanup levels for 21 most licensees. I mean, if the level is relatively high and. .-
22 you look at it and you think that it's cost effective to 23 cleanup to that level, that's fine, but that's not what 24 they're intended for.
25 And so one of our concerns is that they'll do the ANN RILEY & ASSOCIATES, LTD.
Court Reporters 1250 I Street, N.W., Suite 300 Washington, D.C. 20005 (202) 842-0034
257 1 process that you're talking about, they'll take these 2
screening numbers, they'll do the evaluation through MARSSIM 3
and decide that -- I mean, that kind of orients everybody 4
toward saying these are cleanup levels and I have to set up 5 my whole program in order to meet these numbers. And that's 6 not what we want to achieve here.
7 So I understand that you want to have a place to 8 start, and I understand that everybody is oriented toward 9 looking at concentrations at their site, but what we would 10 like to do is have you step back one step behind where you 11 were starting and first ask the question of whether or not 12 you meet the dose criterion. Part of the cost benefit or 13 the analysis of deciding to go and do a better evaluation of 14 your site may be that you think that it would not be cost 15 effective to demonstr' ate that you're meeting those screening i
16 levels. But we don't want people to start from the l 17 assumption that you want to set up something in order to be l
18 able to clean up to those levels. {
l 19 MR. ABELQUIST: Well, to answer that real quickly, {
20 I think then we have'a major disconnect where we're starting i
21 with MARSSIM that we do not address a good, courdinated way -
22 of getting that information initially to come up with a 23 site-specific DCGL so it's which comes first, the chicken or 24 the egg? But we have a real concern here as to how we're 25 going to get to the end point? i i
ANN RILEY & ASSOCIATES, LTD.
Court Reporters 1250 I Street, N.W., Suite 300 Washington, D.C. 20005 (202) 842-0034 t
l l
258 l
1 And the other point I'll make real quick is even i 2 though you keep mentioning time and again that screening 3 values are not the cleanup values, we should not suppose ;
4 that once we go through the modeling and come up with 5
exactly what the better DCGL is, is that still going,,to j 6 relate the some dose anyway, and how often are the 7 assumptions that we put into the model actually going to 8 turn out to be true? And so the screening value may not be 9 that bad of a place to start. And I think it's much better l 10 that an having no place to start.
11 MS. DAILY: No, I agree. There is a very useful 12 purpose to having those screening levels. But I think one 13 of the points of confusion here is that'MARSSIM has been put 14 forward as the final status survey and you didn't haven an 15 opportunity to look at things like characterization. The 16 dose modeling really, starts at characterization so some of 17 the confusion here is, where does this process start versus 18 where does MARSSIM start. And I know there's going to be 19 some more work done looking at characterization j i
20 specifically, and that may help make this interface a little 21 clearer. That may be one of the major points of confusion. . .
22 MR. ABELQUIST: Well, that's good to hear. ,
23 MR. GOGOLAK: Just for a point of clarification, I )
I 24 think part of where some of the disconnect is coming here is 25 that you have to recognize that the MARSSIM process and the i i t
ANN RILEY & ASSOCIATES, LTD.
Court Reporters 1250 I Street, N.W., Suite 300 l Washington, D.C. 20005 l (202) 842-0034 1
i
i l 259 1 survey process is not necessarily something that's i
2 envisioned in every single case. There is an Appendix B of 3 the MARSSIM, a section called " Simplified Procedure for 4 Certain Users of Sealed Sources, Short Half-Life Materials 9 in Small Quantities." I have a strong feeling that tpe 6 people who are covered under this Appendix B of MARSSIM are 7 the very same people who would be able to estimate their 8
source term and go through level run screening. So I think 9
when you're talking about Level 1 screening and somebody 10 estimating their source' term, exactly the same people'we're 11 talking about here who will not even enter the MARSSIM 12 process. Okay.
13 So I think we've been starting from different !
14 points here, or at least we thought we were starting from l 15 different points, but we're not. MARSSIM really starts at i 16 the same point that you have, except that we've sort of 17 said, okay, look, we don't need to spend a lot of time and 18 detail about this, you can estimate your source term and 19 meet a screening model, you can document that you had a 20 sealed source, and you leak tested it, you had the right 21 activity with decay at the end, then clearly you can't have . -
22 any contamination left on mass balance, so you're done.
23 This is, I think, the same sort of thing you're 24 talking about where we have a real problem is, is that 25 that's all fine, well, and good at that point, but then we 1
ANN RILEY & ASSOCIATES, LTD.
. Court Reporters 1250 I Street, N.W., Suite 300 Washington, D.C. 20005 (202) 842-0034
{
260 i 1
start getting into the point where you do need to make some i 2 kind of survey, okay, to estimate a source term. And I 3
think that's where we should look and see that it seems that 4
DandD has spent a lot of time developing the part that 5
MARSSIM really relegated to Appendix B and really the,re is a 6 disconnect in that you've got the screening level models and 7
that's fine, maybe that's what it is, it should support this 8 Appendix B.
The part of DandD that's missing is the part 9 that supports the rest of the MARSSIM.
10 But we really don't start at different places.
11 MS. DAILY: Okay. Henry?
12 MR. MORTON: From my perspective -- Henry Morton 13 -- we have a disconnect only if we look at this as one .)
14 process.
But when I think the disconnect occurs is when you l
- 15. have it looking at it from quote, "the MARSSIM process" 16 versus looking at.it.from the NUREG-1549 process. And my 17 point earlier was that I would recommend firmly that we find 18 each of these two processes to be acceptable at the choice 19 of the user. And once that is done, they can reside 20 together and it gives some flexibility to the licensee to 21- make the choice. It utilizes the survey portion of the - -
i 22 MARSSIM process, that is survey design, and collection of 23 the data. So in thinking of the quote MARSSIM 1505, 1507 i
24 process" we need to, in terms of that, think of the parts of 25 it, survey design, and. performance of the survey, the t-i ANN RILEY & ASSbCIATES, LTD.
Court Reporters 1250 I Street, N.W., Suite 300 Washington, D.C. 20005 (202) 842-0034 I
261 f
i-1 analysis of the data, then the portion that's the derivation 2 of the DCGL, that process.
3 MS. BROWN: I think that's a very important point, 4 that you can't view these as two separate but equal 5 processes, they both need each other. Ycu have to do'the 6 dose modeling in order to do your derived concentration 7 guideline. '
8 MR. MORTON: Yes, but they work differently.
9 MS. BROWN: They work very differently.
10 MR. MORTON: They work differently and it would be 11 redundant to impose on the licensee the requirement to both 12 get the DCGL approved beforehand in that process and then in 13 the 1549 process to require that the licensee after doing 14 his measurements and his dose assessment comply with the 25 15 millirem. Those are two different things. One is done 16 before the fact and one is done - -the two calculations are i
17 done at different places in the process.
18 MS. BROWN: In my mind, the verification that you 19 meet.25 millirems, part of that process is a survey that you 20 have concentrations that are low enough that you -- that 21 you've met this in this modeling and you can demonstrate by . .
22 modeling that you've met the 25 millirems.
23- MR. MORTON: I fully accept that for the 1549 l
24 process.
25 MS. BROWN: I don't think the two should be ANN RILEY & ASSOCIATES, LTD,.
, Court Reporters 1250 I Street, N.W., Suite 300 Washington, D.C. 20005 (202) 842-0034
- I i
262 I i
1- decoupled. I think in order to optimize this.
2 MR. MORTON: To couple them is to be redundant.
3 MS. BROWN: No, that's not true. I think - ' {
4 MR. MORTON: Because -- because we had -- we had a i 5 description of a' process until we read 1549. Wehadjhe-6 description of a process that looked at, from that 7 perspective,'would satisfy the rule. In short, if you 8 derive a DCGL, get it approved, go cleanup to that DCGL, do 9 the survey, final status survey, evaluate those survey data 10 with the MARSSIM-type statistics and draw the conclusions 11 that it draws, that's -- basically that's sufficient
. 12 demonstration according to that process.
13 MS. BROWN: And I'm saying they have to be coupled 14 because somehow you have to derive that DCGL in the first 15 place.
16 MR. MORTON: That's correct.
17 MS. BROWN: And that's why --
18 MR. MORTON: You derive that before the 19 decontaminate, so that you have a decontamination goal.
20 Some licensees will prefer that.
1 21 But that's not the 1549 approach. . -
' l 22 MS. BROWN: I think it is. I
- 23 MR. MORTON
- In the 1549 approach you basically 24- are led through a decision process analyzing options, I 25- deciding whether or not we gather more data to demonstrate i
1 l
~. ANN RILEY & ASSOCIATES', LTD.
Court Reporters l 1250 I Street, N.W., Suite 300 I i
Washington, D.C. 20005 (202) 842-0034 I i
263 i so that we can comply, whether we excavate, whether we do i i
2 both, whatever. And we, the licensee, should be able to, at j 3 that point, select a goal. Or in Item 10, select the 4
preferred option and then go implement that option just 5 following the process through. Then go design a survey 6 according to the MARSSIM process, design a final status i
- 7 survey, go collect the data, and then at that point I 8
presumably -- presumably you would take the data you have, l 9 generate the source term for DandD or whatever program is 10 approved for that case to put those data into the program, !
11 follow this decision tree down, and for those data show that !
l 12 you meant 25 millirem. !
And that shoulil be sufficient i 13 demonstration of compliance. !
14 MS. BROWN: The data themselves are concentration 15 data and you have to have the model, again, to demonstrate 16 that you meet the 25, millirein. This is why I'm saying --
17 MR. MORTON: Thac's what I said.
- 18 MS. BROWN: -- that they're intertwined, and it's i 19 part of the process.
20 MR. MORTON: You collect the data, generate your )
21 source term, you feed it into the DandD code -- , l 22 MS. BROWN: Or any code.
23 MR. MORTON: -- or whatever is approved, you 24- calculate the dose and you compare that with 25 millirem.
, 25 MS. BROWN: I think we may be saying the same ANN RILEY & ASSOCIATES, LTD.
Court Reporters 1250 I Street, N.W., Suite 300 Washington, D.C. 20005 (202) 842-0034
264 1 ~ thing.
e
.2 MR.-MORTON: Sure, I think we are. What I 3' interpret to be occurring is that there's one group who is 4 looking at it from the 1549 viewpoint. And then we have
'5 - seen'an other representation that's looking at it from 6 basically the historical way, the momentum or the' amount of 7 ' attention that'the MARSSIM approach has had. And that's a 8 different process. And each is sufficient to demonstrate
- 9. compliance. You don't have to heap the two on each other, 10 each should be acceptable, each' permits flexibility, but 11 they're not the same.
12 MS. DAILY: Okay. There-is something --
13 MR. MORTON: They're not quite same.
14 MS. DAILY: -- that I don't quite understand, 15 Henry. When you're demonstrating compliance, you're 16 ' demonstrating compliance with a dose that you calculated --
17- I mean, you've calculated a. concentration that results in 18' that dose or that value that's comparable to the-dose '
19 criteria. And that's the value that you're using when 20 'you're designing your survey; right?
21 MR. MORTON:
Which process? 4 .
22 In the MARSSIM process we would, before the fact, L23 and I say the fact, the fact of decontaminating and doing-a 124' final status. survey in what we'll call the MARSSIM 1505,
- 25 .1509 process. We would come in with the methodology, the v
1 i
4 ANN-RILEY'&' ASSOCIATES, LTD.
Court Reporters.
1250 I Street,.N.W., Suite 300 Washington,;D.C. 20005 (202) 842-0034
265 l
1 calculations and the DCGLs for approval.
)
i 2 MS. DAILY: But in MARSSIM, where do you get the 3 DCGL? Don't you get it from your --
4 MR. MORTON: Of course, you calculate it. You 5
would calculate it with the same dose codes that it yould be '
6 approved, the same codes, the same parameters, the same 7 everything. The point is, this is done before the fact and 8 you get approval for those concentrations, those 9 concentrations using this methodology and these values of 10 the parameters before you go excavate, before you do your 11 final status survey. Then you use the statistics to test i l 12 the survey data to see if you met that, the DCGL criteria.
13 That's what I interpret in that process. I 14 But in the 1549 process, one by this process would 15 -- would go through the process, set goals for themselves
! 16 for cleanup, then go do the decontamination and go do the 17 final status survey and then take the final status survey 18 data to generate the source term. It' then feeds .zul approved 19 i
computer code model, parameter values, put those together at 20
~
that point after the final status survey to see if he then 21 meets the 25 millirem criteria. If so, demonstration of -
22 compliance. Two different procesces. Each should be 23 acceptable, but they should not be piled one on top of the 24 other and a licensee requests to comply with both !.
25 requirements or all' aspects of both requirements.
I i
e ANN RILEY & ASSOCIATES, LTD.
. Court Reporters 1250 I Street, N.W., Suite 300 Washington, D.C. 20005 (202) 842-0034
266 1 MS. TROTTIER: I think what this demonstrates'is 2 we have not done a very good job of explaining our 3 methodology. So, it's something that we need to go back to 4 do. 1 l
The Reg Guide is in the order of the work we expect to i 5 be done. Dose modeling is first. I Final status survey comes 6 later. We will make sure that we get our message across.
7 It's obviously that at least some people in here have not '
f 8
gotten the message that we thought we were sending. We are 9 not offering up two processes. One does one thing, and one 10 does the other, and it's the way we're communicating 11 obviously that's at fault here and we will fix that.
12 MR. MORTON: In the second process it is clear 13 that I would do the dose assessment and it is iterative.
14 That's quite clear. It's the compliance --
15 MR. POTTER: Before Henry made his comments --
16 this is Tom Potter,,I'm a consultant. Most of.my clients 17 are NRC licensees.
18 Before Henry made his comments I felt almost
! 19 obliged to make a comment for fear that you might to away l
20 this afternoon thinking that the only people you had 21 problems with were your sister agencies or your contractors. .
22 We wouldn't want that to happen.
23 MS. BROWN: Thank you.
24 MR. POTTER: But I'd like to describe -- I'd like 25 l to take a broader view of this and describe a little bit of i
ANN RILEY & ASSOCIATES, LTD.
Court Reporters 1250'I Street, N.W., Suite 300 Washington, D.C. 20005 l (202) 842-0034
I 267 1 a dilemma the way I see it. With the two-phased approach 2
and I don't think anybody has any argument with a two-phased 3 approach, I think that's a very neat way to do it. But with
! 4 respect to guidance for the two phases, for Phase 1, it's 5 quite a' simple process and the guidance is simple.
Here's a
6 table, look up your number, if you go, good.
7 If you have a more complex situation'the guidance 8
is you can be really very creative with about the only 9 constraints being that you've got to satisfy yourself and i
10 your project reviewer that the scenarios that you've 11 described and the parameters values you've- selected are 12 within the bounds of reason, well within the bounds of 13 reason, let's say. That's not real specific guidance. And 14 most of the licensees that are really going to have problems-15 in decommissioning definitely fall into that. Actually
~
1 16 you've put us there .:Ln the NUREG-1549. Basically for'the 17 uranium and thorium people you've basically outlawed option 18 1 of Phase 1.
19 And I'm a little bit nervous about hanging out c 20 there with such little guidance on the creative aspect of 21 things. I'm more nervous, for example, when I see the . .
22 illustrative example that was shown this morning where the 23- guy went out in the field and collected a bunch of data, l 24 several dozen numbers and then wound up using the worst of 25 them in his evaluation to get his -- to do his dose ANN RII7Y & ASSOCIATES', LTD.
,ourt Reporters 1250'I Street, N.W., Suite 300 .
Washington, D.C. 20005 (202) 842-0034 j
'l
268 1 calculation. A project reviewer is -- given that he doesn't 2 have much other guidance is going to look at that example 3 and we're going to have some battles over that.
4 So I'm nervous about the ambiguities out there' on 5 the creative end of things, although I have.to say, the ,
6 feelings are mixed because it's a definite advantage to have 7 the -- at least the theoretical capacity to be creative.
~
8 One thing that would make me a 1ittle less nervous 9 would be a sense that the cutoff between the two, and that 10 is to say the table, be set -- I think you've set those 11 values or you've been inclined to set'those values at a 12 level that makes you folks feel real comfortable. It would 13 be advantageous, I think, to push those numbers up a little 14 bit to the extent -- to the point that people who get pushed 15 into Phase 2 really out to be there. And that may not be 16 possible with uraniugt and thorium, we can't tell really, you 17 know, based on what we've got to work with so far.
18 So, that's my comment.
19 MS, DAILY: Okay. Let me give you a two-part 20 response. First regarding the example that we showed with 21 the uranium, one of the interesting things that we were ,,
22 hoping to point out is, in selecting a value to represent 23 your Kd, for example, of course different radionuclides have 24 different sensitivities to that parameter. And even a 25 radionuclide that is sensitive to that parameter is i
ANN RILEY & ASSOCIATES, LTD.
Court Reporters 1250 I Street, N.W., Suite 300 l Washington, D.C. 20005 (202) 842-0034 l
l l
l 1
269 1 sensitive over a certain range. And what we wanted to show 2 was, if you had used the mean of that distribution you would 3
have gotten.the same number as you got by choosing the 4 minimum value that you measured.
5 MR. POTTER: I didn't pick that up.
6- MS. DAILY: Yeah,andI'msorry,weshouldbave 7 made the.t clearer. When you look at the information that 8
generated on a radionuclide-specific basis, you can get some 9 information about where that sencitivity really starts l
10 becoming important. And for uranium, it looks like when the 11 Kd gets to one or less, that's when you run into some 12 difficulties. And that also has an effect on the question 13 of where you would set that level of conservatism.
14 Obviously there's impacts there.
15 In terms of limiting reasonable ranges for 16 replacement values for parameters, I think that's a good l 17 comment, we need to make it clear what the range is that 18 would be reasonable for people to suggest alternate values 19 within the bounds of this particular analysis. It's going 4 20 to be -- it may be different for different models. And that 21 needs to be looked at. But for this specific model, when ..
22 you look at the way we have developed those distributions 23 that gives you information about what would be a valid range 24 for that parameter as well as what the impacts of getting l 25 information across that range might be.
l ANN RILEY & ASSOCIATES! LTD.
Court Reporters
- 1250 I Street, N.W., Suite 300 i Washington, D.C. 20005 (202) 842-0034 l
l t
270 1 MR. POTTER: Yeah, I think my question had more to 2 -- or my comment was more broadly addressed. It was really 3 how to get the acceptability for creative approaches over --
\
4 actually, the use of the guidance for NRC staff as well as 5 for the licensees. That is to say, how to foster creativity 6 in a punch-list world.
, 7 MS. BROWN: I think you use dollars to do that. I 8 i You say you will spend less money if you look at all your 9 options and use the option that,gives you the greatest 10 change in dose -- in simulated dose, that's still defensible 11 for the least amount of money. And that's the impetus'for 12 change.
13 MR. POTTER: Well', we certainly feel that impetus.
14 (Laughter.]
15 MR. NARDI: Joseph Nardi with Westinghouse. I 16 would like to endorse Henry's comments, but I'd also like to 17 add one other factor,. discussion of what the sequence of 18 events was. He did not discuss the characterization phase, i
19 And that phase, if you put that in front of everything else, i 20 cannot be done really adequately unless you have a very l 21 clear concept of what kind of levels you're shooting for. !
I 22 If you take your characterization phase to one guidance '
23 level and find out later that the guidance level should have !
j 24 been something lower than that, you'll find out that you 25 didn't sample enough ground, or you didn't core deep enough, i l
ANN RILEY & ASSOCIATES,' LTD. I Court Reporters 1250 I Street, N.W., Suite 300 Washington, D.C. 20005 (202) 842-0034
271 1 _ whatever you're going on. And your characterization phase 2
really then would be totally inadequate to define what you 3 need to do for decommissioning.
4 Now, granted, that is for the more complex 5 situations, not the simpler ones. But you cannot start this 6
process on any complex decommissioning project unless you 7
know up front what your guidance level ad DCGLs are, 8 otherwise you're shooting blind. You're going to get 9 yourself into a box that you find out later was later was 10 totally --
11 MR. SAITO: Earl Saito, Combustion Engineering.
12 I'm kind of following the same point that Henry and Joe made 13 here. The way I am viewing this process now and we're 14 currently trying to perform it, is the first thing we're 15 going out and supplying hydrogeologic data so that we can 16 then use the codes t,o then figure out what the.DCGL is going 17 to be. I don't see how you can do anything until you take 18 that first step of having some input parameters to encode.
19 And the people at the NRC have agreed with that, the actual 20 measurement of what the residual concentration is, is as far 21 away from that first step as I think you can get. I think 1 22 there is a lot of confusion here as to what we need to 23 measure to run a code.
24 Am I missing something in that we need to know
- 25 residual concentrations to run this code?
l I thought this l l
ANN RILEY & ASSOCIATES, LTD.
Court Reporters 1250'I Street, N.W., Suite 300 Washington, D.C. 20005 (202) 842-0034
272 1
code was going to tell you what the residual concentrations 2
were that gave_you a certain dose, not that you needed to 3
know residual concent.- stions in order to run the code.
4 MS. BROWN: You need to be able to estimate your 5- source term at some point during this process in order to %
6 demonstrate that you meet 25 millirems.
7 You don't have to have any source term to run the 8 code. I mean, you can just use the unit concentration if 9 you know your isotopes, for example. And derive your own 10 dose conversion factors, for example, if you have 11 site-specific information and put this into the model.
12 There are a number of ways to start this process.
13 I would suggest that you start this process with
- 14. an estimate.of what your source term could be. If you can 15 bound it in any way, do that. And I'm not sure if we have l
16 enough guidance in 1,549 to tell you the different methods of 17 starting this process with different levels of knowledge 18 about what the source term might be and so I think that's a 19 valid point.
20 MR. SAITO: Is that as important, though, as 21 knowing the hydrogeologic inputs? ..
22 MS. BROWN: It depends on the isotope. On what
> 23 the major pathway is. I mean, if it's a direct exposure or 24
. gamma, maybe the groundwater pathway isn't the pathway you 25 really need to be concerned about in the first place.
ANN RILEY & ASSOCIATES, LTD. '
Court Reporters 1250'I Street, N.W., Suite 300 Washington, D C. 20005 (202) 842-0034
273 1 MR. SAITO: Well, it's uranium and the groundwater 2
pathway would be the dominant pathway and that's where I'm 3 putting my efforts in.
4 MS. BROWN: Right. I would look very careful at 4
5 absorption processes along that pathway just based on what I !
6 know now from running the models and looking at 7 sensitivities. That's where I'd put some of ray effort, I '
8 think. But, you know, I think you should run the modelt and 9 look at the sensitivities analyses or just look at the i 10 sensitivicy analysis for your particular isotopes to help 11 prioritize which parameter values are going to make a 12 difference. If you're out collecting data and spending a l
13 lot of money collecting data fo'r parameter values that the 14 model is not sensitive to, then in my mind that's a wasted 15 effort.
16 MR. SAITO: , Yeah, and that's where we're not 17 getting guidance. I mean, we're going down this path and 18 we're, you know, you're not doing it in conjunction with NRC l j
19 reviews. We think we're on the right path, but listening to 20 this I was getting -- I had not had the warm fuzzy feeling 21 that I knew what was going on.again. So it's one of things -
22 that I think we need more guidance on that front end of how 23 to -- what do we need for inputs to these codes and how are 24 we supposed to gather that information. What would be 25 acceptable?
1 ANN RILEY & ASSOCIATES, L'TD .
Court Reporters i i
1250'I Street., N.W., Suite 300 l Washington, D.C. 20005 (202; 842-0034 l
L
(
l 274 '
1 'MS. DAILY: I think from what you said and what I 2
know a little bit about your situation, I'm going to assume 3 i that you've essentially already gone through the process of '
4 you have a general idea of what your source term actually 5
is, you know that you're not going to pass screening, you 6
know that the model is not especially useful for you in i 7
terms of the groundwater portion especially. !
8 You've already got that information. You have 9 that knowledge already. And to our mind when you had that 10 information you would be moving through the decision 11 process, you would be incorporating that information in your 12 decisionmaking. I And since you have uranium and we have !
13 information about the sensitivity analysis for that l 14 !
radionuclide we know that it tends to -- the groundwater 15 pathway is important for that radionuclide, then the 16 !
response is, okay, then would it be useful to me to develop !
17 a better representaticn of the groundwater pathways at my l 18 site for this radionuclide and what should I go looking for l l
19 specifically. And you know that there are certain costs 20 associated with getting different kinds of information ad 21 you bring that into your evaluation of options and you've -
22 obviously decided that it is worthwhile in your particular 23 situation to go out and gather site-specific information to 24 help you better represent that pathway. I think that's 25- exactly what we're talking about you doing.
I ANN R'ILEY & ASSOCIATES, LTD.
Court Reporters 1250'I Street, N.W., Suite 300 Washington, D.C. 20005 (202) 842-0034 l
r-275 1 MR.'SAITO: Well, because in the past we were told f :2 if we didn't have site-specific information that we didn't 3 have any choices. I mean, we're stuck with the defaults.
4 MS. DAILY:
We're in a transitional time and it's 5 very difficult for everybody. There's things that people 6
have traditionally.done up until now, there are things that 7
we're trying to do go get ready for this implementing this 8 new rule. And you're in the -- as you know, the' interim 9
time for development of this guidance and getting it tested.
10 MR. SAITO: Okay. Just one other quick question.
L 11 I don't know how much discontent this is going to stir, but 12 from this morning's conversation I still haven't come to a 13
' clear understanding, is RESRAD an acceptable code with the 14
-- as a code itself, but then all the input parameters have to be justified? Is that where we're at now, or as a 16 licensee, if i decide to use RESRAd, am I still going to 17
-have to come in and justify the code and then hang all the 18 parameters on?
19 MS. DAILY: You have to justify the code' insofar
- 20 _as the same as you would for DandD say that-it is 21
_ encompassing the pathways that are of importance to your. .
22' site, that it represents your' site in a reasonable way and 23 in terms of the parameters, that the parameters that you're 24- choosing to put in there have a defensible basis, either 25 you've done site-specific data gathering, or regional data, ANN RILEY & ASSOCIATES, LTD.
Court Reporters 1250'I Street, N.W., Suite 300 Washington, D.C. 20005 (202) 842-0034 l
,t '
(
276 1
or something that' demonstrate that those are apprcpriate for 2 your site.
3 We're not talking about an intense research
-4 proj ect, we're talking about reasonable assurance that these I
5 are appropriate for your application. ,.
! 6 MR. SAITO: Okay.
t So quality assurance and the 7- source code and it's doing what it.says it's doing, that's 8 nothing that we have to -- as a licensee would have to-l 9 defend?
10 MS. DAILY: Right now I can't tell you exactly
! 11 what kind of paperwork the licensing' office is going to l
12 require, so I can't respond to you right now on that i
[ . 13 ' specific question. But.obviously we have a lot of 14 information about RESRAD at the moment.
t It has been used in y 15 the past, so it's -- you're not' starting from scratch.
{ .
l 16 MR. SAITO:. Okay. That's a lot more -
- an answer 17 I'm much' happier with than what I heard in November. It's a l
18 very scary prospect to me. Thank you.
'19 MR. WILLIAMS: This is Alexander Williams'in DOE.
20 During our lunch period I did.look at'page D-6'and-D-7 of L 21' .your' document and there seems to be.some misunderstandings . . . ~
22 because asLI read.yourfdocument what~would be required if
'someone comes in with another code would be all of Section-B 24 at.the bottom of page D-6 which includes-defending the code, 25 defending:the input parameters, defending the modeling ANN RILEY & ASSOCIATES; LTD.
. Court Reporters ;
.1250-I. Street, N.W., Suite 300 i Washington, D.C. 20005 ;
+
i-(202); 842-0034 '
i 277 1 assumptions, providing NRC with a copy of the code, the 2 user's manual, output file, the input file, then there's a 3 i' note at the bottom saying that there also has to be QA/QC 4 for the code used. That's the note at the bottom of page 5 D-7. !
And in the case of RESRAD there's -- you know hg,is is 6 the sort of thing that was our rather lengthy discussion 7 this morning. And what you seem to saying to me doesn't 1
(
8 square with what I heard back in November and it doesn't \
9 i seem to square with what's written on these particular pages l
(
10 either. And I just thought I would point this out.
Am I i l
11 missing something or is there some misunderstanding on my 12 part? Or is there something that I'm not picking up?
1 13 Sometimes I'm slow on the uptake. I 14 MS . DAIL'_': I think it's saying the same thing I
15 that I have been saying. You have to demonstrate that the 16 assumptions are appropriate for your site. It-characterizes 17 your site correctly. In the case of demonstrating that 18 there's a appropriate QA/QC we've talked about that before.
19 I think for RESRAD we have information. We have information 20 from the November workshop. We have information that you've 21 provided to us. -
22 MR. WILLIAMS: That's true. i 23 MS. DAILY: The point of QA -- I mean, this I
24 guidance has to cover all of the models that somebody might 25 bring in. So what we're ';rying to do is give information l
l l
ANN RILEY & ASSOCIATES,' LTD.
Court Reporters 1250 I Street, N.W., Suite 300 l 3
Washington, D.C. 20005 i (202) 842-0034 l l
1 1
l 278 1 for more than just RESRAD here.
2 MR. WILLIAMS: Oh, I fully understand that. But
, 3 RESRAD is the only other that is named by name here in the 4 middle of page D-6 and so I keep getting all sorts of 5
different signals and different messages and it's very 6 confusing to me when one code is named and then you are on 7 the hook to demonstrate that the implicit assumptions 8
. associated with the code, in other words, someone has to go 9 back through the RESRAD manual and show that that meets or i i
10 matches or how it affects the NRC requirements. And then go 11 through the default parameters and defend those if they're l 12 used, or if you use some other parameters, defend those if 13 they're used. And then defend the model assumptions and 14 then provide a copy of the code and then provide a QA i 15 package of some sort.
16 I mentioned this morning that, you know, there 1 '
17 seemed to be one set of requirements for DandD and another 18 set of requirements for everyone else. And I'm sort of i
19 wondering as to whether I didn't hit the nail on the head at 20 that particular time.
1 21 MS. DAILY: I don't think so.
22 MR. WILLIAMS: Okay. Well, what is in your 23' do~cument though, I don't see it as being entirely. consistent ~
! 24 with what is being said at this meeting.
- 25 MS. DAILY
- We'll look at the way that it's i
9
. ANN RILEY & ASSOCIATES, LTD.
Court Reporters 1250 I Street, N.W., Suite 300 Washington, D.C. 20005 (202) 842-0034
279 1 written.and make sure it get clarified.
. 2 MR. WILLIAMS: Okay. I would appreciate that.
3 And if you would like some help in rewriting it, please feel 4 free to give me a' call.
5 MS. TROTTIER: I'm going to make a manageme,nt 6 decision right now. We're just going to remove the EG 7 RESRAD and that will solve the problem, I think.
8 MR. WILLIAMS: That doesn't settle the issue of 9 people coming in with an approval package or whatever.
'10 Which is -- we went over previously.
11 MS. TROTTIER: We cannot speak for the licensing
- 12. office. It's a licensing office decision. If they choose 13 to endorse the use of RESRAD, they may do so. They can do 14 it in a forum other than this. There are lots of vehicles 15 that the licensing office has for communication. They may 16 choose to tell us to,put this in here that RESRAD is an i
17 approved code. They have not done that at this point. So 18 until that's done we cannot tell the licensing office what 19 to do.
20 We are developing a tool for the licensing office
'21 to use.
And we will make them aware of this concern and .
22 hopefully it will get resolved. The guide is not being 23 published tomorrow.
- 24 MR. COOPER: Hi, my name is Bill Cooper, I'm from l
25 the Oyster Creek Station. The longer I. listen to this l ANN RILEY & ASSOCIATES, LTD.
Court Reporters i 1250 I Street, N.W., Suite 300 Washington, D.C. 20005 ,
(202) 842-0034 l
280 1 conversation -- this meeting, the less I understand what I'm i
'2~
supposed to do. It's my opinion that when I begin my 3 decommissioning planning I should be able to sit down with a 4
computer code and the geometry of my station and come up i 5 with surface concentration limits, DCGLs from a compu,ter 6 code with no knowledge whatsoever of what my current source 7 term is and be able to plan the decommission to that status 8 without having to do any feedback back into another -- into 9 the computer code to do it again. I should not have to have 10 extensive information on my source term and feed that back 11 into the computer code.
12 It seems to me the purpose of a decommissioning, 13 dose-assessment, computer code that I have to have to design 14 my surveys and to design my decommissioning would be to a 15 priori tell me what I have to decommission to using the 16 geometry.of my station. Right now I do know a. lot because 17 I'm a licensee and I have to take surveys. But besides 18 that, I should be able to know, if I could sit down wii.h 19 DandD tcmorrow, I should be able to know what surface 20 contamination I can leave in the building. I should kn)w 21 what soil contamination I can leave or not leave, without ,
22 having to have to go characterize the station first.
23 Characterization doesn't tell me what produced 25 millirem.
24 Computer code is telling me what produced 25 millirem and 25 that's all I should need to do.
ANN RILEY & ASSOCIATES, LTD.
Court Reporters 1250'I Street, N.W., Suite 300 Washington, D.C. 20005 (202) 842-0034 l
i
l l-281 1
( I think this whole comparative thing where I do a 2
dose assessment with the computer code and then decommission 3
and then do another~ dose assessment'and then do some more 4
decommission and then do another dose assessment doesnt 5- make any. sense to me.
You know, the computer code I ,should 6
have to run once with no knowledge other than physical 7
parameters, chemical par meters and the design of my 8 station.
l i
But I keep hearing 1that you need to know source 9
. term to feed back into the code and all that' sort of thing..
10 Either that's not a correct statement or we've designed the {
11 process-wrong.
12 MS. DAILY: Okay. I think I have a semantics l
13 question for you. When you say that you don't have to know 14 p
your source termL flo you still assume that you have to know 15 what radionuclides you're dealing with?
l 4
l 16 MR. COOPER; Yes.
17 .
MS. DAILY: Okay. See, that's what I call a 18 source term --
19 MM. COOPER: A source term is activity.
20 MS.. DAILY:
r is the radionuclides. That's part l 21 of it. -
I j 22L MR. COOPER: Right.
j 23 MS. DAILY: But the activity is based on --
E 24 MR. COOPER: You've said that you need to know the 25' l- source term and the source term is nuclide and activity. ;
L 1 l
ANN RILEY & ASSOCIATES, LTD.
Court Reporters 1250 I Street, N.W., Suite 300 Washington, D.C. 20005 (202) 842-0034
282 1 And that once you know that, having run che code once, 2 started your survey design and your decommissioning and then 3 you have to feed that back into the code again. And I don't 4 see why you should have to do that. The code -- I should be 5
able to sit down and say, okay, I'manuclearpowerspation 6 that's been shut down for three years, I'm not yet, but I 7 will be some day. And, therefore, I don't have any I-131 8 and I don't have any Iodine-125 and go down a list of 9 isotopes that I don't need to fool with. Then on a 10 one-isotope-at-a-time basis or however I feel like doing it, 11 saying how much cesium-137 can be on the surfaces at this 12 station when I'm done; how much cobalt-60 can be on the 13 surfaces of this station when I'm done; and if I have both 14 then I do a sum of the fractions and that should give.me a ,
1 15 number. And i don't have to know what the mix is, what the I
16 ratios are, or what the absolute activity is when I start 17 that process.
18 But we've repeatedly heard today that you do have 1
19 to know that and I don't understand why.
20 MS. DAILY: I think our process would allow you to 21 take the approach you are talking about. You would have to -
22 know what radionuclides you are going to have remaining at 23 your site obviously so that you can know what it is you're 24 trying to deal with and what will be causing any dose -- any l
i 25 result in dose from residual radioactivity. You don't ANN RILEY & ASSOCIATES, LTD. i Court Reporters !
t 1250 I Street, N.W., Suite 300 I Washington, D.C. 20005 j
(202) 842-0034
~
i l
283 1, necessarily have to know when you release if you don't want 2 to get this information exactly what the ratios are between 3 the radionuclides exact activity levels of the different 4 radionuclides. You do need to know what the radionuclides 1
l 5' themselves are.
u 6 You can, if you want, and this is a more j' 7 _ conservative approach, it doesn't give you as many options, 8 but you can certainly say that I have cobalt and_I have 9 cesium, take unit concentrations of the -- based on unit 10 concentrations of those you look at the screening 11' concentration tables, you do your mixture rule because you 12 have some of both of them and you will have concentrations 13 that you can leave in order to meet 25, 14 MR. COOPER: I'm also confident that -- at least 15 personal opinion, that no nuclear power station is going to' ,
16 use those default tables because it isn't going to be 17 particularly useful. So the document and the discussion 18 that we've had earlier where it's got this feedback and 19 iteration and all that is not required.
20 MS. DAILY: You don't have to iterate, you don't 21 have to try to make things efficient, you don't have_to go -
22_ for an optimization in your approach if you don't want to.
23- MR. COOPER: Well, optimization I full intend and 24 expect to do. I just don't know why you need source tera 25 information t'o do -- to derive your DCGL. That's a L
ANN RILEY & ASSOCIATES, LTD. _
Court Reporters 1250 I Street, N.W., Suite 300 l Washington,-D.C. 20005 (202) 842-0034
I l
284 l 1 before-the-fact thing that you need to do the rest of the I
2 process.
3 MS. DAILY: I think if you think of your source 4 term in terms of simply identifying all of the radionuclides 5 rather than source term as identifying all of the 6 radionuclides plus their associated activity it might 7 simplify this.
8 MR. COOPER: Okay. A second comment'is on the 9
parameters and the necessity of validating and justifying 10 parameters that you use, there are no published -- these are 11 acceptable median-ranged default parameters that you can i 12 use. The example that I refer to is Reg Guide 1.109. That 13 provides you a set of these -- if-you use these parameters, 14 these.are acceptable to the NRC. We don't have that. We've 15 got a list of highly conservative default concentration l 16 limits that are probgbly not going to be particularly useful 17 to at ler. ; a nudoer of licensees, but you don't have these 18 are default parameters that are reasonable and usable that 19 we don't have to -- we don't have to justify.
20 Because Reg Guide 1.109 says you can change these 21 parameters if you want to, but I sure don't know anybody
~
22 that has. You know, 10 CFR 20 says you can change the DAQs 23 if you want to, but I don't know anybody that has. And 24 we'll wind up in the same situation here that the research 25 project required to change from the defaults are going to be l
} ANN RILEY & ASSOCIATES, LTD.
Court Reporters 1250 I Street, N.W., Suite 300 Washington, D.C. 20005 (202) 842-0034
i l
I 285 l 1 very unlikely to be done unless we have -- we should start i
2 off with a published set of consistent -- in a similar j 3 manner to what we've got in 1.109.
l 4 MS. DAILY: Our intent is actually to develop 5 something similar to that.
l w l
6 MR. COOPER: Okay.
i 7 MS. DAILY: We intend to have default parameters 8 when we publish the guidance. We've gone over a couple of 9 the issues this morning that are why we don't have them 10 right now. But we do intend to develop those.
11 I think we are hoping that people can shift from 12 where they have been in past years to realizing that with 13 this dose-based rule a different approach should be 14 supported in terms of being able to change values to do a 15 better assessment of your site as appropriate. And we are 16 trying to incorporate in the guidance as much information as 17 we can to make it clear that it's not a research project to 18 change parameter values, that there are a lot of sources 19 information that would be reasonable to use to support those 20 changes and what kind of information we think you should 21 bring forward, where you can get that information. So in .
22 terms of improving the guidance I think it's clear that we 23 may need to emphasize even more that we expect people to 24 change from default values and what kind of information they 25 should use to do that and where they can go to get that kind ;
I t
ANN RILEY & ASSOCIATES, LTD.
Court Reporters 1250 I Street, N.W., Suite 300 Washington, D.C. 20005 (202) 842-0034 t.
F.
286 1: of information.
'2 MR. YIM: Man Sung Yim, NC State. At'my quick 3 glance'at your Volume 3 to the 5512, volume 3, it seems to
.4: me, I mean, NRC has done a very innovative-work to support 5
industry.. And my feeling is NRC is not getting that ,.
~6) appreciation at all.
p 7 '(Laughter.] ~
8 MS. TROTTIER: What ever gave you that idea?
- 9. (Laughter.]
l' 10 MR. YIM: I looked at Volume.3, I mean, to be, 11- Attachment 1, some of the parameters that it seems to me are H
12 going to be a default value.are much less conservative than- )
13 what we see as our current default-values for example RESRAD 14 or some other things. It seems to me right now peo.ple are R-15 saying DOGL, okay, now you use all this default DCGL values 16 - ,I mean, use default values to' derive this.DOGL'and it's
'. 17- .
based on all'this conservativism and so on. And then you I 18 come up with a value.- Okay. - It's easy that you have a. !
1 19
~ value and you try to decontaminate everything but you're 120' spending just billions of dollars without having any success l 211 or achieving nothing.
x Just make -- I mean, you just made ...
-22 your compliance business easy. But other than that, you_ re !
C23 spending a lot'of. money.for nothing.
24-Now, I-think NRC.is trying to help us that you '
,, 25: ' don't'have,to do that. . We are puttingLin a reasonable-l ANN RILEY & ASSOCIATES, LTD.
, Court Reporters 1250 I Street,:N.W., Suite 300
- Washington, D.C. 20005 (202)'842-0034
, w .
287 1 degree of conservatism, but at the same time it is 2 reasonable enough. Some of the data that I see in the 3 Appendix -- oh, I mean Attachment 1, clearly one of the 4 first kinds of data I am seeing actually. Some of the work l
5, they have done, I guess for the last several months or 6 years, is really a new data that is really useful. It 7 appears to be very useful and if we can accspt it as a
- 8. default, that's going to help the industry a lot, I think.
i 9
l We don't have to exercise all this conservatism. We can 10 play with more reasonable values that we can target for DCGL 1 11 as -- I think we can use the same approach as we've been 12 doing just you say, use this code and use these parameters 13 as default and then you add this DCGL which is much more 14 reasonable and we don't have to spend all this billions of 15 dollars for nothing.
16 I am just anxious, you know, that - I just --
17 don't understand why we don't get this message. We're 18 getting.a lot of benefit we do not appreciate. If you look 19 at the table, I think you will see the difference. !
20 MS..TROTTIER: Thank you, we needed that.
21 [ Laughter.] ,.
22 MR. GOGOLAK: I heard a little confusion a while 23 ago which made me think that it might be worth repeating 24 that in the case where it was mentioned that you try to 25 bound the source term. So, you know, in that case it's ANN RILEY & ASSOCIATES, LTD.
Court Reporters 1250 I Street, N.W., Suite 300 Washington, D.C. 20005 (202) 842-0034
r-288 i l 1 sometimes been used as activity and quantity bounded, in the i
2 printed version and just verbally before. t j
3 And I think that there is a fairly close i 4 responsibility between the cases where one can do that and i 5 !
the cases that were envisioned as being a quick dismi,ss '
6 thing in the MARSSIM process that I described earlier in i 7 Appendix B. '
i 8 I would encourage you to kind of look at that and 9 see if that is a connection point that can be made to avoid l 10 some of the confusion.
11 I think I've heard some confusion on the part of
, 12 some people today thinking that the MARSSIM process somehow 13 insists that everybody always has to have a survey. That's 14 not it. It's just that it was figured, well, this is a l 15 survey manual and those cases that don't involve survey
- 16 really should just be dismissed quickly in an appendix. '
So 17 it's not that -- that's not the case, but I think that you 18 if you try on the other hand to say that your expedited 19 first step where you would maybe not even want to take any 20 survey or measurements is those cases where you could bound 21 the source. If you make that connection it might be -
22 clearer.
23 The~ comment about the DCGLs, I think there were .
24 several sets of DCGLs where I looked at radionuclides, and 25 .if you compared the surface activity limits for many, many ANN RILEY & ASSOCIATES, LTD.
Court Reporters l 1250 I Street, N.W., Suite 300 j Washington, D.C. 20005 (202) 842-0034
L 289 1
i of the radionuclides compared to the old default limits they 2 are indeed higher. And so that may be what the gentleman 3 was just referring to. But there do remain the problem 4 cases and they keep bubbling to the surface and it's the 1
5 uranium and the thorium and the radium. And, again, I would 6 say that if you could use some of the information that you 7 learned by doing all these parametric analyses to, you know,.
8 give some kind of default. Because what I'm really afraid 9 of is that if -- unless it's made clear to people that, you 10 know, look the screening model has a very, very conservative 11 Ks of B, but we know.that in 90 percent of the country it 12 could be the others if there could be some quick way i
13 through.
14 on the other hand, you know, in a related matter j
15 that's on the parameters. But on a related matter I 16 understand that you have to have some guidance for somebody I j
17 who comes in with a homemade, did-it-myself-in-my-garage 18 model, and that's fine. But I do think also, rather than 19 having maybe 100'different licensees all bringing you the 20 same QA package if there are a couple of codes out.there j 21 that are identified and used I think it would be cost .
22 effective if you indicated in what way the use of the code 23 would be acceptable and in what way it wouldn't. If there 24 are things that you know that are in that code that you I 25 don't like, explicitly, and you know about it, I think it !
l l l ANN RILEY & ASSOCIATES, LTD.
Court Reporters 1250 I Street, N.W., Suite 300 Washington, D.C. 20005 (202) 842-0034
290 1
would be useful to share that information so that people 2 would know not to try to justify that. And so it would be 3 -- I think save time all along if you guys could just look 4 at a couple of these codes and decide what use of them is 5 acceptable and what isn't. And then, I mean, I understand 6 you still need the generic. thing, but this other would save l- 7 time for everybody, I think.
! 8 MS. BROWN:
l One of the things that we probably j 9 haven't made terribly clear is that the intent of the L
10 framework is to have the licensees start to work with NRC l 11 very early in this process. And maybe by not telling them 12 about the QA package we force them to down to NMSS and ask 13- and they start this dialogue, I don't know.
-14 MR. WILLIAMS: [Off mic.) -- find out what that 15 package --
16 MS. BROWN:. They'd have to talk to NMSS.
17 MR. MANN: Bruce Mann, Commonwealth Edison. Just 18 to follow up on your' comment, Theresa, and then I have some s
19 other comments.
20 The recommendation throughout 1549 is to consult i 21 .with NRC, get approval for this and that, and I expressed -
l 22 the concern yesterday and I'll express it to this group too.
23 We,'I1 feel, would be much more comfortable if -- in_ support I
24- of.this and perhaps in some other vehicle though that the. ,
25 licensing process be laid out'. It talks about the new I I
ANN RILEY & ASSOCIATES,' LTD. !
L Court Reporters 1250 I Street, N.W., Suite 300
' Washington, D.C. 20005 (202) 842-0034
291 1 decommissioning process, well, it's for those of us in the 2 utility business it's a long, drawn-out process and it's 3 equally important to understand and allow us to plan'for how 4 to interact with the staff. It's fine to say, well, go --
5 that's not my -- that's not in my territory, go talk to 6 licensing.
7 But from our perspective we would really 8 appreciate some guidance on this for yourselves as well as 9 us so we can develop agreements basically so we can get 10 through this process together in an effective manner and 11 give us a little bit of certainty in how to approach it, 12 otherwise we're back to where we have been for the last four 13 or five years in the license determination process. They're 14 all ad hoc continuous negotiations between the licensee and 15 the staff and it continues from day one right through to the 16 end of the final site survey, perhaps that will continue to 17 some extent, I expect that it will. j But it would be nice to j 18 be able to help us with how the process is going to support 19 this review of DCGLs and how you gain approval and what kind 20 of submittals because there's this issue of the docket and 21 the public record and so forth.
22 MS. BROWN: And I would hope in the writeup on the 23 results of the parameter analysis, essentially, that that 24 helps you have something to bring to the table to say, look, l
25 this is what shows up in your model as the sensitive ANN RILEY & ASSOCIATES; LTD.
Court Reporters 1250 I Street, N.W., Suite 300 Washington, D.C. 20005-(202) 842-0034 t
l t
(
292 1
parameter for the isotopes that I'm concerned abo u~t, and 2 this is the direction that I think I should go. It gives 3 you something concrets. It gives you something that NRC has 4
already looked at and reviewed, and it also gives you an 5
idea of how best to spend your money to go out and co31ect 6 data. It also gives you an idea of whether or not you need 7 to look at a certain pathway in more detail and whether that 8 modeling is going to be valuable and let you say 9 straightforward, up-front, look, this model does not 10 represent my site adequately for this specific reason and i 11 this is the area where I think we can get the most 12 information, the best information to provide moving to the 13 next step.
14 And so I think it can be a useful tool in that 15 respect for both sides.
16 MR. MANN: , Oh , I'm not arguing that at all. I'm 17 jus concerned about the communication process between the 18 licensees and the NRC staff. At what stage in the process, 19 lay out the decommissioning timeline on a -- maybe not a 20 real calendar basis, but there are a lot of submittals. A 21 lot of this is spelled out in the regulations, but there are. ..
22 a lot of gaps also. You complete the licensing support and 23 documentation-required to get these approval and resolved to 24 show your compliance. I'll get off of that. I don't need 25 to beat it to death.
l ANN RILEY & ASSOCIATES, LTD.
Court Reporters 1250 I Street, N.W., Suite 300 Washington, D.C. 20005 (202) 842-0034
293
- 1 MS. TROTTIER
l I will say, I think I said this, 2 this morning. My understanding is that NMSS is producing a 3 document, basically a standard format on content which 4 should walk through the process. That is their plan to do 5- that.
6 MR. MANN: Good. I'd like to get back the to the 7 draft Reg Guide and discuss a couple things further. '
I 8
i think Section C, the regulatory position, Chapter 1, I'm l 9 still a little bit uncomfortable in'what I thought I heard 10 this morning. This is about two pages of discussion and L 11; there's just a huge disparity in that regard in the 12 treatment of this so-called " module" with respect to the 13 survey module. And I heard the comment that well, MARSSIM 14 if a multi-agency document and so we felt we needed to have 15 more explanation of MARSSIM in this Reg Guide.
R16 But also in the handout this morning it's clearly 17 identified that 1549 was a technical support document, not 18 the guidance.
L I really am concerned that the regulatory 19 position should be the philosophy, theLpolicy. The key 20
- issues should be addressed here in this document. And 2 11 reference 1549 in a similar fashion that it's been done with. -
22 respect to the survey and the survey support documents. I l
ll -23 heard that -- well, it's in black and white, and if I l 24 understand that, I think that has important implications.
25 This is the policy guidance, this is the place that we ANN RILEY & ASSOCIATES,' LTD.
. Court Reporters 1250 I Street, N.W., Suite 300 Washington, D.C. 20005 (202) 842-0034 i
294 1
should be looking for policy guidance on how to comply with 2 the new rule.
3 Dose modeling is the linchpin of all that follows 4 from our perspective with regard to complex sites. I second 5 Eric Abelquist's comments on that.
6 Another somewhat related issue out there, more 7 kind of a licensing topic, but it's important here. There's 8 a. legacy of staff positions in the public record, varioue 9 forms of documentation, SECY papers, PG-08, if I have the 10 number right that relate to topics that have been dealt with 11 here and discussed, and at some point I would like to see an 12 explanation of what happens to those -- those issues. Now, 13 if we're going to a dose-based criterion for instance and I 14 assume the regulatory guide 1.86 will be put to bed at some 15 point, right now we're in the transition period until 16 August, I guess. But it's not just Reg Guide 1,86, there 17 are other documents out there. And it's quite a mixture of 18 them and if we go back to cases, some of them are letters to 19 licensees that have referred to in draft -- not draft, but 7 20- these GEISs, a whole raft of paper out there that contain 21 criteria of one type or another. -
l 22 It would be very helpful either in the form of a 23' blanket statement that all of that is history, or 24' specifically identify those that are still perhaps relevant 25 or useful. And I don't know if it can be done in this 1
AN'N RILEY & ASSOCIATES, LTD.
Court Reporters l 1250 I Street, N.W., Suite 300 l Washington, D.C. 20005
!. (202) 842-0034 t
L'
295 j 1 regulatory position or not, that may be a little too 2 cumbersome. I think that's an important source of confusion 3 to some of us out there.
4 MS. TROTTIER: I will say that we could easily 5 make that statement. And just for the record Reg Guide 1.86
- 6. will not go away in August. I want to reiterate, I think 7 Steve said this yecterday, his guidance is only for 8 structures and land that remains in place. Materials that 9 leave the site are not covered by this regulation. This is 10 residual radioactivity.
11 So, until or if ever the NRC publishes another 12 rulemaking that addresses clearance of material from site, I 13 presume the staff will continue to use some of the existing 14 ~ guidance, at least that table in Reg Guide 1.86.
15 On these other staff positions that have been 16 issued over the years, it might be worthwhile on the guide 17 to make mention of it. Rules always supersede all other 18 guidance for ever. But we'll make it clear that it does 19 supersede it.
20- MR. MANN: Well, even with respect to your comment 21 'about use of Reg Guide 1.86, I think there's a lot of - i 1
22 disagreement on that out there in the licensee world. And !
23 even in the region, it's the appropriate use of Reg Guide 24 .1.86, Table 1 limits with respect to -- with all due 25 respect, but with respect to clearance of materials from ANN RILEY & ASSOCIATES,' LTD.
Court Reporters 1250 I Street, N.W., Suite.300 Washington, D.C. 20005 (202) 842-0034
i L
L 296 1 sites. Just an aside.
2 Another comment in regards to Section C, Part 1, 3 dose modeling, in the second paragraph just to amplify on l
4 concern about use and derivation of the DCGL,s this 5 paragraph identifies the purpose of 1549 and it addresses 6 quite a number of situations. I haven't read 1549, but I'm 7 asking the question, does it really provide guidance on l
8 derivation of how to derive DCGL,s for all these other 9 circumstances, residual radioactivity, sewers, waste 10 plumbing systems, floor drains, ventilation ducts, imbedded 11 piping and handling special circumstance soil. Because 12 those are the issues that are troublesome to those of us 13 that have sites that have all these circumstances at 14 decommissioning?
15 MS. DAILY: At the moment it doesn't directly in .
16 detail address every, single one of those. There is some 17 mention of those issues and in the -- probably in the 18 interim coming year we're going to be putting in more detail 19 on those specific iss6es. Right now they're handled more in 20 a general sense in terms of your conceptual model would be 21 different for certain things. There's ways of handling . . -
4 22 these within the screening. Some of these it needs to be 23 clarified where we're assuming that 1.86 would apply versus 24 where the dose modeling applies. And that does need to be !
,' 25 clarified. '
l :
I I ANN RILEY & ASSOCIATES', LTD.
. Court Reporters 1250 I Street, N.W., Suite 300 Washington, D.C. 20005 (202) 842-0034 i
297 1 MR. MANN: That would be a big help to me.
l 2 MR. LITTLEPIELD: Pete Littlefield, Duke l
l 3 Engineering. Just to follow up on what Bruce is saying, one :
j 4 of those areas that we all have on our sites is !
5 contamination that's below the surface. That might b,e j 6 easily addressed, you know, in the current guidance and it's 7 a situation that we're all going to have to face in terms of i 8 performing some dose assessment. !
It was noted that the i 9 regulatory guidance for final survey was pointing towards 10 this document at one time as providing that guidance.
11 Now, I don't know if there's something that could 12 be done in the short term that would just help us in terms 13 of providing whether our dilution factor is applicable to j 14 that situation in which case we could use DandD directly 15 with just the addition of a dilution factor for buried 16 soils, but something.like that would be fairly. easy and it 17 could be put into the guidance at this time.
18 MS. DAILY: I don't think that a dilution-factor 19 approach is necessary consistent with what we're trying to !
20 achieve with the dose modeling, but we can try and add more 21 information about how you would handle subsurface -
22 contamination. I think it's probably been mentioned the 23 survey and I'm sorry, but I wasn't here for yesterday's 24 . workshop, but we still need a fair amount of work on 25 developing approaches for surveying for subsurface ANN RILEY & ASSOCIAT5S,' LTD.
Court Reporters 1250 I Street, N.W., Suite 300 Washington, D.C. 20005 (202) 842-0034
298 .
i 1
contamination and we do need to add details about how to
! 2 handle it in the modeling.
3' MR. LITTLEFIELD: But I guess I don't know'what
{
4 the details would be in terms of the dose modeling. If all 5 I'm going to assume is that the contaminated soil beneath 6 the surface is now brought to the surface in tact-so that 7
i it's the same concentration, then_I really don't need any 8 guidance for dose modeling because I can use the dose models 1 9 that are already available to me.
10 MS. DAILY: Yeah, what I'm thinking about it you 11-could assume that the entire source term is in the top 12 layer. That would be a fairly restrictive assumption. Just l
13 take your whole source term --
14 MR. LITTLEFIELD: Right.
15 MS. DAILY: -- up to the-top 15 centimeters.
~
16 MR. LITTLEFIELD: Right.
17 MS. DAILY: And go forward with the screening.
18 MR. LITTLEFIELD: Right.
19- MS, DAILY: And that may be perfectly reasonable 20 for some licensees to do. The model-also allows 21 contamination to move down through deeper layers. And can- .
22- have multiple. layers in the unsaturated zone. And what we 23 need to do is explain to people how that works and whether I 24~ or not you can start with the source distributed in that 25; way.
. I ANN RILEY & ASSOCIATES, LTD.
, Court Reporters 1250 I-Street,_N.W., Suite 300 Washington, D.C. 20005
-(202) 842-0034 L
l 299 1 MR. LITTLEFIELD: Okay.
2 MS. DAILY: There must be more questions.
L 3 Comments?
l 4 MR. SAITO:
t I'm requesting a comment from you on l 5 that is one. Earl Saito, Combustion Engineering. Looking 6 !
at your draft values here, default concentration values, you 7
give the 50th, 75th and 95th percentile, how are we going to '
8 apply these in relevance to the 25 millirem limit? Are we 9 going to go with that 50th percentile, we go with the 95th, 10 and then how does that turn back with survey uncertainties?-
11 That's one thing that the draft guide I expected to find and 12 I haven't found that yet.
13 MS. TROTTIER: That's because that's the h
14 unresolved issue. You're getting this.in the middle of 1 i
15 staff resolving issues. And so it's likely that this is an 16 issue we will give to the Commission to resolve, what !
17 percentile they want to use in these tables. And so this 18 was our interim approach to provide information primarily 19 that we're using in house with a staff review. What we're ;
20- doing is just providing staff drafts. So what you see as a 21 staff draft in the form it's in now. -
22 When this document is published that issue will be 23 resolved. There will not be this level of confusion over 24 which percentile we're using.
25 MR. SAITO: Okay. And then also there will be a I'
ANN RILEY & ASSOCIATES, LTD.
Court Reporters 1250 I Street, N.W., Suite 300 Washington, D.C. 20005 (202) 842-0034
300 1
discussion, I suppose, on how to factor both modeling 2 uncertainty and measurement uncertainty together when 3 demonstrating compliance? That's something that hasn't i
l 4 really --
5 MS. DAILY: I think that's going to be part of our 6 discussions in trying to get a better interface between the 7 dose modeling and the surveying over the coming yeaf. j 8 MR. SAITO: That is an issue that is going to be 9 resolved in the next year?
10 MS. DAILY: It's a very important issus. I think 11 we're going to have to resolve it in the coming year.
12 MR. SAITO: Thank you.
13 MR. CHAPMAN: Greg Chapman, Nuclear Fuel Services. !
14 I've just got a concern I would like to voice regarding 15 those screening values that were in the Tables A-1 and A-2 16 especially as it pertains to sites or portions.of sites that 17 have previously been released. In the rule it states that 18 NRC would come back and only require additional remediation 19 at these sites, if new information showed there to be 20 significant public risk. And if you go back to the SDMP 21 guidance right now for thorium and uranium, at least for . .
22 thorium in soil the criteria would correlate to more than 23 100 millirem per year, the public dose limit, and 24 instructors it would go up above the occupational dose
!' 25 limit. And I'm afraid that any site that has processed l
ANN RILEY & AS50CIATES, LTD.
Court Reporters 1250 I Street, N.W., Suite 300 l
Washington, D.C. 20005 (202) 842-0034 I
f t
r 301
-1 uranium and thorium will automatically have to go back and 2
start reevaluating even if they're closed out and no longer L 3 even in operations.
l 4 MS. DAILY: I think once again we need to make a l 5 clear distinction between screening numbers and what is an 6 actual dose at an actual site. And while I can't answer all 7
l of your question, I can't even address it at this point 8 because that's really something for the Commission to do.
9 Just in terms of looking at numbers in the screening tables i 10 {
and assuming ~that that means that a dose at an existing site i 11 would exceed our dose criterion is invalid. And we need to 12 make that as clear as we possibly can.
13 MR. CHAPMAN: Well, all these sites aren't going 14 to have access to additional evaluation as a part dose 15 model. Like I said, for the most part they're going to go 16 to that table and see if they're an immediate concern.
17 There might be something that NRC has to do to past 18 licenses.
l 19 MS. TROTTIER: Yeah, unfortunately you have the l l
20 wrong group here. We can't make that call. That, again, is 21 a licensing office call. , ,
22 MR. DUVALL: I'm Ken Duvall, DOE. I'm a little 23- surprised that the decision framework for the release of the 24 site would seem to be more appropriate to be located in the 25- implementation --
ANN RILEY & ASSOCIATES' LTD.
~
Court Reporters 1250 I Street, N.W., Suite 300 Washington, D.C. 20005 (202) 842-0034
l 302
- 1 i l
THE COURT REPORTER: Excuse me, pull your l 2 microphone closer. 1 3 MR. DUVALL: I'm speaking of the decision 1 4 framework that's indicated in this module, and it seems more 5
appropriate that this type of information would be developed l 6 and constructed in the implementation guide. !
I 7 MS: DAILY: What do you mean, " implementation 8 guide"?
1 9 MR. DUVALL: The implementation guide --
10 MS. DAILY: You mean the regulatory guide?
11 MR. DUVALL: Yes, the regulatory guide, I'm sorry, 12 the implementation of the rule would -- it would be more l 13 appropriate there to describe a decision framework for 14 releasing the sites which would include all of the 15 components including modeling and also the survey 16 requirements. ,
17 MS. TROTTIER: There wasn't any real reason other 18 than the fact of size. They weren't being put together on 19 the same schedule. During the year, since other people have 20 made comments, we will go back and look and decide whether 1
21 or not to incorporate it and make it all into one document. .
22 I mean, unfortunately the Reg Guide is already too darned 23 long. So, you know, we're now going to have a 200-page 24' regulatory guide by the time we get done. But we will 25 evaluate that. It's not going to happen tomorrow, though.
l ANN RILEY & ASSOCIATES, LTD.
Court Reporters 1250 I Street, N.W., Suite 300 i i
i Washington, D.C. 20005
) (202) 842-0034 i
\
l 303 1 i l MR. DUVALL: One of the problems that was alluded l j l
2 to earlier indicating that there were two processes -- two l l
-3 decision processes in place, the MARSSIM decision process --
4 l
l
.the MARSSIM is not only a guidance for survey conducting, l 5 l
planning and conducting and assessing surveys, but it,also 6 provides guidance for deciding certain sites. So the 7
MARSSIM is a decision framework, it's the decision process l 8 that you go thrcugh and the final decision is to release the 9 site. Here you also have a decision process or decision 10 framework which the final process is to release the site.
11 You have -- essentially as alluded to earlier,-you 12 have two separate processes for determining whether you i
13 release a site, and the question becomes, do you have to -- i 14 in order to meet requirements to release a site, do you have-15 to pass both~ processes. I think that somehow we have to l 16 integrate the two processes so that we have one decision l
17 being made.
18 One of the things that disturbs me as being a l 19 member of the MARSSIM work group we depended heavily on DCGL l 20 values coming in from somewhere outside, that was outside of l l 21 the scope of our work. These DCGL values were required in -
22 the planning, in the conducting, and in the assessment state 23 they were required at all phases of the MARSSIM, DQO, and 24 DQA process. There's a responP1bility out there somewhere 25 for providing this information and also in order to make t .
ANN RILEY & ASSOCIATES,' LTD.
, Court Reporters 1250 I Street, N.W., Suite 300 ;
Washington, D.C. 20005 '
i
!- (202) 842-0034 f
304 1 MARSSIM successful.
2 I was hoping that since NRC is sort of 3 spearheading this approach of incorporating the whole 4 process, not only survey, but all the other aspects of 5 releasing a site that NRC would heavily endorse MARSSIM ask 6 consensus -- interagency consensus document and build on I 7 that platform. It doesn't appear that NRC has built on the 8 platform, it appears that they've developed a separate 1 l
9 platform at which at some time they would have to be --
10 these two platforms would have to be integrated.
11 MS. DAILY: Thank you. Are there any other i
12 questions or comments people would like to make on the
! 13 guidance?
i
. 14 MR. MORTON: Henry Morton. Could you give a bit 15 of explanation about the derivation of the tables in l 16 Appendices A and B, these screening tables? What more 17 specifically I'm interested in knowing is how the Monte I i
18 Carlo or the Batton sampling shell feeds the code to get the i 19- data out? Let's see if I can possibly give a couple of 20 examples of what I'm thinking about. In one process I can 21 see that you might separately and independently of the -
22- deterministic code generate the distributions and the value i l
j '23 of each of the parameters, then go and choose a percentile, 24 let's say the 75th percentile of the value of each of those i
'25 ' parameters. And those become, in effect, the parameters ANN RILEY & ASSOCIATES, LTD.
. Court Reporters 1250 I Street, N.W., Suite 300 Washington, D.C. 20005 (202) 842-0034 i
[
i i
305 !
I that go into the deterministic code to generate the output. '
2
! And whether that would be called the 75th l 3
l l
percentile of the output or the other process basically of !
1 4 having the distributions all fed to the program or the
- 5 process such that-the -
l in effect the 75th percenti%,e of 6
this mixture of distributions generates the 75th percentile, 7 which of these is it?
I 8 MS. BROWN: It's the second. It's the latter of 9 the two that you mentioned. You sample for each one of the 10 parameter values from the distributions, you run many, many 11 simulations with different parameters -- sets of parameter l
12 values. The distribution that we're talking about, the 95th >
13 percentile is of the output with all those different 14 parameters sampled in different combinations for each one of 15 the isotopes.
16 MR. MORTON.: Sampling each parameter distribution {
l 17 essentially simultaneously to feed the code?
18 MS. BROWN: That's right.
19 MR. MORTON: With many iterations of the code?
20 MS. BROWN: That's correct.
21 MR. MORTON: Okay. Thanks. . -
l 22 MR. POTTER: Tom Potter, could I get some I
l 23 clarification on units in the tables in Appendices A and B.
24 Are they on the same basis as the tables in NUREG/CR-5512, (
25 Volume 1? Specifically, for example, when it says millirem ANN RILEY & ASSOCIATES, LTD.
Court Reporters 1250 I Street, N.W., Suite 300 Washington, D.C. 20005 (202) 842-0034 1
l 306 1
l per year per picocurie for gram for a nuclide, it's not per 2 picocurie per gram of the parent nuclide of a chain, but the 3 total activity of the chain?
4 MS. DAILY: In the case where there's no plus C, 5
there's no chain indicated, it's the activity of the parent.
6 And the ingrowth is assumed. So the millirem per year per 7 picocurie per gram is for the peak year. So that includes '
8 -- a picoeurie per gram when you start of the parent.
9 MR. POTTER: Yeah, I'm not talking about the dose, 10 I'm talking about the activity basis.
11 MS. DAILY: Yeah, activity is at the start of the 12 simulation you have a picocurie per gram of the parent 13 unless it says that it's a chain in which case you have a 14 picocurie per gam of total activity of the chain.
15 MR. POTTER: Oh, okay. Right. So it is the same 16 as Volume 1 of NUREG/CR-5512, same basis?
17 MS. DAILY: Yes. And actually the DandD software 18 will' allow you to use either approach. You can use total 19 activity of the entire chain or you can have -- you can look :
20 at it in terms of ingrowth.
21 MR. POTTER: Yeah, I think a footnote to that -
22 effect would be useful just to make sure everylvly is clear 1
23 on it.
l =
24 MS. DAILY: Good point.
25 MR. PETERSON: Earl Peterson, DOE. I'm somewhat I !
l ANN RILEY & ASSOCIATES, LTD.
Court Reporters 1250 I Street, N.W., Suite 300
! Washington, D.C. 20005 (202) 842-0034
f i
i 307 i
f concerned about the use of the Monte Carlo approach simply i
2 because I don't think on the basis of the data that we have 3
i on each parameter that we really know the distributions very 4 well. And, hence, I'm concerned because the codes will take 5 three or four points, fit them to say distribution and then 6
generate your percentiles, your means and everything else 7
when in fact the output is ridiculous particul'arly if you're 8 going to try to regulate on a percentile basis such as 9 you're talking about.
10 MS. DAILY: I think it would help if you had some 11 time to look at the Attachment 1 to 1549 where we tried to 12 put in how we developed those parameters, what we used as 13 information sources because we were concerned about the same 14 thing.
And as far as we could we have dealt with that.
15 MR. PETERSON: And the second thing is that most 16 of those variables are not random variables they're 17 determined by environmental conditions. And if we knew what 18 the environmental conditions were and could segregate the 19 parameters we could reduce the overall uncertainty quite a 20 bit.
21 MS, DAILY: And I agree. That was what I was ..
22 trying to get at when I said it might be more reasonable to 23 have a regional approach to the screening, it would take l 24 care of that concern.
l 25 Do you have any more comments or questions?
l l
7001 RILEY & ASSOCIATES, LTD, !
Court Reporters l 1250 I Street, N.W., Suite 300 i Washington, D.C. 20005 (202) 842-0034 L
308 1 (No response.]
2 MS. DAILY: It's not Friday, and you want to leave 3 early? Oh, one more. Okay.
I 4 MS. GOOD: Bev Good, GPU Nuclear. In summary from 5 everything that we've been hearing today, I would like to 6 reiterate that we do appreciate the approach with the 7 rulemaking and the attempt'to provide some flexibility in 8
how we go about doing our business and maybe that has gotten 9 lost in all of the comments that were eing made today. We 10 are stausgling and we understand your time constraints with 11 1549 and other documents. We're struggling to go through 12 them and understand them and try to interpret how we go 13 about implementing them. And therein lies some of the 14 problems that we're talking about, the discussions between 15 MARSSIM and 1549. If it is clear in the Reg Guide that it's 16 ordered in the way that you're supposed to proceed that 17 you're supposed to start with dose modeling, maybe -- I 18 think you mentioned you're going to provide some additional 19 discussion that.this is the approach that you expect us to 20 proceed with.
L 21 And-in fact with that dose modeling we should be~ . .-
22 able to look at the various scenarios that we could use in 23 planning our decommissioning. And we think that in looking 24 at it initially that there is a lot.of flexibility which 25 could allow us to do our decommissionings in a more cost I
ANN RILEY & ASSOCIATES,' LTD.'
Court Reporters 1250 I Street, N.W., Suite 300 Washington, D.C. 20005 (202) 842-0034
309 1 effective manner. But then as we go through we're 2 struggling with the dose modeling and then go through and I l
3 we're going to be doing our site characterization, then the 4 decommissioning or how much decontamination and remediation 1 5 do you do, and then what kind of surveys do you do and then 6 do you get to the end, you know, do you have to do the dose 7 modeling again or do you follow the Reg Guide and it says 8 you've done your ALARA and everything else and so you're ;
i 9 finished after you've done your final release survey. And i
10 we're still struggling with that -- with the implementation 11 of it.
12 But I wanted to at least let you all know that we 13 do appreciate that you are trying to provide some 1' 14 flexibility versus historically the way we used to have to 15 do things. So we appreciate that.
16 MS. DAILY: Thank you. I think it's. clear that we 17 do need to add additional information about how the process 18 is going to be performed and how. things are linked together.
19 We'll be working on that.
20 Any more questions or comments from anyone?
21 [No response.] -
22 MS. DAILY: Okay. I guess we don't need to take a 23 break if we're all talked out here. I really appreciate 24 everybody showing up today. I appreciate any effort you put 25 into review of this.
l ANN RILEY & ASSOCIATES, LTD.
Court Reporters 1250 I Street, N.W., Suite 300 Washington, D.C. 20005 j (202) 842-0034
310 1 And we will be taking comments on this. We 2 definitely want to have feedback from you. So, if you would 3 give us comments, they will get. incorporated either before 4 publication or during the coming year.
5 Thank you very much.
~
6 [Whereupon, at 2:56 p.m., the workshop was 7 concluded.]
8 9
10 11 12 13 14 15 16 17 18 19 20 21 22
- 23 24 25 ANN RILEY & ASSOCIATES, LTD.
Court Reporters 1250'I Street, N.W., Suite 300 Washington, D.C. 20005 (202) 842-0034
r-REPORTER'S CERTIFICATE This is to certify that the attached proceedings before the United States Nuclear Regulatory Commission in the matter of:
NAME OF PROCEEDING: WORKSHOP ON DEMONSTRATING COMPLIANCE WITH RADIOLOGICAL i u
CRITERIA FOR LICENSE TERMINATION -
i L
METHODS TO CONDUCT A FINAL STATUS i'
SURVEY AND DOSE MODELING 1
)
l ' DOCKET NUMBER:
l' PLACE OF PROCEEDING: Rockville, MD were held as herein appears, and that.this is the original i i transcript thereof for the file of the United States Nuclear Regulatery Commissio,n taken by me and thereafter reduced to
! typewriting by me or under the direction of the court reporting company, and that the transcript is a true and l accurate record of the foregoing proceedings.
l - - ,
f\ YO _ - ;W Cynthia D. Thomas Official Reporter Ann Riley & Associates, Ltd.
I
v u s
[
Y s
0 f
_ s e
4 i r 4 [
4 .
p -(g5: 8 r
s t, /
d & 4 -
d C />
5- ( J(
o A A x f 3
W 7_
g e 3 l
.T ia u 5 e ; 4 -
s m- 6 1 9 e c
- t
- y y r X o _
.d E 4 ?
'- S n: W g g
.o )
5 //
g s_s h /r t
e e 0 I o i i - y_ - 3
_ b r Y' f, f / g M m "A A r p c
- u 3( n e f 0 g ,
n o
N - : : : : :
3 :
it e il:
n a e
n l
ia e n
i ia e l:
i a e i
l ia
- l i
e l i
e ii e _
a o m n n
- a n c n a n n h h o m- h o m- h o m- o m-
- m- o m- o m- o .
i - h h h h h .
m r
P E P E P E P E P E P E P E P E P ~
e (
T e
s n
e i
c -
L .
r f
o
- i a s q
_ r s p t
e e r
r -
C i
r d d
n P t l
a c
A s rn T y -
_ i g s T -
o e g.
l o i n j
/
s i
d R
a B u K W n f M h A
/ .
/c -
d ,eg 2
/
e /D u N g
- y. g hin t l h d e m a /
D 4 D f. g .
m .
c,, [ .
-% C t
io
. WM N C -
f / #
ee s
s A fC 5t p_
j g g,
6 m G-mc s e K 9 (L yj d 6
u iaD rl o pd dmn i
t i
n o in B
s u
M N A O g g
h a m
)[ _
L u oa 1 1 2 1 2 1 2 1 2 1 2 1 2 1 2 1 2 ACy N Ngvnre Fiu t
_ WaS Tt r
-. ,ss u 8 oa 9 nt 9mS 1 e t
l
,D a 1 ni n 9
y oF f e
g g m -
T-
,m r pa o '- ' e aot , f uhc r
bsu kd f
/f Q A.
h
( [
er no Me ed e, _
Fo i
L tWC s
E
~4 s
/ _f 3
/
6 o- & h
/
a )w -
ds e:
cp no ah h
1 k
h / C- g y
p u
('
nk ero o , 4 /
m p h t
3 W y
6 )( W
[;
h t
AW e E m
a . - .
$ E <
N 1 2 3 4
- 5. 6 7
8.
_ n m w w p o , .
n ,
m s p f, j F-r
, d.,
- m. . g m -
s y y 3 y g 55g s t c et y 5 s
e r j o
L
(,
7 K k L,
( e a
e t
3 y
g 2 d
r g d 4 s 2 y g 7_
o A g g 3 7
3 y n T
l i
a ;
j 2
6 1 y 7 0
L., m 1
% c p 0-g 3 ;
w+
s m- g 2 6 .
,h g g
d o E
/r g .
y 4 h. )2 / t -
o ,. y h s e
i e (o )7 t
f-o e u o 4 m (t -l:( m b .
y M S 3 t e 3, g
m 3 1 ,
n N
u lc (7 j 4 a c f_
o e l e il:
e l e l e l e l: e e
il e
it a n i
a n a n i
a n i
a n i
a n i
a n ia n a n n o m- o m- o h m- o h m- o h m h o m- h o m- h o m- o h
i h h
_ m P E P E P E P E P E- P E P E P E P .
r G ,
e 3 T t 9 W
e s
y 9 _
n o M g
4 e
c V S, _
f _
A -
f_
i L f y i o
2 M r( Sz v r j 2 s f
o n f J 7 d - t t
a a s o o f h ? o o o x
E c e, v
- i / g e
r s c 2- i - l t e ,
l s g i r n
b
_ r d w t y C d g 4 A s )
] o g g g g n l
a i
c s s J t p 7 4
g M o
o 0
0 r g l
g o
o i e
n s
d s
i n J
}3 dg t
a M f l e ,
% fk i
d u a e e v ,
a B K w hg h t
o .
s R_ . -
o M. g%,, L (,,,
egn hi.
t le h d 2/
e Bu d m
a a
n u Q N y r h
n A
g y pg g
C o
[p C
, P O-s f
I C g e f .
e-gg 8
),
_ t io N 9 ke 2 9 -
i
[p WM ee s
s 8
v 6 I.
A.
b _
k h Q $
M % is.
l e g + f, n
mc u an o s n
{
t R Y <
4 $ _
h m t
)
c .
o m / _
i rid i
s u l G-P F 5 u K o W@ 41 U o o g
i Y. _
C 2o l ._
io pd B / p _
dmn t
u oa . . - . . . . . . . . . . . .
1 1 2 1 2 1 2 1 2 1 2 1 2 1 1 2' .
A Cy Ngvnre Fiu t WaS r _
Tt s s _
_ 8,nt u .
9 oa m 9mS 1 e l t
o . S 6
,D a t 9
1 ni n g oF y
r pa j p; _
g aot j g
uhc r
bsu l
Q t
. o ,
t _
6 ekd For on ,
L
- W g c
- c. g -
y m
t :WC s 'Q A
/
4 w e
).,
?>
i L
e:
cp no f e 1
'g (
M W .
ah g _
ds n 4 % g _
nk ero \ e e y T f
t t
AW { c
(
Q t
e m g e
m a .
6 . . l. 3 K
g f .
N 1 2 3 4 5 6 7 8 _
, m
/ , m e y [f .
p (
c
. s s
'y 56 p
f k , n s
a ,
g A
l_
j 7; g c 0 r d
e r 4 sg ~
Q)g _ /
7 p f
g 9^
y-c r y - -
d 4 ,
y g w /
_ o A
l i
a g
. [s C P~ g fc f, P L
gm 7 5 5 q
g g g gh
- T m-7 y j,, T 9 7
(, g 9 d
s s g M
(,
g 2 p g-E Y gf g7 o 3 7 ,
2 h
t e
/r b
e g gi 3
' )
'6
)
3 n e
o 2 f e Qy M o m
u i
[
mR 3 [g/
0 t 1
1 lf / [g [6 (d k',
n o
N e l:
i e l i
e l i
- /
g'ia i e l: [3e il: e l: :
e l e
. it a n a n a n a n ia n ia n ia n ia n n h o m- o h m- o h m- o m- o m- o n- o m- o m- o h h h h h h
. imr P E- P E P E P e P E P E P E P E P
'e '
V 2 T
e ,
Ae D f f -
3 s 0 n
To 1 o o 2 *u e
c 1
~
y f J C 0 'g i
L b 4 '
9 mL H g gg r 7 o
G n'- 4, f
ia r s s I e}r-h 6 K n
e c 4 e
LN 3 N # O a -
d M
t r i
C r d d
A :
~
7 2
s m @
R J a l
a C 0 7 ' _
i c
g s
s /, <l f ) .
4 4 y l
i d
o o
a is B
e n
u sr s a e
N
%k }C f C
'u J'
e m
a c_ M o f x
(
s v
6 mWJ24 g
R egn 2 N k D m
m C
s I T. FN, 6 (
Q
/.
C /g hi t le
/e m
se & h h fm c i )
/' % 't b- ,-
h it d 4 $
mc WM ee o
s N
a s
s e
a w
s 1
S 3
)A v
fA
[
%d h, s_7 0 1
a 3
E 2
/
/
8 0
- )
v 7
{
c o
C, O(A 't i
u no iaD n
u A( db es c
x 5 0 o B l)f d
~
- rl o pd i is B
u O }
O c.
c 0 r % f b g dmn t
i
_ u oa ACy 1
1 2
1 2
2 1 2 1
2 1
2 1
2 1
2 Ngvnre Fiu t WaS Tt s r
s u 8
9 nt oa
_ 9mS 1 e t
J s f
Da l
e r- :
1 ni n 9 <
h
_ o ~
y oF o n
$0 r pe aot uhc ls bc n t_a M
- C e
r bsu (
ekd For on 9 K z W & '
tWC U
s f^ h A % Y' '
e:
cp no h $d b e c
a ah e N W ds nk h/ D J ( t r
V ero
$ $ L h I
t t D AW e m
a 8 .
g,,
Ye l bI N 9 1 0
. I.
I A l f Y-l
$l
e
, g <,
, e,
_ o 7 _
4 s
w.
c, 5 g 6 ,
A.
L n 2
- u. Q s p b .,,
d e
r 9 T 7
~
i S
c r
t i
- e. g S '
s o r o
c a
k >
f 5 _
- d o 7
~ n 2 c y o 3 e - _
A t z
2 / - u 2 g
T o l i
a a ( s 7
g 2 9 ,
, y 5 Q 2 .
s m- 5 -
7 g 3 f 3 &s 7 u S ra 3 g _
d o E _
3-5- e rf .
3 . L r
)
J u )
y
/r f,
_ h -
g
_ t M
e b e 3 0 s u
7 r g V.
7 g
y 4 e pa
, L )s- t i
c
\
e 2 m f n
N u % 5 S P 7 (S a v (
/-
o e l i
e l: :
e l: :
e l: :
e l: :
e il: e l e l e
ia i i i it a a a n o
n o n a n n a n a n ia n i
a n n h m- h m- h o m- o m o m- o m- o m- o m- o i h - h h h h h m P E P E P E P E P E P E P E P E P r
'e _
T 'J _
e s w c f f W _
n Mc r c. Y _
e e J v a S _
c o 7,
% s n
i L -
( _
'WM r
o e :
e a c kc 3 T
?e W w& h C F/
f
'o 2
a C. $ P i
r e
s s -
3
~ D / r 1 1 4 ^
e /s
& / r c w ,
W
. t e 0 i
r d r
E i L L, w v C
5 _
C d n
f r F f $ 5' W T n,c 4
_ 4 e s.
l a A r /s c I D
e c 4 t r c s s
I o e C
/ O l
i g
e
. t J M N k
e E i o n r e o - S 'C lo i s t, ic i v- , D c i
N ,. d. M, id u <
Jg E 1
R a B 6
e i k,
o>, f # v l ,
1 g wc %
e ff T
/- a/C egn 2 hC % u 5
- /e 6 t v n ' e /
hile t
m d o i.
'M R a e o y $ w, GM c>
h d it o N a , di 'r o r,/6 N t
s e e
M r WM l$. u .
a >
r
- ee s
s e 7r 9 g I S 4 M ^/ 7rY S c 0 M w 7 8 mc u no s n d % r M?
e c; 5 iaD ri l is u ?e t P
o
'l' C f
- b f' / t r L A q 7 c d
_ o pd B
~7 - /v E, 1 J '/
dmn t
i N
u oa ACy 1
1 2 1
2' 1 2
1' 2 1
2 1
2 1
2 1
2 Ngvnre Fiu t WraS Tt s s
8,nt u .
- 9 oa 9mS 1 e l t
,D a 1 ni n 9
y oF : u -
r pa A r -
y C a ot uhc G s t $
Fo r
bsu erkd no E k 0 C v
=
e s.
W s_
o D t
s WC M E. b W- o
- 4 0
N i
L' e: M ..
N>
cp no ah n
/e Nc 1
~
dJ b
i a
MV 7
e -
ds nk r e 7 y/
e o b
r A
t t
AW
~
e d l } (p m , .
a f. 9, c
3 f l
N 7 1
f I 1 M b I
M h A
- o. ., _
c c l
a q, /
c, g
a x
G.
e.
g s
f g".
m
/,. x s
O P m
(.
c
- A f, o (
- o
7 G O c j w
- s j, ,
Cs u g A c f g @ x y o e j 1 9 /
f f r
[g g
"[
d 2 2( 7 /
%6 x }y, p /
d e
c ff 1
u A be 4 m g r-g Z.5 8-
@ 6(c- f(g Q Qa k
_ 6 -
o l ia i
- r ,
T S {
s m- m G. . 4 g 7 d
o E Q 2 2 g
i p 8 f i 3,
% g. ,a 3 6
g 34, h D ky 7 t
h /r i - - r ,
t e .
e R 7, e b
{g n br n y f,, y 4 2-M n
N m
u l
}
y yt e H s o
W (7 W ,
h Q:
W b fu7 fe 20(.m o e l i
e il:
e l e l e il:
e l e l e l: :
e ia i
ia i
it e- n o
a n a n n a n a a n n ia n n h m- h o m- h o m- o m- o m- m m- o m- o m- o i h h h h h h n P E P E P E, P E P E P E P E P E P h
e t o
T e
s (as t
i 3 5 (4 n
e c m u
- b. k
@ 1f '
M t
A i
L r ( e M. g _
o < (,
f i
a r
e s
s e
Ya 5 7
G a.v p l
% D
- w m ;
y L.
e r
r itr r
g h .
dr e E $& q n d
C l
a d
A _
IA o
a0 2 h4 Se e
N t
w m
c s en d e w $]
o w 7s i
g-o in s
e e Y 5 w e0 e + S y S, s g k Y\
l l
o s s h4 n. 9 J u i
u n di e cb M 6u fi + D d
a B Y
r o / 0 y 7 9 w d5 1
R n
egn 2
/
e.
E. h < 9 a &,d, yi e e,h C e hi t l h d e m a
m_
s =u.m 4
f o f e K :r 9
4 h .
t io WM N n u
Gl Jr J c v 9, g
9 d so 4
mc u no ee s
s s
e n
K R h. C M 5h 0 b O U
0 3 ex f
R w fl w c
5a 4 t
p ~h i
iaD ri s P & )l. .
L u 5 $ b 6 l
t opd B %_ 5G Of
(
9 %
imn
. d u oa ACy 1
1 2
1 2
1 2
1 2 1 2 1
2 1
2 1
2 Ngvnre Fiu t WraS Tt s,s u 8
9 nt oat.
9mS 1 e
,D a l
O' - 'a 1 ni n 9 e f ' D '
n y
r a ot oF pa k u e M u
R h uhc r k< % J A
@ C bsu ekd % Y 0 o "A Fo t
i L'
- WC s
r n o
T 3
+r P
3 G A
2 N
/ n .
4 f/
5 s e: E t e
1 cp C .
. u =.
no n b 4 t <
'n. d ds ah r-m 6 Y )n v o et h
nk > T.
ter
, t o
A t
I .
a o t
AW e m l-
/
$ $ ~
Q $
]
/
a i- 7 0
N 2 4 3 O' N' 3
/*
3 M
g, ne -
q x n .
f.
e o o
, Mr -
c t, c a 6 ,
o s
y a ,
6 sy T n 2 s
s B sa f. 3 s g 7 -
c g o O g O e ? m ~/ p- -
r 3 r l a R d t s. g -
d ,
3 @t / e 3 b F Q , g 2 y
_ A ,,
$s ) & o p_
- o l ia 2 H 5 - Jc e - 2
(
y T'
d s
o E m-m ,
'/
6 i
c er p
f-f f.
n e
9
/ 7 f.
r d a w.,._
9 u O
M 3
g 3
y g
))d(
h /r ,
e - ,
e p m .
4, g i
)o.
t e g e ) p M
b c n e v o g 3 g 3 o
g (g 3 g u m {
n N
u g, V S t
($ 4 c g i
t
.an o e n
i l
a e
n l
ia e
n l:
i a
e n
l:
i a
e n
l i
a e n
l:e i
a d[e n a l:
i e
n l
i a
e n
h o m- o m- o m- o m- o m- o m- o m- o m- o
_ i h h h h h h h h P E P E P E P E P P m E E P E P E P -
'r e .
T 2 e'
s 2 ~
5 A
. n e r 2 9 / C-c e d S T iL f 4 m
r / u F 0 TC f.,
o / p 4 f
a s
~ o 7 c y i
t r
e s e
r E
,, 'd a y C, 5 bA i
r d c, s r l
C d E( k e l
a A g
4 r ln r
W d -
i
.g c s s c L k )/
f % o A g.
- o e n /r A A b tJ.
l io {, , c, is s.
u 4 g t
d /. r n
a B R
1 D e hs c 0 ke
. 1
'n p s l t egn 2
/ v A
/ a I e .
.hi e l; .
t le m ], ll a
/; /a M .
g g
P h d t
N a
k e u. m io WM s I, /
y J /o % m Go t
C .
mc ee n os s
e s e h 0 C c
2 d.
e e ,
h g f
u iaD rli o pd is B
n u UL S 0
/
/ 6 a
- c 4
0
/o 6
g 00D
/ 9 M n 2
3 O a g 2
dmn /
t i
u oa ACy 1
1 2
1 2
1 2
1 2
1 2
1 2
1 2
1 2
Ngvnre Fiu WaS t
r -
Tt s s -
8,nt u 9 oa k .
9mS t
1 e l
,D a 9
1 ni n % -
y oF r pa
/ b ._
aot uhc n e. e r a \
e bsu ekd For on W h'ts
- fy K..
y
< r t
n i T A s WC o t
e s e L c dr s A U
e:
cp e
h 4 N hf )
A 4 D no ah br y R p \ # \
ds t e f 7 w nk ero LL v 1 1
t 6 o A t
t AW C
% 8 d We Y
- e D(
m a .
- t. 5 T- t N B S 3 % O. 3 F'
m
- o. -
_. e so g 4 C w.
' b qv o.
3q "' c o 7
< A, , i p -
P ds s 7 ,
2 y /5 s
e o O 2 W &
c s e<
v 2
y #'
r g d y .
S c.
4 g y f 2 V
b #"
d A m lg f " hs a G o- 2z o
l sy g . J T ia Y s Ae k v>If 4
s m- g S "S - d M o, w
t s
d h
o E
/r e llc y
. M "" A s {w 4 m_ kv 1
3 (9
@ kp <
g s
ee 2-v t t e b e
- s o 1
g 6 i
3 o _
M m u
z 1 1
'" f p-0 1 g 3 c p >
o 5 J n
o N :
4
": M : :
[%
- 7 e l i e l ia e l ia l
ia e l:
e il: e l: :
e i: :
e it a n a n n N n ia n a n i
a n ia n
- n o m- o m- o m- O m- o m- o m- o m- o m- o
_ i h h h h h h h h
_ m r
P E P E P E N E P E P E P E P E P e . -
T c l s e
s c f' y- '
E i
n g .
o e >r 0 A C
/
c h -
1 y
,VA w 3
_ iL_
r o
2 e u.
n p us
_ v f
i a
r e s s
e N
Hs dw /r-o M w
f 94 A
- t r a c c i
r d l o . k J C d e 5 t, c C R l
a c
A s -
UA E A
t i
u k M h A s
s o ut i
g
_ lo io is e
n ?
(
L w t 9 c (m h t
Jv d u A k i e 4 g [c S u k 2 ?_
f a B i' z
(.
R k a egn 2
/
0 4
l 7 M2' R u
t e N. k 6 L s 3' _
_ hile t
m i w
E v
A h d -
u 6e a /e s
e O t
- f
- N c e io
% A WM s e L H s
D (Qb ee s 0, b @ C E C s
4 mc s e t i d- W u n o n 0 T o i iaD is O - r t 0 /
A /
dk )p 0 W
s A
D. %
0 e h
- i rl u 3 o pd B / 4 S C 2
6 dmn t
i u oa
. ~ - . . . . . . . . . . . .
1 1 2 1 2 1 2 1 2 1 2 1 2 1 2 1 2 ACy
- Ngvnre Fiu t WraS Tt s,s u 8 nt 9 oa . $
9mS 1 e t
l
,D a 1 ni n 9
y r pa a ot oF W M d uhc r
bsu ekd N n
o k & b [
q -
u .
e Fo r n m
o v p kS
[..
tWC % O k
e, m
i L'
s
/~
t a [ L e:
cp 3 M /
no l e t 6 - k
- ah $ -
y w Q l. c ds y t
nk
~ w s d t
t AW e or e b (/
e O b S1 hc
[
d h d --
m w
a g
g
- y. E, 2 r.
N t i a a t 8 4 Y V y
J
,mp.
t w
h
/
s b a c_ )
g w i
s e 4 j d o s
e r o 7'3, V e C W *^ 7 e
- m. s se r
D l a
h e, d "
d a / -
6 c.
, /'
n A, C 3 h - t I 7 5 4 g y , e o l c r- c 5
/ A 'T
~%
e4 i
p T
s a
m j
s g 6 n /
Z 5 8
/'
- G N t
c 9
Q 5
- 5 d E v -
o /r w b i
s 3 -
- P ?
J- .
w5 s
h M
t e
b e h m
u yh e
2 4 4
f j
y p, m 3 D'
3 '
)
3 2
f 3
5 o w t
m Y
7 r-.
2 O 6 1 1 n
o N : :
i: :
l: :
3 l:
~~
5 l: : : :
it e n ia l
e n
i a e n ia e n
i a e n ia e n
l:
ia e l:
i a e l:
ia e a o m o o n n n n m- m- o m- o m- o m- o m- o m- o i h - h h h h h h h h m
r P E P E '
P E P E P E P E P E P E P e 7g T
d u
e 3 7 r c s t n a (s t i
L e
c %
P r- e 7 z
g 5,
c c.
g L G r
o dv g A y/ t t
N r
f a s #
g L rc i
r s l z /
i t
e e r
y kx k zw 'h J' S.
L %
r d G C
g s l
a c
d A
s 2
5 1
1 4 d ,
4-s h Ry ig s g l o in e - t
(
p A 2
C x l
io s ,
U d
R a
egn B
2 u
l A <
7 $"Y
)
l
< (
Ci b %
d j
+ G A b
/
hi e n
t le m '_ 6 g C ha o N
C E +
h d it a 9 ,
o N o e, w.
WM s g k 0 E s n ew mc ee s
s e 6 E v FS 7A e k /, / c u n o in iaD rli o pd B s
u MD O f6 w 0^0 O Vj D N l
- f o E g s D E w dmn t
_ i /
u oa ACy 1
t 2 t 2 1
2 t 2 1
2 1
2 t 2 1
2 Ngvnr e Fiu t
- WaS Tt s r
s 8,nt u
- ~.
9 oa b 9mS 1 e t
l o
,D a 9
ni n * '
1 y
r pa oF
,b 7
5 A f 4
1 Wv n
o e
a ot s uhc r h bsu 6 's er kd n Q. i m Fo o % G ~, $
t :WC s ph S ") P %
M I V \
m
^, y i
L e: E' I
\
cp no h" o (4 d ds ah fr c
& M f e e R-h 4 nk r s E 4 h b d)e e o t
t F r T v 4 AW e S m
a - d.I 6
& V.
N rf g 9 I S g f
' T $
y v
n na g m Q p 4 c g 7 / , c [O I S
f.
s y
. a.,
4 6
7 0
g g f ) L G
N
]/. (,
s g 0 3 9
e - -
d r g 7, e.
a f
t g g
/ 7 6 9
g e
~
l
(
d m,, g e
s c/ ,
g e
@ S 7 T
o s
A l
i a
m- Q f g
q s
s r
e g
)
g
, g mh e b q .. T T
I V
5 g p ]
H m so & k)2 Jy
- d E 9 Ae 7, h
o /r g, 7 7 g d 1
@f 37 t
e b e 7 ff //
dp 5 Af (s g g 2 n B 4
v 1 -
M N
m u
(
r /
g Y, "
3
$ 3 2 r n
I o $ M l j
t
(
{
n o e l a l: :
e l: :
e il:
e e l:
(Je l:
t i
t n ia n i
a n i
a n a n l
ia n ia n l
ia e ia e a o o o o n n n h m- h m- h m- h m- o m- o m- o n o o r o i
h h h t
- h - h m
r P E P E P E P E P E P E P E P E P e g T s
- e J ,
s . c n L,
- t. 7 t
%,, ,M i
,ce i ,
i L 5' R, D, _
h 11 f o r
a l
? hr g y
8 s
bb Q y i
r s a ,
e ' g e L _
i C
t r
d d
r d ,
% ) N n
W f l
a A J, , e. vv _
c s /n t ,
K v s ds s M
ig lo o e n v h _
t s
ly l a 4.
/
v L( O e
i d
a is B
u 0 .
s c
S e
W R r J @ A. n' T m.
. eg 2 s U L
hi n /
e er t o
J c
- t le
.t h d m
a h
t t h/
C N e
c )fp C, 3 l-o t
M Qb i o N m c 8 c.
. WM s $c a r
G 4 ee s i
'g . 3 f d mc u n o s e n h
/v Tr 9 o
- j , c F/
6 N ((
iaD A rl o pd dmn t
i i is B
u & O-xt 4
l D QG 6 (M
. u oa ACy 1
1
~
2 1 2
1
- 2. 1 2
1 2
1 2
1 2
1 2
Ngvnre
. Fiu t WraS Tt s s u 8 , nt
\
9 oa .
9mS 1 e l t
Da 1 ni n 9
y oF V
r pa L a ot 5 ( \
uhc r
bsu erkd n 0
G
/
P e j s y
k Fo o
k M tWC
- 7 M s
}
iL e:
cp no ah d s S
w T
M L W az l
I/
e y #
)
E v
A
[/
Q W,
(
r l nk r te o m_ o
& N t
AW a
e m !
n>
f i
e b b A!
3o N Yt 09 }lg 0 (f
/,
h l f
[/
(
E
4 e _
C t.
t _
6 f g e m W .
G o,
s M y d 8 o % 3 s b g 6 e $
s w 2 S p v a
r 5 0 G- 5 (_ e S cw h d 3 w 3
?
d A w t 6
-t 6 *
/
g - e <
i i
l T
o l ia t
s - ^ [
?
~
s m- i sp 2 g
C @ G
% G. r d
h o E
/r f,
e gf, )
T z
a % 3 g
9 M f a
e w
e g
t o N )D I _
e b 4 0 0 _
M m Lr e
o ( g 7 9 r P _
- u d
[
P m 3 3 (, (3 A n N f _ _
o e l: (e l:
e ia l: :
it n l
n e i a
i ia e l:
i e l e l e l: :
e n a n n ia i
a n a n a n i
a n n h o m h
o m- o m- o m- o m- o m- o m- o m- o i - h h h h h h P E T
m r
e 7
P E P E h E P E P E P E P E P m
e 6 Mc s
n e
c
/
/-
e d d C a
0 n
/ iL 6
0 w k
{4 0
/ r 7 , c o 4 n c
-.. f a s 2
h >-
R h
w i
r s it r
e d
e r WV g .
M 9
1 _
.C d ,
t u e 1 A t- e v ,
l a s se p P 8 c f.
ic g s e B E w m d _
/( lo o n is g /
t E u e f- _
id
.R a B u 3, y g L us m
s '
A C .
-r u
co r.
j o
/
S E
eg 2 o R o _.
hn
/
e p ~
' 's v A _
Q i
t H j( 5 e _
h d d e m a . b w a, n N _
WM o N s m I
l i
i_
t.
iO y t-i
/
ee s s a o h d
'm cs e t m p C u n o iaD rli n
is u p i
h e
u Y M (g y j w m R-a G T M o pd d
B S dmn t _
i f I -
u oa ACy 1
1
- 2. 1 2
1 2
1 2
1 2
1 2 1
2 1
2 Ngvnr e _.
Fiu t _
WraS Tt s s
8,nt u 9 oa A . e .
9mS t Jp O 1 el,D a L 5 q m _
1 ni n b
9 f
/ 1 y oF W
ll A f pa (
- r a ot T _
Ny 1 m S _
uhc M ;
r bsu f as E _
kd er on S -
S Q 3 N Fo b m . _
- WC
- I i
t s f .
. 1 O .v
]
L L . _
v 1 I'
[ _
e:
% E .
cp no A Jr ah ;- dr/ M m he d s <- w .
nk r f o f c .
te o Nl E- _
t AW e k O-M % k b _
m .
a .
X . .
lb 7 p
h 9 0 f N k 0 & 4 7 9
'