ML20216E168
| ML20216E168 | |
| Person / Time | |
|---|---|
| Issue date: | 06/23/1999 |
| From: | NRC |
| To: | |
| Shared Package | |
| ML20216E138 | List: |
| References | |
| NUDOCS 9908020091 | |
| Download: ML20216E168 (414) | |
Text
{{#Wiki_filter:.ae. dear.amm aaoawn ena anaannrum ma., wa,em,,assk%emm a n
.w%,gr a, e 6 wfgswedNka eA E._c4,86.&
d hI po 'q f;j! g[d
#4 9900020091 M YW)i PDR ORG a
990727 NREB PDR hhkhkh!!hfbbkh$%d4 N fR94f 5fN e - 1pMd%7 M 3% k%_d%gg d
$f M M g h da dM4 esf $p! J %M Myf6%A4fM.7;$$CdjMagd@Yf
^
o
]
' O OFFICIAL TRANSCRIPT OF PROCEEDINGS UNITED STATES OF AMERICA NUCLEAR REGULATORY COMMISSION
Title:
DECOMMISSIONING WORKSIIOP v l i 1 Case No.: l 1 Work Order No.: ASB-300-835 i 4 LOCATION: Rockville, MD DATE: Wednesday, June 23,1999 PAGES: I - 255 ANN RILEY & ASSOCIATES, LTD. 1025 Connecticut Avenue,NW, Suite 1014 O Washington, D.C. 20036 (202) 842-0034 9908020091 990727 PDP ORG NREB PDR
1 1 UNITED STATES OF AMERICA 2 NUCLEAR REGULATORY COMMISSION 3 4 DECOMMISSIONIFs WORKSHOP 5 6 7 U.S. Nuclear Regulatory Commission 8 Two White Flint, North, Auditorium 9 11545 Rockville, Pike 10 Rockville, MD 11 12 Wednesday, June 23, 1999 13 14 The above-entitled workshop commenced, pursuant to 15 notice, at 8:34 a.m. .O 16 17 18 19 20 21 22 23 24-25 ANN RILEY & ASSOCIATES, LTD. O Court Reporters 1025 Connecticut Avenue, NW, Suite 1014 Washington, D.C. 20036 (202) 842-0034
2 ! 1 PROCEEDINGS g- 2 U 3 MR. NICHOLSON: (8:34 a.m.) Good morning. I'd like to begin 4. the meeting and I'd like to begin the meeting by introducing 5 John Greeves, the Director of the Division of Waste 6 Management of the Office of Nuclear Materials Safety and 7 Safeguards, of the US Nuclear Regulatory Commission. 8 John? 9 MR. GREEVES: Good morning. It's good to see so 10 many of you here for an early start. I also know it starts 11 to fill up a little bit as time goes on. So a couple of you 12 might be here, a couple of the people might be grabbing 13 coffee. 14 So good morning. This is a fourth workshop that 15 the Nuclear Regulatory Commission il sponsoring, in part, to 16 help us develop a-standard review plan for the 17 decommissioning program. 18 The standard review plan would support 19 decommissioning of nuclear facilities, and this would i 20 include the reactor facilities. We've had a lot of contact 21 with the reactors that have shut down recently, fuel 22 fabrication facilities -- I see a number of familiar 23, attendees in the audience -- and large nuclear materials 24 licensees. 25 There's quite a stake that you have involved in ANN RILEY & ASSOCIATES, LTD. [ Court Reporters ( 1025 Connecticut Avenue, NW, Suite 1014 i Washington, D.C. 20036 (202) 842-0034
3 1 this process and I'm pleased to see you participating in
'2 these workshops.
N 3 As I said, this is the fourth in a series of six 4 public workshops, for those of you who may be here for the 5 first time. The first one was held in December of last year 6 and then we had follow-ups in January and March. These 7 covered topics ranging from dose modeling, restricted use 8 criteria, and ALARA analyses. 9 This workshop, as you know, is focusing on ground 10 water modeling. A number of you were at an off-site meeting 11 the last couple of days focused in detail on this. So I 12 welcome you. I also want to welcome our regions. We have 13 all four regions plugged in. I understand we have some 14 technical difficulties in terms of hearing what they have to (S 15 say. \_) So we'll get that fixed by lunchtime, but it's my 16 understanding they can hear the presentations here. So I 17 send my welcome to all four regions. 18 I'd like to thank the people here from industry, 19 government and some of the academic speakers for taking time 20 out of your busy schedule to join us. There's a number of 21 interesting presentations. I'm going to try and stay for as 22 much of the workshop as I can for the next two days and try 23 and visit with you, so if there is something that comes up i 24 you need to talk to me about, I will be around to 25 participate as much as I can. l ANN RILEY & ASSOCIATES, LTD. ! \- Court Reporters 1025 Connecticut Avenue, NW, Suite 1014 Washington, D.C. 20036 (202) 842-0034
\ : i l 4 1 l Before I turn over the meeting to Tom King from 2 the Office of Research, I've got a few admini.strative items. 3 We are planning topics for the future workshop on 4 the 18th of August. The first day, we would look to have 5 the comments on the current set of standard review plans 6 that we have up. We have a number of modules posted. We've 7 also, in this workshop and other venues, gone over the D&D 8 screen code. We'd like to get some feedback on that, and 9 the draft guide for 006, which we've put out for comment. 10 So those are some of the topics. 11 We could spend a day on issues associated with 12 surveys and other issues identified by the states. The I 13 states have a large role in this process, because many 14 licensees report to the agreement states. 15 So we'd like some ideas on that front. i 16 l I'd like for you to contact Robert Nelson, if , 17 you've got other possible topics for the August workshop. 18 Copies of the draft standard review plan modules 19 are up on our web site. Please take a few copies of these 20 back for your colleagues. ! l We are looking for comments on j 21 t these till the end of the year. i 22 Nick Orlando is the point of contact. I think 23 most of you who have been to our meetings know Nick. So he 24 is the point of contact for these comments. This is your 25 chance to provide input on the standard review plan process. ANN RILEY & ASSOCIATES, LTD. O Court Reporters 1025 Connecticut Avenue, NW, Suite 1014 Washington, D.C. 20036 I (202) 842-0034
i 1 5 Wa've gotten~a lot of good feedback and some help in solving } 2 f~% some of these difficult problems. So we appreciate that. k-s 3 4 We do have a new web address and it's on that single slide that indicates which ones of the modules are up 5 there. We're doing better. , 6 The address is shorter than it 1 was before, but I'd like to challenge us to even make it 7 easier to recognize. 8 So as I said, ten of the 16 modules are up on the 9 web and we would appreciate those comments. 10 The rest of the modules we expect to be up on the 11 web by the end of July and we have a full schedule today, so 12 I don't want to really take much more time. So at this 13 point, I would like to turn it over to Tom King. 14 Tom is the Director of the Risk Analysis and k 15 j Applications Division in Research. O 16 MR. KING: So, Tom, please join us. l l ( Thanks, John, and I want to add my 17 welcome to all of you, also, to the workshop. Probably at 18 the last workshop you had a different Division Director from
.19 Research up here, John Craig.
We've reorganized recently 20 and all the work on radionuclide transport and dose modeling 21 has been relocated into a new division, which is my 22 division, in the Office of Research. ! 23 ! So the people working on it are the same, but the 24 name of the division has changed and the division director 25 has changed. 1 ANN RILEY & ASSOCIATES, LTD. O Court Reporters 1025 Connecticut Avenue, NW, Suite 1014 Washington, D.C. 20036 a (202) 842-0034 1
i 6 1 I'd also like to express my appreciation for all g-g 2 of you who are participating in the workshop, particularly ('j 3 those of you who are going to give presentations, our 4 contractors and other Federal agencies, people from Sandia, 5 from Pacific Northwest National Laboratory, Argonne National 6 Laboratory, University of Arizona, DOE and EPA, and, also, 7 appreciate their helping to organize this workshop. 8 The workshop moderators are going to be Tom 9 Nicholson, from the Research staff, who is in my division, 10 and Jack Parrott, from the Office of Nuclear Materials 11 Safety and Safeguards. 12 The objectives of the workshop are to discuss 13 ground water modeling using dose assessments, for 14 demonstrating compliance with the radiological criteria, for 15 decommissioning and license termination. 16 ] Today and tomorrow's workshop follows a two-day 17 workshop we had Monday and Tuesday of this week out at the 18 Department of Agriculture Center in Beltsville, Maryland, 19 where the details of the research work in this area were 20 discussed. Probably a number of you were there. 21 NRC staff and contractors are going to present 22 their findings and discuss their ongoing studies in the area 23 of ground water modeling. There are several places in the 24 agenda where there is going to be opportunity for discussion 25 of the presentations and the technical subjects, and I 7- ANN RILEY & ASSOCIATES, LTD. Court Reporters 1025 Connecticut Avenue, NW, Suite 1014 Washington, D.C. 20036 l (202) 842-0034
E l 4 7 1 certainly encourage you to participate actively in those g3 2 discussions. ' N.] 3 To facilitate these discussions, we've attached, 4 at the end of the agenda, a set of questions that might be 5 used to stimulate some thinking and stimulate some of the 6 discussion. 7 ! The staff is going to use the information from 8 this workshop to help identify technical and licensing 9 issues that need further work and help improve and continue 10 to develop the guidance and the SRP, standard review plan, 11 content for implementing the license termination rule. 12 It will also be used to inform the Commission of 13 the status of the program. Because of that, we're making a 14 i transcript of the workshop proceedings today and tomorrow. ' , (' 15 We intend to publish these proceedings similar to previous N.)] 16 workshop proceedings we've published on the same topic. 17 Like John, I'll be in and out the two days. So if 18 any of you have any subjects you wanted to talk about, if 19 you see me around, feel free in come and approach me and we 20 can talk about anything that's on your mind. 21 With that, I'd 2_ke to turn it over to Tom ! 22 Nicholson, who is going to talk about the agenda and the 23 objectives of the workshop. 24 MR. NICHOLSON: Thank you, Tom. I also want to 25 welcome all of you to the US Nuclear Regulatory Conaission. e ANN RILEY & ASSOCIATES, LTD.
\~s'* Court Reporters 1025 Connecticut Avenue, NW, Suite 1014 Washington, D.C. 20036 (202) 842-0034 i
L
F, 8' 1 . Ws're very pleased to have you here. I want to tell you ' l 2 r [b N/ just briefly about how we're going to run the meeting, so 3 everyone understands the ground rules. 4- 1 First of all, as Tom pointed out, we are going to 5 have a transcript of this meeting, which will go to the 6 public document room. This morning, Jon Hundley, to my 7 i left, your right, from Ann Riley & Associates, will be the ! 8 courtroom reporter, and in order to help Jon do his job well 9 today, whenever you ask, raise your hand and we will call on 10 you to speak, if you could go to the mic and speak very 11 distinctly into the mic, identify yourself and your i 12 organization, that would help Jon an awful lot trying to 13 keep a transcript of today's meeting. ( 14 As Tom pointed out, there are going to be five (- 15 group discussions. (_}/ It's extremely important that all of you 16 participate in those group discussions and to help the 17 discussions begin, we've proposed a series of questions. 18 We're not limited to these questions. 19 ! If you have questions that aren't listed here, 20 you're more than welcome to bring them up and we'll 21 entertain them. If we run out of time at the end of each 1 22 one of those group sessions because of time issues, tomorrow 23-1 afternoon, we're going to revisit the questions that have 24 not been covered in the group discussions because of time 25 ' lind tations . ANN RILEY & ASSOCIATES, LTD. [\ -} Court Reporters 1025 Connecticut Avenue, NW, Suite 1014 Washington, D.C. 20036 (202) 842-0034
1 9 i 2 So hopefully we'll have plenty of time for group
/
discussion throughout the meeting. 3 Now, I'm asking all of you, as the person makes 4 the presentation, if you could let the person make the 5 presentation without 6 interruption, but at the end of every presentation, you're more than welcome to ask clarifying ( 7 questions. But if the questions become more of a discussion 8 topic, then we'd like to defer that to the group discussion 9 . So with that, I'd like to introduce our first 10 speaker, Bobby Eid, ( 11 from the Office of Nuclear Materials i Safety and Safeguards. Bobby? 12 MR. EID: Good morning. ! l My name is Bobby Eid. I ' 13 came here to talk about a very important issue, which is the 14 ground water modeling and the dose assessment. I - 15 j The major issue is how to integrate ground water 16 modeling with dose assessment. My name is Bobby Eid, and 17 18 I'm a Senior Technical Staff with the Division of Waste ! Management, and I lead the group who does modeling for ! 19 development of the standard review plan. 20 ; t'
- name, phone number and e-mail is on the lef t 21.
corner of the first slide. So please don't hesitate to 22 contact me if you have any questions. 23-The title of my presentation is ground water 24 modeling issues and dose assessment of decommissioning 25 sites, critical group receptor, scenarios on site-specific (N ANN RILEY & ASSOCIATES, LTD.
\s,) Court Reporters 1025 Connecticut Avenue, NW, Suite 1014 Washington, D.C. 20036 (202) 842-0034
I l 10 l 1 conditions. gs 2 My presentation outline will cover the following 3 areas; generic dose modeling issues pertaining to 4 implementation of the radiological dose criteria under 10 5 CFR Part 20, Subpart E. 6 The next topic I will be talking about is the 7. ground water models for screening analysis, focusing on 8 generic assumptions and applicability of the screening 9 models. I 10 Then I will be talking about generic performance 11 assessment issues for integrating ground water models in 12 dose assessment. Specifically, the source term abstraction, l 13 compatibility of site conceptual models, critical group 14 receptor, and exposure scenarios and pathways. ! js 15 N Also, I will be presenting some examples on 16 conditions for pathways and scenario modification. ! 17 l Then I will touch base on generic approaches for
- 18
! selection of input data, providing some thoughts on modeling 19 of complex sites, and hopefully I will make some 20 recommendations and some conclusions. L 21 First, l I would like to talk about the generic dose modeling 22 l issue pertaining to implementation of radiological dose 23 criteria under 10 CFR Part 20, Subpart E. 24 10 CFR 20.1402, which is the restricted release 25 criteria, first, it is required to have restricted use Adai RILEY & ASSOCIATES, LTD.
,Os Court Reporters 1025 Connecticut Avenue, NW, Suite 1014 Washington, D.C. 20036 (202) 842-0034
(
N 11 1 conditions; in other words, there would be no controls in i 2 place. Then the TEDE, which is the total effective dose 3 equivalent for the average member of the critical group is 4 25 millirem, or .25 a millisievert per year. 5 The average member of the critical group scenario 6 on pathways, as required under 20.1402, requires initiating 7 or modeling using resident farmer scenario, specifically 8 when we are talking about soil and subsurface contamination. 9 Then the pathways, they may be modified or 10 eliminated based on site-specific conditions. And as you 11 know, we have DG-4006 and NUREG-1549 to give you more 12 guidance on this issue. 13 Also, under this regulation, you know that the 14 ground water pathways are included. So you have to include 15 starting with the dose modeling, you need to consider that, 16 of course, if it is applicable to site, to include the 17 ground water pathways. 18 In addition, you know that we do not have separate 19 drinking water dose criteria, as possibly other agencies. 20 So the 25 millirem will include, as well, drinking water 21 pathways. The performance assessment timeframe, as you 22 know, is 1,000 years. This is in accordance with NUREG-1496 23 and DG-4006. 24 Also, under the regulations, there are ALARA 25 requirements. For 10 CFR 20.1403, which is the restricted ANN RILEY & ASSOCIATES, LTD. ( Court Reporters 1025 Connecticut Avenue, NW, Suite 1014 Washington, D.C. 20036 (202) 842-0034
l 1 12 1 use, the regulation requires that you need to demonstrate i l
~
2 that further reductions and residual radioactivity to comply 3 with 20.1402 would result in net public or environmental 4 harm or they were not made because the residual levels 5 associated with restricted conditions or ALARA. 6-In this case, under 1403, the average member of 7 the critical group scenario is a modified resident farmer 8 scenario based on site restrictions. The critical group i l 9 ! member or the average member of the critical group may l 10 actually reside off-site, rather than on-site. 11 The pathways may be modified based on site 12 conditions and restrictions and the total effective dose 13 equivalent also here is 25 millirem. However, the 14 conditions now we consider restricted use conditions, like 1 ! 15 institutional controls, are in place. 16 Also, under 20.1403, you need to know that the i 17 average member of the critical group under unrestricted use 18 i conditions, so there will be two critical groups, one group 19 under restricted conditions, other group under restrictive l 20 conditions, this means when institutional controls are no 21 l longer in effect, the TEDE dose to the average member of the l 22 critical group here in this case now is 100 millirem rather 23 than 25 millirem under 1402, or 500 millirem provided that 24 further reduction in residual radioactivity necessary to 25 comply with 100 millirem are not technically achievable, I ANN RILEY & ASSOCIATES, LTD.
\- Court Reporters 1025 Connecticut Avenue, NW, Suite 1014 Washington, D.C. 20036 (202) 842-0034 i
T ! 13 1-would be prohibitively expensive, or would result in net 2 public or environmental harm. 3 The ground water pathways are included under 1403, j 4 but may be eliminated based on site-specific physical 5 conditions. i This is an addition also to the -- for 6 consideration of restrictions under 1403. 7. Again, no separate drinking water dose criteria 8 and the performance assessment timeframe is 1,000 years. 9 { Those are the generic outlines under our 10 regulations and, in general summary, what is needed to be ' 11 done. Now, in dose modeling, we'd like to make things 12 easier fe' you. One way is by not collecting so much data 13 or to go into much detail, by doing screening analysis. For 14 screening analysis, we develop D&D code and models in order
/ 15 to demonstrate compliance with the dose criteria, without 16 having much site-specific data.
17 The screening models currently we are having, they 18 have special characteristics and they have special 19 assumptions. The main characteristics and assumptions are, 20 in a generic way, they are generally simple and 21 conservative. For example, on D&D screen, we have what's 22 called the three-box model. I guess many of you are I 23 familiar with this model, where you have the contaminated 24 zone in the 15 centimeter, then the unsaturated zone, and 25 then, of course, the aquifer or the saturated zone. ANN RILEY & ASSOCIATES, LTD. ( I)
' Court Reporters 1025 Connecticut Avenue, NW, Suit: 1014 Washington, D.C. 20036 (202) 842-0034
14 1 Also, these models assume a uniform surficial soil 2 source term, typically, in the top 15 centimeter, but this i 3 could be expanded to be a little bit thicker. 4 Also, they assume the critical group receptor is 5 on-site rather than off-site. So specifically for small 6 sites, when you do dose modeling for ground water, you think 7 about the receptor on a small site residing directly and 8 performing all the following activities. Such that you 9 think about what kind of transport under these conditions. 10 f Also, typically, they assume that activity, i 11 leaches in the contaminated area via infiltrated water 12 through the unsaturated zone to the aquifer, with no 13 significant dispersion. They typically assume no 14 retardation in the aquifer as we have in the D&D screen. 15 Typically, the infiltration volume is dependent on 16 infiltration rate, the area of irrigated land, and the 17 infiltration period. 18 ! The unsaturated zone and the aquifer are initially 19 free of contamination. So those are the major assumptions 20 for the screening models. 21 When you have your site, you need to do the dose 22 assumptions and see your conceptual model for your site 23 integrates well or is compatible with the conceptual models 1 24 in the screening analysis. 25 Now, I would like to talk about the screening l
\ ANN RILEY & ASSOCIATES, LTD.
Court Reporters 1025 Connecticut Avenue, NW, Suite 1014 Washington, D.C. 20036 (202) 842-0034 L-
i 15 1 models and the default tables and their applicability. We 2 are currently having D&D screen version 1.0 on the web site 1 3 for use as a screening tool. We are testing and evaluating 4 the current code and we may modify the screen numbers in the 5 code. 6 We are testing the parameters, as well as 7 developing default tables for soil. As you know, in 8 November 1998, we developed a default table for beta and 9 ! gamma emitters for surface contamination and we are now in i 10 I the process of developing a default table for soil. 11 So the numbers in the current version 1.0 may 12 change. So please note -- try to look at what we are doing, 13 { so in the near future, it we'll be publishing revisions. t I 14 Just our notice for modification of those numbers for 15 screening. i 16 So the screening analysis using D&D code or 17 default tables, they typically apply under the following 18 conditions. First, unrestricted release conditions. i If you 19 have site restrictions, this means you will be eliminating 20 the pathways, this means you will be moving away from the 21 resident farmer scenario. 22 Number two, the source term is compatible with the 23 source term assumptions of the screening code. So when you i 24 use screening codes and models and default tables, you will 25 need to ensure that your source term is compatible with the ANN RILEY & ASSOCIATES, LTD. O Court Reporters 1025 Connecticut Avenue, NW, Suite 1014 Washington, D.C. 20036 (202) 842-0034
r 1 16 source term assumptions in that code. 2 () 3 The site physical conditions are compatible with the screening code assumptions. The unsaturated zone and 4 the aquifer at the site would be free of contamination . If 5 you have already contamination in the ground water, you need 6 to think about using the screening codes before you try to 7 use them to see how they are compatible with your site. 8 Also, in the screening analysis, all the code 9 default scenarios -- you cannot use different kind of 10 scenarios or default input parameters. As soon as you 11 change input parameters, this means you are moving from the 12 screening mode to the site-specific mode. 13 Only code and default input parameters are used 14 unless approved by the NRC, who are in the process of 7_, 15 modifying some of the parameters in the D&D code and then 16 j will inform you about these changes. 17 The next topic I would like to talk about is the 18 generic source term abstraction, it's the generic ground 19 water performance assessment issues for integration of 20 ground water models into dose assessment. As I said, it is 21 extremely important that the performance assessment issues 22 for ground water and dose -- and for dose assessment, they 23 must be integrated together. Otherwise, we may have i 24 inconsistencies in the models and the codes. 25 l The first issue that you need to adarues to ensure ANN RILEY & ASSOCIATES, LTD. ( Court Reporters 1025 Connecticut Avenue, NW, Suite 1014 Washington, D.C. 20036 (202) 842-0034 L
17 1 that you have integration, proper integration between ground 2 water modeling and dose assessment is the source term 3 abstraction. You need to assess, first, the source term 4 configuration, to look at the contaminated area, the depth 5 and vertical and horizontal extent of contamination, to look 6 at the residual radioactivity profile, distribution, 7 homogeneity, and if you have multiple sources or you have a 8 single source, and then to look at the chemical form, 9 specifically the solubility issue in this case applies for 10 performance assessment. 11 The next issue to ensure that you have proper 12 integration is to look at the compatibility of site 13 conceptual model. The code conceptual model assumptions and 14 mathematical computations should be compatible with the site 15 physical and environmental conditions and the exposure 16 pathway scenarios. 17 The model should also account for site 18 restrictions and instituti,onal controls. ! 19 The third significant issue to ensure integration 1 20 1 between ground water modeling and dose assessment is the ! 21 critical group receptor and the exposure scenarios and 22 pathways. Under restricted release conditions, we initiate 23 l ! dose modeling using on-site resident farmer scenario and the 24 exposure pathways. Thus, in ground water modeling, it's i i 25 extremely important to consider where that receptor is l ANN RILEY & ASSOCIATES, LTD. O Court Reporters 1025 Connecticut Avenue, NW, Suite 1014 Washington, D.C. 20036 (202) 842-0034 i
I' l 18 1 located. It is on-site. ! - 2 l The pathways that are considered for that specific l 3 scenario include external exposure pathways, indoors and 4 outdoors, the inhalation exposure pathway from the suspended 5 , soil for outdoors, indoors and surface sources, and from the l 6 ingestion pathways, this is a lengthy pathway, which ! I 7 i includes direct soil ingestion, soil tracked indoors, I { 8 drinking water, plants grown in contaminated soil, plants 9 irrigated with contaminated ground water, and animal 10 products grown on soil and fish products from contaminated 11 surface water ponds. 1 l 12 Then you could modify the pathways and scenarios 13 as necessary based on site-specific conditions. You may use ! l 14 more than one critical group or scenario, to multiple 1
'")
[G 15 sources. In some cases, you may segregate the contaminated 16 areas into more than one source area and you may need to 17 remediate one area in a different way than remediating a 18 second. i 19 So you need to consider the possibility of having i 20 multiple sources on the site and the impact of one source I 21 t area on the other, in order to account for possibility for 22 having multiple remediation options. I 23 The next issue is the critical group and expcaure 24 scenarios. Under restricted conditions, initiation of dose 25 ! modeling should be through the use of on-site resident t ANN RILEY & ASSOCIATES, LTD.
]) Court Reporters 1025 Connecticut Avenue, NW, Suite 1014 Washington, D.C. 20036 (202) 842-0034
I 19 1 farmer scenario and the exposure pathways. As we indicated. f- 2 and I explained, the pathways included under the scenarios. 3 Then modify the pathways based on site physical 4 i conditions, modify the pathway scenario based on site 5 restrictions. We have here addition of site restrictions 6 and institutional controls. So you could eliminate the 7 pathways based on these restrictions. 8 Then perform dose analysis based on failure of 9 these controls, because you need to meet two criteria; one 10 criteria under 2 :strictive conditions and another one under 11 unrestricted conditions, under the same criteria, when the 12 controls are removed. 13 You may need to perform, also, off-site dose 14 assessment using a critical group receptor at the boundary ()
\J 15 of the site, because of the potential for off-site releases, 16 and in this situation, you may require more advanced ground 17 water modeling. However, we anticipate that for on-site 18 resident farmer, specifically if the site is relatively 19 small, ground water modeling would be simple.
20 Many of you may question about how can we modify 21 the scenarios, how can NRC accept modification of the 22 scenarios. Do we need always to use the resident farmer 23 scenario? The answer is no, we do not need always to use 24 the resident farmer scenario. You could modify the scenario 25 and you could modify the pathways. l l l I \ ANN RILEY & ASSOCIATES, LTD.
~/ Court Reporters '
t 1025 Connecticut Avenue, NW, Suite 1014 Washington, D.C. 20036 (202) 842-0034
20 1 If you modify the scenarios and the pathways, you 2 need to construct the local and regional characteristics of 3 the soil. Maybe these characteristics, they do not support 4 the agricultural activities or the pathways associated with l 5 that scenario. For example, the soil in the unsaturated 6 zone, there are salty deposite and suitable for growth of-7 plants and crops. 8 In this case, the agricultural pathways may be 9 modified or eliminated under these conditions. 10 Another example, the water quality or the volume 11 of the aquifer do not support ground water exposure pathways 12 or drinking water pathways. If you cannot have -- if the j 13 ground water is not suitable for drinking or for irrigation, 14 this means the scenario, resident farmer scenario is (~ s 15 inapplicable. ty You cannot apply the resident farmer scenario 16 if you have quality of water that's not suitable for 17 irrigation or for drinki.;7 18 l So in this case, irrigation and drinking water 19 pathways may be modified or eliminated. In this case, you 20 may need not to address the issue or to do ground water 21 modeling in this case. 22 A small site -- another example, a third example, 23 when you have a-site, a small one, situated in a highly l 24 populated urban area and historical and future planning 25 records support the assumption that the site cannot be used s ANN RILEY & ASSOCIATES, LTD. Court Reporters 1025 Connecticut Avenue, NW, Suite 1014 Washington, D.C. 20036 (202) 842-0034 i
l i 21 1 for farming activities. ! 2. ( Like in the middle of a major city, and historical ) 3 L and future planning shows that this city is very unlikely to 4 be developed into farming. So in this case, the scenario 5-may be modified to a resident gardener scenario, for 6 example, and instead of having farming, where you grow cows 7 and crops and grains and all of these kind of activities and 8 fish ponds, so you could eliminate those pathways. 9 So meat products, grain crops and fish pathways 10 may be eliminated under this specific situation. 11 One last thing I would like to touch base is the 12 generic approach for assessment and selection of input data. 13 For input data, they are extremely important because the 14 effect calculation of the dose. If you have highly 15 conservative input data, as currently we have, in the 16 screening analysis, of course, you expect to have 17 conservative dose. I But you could modify those based on j 18 site-specific conditions. i 19 So for screening analysis, you need to use the I 20 current input data in the code or when you use default I 21 tables, of course, you do not need to have input parameters, 22 without any modifications. 23 Then you could use the mean values of metabolic 24 and behavior parameters for your site. So you could assume 25 a range of metabolic and behavior parameters or preferably
"'s ANN RILEY & ASSOCIATES, LTD.
f ( ,j/ Court Reporters l 1025 Connecticut Avenue, NW, Suite 1014 Washington, D.C. 20036 (202) 842-0034 i m
[. 22 1 you could use NRC's values for metabolic and behavior i 2. {"% parameters in NUREG-5512, Volume 3, and 1549. (m-) 3 You could use also site-specific physical 4 parameters based on available historical and environmental 5 monitoring data. This data could be valuable, such that it 6 7 could reduce the amount of effort of characterization that you may need to do for the site. 8 Also, you need to take a look at the regional data 9 and to see applicability of the regional date to your 10 site-specific conditions. There will be a great deal of 11 information related to soil, data, climate, and hydrological 12 data, which one of the speakers will address this morning. 13 Then conduct site-specific measurements focusing 14 on sensitive parameters. You need to assess in advance what 15 O 16 are the sensitive parameters in modeling, whether in dose analysis or in ground water modeling, and focus on dose 17 parameters and see if you need to do measurements or you 18 could -- you could have sufficient data available based on 19 regional data. 20 So before going through much detail of 21 characterization and analysis, you could focus on the 22 sensitive parameters and if you need to do measurements, you 23 go and do the measurements. 24 Then you need to establish reasonable ranges of 25 parameter distribution and assess the uncertainties. We
/~'\ ANN RILEY & ASSOCIATES, LTD.
(%,/ Court Reporters 1025 Connecticut Avenue, NW, Suite 1014 Washington, D.C. 20036 (202) 842-0034 L
F l 23 1 like to see the uncertainties and how conservative you are
.-s 2 \-'
in your assumptions and how these assumptions they fit your 3 site-specific conditions. 4 For site-specific analysis, the staff's current 5 approach is to'use the mean of dose distribution. So when 6 you move to screening analysis to site-specific analysis, 7 currently, we are proposing and most likely will adopt the 8 use of the mean of the dose distribution. For screening 9 analysis, we are not using the mean of the dose 10 distribution, but this is not final. We are looking into 11 that issue. 12 The dose modeling, you need to perform iterative 13 dose modeling and assess the data needs. So before going 14 through much details of characterization and collection of r" 15 data, k_T ) try to do iterative dose modeling based on the 16 available data and assess how much data that you need to 17 optimize your dose modeling. 18 In some cases, certain sites could be quite 19_ complex and you need to have more advanced ground water 20 modeling and analysis, and these will be based on site -- on 21 case by case. It's very difficult to explain these 22 conditions specifically for each site and to make a list of 23 those conditions. 24 However, I would like to give examples for these 25 complex cases that you may encounter. I would like to l [s ANN RILEY & ASSOCIATES, LTD. Court Reporters 1025 Connecticut Avenue, NW, Suite 1014 Washington, D.C. 20036 I - (202) 842-0034
24 1 - emphasize that those are very, very few sites. 7s 2 So more advanced water modeling is needed for dose 3 assessment for complex sites. The first example, when you 4 have extensive on-site /off-site ground water and/or surface 5 water contamination. 6 Also, the other example, you may have 7 on-site /off-site complex engineering barriers that are used 8 in the performance assessment analysis. In this case, you 9 need to address the performance of the barrier and the 10 transport through the barrier and the integrity of the 11 barrier. 12 So these are becoming now more complex. In other 13 cases, you may have complex restrictions, maybe you propose 14 very complex restrictions and conditions and you have 15
) multiple on-site barriers, but, in this case, you need to 16 assess the source term issue and how these restrictions will 17 apply and the critical group could be more complicated. It 18 is not only a single critical group: rather, there could be 19 two or three or it could be more critical groups that 20 represent the site in order to derive the dose.
21 Also, we have complex site review of -- but 22 briefly, the way to deal with the modeling of complex sites 23 is not exactly can be described in a plain language to you. 24 However, the staff will be reviewing those complex cases 25 under case by case conditions, will review all the details ANN RILEY & ASSOCIATES, LTD. (s Court Reporters 1025 Connecticut Avenue, NW, Suite 1014 Washington, D.C. 20036 (202) 842-0034 ( .. .
=
25 1 pertaining to these complex sites. s 2 Finally, I would like to make some conclusions and 3 recommendations for ground water modeling and dose 4 assessment. First, l there is no need for additional ground 5 water modeling for screening analysis, because it's already 6 taken care of. 7 Default tables and NRC approved screening codes 8 could be used in this case, but the users, they need to 9 verify the compatibility of the site conceptual model with 10 the conceptual models and assumptions of the screening 11 codes. So we need to verify how compatible these models and 12 the assumptions in the screening code are with your site. 13 The critical group receptor exposure scenario and i 14 the environmental pathways are significant factors. The ("% 15 (l ground water modelers need to consider in their modeling , 16 l analysis and they need to be addressed and integrated with 17 the dose modeling. The ground water and dose modeling may 1B be initiated using the default resident farmer scenario. So 19 without having any information about the site, this is the 20 first thing you need to assume, the resident farmer on-site, 21 and you move from there. 22 The transport and dose impact pathways could be 23 modified or eliminated based on site-specific physical 24 conditions. As we gave an example, for the soil and the top 25 of the aquifer. Site institutional controls or ANN RILEY & ASSOCIATES, LTD. l [\- I Court, Reporters I 1025 Connecticut Avenue, NW, Suite 1014 Washington, D.C. 20036 (202) 842-0034 I
26 1 restrictions, potential off-site release conditions, 2 source-term conditions, potent $al current and future land 3 uses, as I explained, and, finally, the conclusion, the 4 recommendction that dose modeling is an iterative process, 5 initiated using limited data and conservative assumptions 6 and subsequently optimized using site-specific data and 7 nctual conditions. 8 If you have any questions. 9 MR. NICHOLSON: Thank you very much, Boby. Are 10 there any questions from the floor for Boby? If you could 11 just go to the microphone and identify yourself and your 12 organization. 13 i MR. ROBERTS: Rick Roberts, Rocky Mountain ' 14 Remediation Services. A couple of clarifications. On your 15 earlier slides, you talked about the D&D screening code only 16 being applicable to soils, that's the top 15 centit-ters of 17 soil. 18 Does that mean if you have contamination below 15 19 centimeters, down to ground water and within ground water, 20 you would expect there to be some site-specific modeling 21 involved and the D&D screen model really would not be 22 applicable? 23 MR. EID: I believe if the source term is 24 different than your side, then you are, in a way, moving to 25 site-specific analysis. The way you may have your 4 Alni RILEY & ASSOCIATES, LTD. O Court Reporters 1025 Connecticut Avenue, NW, Suite 1014 Washington, D.C. 20036 (202) 842-0034
f 1 27 conceptual model to fit with the top 15 centimeters, that 2 \ () 3 Mark Thaggard will talk about, I will not talk about it right now, and if you convince us that you could have a 4 conceptual model similar to the source term that we are 5 having, which is in the top 15 centimeter, it could be more, 6 it could be down from 15 to one meter, possibly, depending 7 on the conditions of the site. 8 This means we could -- we may entertain to 9 consider screening analysis, and Mark will talk about this. 10 MR. ROBERTS: And the second question is, you had 11 said, in one of your overheads, that it would be acceptable 12 to the NRC if you did a site-specific analysis to take the 13 mean dose from the output distribution. 14 I just wanted to clarify that's from all pathways 15 . / and not just from the groundwater pathway. That would be, 16 if you had an output distribution and you took the mean off 17 of there on a site-specific analysis, the NRC would find 18 that acceptable. 19 MR. EID: For the all pathways applicable to your 20 site, in case you eliminate certain pathways because of 21 certain conditions at the site, still you need to calculate 22 the dose from all pathways applicable to the site. 23 MR. ROBERTS: Thank you, 24 i MR. NICHOLSON: Thank you. Yes? 25 MR. CHENOWETH: Good morning. Paul Chenoweth, ! I ANN RILEY & ASSOCIATES, LTD. () (~5 - Court Reporters 1025 Connecticut Avenue, NW, Suite 1014 Washington, D.C. 20036 (202) 842-0034 I
l i 1 1 28 NEI. I jrst had one question. On Table 7, you were talking 2 about screening models and default 3 table applicability, and you indicate, at one point, 4 only code input -- default input parameters are used and then, in parens, unless approved by 5 the NRC. 6 What is the mechanism that a licensee would gain 7 that approval in moving forward, if they decided to change a 8 specific input parameter? 9 MR. EID: Currently, we have a code with input 10 parameters and the probabilistic analysis for the input ! 11 parameters has been established. We are reviewing those 12 parameters and certain parameters will be changed. 13 . An example of a parameter that may change is the 14 mass loading factor for plant deposition. That's an 15 example. We realize that O' 16 need to change it. it is highly conservative and we 17 Now, 1 l when we change this parameter, we will let 18 ; you know that we are changing that parameter, but if you 1 19 have a proposal for a parameter that indeed generically that 20 should be acceptable, you need to talk to the NRC in order ! 21 .to modify that parameter. 22 . For example, before, we were talking about using 23 the screening analysis for the current code about cesium and 24 i we realized we have soil to plant transfer factors, they are 25 j I rather conservative, that they need to be modified. ANN RILEY & ASSOCIATES, LTD. Ox _ Court Reporters 1025 Connecticut Avenue, NW, Suite 1014 Washington, D.C. 20036 (202) 842-0034
29 1 So we are dealing with this and then we'll let you 2
,_ know about dose parameters.
( ( But, also, you propose to us 3 that dose parameters, they need to be changed. 4 So this interaction, it is very fruitful and we 5 are trying to modify now dose parameters. 6 MR. GREEVES: Just to amplify, Paul, I think a 7 l licensee should contact his project manager and work through i
'8 them.
It could be a public meeting, discuss it, and then 9 whatever comes out of that. So work through -- the licensee 10 would have to work right through his project manager on the 11 record. 12 MR. CHENOWETH: Thank you. 13 MR. NICHOLSON: Are there any other questions for 14 Boby? j-% 15 [No response.) 16 MR. NICHOLSON: Thank you very much, Boby. And 17 thank you for your questions. 18 Our next speaker is Mark Thaggard, also from the 19 ; Office of Nuclear Materials Safety and Safeguards. Mark is 20 going to talk about decommissioning decision framework 21 discussion, focusing on ground water. This is from i 22 NUREG-1549. 23 MR. THAGGARD: As Tom mentioned, my name is Mark 24 Thaggard. I'm in the Division of Waste Management, here at 25 NRC. I'd like to say good morning to everybody. i l ANN RILEY & ASSOCIATES, LTD. {w_/} Court Reporters 1025 Connecticut Avenue, NW, Suite 1014 Washington, D.C. 20036 (202) 842-0034
F l p 1 30 I'd ask your apology, I do have a head cold. Can 2 everybody see the overhead or do we need to dim the lights? (O,,/ 3 Okay. I'm going to talking about our 4 decommissioning decision framework. 5 I'd like to acknowledge my colleagues at Sandia National Laboratory, who actually 6 put this framework together under contract here at the NRC. 7 I have included my telephone number and e-mail 8 address in case anybody needs to get a hold of me after the 9 workshop. 10 The purpose of my presentation is to go over our 11 decommissioning decision framework and, in particular, to 12 talk about how ground water, analysis would fit into this. 13 This schematic here on the right is a diagram of 14 the decision framework. I'm not going to go over it right 15 now, O 16 because I am going to be talking about it as I go throughout my presentation. 17 As was mentioned earlier, it's documented in 18 NUREG-1549. NUREG-1549 is referenced in our draft guidance 19 document 4006. l So some of you may have already -- may be 20 l familiar with this already.
- 21 l In terms of framework, I would like to talk about 22 a few important aspects of it.
The decision framework is
- 23
! similar to the hydrogeological decision framework put 24 together by Freeze and Massman and some others in a series 25 of ground water articles in the early 1990s. Some of the
<^ ANN RILEY & ASSOCIATES, LTD.
I Court Reporters 1025 Connecticut Avenue, NW, Suite 1014 Washington, D.C. 20036 (202) 842-0034
~
1 i 31 i ground water people may be familiar with that work. 2 I would also like to emphasize that the framework 3 is still under development. So because 1549 is linked to 4 our guidance document, which is a draft, we may see the need 5 to change this framework at some time in the future. 6 We've been testing the framework out on some test 7 cases and we also currently have Sandia testing the 8 framework out on a fairly complex decommissioning site. l 9 ! The framework should facilitate decision-making 10 for a range of sites from simple to complex and it should 11 facilitate the decision-making by providing us a structured 12 approach for evaluating a range of decommissioning options 13 and these options can include the remediation, land use 14 restrictions, site characterization or some combination of - N 15 these.
.16 \
Ultimately, we intend to implement the decision 17 framework in the computer code called SEDSS, which is being 18 developed through our Office of Research. 19 Over the'last couple of months, as we've gone 20 through these workshops, we've kind of given you guys a lot 21 of information and some of it may be a little bit confusing. 22 So I'd like to really step back for a moment ar.d 23
, say how this framework fits into the analysis for these 24 various sites. Some people may take a look at the framework 25 and say, well, gee, what if I've got a simple analysis or a ANN RILEY & ASSOCIATES, LTD.
( Court Reporters 1025 Connecticut Avenue, NW, Suite 1014 Washington, D.C. 20036 (202) 842-0034
32 1 site that doesn't really ought to have an analysis, do I l ~s 2 really need to walk through this maze or this complex U 3 process. 4 So if we group our decommissioning sites based 5 upon the type of analysis that we anticipate, we can see how I 6 these various analyses would fit into using the framework. i 7 ! Some sites can be decommissioned obviously with B doing no analysis. For example, if we've got a site that 9 has sealed sources, where they can demonstrate that the 10 sealed source has not leaked, then they decommission without 11 doing any kind of analysis. These sites will fall outside 12 of the decision framework. So a licensee for one of those 13 type of facilities wouldn't even need to be bothered with 14 the framework. And obviously, because there is no analysis, i 15 there would be no ground water analysis for those sites. 16 Some sites can probably be decommissioned using 17 screening analysis, as Boby went through earlier. For our 18 i screening analysis, we assume that the only information that 19 you have for doing the screening analysis is information on 20 the source term. So in step one of the framework, it's kind 21 of difficult to read here, you would gather information on I 22 your source term. 23 For step two, which is defining your scenario, 24 we've defined two scenarios for screening analysis, the 25 building occupancy scenario and the resident farmer I \ ANN RILEY & ASSOCIATES, LTD.
\s l Court Reporters 1025 Connecticut Avenue, NW, Suite 1014 Washington, D.C. 20036 (202) 842-0034
o 33 1 scenario. 2 For the building occupancy scenario, we assume
\ 3 that there is no ground water exposure for the resident 4 farmer scenario. As Boby mentioned, we do assume that there 5 is ground water exposure.
6 In terms of our conceptual model, for the 7 screening analysis, we've already defined the conceptual 8 model. So it's a little bit incumbent upon'the licensee to 9 make sure that that conceptual model that we've developed is 10 applicable for their particular site, although we've 11 developed these conceptual models with the goal of trying to 12 have wide application. 13 As we move into step four of the decision 14 framework, where we're actually doing the dose assessment, 15 this can be done -- for screening, this can be done using 16 look-up tables or actually running the D&D code. 17 I would like to point out, also, that for the 18 screening analysis, all the parameters have been 19 established. So you can use default parameters. 20 In step five of the framework, you basically make 21 a decision at that point. You look at your results from 22 your analysis and compare it against the dose limit. If you 23 pass the dose limit using the screening analysis, then you 24 move on to steps six and seven. 25 Step six is where you carry out your ALARA ANN RILEY & ASSOCIATES, LTD. O Court Reporters 1025 Connecticut Avenue, NW, Suite 1014 Washington, D.C. 20036 (202) 842-0034 l [ i I
l 1 34 onalysis. On the other hand, if the screening analysis 2 indicates that your site don't pass, then you'd probably 3 want to look at some other options on ways that you might be 4 able to demonstrate that you pass. 5 You obviously wouldn't stop just at the screening 6 analysis. So you would move into step eight and start 7 identifying some other options, and one of those options 8 could be the use of more site-specific analysis. 9 , So let's talk a little bit about site-specific ) 10 analysis. A site-specific analysis, we assume that you have 11 a little bit more information about the site. The step one, 12 we recommendation that you gather all available data that 13 you have about your site, and the reason I underlined 14 available there is to indicate that at step one, we're not
. 15 recommending that you go out and do a full-blown 16 characterization program. It may be possible to do some 17 preliminary analysis based on just the information that you 18 have available and doing those preliminary analyses may
, 19 steer you in terms of what information might be most 20 fruitful, 21 Then as you move down to step two, similar for the 22 screening analysis, the ground water pathway is generally 23 included for the resident farmer scenario. But when you get 24 into a site-specific analysis, as Boby indicated, you might 25 have some data that could allow you to either determine the l (g ANN RILEY & ASSOCIATES, LTD.
\ Court Reporters 1025 Connecticut Avenue, NW, Suite 1014 Washington, D.C. 20036 (202) 842-0034
35 1 specific type of use or maybe even to justify excluding tha 2 ,_ ground water pathway all together. k) m 3 When you get into step three, where we're looking 4 at developing our system conceptual model, generally, for a 5 site-specific analysis, we assume that you have enough 6 information, more information about your site so that you 7-can develop a site-specific conceptual model. 8 Generally, what we're talking about here is that 9 for most analysis, we're using computer codes where they 10 have predefined ground water conceptual models, and the 11 em.phasis here is that we need to demonstrate that those 12 pre-defined conceptual models are appropriate for our 13 particular site, based upon what we know about the site 14 features and processes. 15 In step four, for the dose assessment, i you would \/ 16 use -- you can use the D&D code, the RESRAD code, or some 17 combination of other codes. There is wide latitude. 18 We're going to be -- we're going to have some 19 presentations throughout the workshop on the conceptual 20 models in D&D and in RESRAD. Also, you may be able to use a 21 more complex ground water code to do some more sophisticated 22 analysis, if you think that's necessary for your site. 23 However, if you do that, then there are some issues that 24 need to be addressed in terms of how you link the ground 25 water analysis with the dose analysis, and we're going to / ANN RILEY & ASSOCIATES, LTD. ( ,)s Court Reporters 1025 Connecticut Avenue, NW, Suite 1014 Washington, D.C. 20036 (202) 842-0034
i 36 1 clso have some presentations on that tomorrow. 2 10 'Again, at step five, you have a decision. You
\s / 3 look at your dose assessment, the results from your dose 4'
assessment, compare it against the dose limit. If you pass, 5 then you move on to steps six and seven. If the dose 6 analysis indicates that you don't pass, then under step 7 eight you would start evaluating various options that might 8 allow'you to be able to pass the limit. 9 And then on step nine, these various options would 10 be evaluated, and these options could include, as I've 11 indicated, characterization, remediation, the use of land 12 use restrictions or some combination of these. You could i 13 have actually part of the site where you're proposing to 14 have land use restrictions and part of the site you plan to 15 remediate. So there could obviously be some combination and 16 there could actually be another part of the site where you i 17 may think it's pertinent to gather some additional data so 18 that you can refine your analysis a little bit more. 19 When we talk about characterization, most times 20 people think about collecting data to change the parameters 21 in the analysis. But there's actually other information we 22 could also gather as part of characterization. Obviously, 23 the most important one would be to collect data to support 24 changing parameters. 25 I As I've indicated, at step one, when we initially l ANN RILEY & ASSOCIATES, LTD. O() Court Reporters 1025 Connecticut Avenue, NW, Suite 1014 Washington, D.C. 20036 (202) 842-0034
r 37 1 gather the data to begin the assessment, we use whatever 2 available information we have and some of that information 3 may not be all that relevant or tied to our particular site. 4 The preliminary analysis may help us focus in t 5 terms of which parameters are most important for the-6 analysis. So for the Kd, we may realize during our 7 preliminary analysis that the Kd is an important parameter. 8 Well, one of the options that we would consider under step 9 eight here would be to go and get some more site-specific 10 data on Kd. 11 One other thing I should have mentioned, in step 12 one, when you're gathering available data, there's a lot of i 13 data that's readily available and we're going to have some 14 discussions on this throughout the workshop. There's data ('~ 15 I V} available on the internet, there's data you can get from 16 your local soil conservation service or even your nearby 17 university. 18 But some of that data may not be particularly 19 relevant for your particular site, but it may be adequate to 20 at least begin the analysis. And as you get through doing 21 the preliminary analysis, it may help you focus on which 22 parameters you actually need to think about collecting more 1 23 site-specific data. 24 Another option on characterization would be to 25 gather information to support changing your conceptual model ANN RILEY & ASSOCIATES, LTD.
\ Court Reporters 1025 Connecticut Avenue, NW, Suite 1014 Washington, D.C. 20036 (202) 842-0034
38' 1 of the ground water system. If you've started your analysis 2 using a code like D&D, which has some severe limitations on 3 the conceptual model, or even RESRAD, for that matter, one 4 of the things you may want to consider as an option under 5 step eight would be how you could change the conceptual 6 model. So that would be another type of characterization, 7 would be to gather data to help support the changing of B that. 9 Then another one would be to gather data to 10 support changing the land use restriction, as Boby pointed 11 out earlier. Typically, we include the ground water 12 exposure pathway in the resident farmer scenario and so we 13 may do some preliminary analysis and turn out that the 14 ground water pathway is the most important pathway, where () 15 16 one of the options we may consider under step eight would be to gather some information that could help maybe support 17 eliminating that pathway, as Boby pointed out, things such 18 as the well yield or the water quality, things of that 19 nature. 20 The bottom line is that the framework provides an 21 opportunity or provides a way to help focus the 22 characterization effort. If site characterization is 23 selected as our preferred option in step ten, then we will 24 implement it in step eleven of the framework, and then 25 whatever changes that we need to make in terms of the f- ANN RILEY & ASSOCIATES, LTD. (g) Court Reporters 1025 Connecticut Avenue, NW, Suite 1014 Washington, D.C. 20036 (202) 842-0034
r-39 1 parameters or the models or even the scenario, that would be 7- 2 t]s implemented in step twelve, and then we would basically 3 reiterate back through the framework again. So it's an iterative process. 4 1 5 So that pretty much concludes my presentation and 6 I'll try to answer any questions that you have. 7 MR. NICHOLSON: Thank you very much, Mark. Are 8 there questions of Mark on the framework? 9 [No response.) 10 MR. NICHOLSON: If there are no questions, I'd I 11 like to call on our next speaker, Dr. Phil Meyer, from 12 Pacific Northwest National Laboratory. 13 Dr. Meyer and his colleague, Dr. Gee, are 14 contractors for the Office of Research. We've asked Phil to 15 give you an overview on the ground water conceptual models 16 that are inherent in the D&D, RESRAD, NEPAS and PRESTO 17 codes, and the types of information that would be needed in 18 order to use those ground water models, those conceptual 19 models, excuse me, within those dose assessment codes. 20 Phil? 21 MR. MEYER: Thank you, Tom. Can everybody hear me 22 okay in the back there? 23 I was asked to spend a few minutes talking about 24 1 ground water conceptual models of D&D, RESRAD, NEPAS and 25 PRESTO, all dose assessment codes that involve the ground i t ANN RILEY & ASSOCIATES, LTD.
\~ Court Reporters 1025 Connecticut Avenue, NW, Suite 1014 Washington, D.C. 20036 (202) 842-0034 L
1 40 water pathway. 2 O (_,/ 3 I thought since most of the authors of these codes were going to be in the audience here, 4 that I didn't want to get bogged down in details and finding out that I didn't 5 really know what I thought I was talking about. 6 So I'm going to avoid talking about most of the 7 details and emphasize the similarities between the codes and B similarities in the parameters related to the ground water 9 pathway. 10 So I thought that it might be useful, since we're 11 talking about conceptual models, to start out by asking the 12 question just what is a conceptual model, so that we can 13 have a common understanding. 14 This is one definition of a conceptual model that 15 (O.
\/
was taken from a National Academy of Sciences book on 16 fractured flow or flow in fractured rocks, and they 17 described a conceptual model as a hypothesis that describes 18 the main features of geology, the hydrology, geochemistry, 19 and the relationships between these different components. 20 This idea of a conceptual model, then, is a 21 mathematical modeling that goes on -- is a mathematical 22 modeling, is a process of hypothesis testing, whereby you 23 can propose different hypotheses for the way things work in 24 your system and then use a mathematical modeling to test 25 this hypothesis and reject or accept various hypotheses. I i l ANN RILEY & ASSOCIATES, LTD. p/ q, Court Reporters 1025 Connecticut Avenue, NW, Suite 1014 Washington, D.C. 20036 (202) 842-0034
41 1 As an example of that' sort of thing, I'd present a 2 figure from a report at the Hanford site, prepared for the 3 Hanford site. This is one of the tanks in which there is 4 high level waste, and this figure encompasses various means 5 by which the contaminants from the tank can make it down to 6 the ground water, including things like clastic dykes and 7 breaks in this caliche layer, leaks at the top. 8 So this figure describes a variety of hypotheses 9 by which contaminants may move in the environment. The rest 10 of the report went on to examine several of these hypotheses 11 using mathematical models. 12 As an alternative, perhaps a more conventional 13 version of what a' conceptual model is, it is a pictorial or 14 a qualitative description of the ground water system itself 15-and it involves the same components, hydrogeologic unit, 16 system boundaries, that is the boundary conditions, inputs 17 and outputs, sources, sinks, and all the parameters and 18 properties that are needed to completely specify that 19 system. 20 In this version, mathematical model is a 21 quantitative representation of this conceptual model. 22 Usually, these pictorial representations look 23 something like this. This is also a tank farm at Hanford 24 and the figure just shows various -- a conceptualization of 25 the way the soils vary throughout the site, underneath the
/~ ANN RILEY & ASSOCIATES, LTD. i \,.j} Court Reporters j 1025 Connecticut Avenue, NW, Suite 1014 j l Washington, D.C. 20036 )
(202) 842-0034 j
)
o 42 1 , tank farm, with questions and variations, heterogeneities. 1
,, 2 Here is another example, also, of the soils (m / 3 underlying the tank farms at Hanford, a more particular 4 site.
You can see the same sorts of structures, somewhat i 5 simplified. 6 As a further simplification, we might see 7 something like this, where the heterogeneities have been 8 evened out ! into a series of homogeneous layers, basically ' 9 rendering a 1-D flow system. 10 So these ideas of conceptual models that I have 11 presented always couple a conceptual model with a 12 mathematical model; that is, in general, the analysis to 13 take place involves both. You have a conceptual model, 1 14 which represents your ideas of the system, and a s 15 mathematical model, which implements them in a quantitative 16 way. 17 I'd just -- l I put this slide together to try to 1B illustrate some of the issues there and some of these 19 concepts. 1 The conceptual model is here and informing that 20 conceptual model, we genera.'.ly have a number of inputs. We 21 have regional or -- what I call regional or analog 22 information, which is what Mark referred to as available 23 l l information, perhaps, stuff that you can gather from other 24 sources. 25 Then, also, site-specific field observations about l ANN RILEY & ASSOCIATES, LTD. O Court Reporters 1025 Connecticut Avenue, NW, Suite 1014 Washington, D.C. 20036 (202) 842-0034 l
T f 43 l 1 properties and boundary conditions, i et cetera, for your ( 2 model. 3 These data all go into forming a conceptual model 4 and then in order to either do your hypothesis testing or to 5 do any quantitative analysis, that's coupled with a 6 mathematical model and then there's some sort of iterative 7 process whereby field observations may modify your 8 conceptual model, which, in turn, modifies the mathematical 9 model and changes your ideas about how the system operates. 10 So some basic concepts expressed very simply. 11 Two things to be aware of here. You can take the 12 same information, two people or the same person, can look at 13 one group of information and develop multiple conceptual 14 models from-the same information; that is, there's usually 15 some sort of ambiguities involved and there is no single l 16 conceptual model that will be consistent with the data. 17 Usually, there's more than one. l 18 So in this conceptual mathematical framework, as 19 Dr. Neuman will discuss later, you can end up with a 20 separate mathematical model for each conceptual model, that 21 alternative conceptual model that you can propose. 22 Alternatively, for a single conceptual model, you 23 may implement it in more than one way, in the mathematical 24 model, and that situation is quite similar to the models, 25 RESRAD and NEPAS and PRESTO. They all embody fairly similar l g ANN RILEY & ASSOCIATES, LTD. f5 Court Reporters j 1025 Connecticut Avenue, NW, Suite 1014 Washington, D.C. 20036 (202) 842-0034
r: 44 1 conceptual models, but they're implemented somewhat 2 differently. And the fact that their implementation in a 3 mathematical model is different means that you could run the 4 same -- the code on the same problem and arrive at two 5 different solutions. 6 So given that discussion, I wanted to now move 7 into a bit of specifics about those codes. Just describing 8 what I see as the similarities in the ground water 9 conceptual models for these codes. 10 This is a figure that represents the D&D model and 11 that's already been discussed somewhat by Boby Eid, but what 12 . we basically have is a contaminated zone here, a vadose 13 zone, which is a simple box, an aquifer and perhaps a 14 surface water pond. That 's D&D.
/)
NJ 15 This is a figure here taken from the RESRAD 5.0 16 manual and we see something very similar. There is a 17 contaminated zone, an unsaturated zone, a saturated zone, 18 and perhaps a surface water pond. The representation here, 19 although they're somewhat different, there's a lot of 20 similarities; that is, the contaminated zone is assumed to 21 be fairly uniform. 22 The unsaturated zone is either uniform or a series 23 of layers, each layer which is uniform. This figure could 24 also be used for MEPAS or PRESTO. ' They are very similar. 25 The exposures are also fairly similar. There's p ANN RILEY & ASSOCIATES, LTD. () Court Reporters 1025 Connecticut Avenue, NW, Suite 1014 Washington, D.C. 20036 (202) 842-0034
45 1 usually a well maybe on-site or further off-site, and in the 2 aquifer, the mixing that goes on there is described perhaps 3 differently, but in similar ways. 4 r So just to summarize these, I have a slide that I 5 believe highlights the major assumptions of these models 6 that are held in common by all the models. That is, they 7 have a simple near-surface water budget. So the input water 8 to the system that drives the transport is all described in l 9
-- described in every case by a very simple water budget 10 model.
11 ! The codes all assume that flow in the system is 12 . steady-state; that is, it's constant with time. The flow is ! 13 one dimensional throughout the system and advective 14 transport is also one dimensional, down through the 15 unsaturated zone and one dimensional in the saturated zone. 16 There is a small number of layers, contaminated 17 zone layer, maybe a few vadose zone layers, and aquifer, and 18 each of these layers has uniform properties. In addition, 19 they all assume some fairly simple aquifer mixing to 20 determine what happens in the transfer of contaminants from 21 the unsaturated zone to the saturated zone. 22 So given those assumptions, we can ask ourselves, 23 well, what are the site-specific conditions that might-24 violate those assumptions. I list some of them here. I 25 don't think this list is all inclusive, but I did identify l ANN RILEY & ASSOCIATES, LTD. O, Court Reporters 1025 Connecticut Avenue, NW, Suite 1014 Washington, D.C. 20036 (202) 842-0034 l l t
46 1 thase conditions. 2 Preferential flow, that is, flow in the near ' 3 surface that could lead to enhanced infiltration that 4 wouldn't be considered by the simple water budget models. 5 Transient flow could lead to changes in dose that would be 6 significant and violate the assumptions of these models. If 7 we had significant heterogeneity, this was mentioned by Boby 8 earlier, two or three dimensional flow, focused f] v, fast 9 paths, and fractured formations. These all lead to 10 11 conditions that concentrate flow in small areas instead of spreading it out as the codes all assume, 12 So I'm going to finish my talk just by going 13 through the parameters of these models that are basically 14 all held in common. This will lead into later talks by 15 other people and also by myself that will discuss the 16 parameters in detail, data sources, that sort of thing. 17 The near surface hydrologic input, this deals with 18 the water balance, the water budget at the surface, 19 determining how much water goes into the system. i 20 Specify yearly net infiltration rate. So that's 21 the purpose of this component and all the codes require this 22 net infiltration, which I define as the amount of water that 23 moves past the root zoner so the amount of water that's 24 going to contact the waste and contaminants and produce the 25 contamination that potentially will result in exposure. ANN RILEY & ASSOCIATES, LTD. O Court Reporters 1025 Connecticut Avenue, NW, Suite 1014 Washington, D.C. 20036 (202) 842-0034
47 1 Basic water budget is solved by all these codes in 2 some way or another. Infiltration equals precipitation 3 minus a combination of runoff and evapotranspiration, or 4 water used by plant's. Irrigation may be included, if 5 desired. 6 The parameters here, net infiltration rate, for 7 instance, D&D requires that you just simply input the net 8 infiltration rate. 9 If this is calculated actually from the 10 relationship like this, then it requires you to input ' i 11 precipitation, either as an average annual value or perhaps 12 as a monthly average value. Runoff needs to be calculated 13 somehow, either specified directly as a fraction of 14 precipitation or else calculated using something like the 15 SCS runoff method, which requires a curve number. \~'# 16 Evapotranspiration could be either specified 17
.directly as a coefficient, again, or else calculated in some '
l 18 manner from a variety of meteorological parameters. 19 In the contaminated zone, a similar situation j 20 exists. All the codes basically have -- require the very i 21 similar parameters. The purpose here is to specify the ! 22 contaminant flux tt the unsaturated zone, given the net 23 infiltration rate. 24 i The leaching models vary somewhat between the ! 25 codes, but are relatively similar. The parameters required i ANN RILEY & ASSOCIATES, LTD. {~'g Court Reporters s ,/ 1025 Connecticut Avenue, NW, Suite 1014 Washington, D.C. 20036 (202) 842-0034 ,
48 1 in terms of calculating this contaminant flux are thickness 2 of the zone, the area of the contaminated zone, bulk 3 density, the porosity and potentially the effect of 4 porosity, which are related, and the water content. 5 Now, the water content may either be specified 6 directly or else it might be calculated using a simple 7 assumption on the nature of the flow and the variety of a soil, specific parameters. 9 Then the distribution coefficients for each 10 contaminant are also required. 11 Again, those parameters are required by all the
-12 codes in one manner or another.
13 With the unsaturated zone, the purpose here is 14 simply to continue the transport, contaminant transport to 15 the saturated zone, determine the_ flux. C) 16 The parameters required, again, are the number of layers, the thickness of 17 each one of those layers, and then within each layer, bulk 18 density, again, porosity or effect of porosity; a water 19 content value, which, again, may be either specified or may 20 be calculated using a simple relationship. 21 Distribution coefficients and then in some cases, 22 a dispersivity value may be an option. 23 Very similar in the saturated zone. The purpose 24 here is to calculate the flux to the well or the surface 25 water, and, Charlie, you pointed out to me that the RESRAD 5004 RILEY & ASSOCIATES, LTD. O Court Reporters 1025 Connecticut Avenue, NW, Suite 1014 Washington, D.C. 20036 (202) 842-0034 l
1 49 mass balance model does, in fact, consider -- although all 2 O the contaminants end up in the well, the mass balance model k-s/ 3 does consider the time to the peak -- time to peak dose in 4 the saturated zone itself. 5 The parameters required are basically the same as 6 in the unsaturated zone, thickness, bulk density, porosity, 7 specific discharge. Again, this is like water content. 8 Specific discharge may be either specified directly or 9 calculated using a simple relationship and a couple of 10 parameters. Distribution coefficients and then 11 dispersivities. Dispersivities may be multiple, because 12 some of the codes allow dispersion in the longitudinal 13 direction, as well as transverse. 14 There are just a few other parameters related to 15 the ground water pathway, related to exposure terms; that 16 is, the depth of the well, distance to the well, and the 17 well pumping rate. These parameters aren't required by all 18 i of the codes, but in some manner, they have parameters 19 involved that determine basically dilution in the well, 1 20 whether or not it occurs and the effect of dispersion. 21 So in conclusion, I would like to just point out ' 22 that the way these models may be used, and you might be 23 tempted to use them, is to do the opposite of this, where 24 you actually use the -- accepc the model before you set the 25 conceptual model for your site, and it should be the other ANN RILEY & ASSOCIATES, LTD. O- Court Reporters 1025 Connecticut Avenue, NW, Suite 1014 Washington, D.C. 20036 (202) 842-0034
1 50 1 way around;.you want.to fit your mathematical model to your 2 conceptual model, not vice versa. O 3 So if you determine that your site does not fit 4 the conceptual model as one of these codes, then consider 5 very carefully before you decide to use one of these codes 6 and consider what the implications of doing that are. 7 As I've discussed and mentioned many times here in 8 my short talk, these codes are all very similar. They 9 embody similar conceptual models and the parameters required i 10 are very similar. i 11 Now, that doesn't mean, as I mentioned before, 12 because the implementations are different, it doesn't mean 13 that they're going to give very similar identical results. 14 A number of comparison studies do exist between 15 these codes and I list a couple of them here. ! This is a 16 draft report that I believe is available on the NRC 17 decommissioning web site. Charlie, you also mentioned to me 18 a paper and a report with additional RESRAD comparisons and 19 I can give you that reference or Charlie can give that to 20 you, if you're interested. 21 Also, this NUREG here compares D&D to a variety to 22 a variety of numerical and hybrid codes. 23 So are there any questions about what I've talked 24 about? 25 MR. NICHOLSON: Thank you very much, Phil. Are O '. ANN RILEY & ASSOCIATES, LTD. Court Reporters 1025 Connecticut Avenue, NW, Suite 1014 Washington, D.C. 20036 (202) 842-0034 l i
F I 1 L1 there any questions for Phil on the conceptual models within 2 RESRAD, MEPAS, D&D and PRESTO? A question from John 3 Greeves. ; 4 MR. GREEVES: Yes, to loosen this group up a 1 5 little bit. Phil, either you or other speakers, as you go 6 through the day, both Mark and Boby talked about a ! 7 transition from the fairly simple -- actually, going from a 8 screening case for the lucky licensees who don't have 9 complicated sites, but then you step into the D&D and RESRAD 10 type models and you showed us a quite interesting picture of 11 the Hanford site. 12 My question to you and others, as you go through 13 the meeting, can you give us some insights as to whac the 14 experience base is in going from RESRAD to a more 15 O 16 complicated model and particularly out at Hanford, if you have that knowledge, what type of modeling is used out there 17 and under what circumstances could you select these various ! 18 models? 19 I can imagine, even on the Hanford site, at some 20 locations, you could use screening technique. Others, you 21 might be able to use RESRAD, but for the tank you just put 22 up there earlier, I have my doubts about using some of these 23 simple models. 24 So do you have any experience base on that, you 25 and other speakers? As you present, I think it would be ANN RILEY & ASSOCIATES, LTD. O Court Reporters 1025 Connecticut Avenue, NW, Suite 1014 Washington, D.C. 20036 (202) 842-0034
1 52 useful if you share. And it's really the 1549 methodology. 2 () 3 You walk through that and there are places that it says you can't use this model, it's too simple, it doesn't match the 4 conceptual model, you need to go to a more site-specific 5 model. 6 So, again, if there is some background you have on 7 that, or other speakers, I think the audience would 8 appreciate hearing how that process takes place. 9 MR. MEYER: Sure. I can just offer a couple of 10 11 comments related to the Hanford site and also discussions that we had in the research review meeting on Monday and 12 Tuesday. 13 In general, and this is the case at the Hanford 14 site, in fact, at the tank farms, MEPAS has been applied 15 fairly extensively. In general, you assume the simplicity, 16 unless your -- at least this is the way things work in 17 practice. 18 You assume the simple case, unless you are given 19 evidence to suggest that things are not quite so simple. 20 And in the case of the Hanford site, it was evidence of 21 contamination in the ground water that violated people's 22 assumptions about the way the unsaturated zone behaved, as 23 the contaminants showed up when they weren't expected to. 24 And this caused people, some people -- of course, 25 there were some people who were already thinking this, but L ("N ANN RILEY & ASSOCIATES, LTD. (' Court Reporters 1 1025 Connecticut Avenue, NW, Elite 1014 ! Washington, D.C. 20036 (202) 842-0034 i l
53 1 it caused many people to start asking questions about the i p_ 2 representation of the unsaturated zone and so then, at that k -) 3 point, because there were observations, field observations 4 that didn't match the conceptual model, then you had -- it 5 had to be that the conceptual mcdel was wrong and it needed 6 to be modified in order to explain the field observations . 7 The difficulty, as we discussed yesterday at 8 length, is that in a site characterization, if you don't go 9 out looking for things that violate your conceptual model, 10 I then you will never find them until perhaps sometime far in 11 the future when the contamination is some place potentially 12 where it's totally not expected to be. 13 That is the difficult issue of trying to do an 14 efficient site characterization and still look for things gg 15 that violate your conceptual model. U 16 I'm sure other people here in the audience can 17 offer other ideas. 18 MR. GREEVES: Just as a follow-up, I thought I had 19 heard that Hanford was developing a site-specific model. 20 Now, was that a version of MEPAS or -- I'm a little beyond 21 my depth here, but I thought I'd try -- 22 MR. MEYER: Hanford is a huge site and it involves 23 a lot of smaller sites which are having analyses and models 24 developed for specific sites. As far as the unsaturated 25 ) zone modeling, most of that or all of it that I'm aware of I I l l i ANN RILEY & ASSOCIATES, LTD. } (T (_,/ Court Reporters k 1025 Connecticut Avenue, NW, Suite 1014 t Washington, D.C. 20036 (202) 842-0034 I
- ~~9 1 i
i 54 1 has been on a site-specific basis, small-scale. Some of the 3 2 ground water modeling has been for the entire site. 3 MR. NICHOLSON: Again, a question from the floor. 4 Would you identify yourself, please, sir? 5 MR TAYLOR: My name is Stew Taylor. I'm with 6 Bechtel Power Corporation. A follow-on question. The suite 7 of models you presented, could you clarify what assumptions 8 regarding the initial state of the saturated zone, whether 9 it's contaminated or uncontaminated, what's the underlying : 10 assumption? 11 MR. MEYER: I believe, and some of the authors of 12 { the code could correct me on this, but I believe that with 13 MEPAS, I'm pretty sure you can specify, if you have some 14 information to specify the distribution of contaminants in 15 the saturated zone, you can perform a simulation with the 16 contaminated zone that's actually within the saturated zone. 17 With RESRAD, I think that it's been -- it's 18 capable of doing that, they've done that in a comparison, 19 I'm pretty sure, but I'm not sure how straightforward that 20 is. And with PRESTO, I'm not aware of that. 21 The D&D, it definitely violates the assumptions. 22 MR. TAYLOR: So if there was an aquifer -- or a 23 site with an aquifer that was already contaminated, we'd be 24 either driven to use one of the codes that allows the 25 initial condition of contamination or, alternatively, go to ANN RILEY & ASSOCIATES, LTD. T
/ Court Reporters .
1025 Connecticut Avenue, NW, Suite 1014 Washington, D.C. 20036 (202) 842-0034
I 55 1 a more complex numerical model, perhaps. 2 O 3 MR. MEYER: Definitely, if you have contamination in the aquifer, then you should be using a code that allows 4 that. D&D does not, as Boby pointed out in his talk, so you 5 would not be able to use D&D in that situation. 6 MR. TAYLOR: Thank you. 7 MR. WOLBARS: Tony Wolbars, from EPA. Does that 8 mean that a reasonable way to check your code would be to 9 start the clock running at a much earlier time and see if 10 you end up with a contaminated aquifer? 11 MR. MEYER: Some people have done that sort of 12 thing. I'm not sure if -- this may be a discussion that we 13 want to carry some other time. We're getting a little bit 14 sidetracked. 15 If you haven't satisfied the conceptual model of 16
-- if your site does not satisfy the conceptual model 17 requirements of the code, then doing the sort of thing that 18 you're talking about could lead you to an erroneous 19 conclusion nonetheless.
20 I'm not sure if I'm answering your question, bat i 21 we could maybe discuss this a bit later. 22 MR. NICHOLSON: I think during the group 23 discussion. We're going to take a break in about a minute 24 here. There's a couple of things I wanted to point out. 25 First of all, as Phil points out, the authors of the codes p-~ ANN RILEY & ASSOCIATES, LTD. i' Court Reporters 1025 Connecticut Avenue, NW, Suite 1014 Washington, D.C. 20036 (202) 842-0034 i i L I
56 1 are here and they'll be making presentations, talking about
, 2 their conceptualization.
Dr. Cheng Hung from EPA will be
\'
3 the next speaker after the break. 4 Also, I want to point out that some of the 5 viewgraphs were missing this morning. We're having copies 6 made. They're out on the table out in the foyer. So by all 7 means, pick those up. 8 Are there any other questions before we break? 9 [No response.] 10 MR. NICHOLSON: Okay. Let's take a 15-minute 11 break. Be back here at cbout 10:20. Thank you. 12 [ Recess.) 13 MR. NICHOLSON: As people are sitting down, I want 14 to let you know that some of you haven't been able to sign 15 in.
) We'd really appreciate it, because we're keeping a 16 record of this public workshop and it will be going to the 17 public document in about a week, and it's important that we 18 know everyone who attended.
19 So those of you who haven't signed in, on the 20 table in the foyer of the auditorium, in the back there, ! 21 . there is a signup sheet. So if you could please put your l 22 name down, it will be part of the transcript record, because 23 we want to make sure of everyone who is here. If you'd 24 accommodate us, I'd appreciate that. 25 I'd like to now continue the discussions. This i ANN RILEY & ASSOCIATES, LTD. O g Court Reporters 1025 Connecticut Avenue, NW, Suite 1014 Washington, D.C. 20036 ' (202) 842-0034
57 1 morning, we were lucky to hear from NMSS staff on what their f3 2 view is on decommissioning and on the framework, and then U 3 Phil was kind enough to walk through the conceptual models 4 for the conventional dose assessment codes, and now we're 5 going to talk in particular about one of those codes, the 6 PRESTO code. 7 Our next speaker is Dr. Cheng Hung. He's from the 8 US Environmental Protection Agency, the Office of Radiation 9 Protection and Indoor Air, and, Dr. Hung, you have the 10 floor. 11 MR. HUNG: Thank you. Good morning. First of 12 all, I would like to say thank you to NRC staff members for 13 letting me speak on these topics. As you all know, the 14 1 ground water pathway is one of the most important pathways 15 in the risk assessment model, and more so, the calculation 16 of the well water concentration, pumping out from the pool 17 is also an important process. 18 So my presentation will start with a theoretical 19 background of this calculation and going to the practical 20 modeling. In that section, I would like to present the 21 upper bound model and semi-dynamic model. 22 Following that, I would like to say the difference 23 between the results of the analysis between these two models ! 24 and make the conclusion of that. 25 First of all, the system we are considering is a I l ANN RILEY & ASSOCIATES, LTD. tO Court Reporters 1025 Connecticut Avenue, NW, Suite 1014 i 1 Washington, D.C. 20036 (202) 842-0034 j
58 ! lAndinthesystem,weassume I steady, uniform f]ow condition in the aquifer. 2 O' 3 will be the dominant injection, at the optimum, and then at 4 the boundary, well have a well scheme sink. With this kind 5 of schematic, we could write the equation that's shown here. 6 Because of the steady uniform flow, all we need is a mass 7 balance equation. 8 In this equation, we would include that they are 4 9 diffusion term, convective term, and decay term. For this ! 10 particular case, we used a line source injection and at the 11 sink, we assume that they will screen as a sink. I 12 So as you know, trying to solve this equation, we 13 would need that three-dimensional model. One of the 14 examples that we use in there, flow surface model that the 15
-- we all know that trying to integrate this type of complex 16 three-dimensional model into this screening type of risk 17 assessment model is impossible. i 18 So we at the practical point, we do need a 19 practical model.
In the practical model, all I see are all i 20 these models would divide into two important steps. 21 The first one is the transport through the 22 aquifer. In this calculation, we're trying to calculate the 23 transport through the section of interest or to analyze for 24 the plume. At the second step, we're trying to use the 25 result that we calculate from there, from the first step, ANN RILEY & ASSOCIATES, LTD. O Court Reporters 1025 Connecticut Avenue, NW, Suite 1014 Washington, D.C. 20036 (202) 842-0034
59 1 2 and to calculate the concentration of the water in the well. Trying to calculate the transport through the 3 aquifer, the basic equation will be the same that we 4 originally had, but we don't really need the sink as a 5 boundary condition. 6 So in the practical model, we're trying to 7 simplify our calculation and all the models, like in the 8 previous presentation, different models have different types 9 of simplification. 10 So MEPAS has their own simplification and RESRAD 11 has its own simplification, and the PRESTO model also has a 12 different simplification. 13 We will talk about that a little bit later. 14 The detail, just that the simplification is not 15 the scope of this talk, so I'm going to -- I'm not going to O 16 talk about this. We'll concentrate on the transport through 17 the pumping well. f i 4 18 i As you know, trying to define the equation to 19 calculate the well concentration is quite complex. I 20 couldn't find any governing equation. So the purpose of i 21 this talk, I would use the dimensional technique just to 22 show the functions, it's shown here. 23 This equation simply says that the concentration 24 in the well is a function of Cw, the radionuclide, the 25 concentration in the well water, the transport through the ANN RILEY & ASSOCIATES, LTD. ( Court Reporters 1025 Connecticut Avenue, NW, Suite 1014 Washington, D.C. 20036 (202) 842-0034
1 60 wall section, cnd blah, blah, blah, blah. 2 So this is just to show the function, and I (O) 3 couldn't find the governing, equation for that. 4 5 So we have to based on this to see how all the model transport, I simplified that. 6 model, So as I look into the 7 there's two types of models that are currently being used. 8 The first one is the upper bound model and the 9 second one I call the semi-dynamic model. 10 So I would like to go into the detail of these two types of models . 11 First of all, let's see what the upper bound model 12 do. The upper bound model, trying to simplify it just that 13 the calculation and say just calculate the upper bound It . 14 means the maximum concentration that a person drilling a 15 well at that location, what the maximum concentration you 16 could get. 17 The way to calculate it, theoretically, using the I 18 screen, the well screen, and calculate the moving average 19 concentration when we move in this and get the maximum for 20 that. This is the way to calculate the maximum equation
.21 concentration.
22 So you know that this is a rather simple way to do 23 it,- but you calculate the maximum probable equation or 24 concentration. The scenario of this occurrence will be 25 minimal. l l rx ANN RILEY & ASSOCIATES, LTD. ! t Court Reporters 1025 Connecticut Avenue, NW, Suite 1014 Washington, D.C. 20036 (202) 842-0034
61 1 Let's move into the semi-dynamic model, which is 2 what the PRESTO model used. Before that, I would like to x_/ 3 talk about the transport through the aquifer, how our PRESTO 4 model calculates that. The PRESTO model uses a 5 one-dimensional model to calculate the transport through the 6 section at the well. The PRESTO model uses a dynamical 7 solution that will be a simple equation. This is the 8 boundary conditions. 9 Using the boundary condition, this is the 10 radioactive decay term. We suggest that you could have some 11 sort of -- would have some sort of error. So we needed one 12 correctional factor. So this correctional factor was 13 derived at, and it's in our paper and I'm not going to 14 describe that here.
-- 15 A- 16 but this includes a function -- this correctional factor is site specific data.
You can calculate from the 17 site-specific data, that P is the number, probably people 18 are not too familiar with that number. That's involved at 19 the diffusion term. 20 So we suggest Tna can reasonably use the simple 21 calculation to get the rate of radionuclide transport. 22 Of course, being a one-dimensional model, we don't 23 calculate that the -- we cannot calculate that the plume. 24 So now I -- let's go into how the PRESTO model calculates 25 the transports through the pumping well. s ANN RILEY & ASSOCIATES, LTD.
,/ Court Reporters 1025 Connecticut Avenue, NW, Suite 1014 Washington, D.C. 20036 ,
(202) 842-0034 ! i
l 62 1 The first step is trying -- we have to design the
,, 2 configuration of the well. In designing this, we have a
( 3 concern. It's that the well will be drilled by a licensed 4 well driller. That's the best under typical state 5 regulations. 6 That meant we suggest that if the driller is 7 putting in the drilling machine to the site, it means e additional costs due to incremental well depths is minimal. 9 The increase in the depth is just the drilling machine, if 10 they take a few minutes, and they had a steel pipe, it's at 11 the minimum. 12 So all this -- that the well driller and well 13 owner would like to go into as deep as possible to secure 14 the water quality and the water quantity. 15 So you know -- everybody.know that in our case,
- O 16 the risk assessment considers that contamination is in the 17 vadose zone. So in order to get a better quality, we have 18 to go the well as deep ao possible, 19 The third assumption is that we try to do a -- 1 20 simply by that calculation, we assume just a well. The 21 screen depth -- actually, the screen depth would be just at 22 the bottom parts, but we assume that the whole depths will 23 be screened.
24-The reason for that is a conservative assumption. 25 With this kind of calculation, if you compare it with a ANN RILEY & ASSOCIATES, LTD. Court Reporters 1025 Connecticut Avenue, NW, Suite 1014 Washington, D.C. 20036 (202) 842-0034 ' l
F 63 1 three-dimensional model, it's screen depth at the bottom, 2 fs that would get a higher concentration.
's 3 The reason I say conservative is that it's higher, 4
but if you use that three-dimensional model to calculate the 5 screen depth at the bottom, then you get much, much less 6 concentration, but it's not the maximum concentration that 7 the critical population may expose. B When you move downstream, then this concentration 9 would increase. So we use, in the PRESTO model, this is the 10 { upper limit of the concentration when you move that 11 downstream. 12 So I would say this is a conservative assumption. 13 With this kind of assumption, then the previous 14 function can be simplified into a very, very simple form, 15 and with this form, we can find that they are -- the 16 equation for calculating the concentration in the well is ' 17 ! transport divided by available dilution of water. So it's a ; 18 very, very simple calculation. 1 19 So let's compare the two models. We can see the 20 upper bound model is very easy to calculate the 21 concentration, _there is no question about that, and -- but 22 it would give an overly conservative, compared with the 23 PRESTO model, overly conservative dose values. And the 24 important thing is that the third one, it may unnecessarily 25 boost the cost of disposal or decommissioning or cleaning ANN RILEY & ASSOCIATES, LTD. O Court Reporters 1025 Connecticut Avenue, NW, Suite 1014 i Washington, D.C. 20036 ! (202) 842-0034
64 1 up. e- 2 k_'S I On the other hand, we look at the PRESTO model, 3-the semi-dynamical model, as we discussed it before, it 4 doesn't really add too much process in the calculation. So 5 it's reasonably easy to calculate the concentration. 6 Secondly, it gives reasonably conservative, this 7 still conservative dose to barriers. 'That's a very 8 important thing. It's still conservative. 9 The third one could result in the reasonable cost 10 of disposal or decommissioning or clean-up. 11 With this kind of difference, we would like to see 12 how much is the difference. We use just two models. What's 13 the difference? So what I took one example that's just from 14 the NRC report,
. that's the analysis showing -- analyzed just
(~ 15 V) at the plume. I used that as an example. And then assuming 16 that the well is located at the X=80, that location, and for 17 calculating the moving average, I used the screen depth of 18 one meter and for the PRESTO model, we assumed a depth all 19 the way down to near bottom, about two feet above the bottom 20 of the aquifer. 21 So the screen will be 18 meters for the PRESTO 22 model and the result, we can see that the upper bound model - 23 would be 0.31 units per cubic meter and for the PRESTO model 24 we got .056 units per cubic meter. That's a big difference. 25 The difference is about a factor of 5.5. That's a big fw ANN RILEY & ASSOCIATES, LTD. ( ,) Court Reporters i 1025 Connecticut Avenue, NW, Suite 1014 Washington, D.C. 20036 (202) 842-0034 l l
I 65 l difference. 2 So from this discussion, we'd like to make a 3 conclusion. j 4 The first is that the 3-D model used in the calculation for well water concentration is extremely 5 complex, that everybody should agree with this, and the 6 second is that the simplified upper bound and the 7 semi-dynamical model are commonly used, a practical model. 8 And the third is that the upper bound model is easy to use , 9 but you get overly conservative results and boost up the 10 disposal cost. q 11 The last one would be the use of semi-dynamic 12 model may be slightly more complex, but it gives a 13 reasonable conservative dose, and could result in reasonable 14 disposal costs. 7g 15 So that concludes my presentation.
'~s 16 MR. NICHOLSON:
Thank you very much, Dr. Hung. We 17 are going to have copies of his viewgraphs. They aren't in 18 the back, but we're going to get a copy of these and provide 19 them to you after lunch. 20 Are there questions of Dr. Hung on his 21 conceptualization and calculation of well water contaminants 22 for a given plume? Yes, Dr. Neuman, from the University of 23 Arizona. 24 l DR. NEUMAN: Dr. Hung, I have two technical 25 questions. With respect to your statement that the ANN RILEY & ASSOCIATES, LTD. ("N Court Reporters 1025 Connecticut Avenue, NW, Suite 1014 Washington, D.C. 20036 (202) 842-0034
66 1 calculation is conservative if you assume DW to be the 2 maximum. /
) I don't understand that.
' _/. ( 3 It seems to me that if you had a screen which just 4 touches the plume and does not draw clear water to it, you 5 should get a higher concentration; therefore, a more 6 conservative result. 7 In fact, I'm looking at your equation, CW equal to 8 QP divided by V times DW, C is inversely proportional to DW. 9 Let me also ask the second question and then you 10 can maybe answer both of them, and I'll sit down. 11 My second question is when a well pumps, it 12 develops a. radial flow regime. In a radial flow regime, the 13 velocity is not constant. So V is not a constant. 14 Your calculation seems to be a two-dimensional 15 one. How do the two relate? 16 MR. HUNG: That's a good question. See, the first 17 question, as I mentioned, you use the bottom screen, near 18 the bottom, because you know the plume is on the top. If l 19 you use the three-dimensional model to calculate, they draw 20 most of the water from the bottom. 21 That would go down a bit, but mostly it comes from 22 the bottom. We use the model and calculate it and it's 23 small. 24 The second question is that -- I 25 DR. NEUMAN: Radial flow. ANN RILEY & ASSOCIATES, LTD. '\,, /)N Court Reporters 1025 Connecticut Avenue, NW, Suite 1014 Washington, D.C. 20036 (202) 842-0034
67 1 MR. HUNG: Also, if you look at the plume and the 2 potential line and flow line using the 3-D model, you will
\~# 3 see that it narrows down into that area. So actually, 4 although you get it -- most of the flow is coming from 5 nearby the well.
6 MR. NICHOLSON: Are there any other questions for 7 Dr. Hung? 8 [No response.] 9 MR. NICHOLSON: Thank you very much. And as Dr. 10 Hung indicated, he is going to be submitting a paper for the 11 proceedings and his references in his discussion of the 3-D 12 code you were talking about. 13 He'll have a reference to that. That was 14 developed by Hydrogeologic Inc. for us many years ago in the 15 Office of Research. 16 I'd like to now introduce our next speaker. Our 17 next speaker is from Argonne National Laboratory, Charlie 18 Yu, Charlie is going to be talking about ground water and 19 radionuclide transport model used in RESRAD off-site. So 20 this is the second conceptual model that Phil had talked 21 about. First is PRESTO. Now we're going to hear about 22 PRESTO on-site -- excuse me -- RESRAD off-site. 23 MR. YU: My name is Charlie Yu. I'm the RESRAD 24 Program Manager at Argonne National Laboratory. I will give 25 the introduction of the ground water transport model using gg ANN RILEY & ASSOCIATES, LTD.
\ ')
( Court Reporters 1025 Connecticut Avenue, NW, Suite 1014 Washington, D.C. 20036 (202) 842-0034
l. 68 1 1 RESRAD off-site and then my colleague, Dr. Emmanuel 2 3 Gnanapragasam is going to talk about the details of the
-- 3 models that we implemented in RESRAD off-site.
4 i ! As you probably already know, we have three 5 computer models. l We call it the RESRAD' family of codes. ' 6 RESRAD itself is an off-site model. You can see that the 7 well is located at the site boundary. Although we have two l 8- options in RESRAD. One is a mass balance model, which 9 assumes the well is located at the center of the i 10 contaminated zone, but the default model is that the well is 11 located at the site boundary. We call it non-dispersive. 12 It's considered advection, but dispersion is not considered 13 in the on-site model. 14 If receptors are located at off-site locations 15 away from the contaminated zone, as shown here in this (~)T 16 picture, the well is located off-site. There is a distance 17 from the sell to the contaminated zone. 18 In this situation, if you don't consider 19 ! dispersion, you are too conservative, it will give you a 20 much higher dose or much higher water concentrations. 21 So several years ago, we conducted a benchmarking 22 comparing RESRAD, the on-site model, with MEPAS, the PRESTO 23 code, and MMSOILS code, which is another EPA model which 24 deals with hazardous chemicals. l 25 At that time, we compared the results and we ANN RILEY & ASSOCIATES, LTD. Court Reporters 1025 Connecticut Avenue, NW, Suite 1014 Washington, D.C. 20036 (202) 842-0034 L
69 1
.showed that there is some difference, significant difference 2 if you don't consider dispersion. I'm going to show you 3
some of the results, the benchmarking results that we did a
- 4. couple, two or three years ago.
5 Before I do that, I want to mention that the 6 RESRAD, this is an on-site model picture, if we move the 7 well off-site for the transport, we developed a new model B including three-dimensional dispersion for the model. But 9 for air dispersion, when people are living off-site, the air 10 dispersion pathway is simply used, CAP-80, which is the EPA 11 model, to handle the air dispersion part of the transport of 12 contaminants. 13 This are the results that we did several years 14 ago, two or three years ago, comparing RESRAD, MEPAS and 15
) MMSOILS, PRESTO code, that was three years ago. So by now, 16 probably all three codes -- the other codes are probably 17 already modified, but I'm not quite sure yet.
18 Here it shows that MEPAS and MMSOILS concentraticn 19 profile here, they pretty much overlap, and this is the 20 concentration of strontium-90, a parent nuclide that does 21 not have daughters. The advection dispersion equation using 22 MEPAS and MMSOILS is pretty much the same. So they've got
- 23. pretty much the same results.
24 RESRAD, we took the opportunity when we did this 25 benchmarking, we added dispersion to the model and we can ANN RILEY & ASSOCIATES, LTD. Court Reporters 1025 Connecticut Avenue, NW, Suite 1014 Washington, D.C. 20036 (202) 842-0034
70 1 get this curve here, which the peak dose is lower and the 2 s-) 3 reason is that the definition of the retardation factor in 4 RESRAD is different from the definition of the retardation factor you see in MEPAS and MMSOILS. We used the total 5 porosity and other codes use the effective porosity and so 6 on. 7 So if we modify the retardation factor, we can get 8 this curve. This simply shows that the dispersion model 9 incorporated into the RESRAD off-site, we can get comparable 10 results compared with other models. 11 That's for parent radionuclides. Let me show you 12 some daughter nuclides results. This one in a source, we 13 have uranium-234 and uranium-234 decays to thorium-230, and 14 this is the concentration profile for thorium-230. 15 Initially, in the source, there is no thorium-230. And 16 uranium has a KD-7, in this case, seven milliliter per gram 17 and thorium has 30 milliliter per gram. 18 So uranium has lower KD, is less absorbed in soil 19 during transport and so when uranium transports from source 20 to well, it decays and thorium-230 was generated, and 21 thorium-230 has higher KD, so it's highly absorbed, 22 transported slowly towards the well. 23 And this is the results that we got over here, and 24 the MEPAS results are here, and the MMSOILS is here. 25 I mentioned that MMSOILS and MEPAS, they use
\
ANN RILEY & ASSOCIATES, LTD. s/ Court Reporters 1025 Connecticut Avenue, NW, Suite 1014 Washington, D.C. 20036 (202) 842-0034
1 1 71 pretty much the same dispersion advection equation. So 2 () 3 MMSOILS should get a curve like this, but MMSOILS modelers they adjust their concentration by the retardation factor of 4 daughter and parents to get this height, which is comparable 5 to what RESRAD got. 6 And this curve is another result done by MMSOILS 7 modelers. They used another numerical model, I'll call it B DPM. 9 This numerical'model can handle three-dimensional dispersion and decay or ingrowth of daughters, but they got 10 this shape. So they got this daughter thorium-230 tail, 11 which has a similar shape as what RESRAD models that way . 12 So we expected, because thorium-230 has lower KD, 13 so you expect that they move slowly towards the well, so you 14 expect, with the tail, something like this. 15 And this is the model that we developed at that 16 time for benchmarking studies and then they have 17 l incorporated this model into the RESRAD off. site model. So 18 RESRAD off-site has this model in the code, but this model 19 is not in RESRAD on-site model. 20 If you would like to get a copy of the RESRAD 21 off-site model, we can send it to you. You can send it 22 e-mail to us. Actually, eventually, we're going to put the 23 whole RESRAD family of codes on the web, so you can download 24 it. 25 The RESRAD model, a new model is coming out, t s ANN RILEY & ASSOCIATES, LTD. ( Court Reporters 1025 Connecticut Avenue, NW, Suite 1014 Washington, D.C. 20036 (202) 842-0034 l l i
72 1 varsion 5.9, and will be on the web sometime next month, 2 probably July 1st. I 3 That's my brief introduction and now Dr. 4 l Gnanapragasam is going to talk about the details of the 5 three-dimensional dispersion model, implementing RESRAD 6 off-site. 7 MR. GNANAPRAGASAM: I will talk about the ground 8 water transport model, both the movement of water and also 9 the contaminants. The first two slides show how we modeled 10 the movement of water in the RESRAD. First, in the 11 unsaturated zone, we take the infiltration rate and use the 12 soil moisture, saturated soil moisture, hydraulic 13 conductivity relationship to calculate the saturation ratio 14 that is needed to support the infiltration rate that we have 15 at the site. 16 We checked to make sure that this satisfies the 17 field capacity constraints that may be imposed. Then we 18 applied the saturation ratio that we calculated to the 19 effective porosity that the user might specify to find out 20 the amount of moisture that will actually participate in the 21 movement of water, and we divide the infiltration rate by 22 that moisture content to find out the actual rate at which 23 water moves in the unsaturated zone. 24 For the saturated zone, we get the inputs, the 25 hydraulic conductivity and the hydraulic gradient and we ANN RILEY & ASSOCIATES, LTD. f) \_/ Court Reporters 1025 Connecticut Avenue, NW, Suite 1014 Washington, D.C. 20036 (202) 842-0034 I i I
r 73 1 compute the velocity of movement of water, divided by the 2 effective porosity in the saturated zone, to get the actual 3 rate at which water moves in the saturated zone. 4 Moving on to how we compute transport of
,5 radionuclides, we have two different models, one for
! the 6 parent -- that is, i the nuclide that enters and exits the
?
l zone in the same form -- and a different model for progeny, 8 l where nuclides parent enters the zone and then we're 9 i interested in the progeny coming out. 10 For the parent, we have this transport equation, 11 which considers the dispersion in the longitudinal direction 12 and also retardation of the nuclide. Our definition of the ! 13 g retardation factor is shown here and we have an equation for I l 14 3utput flux. 15 For the progeny nuclide, the situation is a Little 16 different. We have the input flux of the parent. We need 17 to consider its transport up to some length, at which point 18 it will transform into the progeny, and then the progeny 19 l will be transported, this equation, and this distance could L 20 be anything from right at the top here to the very end or 21 bottom down here. 22 Finding a solution for this equation is a little ! 23 difficult and so we make one of two approximations. We l 24 either can consider the transport velocity of the parent and 25 longitudinal dispersion and assume that the progeny will ANN RILEY & ASSOCIATES, LTD. O Court Reporters 1025 Connecticut Avenue, NW, Suite 1014 Washington, D.C. 20036 (202) 842-0034
i 74 1 also move the same rate as the parent or we could say that
,_ 2 dispersion is not important and, therefore, we will use the k- 3 i parent velocity and progeny velocity and account for their I
4 different transport rates. 5 You could also do a third case value and say 6 longitudinal dispersion is important, but this time we will 7-transport at the rate of the progeny and not of.the parent. 8 That could be useful if the parent is short-lived and decays 9 at the top, so it's mostly traveling as a progeny. I 10 In RESRAD, if you input different KDs for the 11 parent and the progeny, it will default to the case of 12 considering their different transport velocities and it will 13 ignore dispersion. 14 If you input the same KDs, it will then do this. f) (_./ 15 If you do input different KDs and you still want to do 16 dispersion, you can set that option, there is a button for 17 that. 18 If you have a case where you think that both 19 dispersion and progeny and parent KDs are important, we 20 could model that, to some extent, by subdividing the 21 unsaturated zone. Right now, we allow up to five
-22 subdivisions for the unsaturated zone, and we will look at 23 that case, considering a two-member chain. If you're having 24 a parent entering here, I've turned the saturated zone 25 sideways -- the unsaturated zone sideways so that you can i
ANN RILEY & ASSOCIATES, LTD.
} Court Reporters 1025 Connecticut Avenue, NW, Suite 1014
' Washington, D.C. 20036 (202) 842-0034 L
75 1 put it on the screen here, if your parent is entering here l 2 and your progeny is exiting here, then we have five 3 different sub-zones, then we have five components to the 4 progeny flux coming out here. One would be the parent 5 entering, transforming in the sub-zone, and then traveling 6 as the progeny to the next four zones. This would traveling 7 as the parent in the first zone, transforming here, and then 8 the next three zones and so on. 9 !
- Looking at the case of the component where 10 transformation is in the fourth zone, if you divide it into 11 five different zones, RESRAD would account for parent 12 transport and dispersion in this zone, same for the three 13 zones. Here it's traveling as the progeny, so it would 14 account for its retardation and dispersion. In here the
-15 i ,/
user has to make a choice, because we cannot do all three. l 16 The user may choose to model the parent retardation and 17 dispersion, or parent retardation and progeny retardation. 18 This one we do not provide at the moment, dispersion and
- 19 progeny retardation.
l 20 And in this manner, we can, therefore, account for 21 -- say, if you choose this option, where you think travel is 22 important, we can account for dispersion in four of the five 23 zones or 80 percent of the region and we would account for l 24 retardation over all of the region. If you go the upper 25 route, you do dispersion over the full zone and in four of i ANN RILEY & ASSOCIATES, LTD.
,O, Court Reporters 1025 Connecticut Avenue, NW, Suite 1014 Washington, D.C. 20036 (202) 842-0034
76 ' 1 the fivo zones, you would account for the correct transport 2 rate. 3 In that way, you could consider both factors. 4 Obviously, it would take a little more time because you have 5 five zones instead of one, the computation time, that is. 6 In the saturated zone, we have a different 7 situation. We have lots of nuclides coming in here and we 8 are interested in a concentration somewhere out here. 9 Again, for the parent, we can follow the equation, 10 considering longitudinal dispersion and also dispersion in 11 the lateral direction. For the transfer dispersion, both 12 vertical and horizontal, we use an approximation and the 13 purpose of that is we also want to be able to model transfer 14 of dispersion of the progeny and this approximation allows 15 us to use the same equation of both parents and progeny. 16 We also output well water considerations in 17 relation to what is in the aquifer and by using the 18 approximation, it allows us to do a semi-analytical 19 ) computation of the concentrations in the well instead of 20 having to get too many points and do a numerical thing. It 21 also cuts down on the computation time. 22 To take the considerations of the well, we look at 23 the well and try to figure out what part of the upstream end 24 of the aquifer contributes to this well. So although things 25 are really at the end of the well, far upstream, it would be ANN RILEY & ASSOCIATES, LTD. O Court Reporters 1025 Connecticut Avenue, NW, Suite 1014 Washington, D.C. 20036 (202) 842-0034
e 77 1 still straightforward and we want to find out what 2 cross-section upstream contributes to the well. O (j 3 And we assume that the depth of the aquifer up to 4 , the specified stream will contribute to the well and we get 5 the width based on the pumping rate and flow velocity in the 6 aquifer, and then we get the average consideration in the 7 aquifer or that cross-section, because we assume that all 8 that water will enter the well. 9 For the progeny transport in the unsaturated zone, 10 we can again go to the same thing. We have to consider two 11 of the three factors, and you can do that, and, if you want, 12 you can, again, subdivide the saturated zone into different 13 zones so that you can model both factors. 14 The consideration is a little different, here 15 because if you want to subdivide, you have to be able to get 16 the expression of input flux this way and output flux there. 17 We have that expression. Then we use the unsaturated zone 18 expressions for flux in, flux out, and then we have another 19 expression for flux in, concentration cut. And we combine 20 all of them and when you subdivide, we can go up to a 100 21 subdivisions, although if you too high, you run into 22 situations where, because of the subdivisions, there is a 23 loss of accuracy. Up to about 25 to 50 subdivisions, the 24 loss of accuracy, so far, for the cases I tested, within 25 five percent. ANN RILEY & ASSOCIATES, LTD. f- x Court Reporters () 1025 Connecticut Avenue, NW, Suite 1014 Washington, D.C. 20036 (202) 842-0034
[~ e 4 1 78 And when you go up to 25 subdivisions, you 2 essentially are modeling both dispersion and retardation in 3 24 or 20 of those subdivisions and then ignoring one of the 4 factors in only one -- say, as many subdivisions as you have 5 progeny, and that helps us model both factors when there is 6 a need for that. 7 That's about all I have on this. Are there any 8 questions? 9 MR. NICHOLSON: Thank you very much, Emmanuel. 10 Are there any questions for Charlie or Emmanuel on the 11 RESRAD code, their presentation? 12 MR. THAGGARD: I'd just like to ask, what is the 13 status on the documentation for that, tPe off-site ground 14 water code? 15 MR. YU: O' 16 off-site. We have used this guide for the RESRAD We have draft documentation describing all the 17 equations available. If you'd like to get a copy for review 18 purposes, we can send you a copy. 19 MR. EID: I have a question to Charlie. What is 20 the possibility for using probabilistic analysis for the 21 off-site RESRAD code? Is it available? 22 MR. YU: It's available. 23 MR. EID: Within the code itself? 24 MR. YU: Within the code. I believe it's also 25 uncertainty adding -- ANN RILEY & ASSOCIATES, LTD. O Court Reporters 1025 Connecticut Avenue, NW, Suite 1014 Washington, D.C. 20036 (202) 842-0034
r . I i i
^
79 1 MR. GNANAPRAGASAM: Same as RESRAD. i 2 MR. YU:
! Same as RESRAD, I'm sorry. But it is \ , 3 available.
l 4 I MR. EID: Thank you. i 5 MR. NICHOLSON: Thank you very much for your 6 presentation. I'd like to invite Walt Beyeler to join us at 7 the table. Walt? i What we'd like to have now is we'd like ! 8 i to have a group discussion now and Mark Thaggard is going to 9 lead us through the group discussion. We've heard these 10 presentations this morning on conceptual models and you've 11 gotten some background from the NRC staff on the framework. 12 We would now like to have a group discussion on 13 conceptual models and scenarios. . These are a series of 14 questions. If you turn to page four of your agenda, on page 15 four, the same thing is on your agenda, and we'd like you to 16 interact with the people here at the table. We'll go around 17 the table and just quickly introduce. Most of you, you've 18 already heard from. But Walt Beyeler, if you could raise 19 your hand, please. 20 Walt is from Sandia National Laboratory and he's 21 been working on D&D. Gene, do you feel comfortable talking 22 about MEPAS? Do you want to join us also at the table? 23 Okay. Gene Whelan from Pacific Northwest National 24 Laboratory is also here and he's worked on MEPAS quite a 25 bit. MEPAS was developed by PNL. !# ANN RILEY & ASSOCIATES, LTD. ! \ Court Reporters 1025 Connecticut Avenue, NW, Suite 1014 ; Washington, D.C. 20036 (202) 842-0034 i
I l i 80 1 We'd like for you to ask questions, as we go
- 2 through these questions ther 'lves, for clarif.ication, and i
! x- 3
~
we're asking the people at the table to contribute, also. 4 So, Mark, lead us on a discussion. 5 MR. THAGGARD: I would just like to say that there 6 ! isn't necessarily a right answer or wrong answer. We want 7 to kind of have an open dialogue here. I think the way 8 we're going to do this is we're going to give people the 9 l opportunity to speak and then if nobody speaks, then as the 10 moderator, I get the free will to start calling on people, 11 and that certainly includes the audience. 12 Why don't we take these questions one at a time? 13 The first question is under what circumstances can the 14 ground water pathway be eliminated. We've had a lot of 15 feedback from some of the previous workshops, people asking 16 what kind of justification is needed to eliminate the ground 17 water pathway. So I think we'd like to get a little bit of 18 dialogue going on that. 19 I mean, we've got some ideas certainly in terms of 20 what we think might be some justification, but we would like 21 to hear and get some feedback from other people. 22 Would anybody like to take a stab at that? I know 23 there are some people here that have got ground water 24 problems. Dave? 25 MR. FAVER: Dave Faver, ISI. I guess a general l ANN RILEY & ASSOCIATES, LTD. [
' Court Reporters 1025 Connecticut Avenue, NW, Suite 1014 l
Washington, D.C. 20036 (202) 842-0034
I V 1 81 1 topic for a conversation is how would the path life of the l g 2 radionuclide come into play on this decision. 3 If you have an area that is essentially 4 exclusively supplied by a local water company and wells are 5 just not typically used ir. that area, for example, if you 6 have a short half-life radionuclide, should that factor into 7 the decision-making in terms'of the probability of the well 8 being dug in that area over the effective time that the 9 water would be contaminated? 10 MR. THAGGARD: I think that's kind of a new j 11 concept, the use of a half-life. Anybody else have any 12 other comments? Walt? 13 { MR. BEYELER: It seems that to a large extent, 14 this question may be a matter of regulatory policy. There T 15
) are certainly physical aspects of the system that will enter }
16 into that consideration, hydrodynamic properties cf the 17 aquifer, can it yield quantities of water that are 18 consistent with the assumption scenario. 19 I think it's largely a matter of policy, I would 20 think, as to how those considerations weigh in, whether it's 21 relevant for the analysis or whether the scenario is meant 22-to capture the physical characteristics of the system, 23 whether it is meant to be a sort of stereotypical surrogate. 24 MR. THAGGARD: I think what we have heard so far 25 is we have heard that the use of the radionuclide half-life, [ ANN RILEY & ASSOCIATES, LTD. Court Reporters 1025 Connecticut Avenue, NW, Suite 1014 Washington, D.C. 20036 (202) 842-0034 u
1 82 if we've got short-lived radionuclides, that might be 2 justification -- of course, there are physical limitations, 3 which I think Boby talked about some of those physical 4 limitations during his presentation. Obviously, if you've 5 got a site where you've got minimum ground water yield, that 6 might be -- that's a physical limitation, so that might be 7 justification, or if you've got ground water that's 8 contaminated. 9 i I'd like to hear maybe from some of the industry 10 people. I know this gentleman, I've forgotten your name, 11 from NFS, you're currently working on a problem with the 12 ground water. Is that correct? 13 MR. KIRK: Scott Kirk, Nuclear Fuel Services. The 14 issue that we are up against is at what depth do you model 15 the ground water. For example, if you have shallow ground 16 water, contaminants that may be in unconsolidated sediments, 17 but site-specific information leaves us to believe that the j 18 most typical well installation practices using state i 19 ! regulations would indicate that the wells would be installed 20 in bedrock. 21 And the transport models indicate that the ground 22 water would not be affected at this depth. Would that be 23 justification to exclude the ground water pathway? 24 MR. THAGGARD: If I understand that correctly, 25 you're saying the use of state regulations. l i ANN RILEY & ASSOCIATES, LTD. Court Reporters l 1025 Connecticut Avenue, NW, Suite 1014 Washington, D.C. 20036 (202) 842-0034 L
~
I 83 1 MR. KIRK: Not only state regulations, but also 7-s 2 site-specific information. I For example, if you did a study, 3 like a five-mile radius from your site, and you have at indications that all wells are installed deep in the bedrock 5 and that you indicate that behavioral traits or the average 6 member of the critical group would comply with state 7 regulations, both of those allow you to exclude the ground 8 water pathway. 9 MR. THAGGARD: I think the current use, if you 10 summarize it that way, you're saying one justification would 11 be current use. If you can demonstrate that people in the 12 area are not using the ground water, then that might be 13 justification that should be considered. 14 Did I paraphrase that correctly? 15 MR. KIRK: O4 That's correct. 16 MR. THAGGARD: Does anybody else have any other 17 comments on this question? Like I said, there's no 18 particular right or wrong answer here. We're just trying to 19 get some free thought going here. Yes? And I should know 20 your name, from Sequoyah Fuels. 21 MR. ELLIS: John Ellis, from Sequoyah Fuels. EPA 22 has a ground water classification scheme that actually has a 23 quantitative number for yield. I believe it's -- say, it's 24 typically ground water would fall into what they call Class 25 3 if it's less than 150 gallons a day. ANN RILEY & ASSOCIATES, LTD. O Court Reporters 1025 Connecticut Avenue, NW, Suite 1014 Washington, D.C. 20036 (202) 842-0034
p -
._a 1 84
,, 2 I was wondering if the NRC has considered trying to develop some sort of a quantitative standard At our 3 site, particularly, when you get down into bedrock, we 4 hand-bale most of our monitor wells dry. 5 It's kind of hard to conceive somebody putting a well in there that would 6 yield enough water to support a household, 7 and that's been the premise of our argument to eliminate the ground water. B. But we don't find anything in the NRC guidance or 9 regs that would allow us to adopt that EPA concept or some 10 version of it. 11 MR. THAGGARD: 12 I think that's one of the areas that needs to be looked at, and this gets back to the 13 14 physical limitations issue again in terms of -- obviously, if you can't get the yield to support the use that you 15 model, then that might be justification. 16 Now, what that number should be, whether it's 150 17 gallons per day, I guess, still needs to be debated a little 18 bit. 19 MR. ELLIS: 20 I might add just one other thing, too. Specific to our area, virtually all of the rural households 21 in a relatively sparsely inhabited area are provided by 22 rural water systems and we believe that that ought to be a 23 strong factor in the consideration. There are highly 24 developed systems for treatment of surface water and 25 distribution all over eastern Oklahoma, where we're located , O g ANN RILEY & ASSOCIATES, LTD.
'V Court Reporters 1025 Connecticut Avenue, NW, Suite 1014 l Washington, D.C.. 20036 (202) 842-0034
1 . 85 l and except for households that are located along the stream 2 bottoms, where there's some alluvial ground water, you don' t 3 find any domestic wells. ! 4 Just the vast majority are served by these water 5 systems and we also think that 6 that ought to be a criteria by which the pathway can be justified to be eliminated . 7 MR THAGGARD: So that kind of gets back to the 8 surrounding land use argument again. Paul Chenoweth. 9 MR. CHENOWETH: Yes, I'm Paul Chenoweth, NEI. 10 This discussicn, I think, is very fruitful and this is a 11 great opportunity in these workshops to highlight these. 12 I guess the question would be, how can you feed 13 back the answers after you've thought about it? Is there a l 14 way for you to, for instance, in a guidance document, 15 O 16 document some of the sources that a person might use that the NRC would find acceptable for determining, for instance, 17 that ground water was not potable or not acceptable for use? 18 Is that merely just the state regulations or the 19 local county regulations in the area or are EPA values 20 already available and so forth? 21 It seems like for salinity, for quality, quantity, et cetera, it would be a very big 22 help, I think, for the industry to know where to go to get 23 that data to document these assumptions. 24 MR. THAGGARD: Yes. Tom reminded me, we're going 25 to be talking a little bit about that during the next ANN RILEY & ASSOCIATES, LTD. l O Court Reporters 1025 Connecticut Avenue, NW, Suite 1014 Washington, D.C. 20036 (202) 842-0034
I I 86 1 1 session, some of the data sources, some of the information f3 2 that's readily available, and ultimately, as we resolve this 3 question, NRC resolves this question, then we would i 4 certainly look to try to provide information on where people 5 can get that type of information to justify. John? 6 MR.-GREEVES: John Greeves. Mark and others, I i 7 think a way to do that is to put it in the standard review 8 plan. If we've got some clear caces demonstrating where we, 9 as an agency, don't think the water pathway applies, we can 10 write-that right in the standard review plan as basically 11 guidance for our reviewers. 12 One example is salt water. We could write that 13 right in the review plan. If the water body is too salty to 14 drink, we're nou going to use that in our review procedures. 15 So I think these kinds of workshops, to the extent 16 we can come up with those examples, they could be written 17 right into the standard review plan and then I think 18 licensees would have some comfort that, well, it's in the 19 review plan, I think that's the place we'd like to locate ! l 20 these things and that's why we have the review plans on the 21 web and we're looking for feedback. 22 l If you, whoever you are, have recommendations on ' i 23 cases where it should be included or cases where it 24 shouldn't be included, we want to hear about that. 25 Actually, the agreement states also have these i ANN RILEY & ASSOCIATES, LTD. .
\ Court Reporters 1025 Connecticut Avenue, NW, Suite 1014 Washington, D.C. 20036 (202) 842-0034 i
1 87 quoctions and these issues. i 2 So this whole series of o workshops, the modules on the web are opportunities to (_ 3 document that 4 and the way to record it is to get it into th e standard review plan. 5 6 So I'd just offer that to all of us as a goal. MR. EID: Yes. 7 said, it's true. I would like to add to what John 8 We will be placing, in the standard review plan, the areas that you are very clear where you could 9
)
10 eliminate ground water pathways and I gave examples in my presentation for three cases, 11 that they are very clear you could eliminate the ground water pathways and the associated l 12 agricultural pathways with the ground water and drinking 13 water, of course. 14 , There are also other areas where you cannot 7-s 15 i eliminate the ground water pathways and those are very clear 16 and it will be explained in the standard review plan. 17 However, there will be gray areas where it is very difficult 18 to decide whether you could eliminate or you cannot 19 eliminate the ground water pathways and it depends on the 20 site-specific conditions and you need to provide other 21 information in order to justify for elimination on the 22 ground water pathways. 23 Those will be explained as gray areas and the need 24 for additional information, of course, would be based on 25 l site-specific case conditions. l l
/' ANN RILEY & ASSOCIATES, LTD.
(,N) Court Reporters 1025 Connecticut Avenue, NW, Suite 1014 Washington, D.C. 20036 (202) 842-0034 L
r - 1 1 88 MR. THAGGARD: Okay. Why don't we move on to the 2 next question here? f~ I think we got a3 lot of good feedback 3 on that first question. i 4 We've kind of touched on this second question a little bit during this morning's session. 5 It has to do with what site features and processes would 6 make the selection of the simple models inappropriate for a 7 given site. 8 I think we would kind of like to lay out some of 9 the -- I mean, Phil, in his presentation, talked about some 10 of the simplification that 11 -- how some of these models are very simple and they may not be completely appropriate in i 12 all cases, and we would kind of like to lay out some of 13 those and maybe get some more discussion on that. ! 14 And why don't we start with Dr. Yu? Can you think 15 of some specific places where you would think that RESRAD 16 may not be appropriate? 17 MR. YU: i Before I answer this question, I want to 18 make a comment on the previous one. 19 There are circumstances that you can eliminate a ground water pathway for certain 20 purposes, maybe in the review plan, you'd want to list them 21 out. If the water quality is not good enough for drinking 22 purposes, you can eliminate the drinking water pathway, but 23 not the feeding livestock and irrigation purposes. You 24 still can use this water for irrigation purposes. 25 tiR. EID: This will depend on the quality of the ANN RILEY & ASSOCIATES, LTD. O Court Reporters 1025 Connecticut Avenue, NW, Suite 1014 Washington, D.C. 20036 (202) 842-0034
c 89 1 ground water. If a certain quantity of the ground water 2 can't be useful.to support the livestock drinking and to 3 support also the irrigation, the salinity could be so high, 4 I mean, certain brackish water, of course, you cannot use it 5 even for irrigation. 1 So it depends on the conditions, the 6 salinity, and depends on other factors. 7 The type of soil also could impact what kind of 8 crops you could grow. So those could be taken into 9 consideration. 10 MR. YU: But I'm just saying that you should not 11
.have just one criteria to eliminate all the ground water 12 related pathways.
13 MR. NICHOLSON: So which one do you think for 14 RESRAD?
/s 15 MR. YU:
t What site-specific features, process, 16 events, would make the selection simple dose codes not 17 conservative. While RESRAD -- the RESRAD model is -- we're 18 trying to make it more realistic. The on-site model is 19 conservative because there is now considered dispersion. 20 What feature is not conservative? 21 ! Do you have any comments on that? I cannot think 22 of features -- it's pretty -- ! 23 { FROM THE AUDIENCE: Or appropriate. 24 MR. YU: Or not appropriate. l 25 MR. THAGGARD: ! What we've mainly had in mind, not I i
;t ANN RILEY & ASSOCIATES, LTD.
s Court Reporters 1025 Connecticut Avenue, NW, Suite 1014 l Washington, D.C. 20036 (202) 842-0034 : i I i r .
I l 1 90 l conservative, but if you can answer it for not appropriate, t 2 that's good, too. \ (~) l (ms/ 3 MR. YU: Well, if the model doesn't fit. i Of 4 course, you can see that -- Phil Meyer presented a 5 conceptual model. If your hydrological conditions are so 6 complex, it doesn't fit the conceptual model, then it's not 7 appropriate. 8 MR. THAGGARD: We've had this gentleman standing 9 for a while. Can you state your name? 10 MR. MAIERS: My name is Bob Maiers. 11 I'm with the Pennsylvania Bureau of Radiation Protection. Pennsylvania 12 13 did a considerable amount of characterization work for our I state in an attempt to site a commercial low level waste 14 facility and we have several automatic disqualifiers which I 15 think apply to this question. 16 Some of the automatic disqualifiers we had were 17 the existence of Karst geology, the existence of flood 18 plains, the existence of underground mining activities in 19 the area, the existence of oil and gas wells. 20 The reason why these are automatic disqualifiers 21 for a commercial facility are basically because it makes it ! 22 very difficult to model what's going on. So I think that 1 23 this should actually be addressed and pointed out to 24 licensees that when they have features like this on-site, 25 these models are not appropriate. 1 l t ( i ANN RILEY & ASSOCIATES, LTD. ( Court Reporters 1025 Connecticut Avenue, NW, Suite 1014 Washington, D.C. 20036 (202) 842-0034 1
r-4 1 91 MR. THAGGARD: Thank you. 2 Before I get to you, Tom, this gentleman. s_ ) 3 MR. BELLINI: 4 Frank Bellini, Duke Engineering at Yankee Atomic. I have one input parameter and I know how it 5 works in RESRAD, 6 codes, I'm not sure how it works with the other that kind of dogs me, and it is the depth of the well i 7 intake, the pump, or I guess it's the depth of intake water 8 level for the well. 9 10 I find it gives me a great deal of variation in my ground water dose by varying that, for example, from a depth 11 ! 12 of one meter to three meters to ten meters to 30 meters. I'm not sure how to deal with it. I don't know 13 how the other codes deal with it. It's not exactly 14 addressed to the question here, but I'd sure be interested 15 in your comments. 16 MR. THAGGARD: Do you have any comments on that? 17 MR GNANAPRAGASAM: 18 If you put the well too deep, you are drawing in clean water and that's why it varies too 19 much. So if you want to be conservative, you should put the 20 well right at the depth of the plume. 21 MR. BELLINI: Sure. I can put it at one meter, 22 but I'm not sure -- that's awfully non-realistic, I'll call 23 it. I'm not sure what the right depth is and whenever I run 24 the code for ground water, I'm really not sure. 25 MR. GNANAPRAGASAM: With the off-site, when there ANN RILEY & ASSOCIATES, LTD.
\ Court Reporters 1025 Connecticut Avenue, NW, Suite 1014 Washington, D.C. 20036 (202) 842-0034
7-92 1 is dispersion, it will be a little less dependent on the 2 depth. With:the on-site code, it's not dispersed in the 3 vertical direction. It is sensitive. 4 MR. BELLINI: So it sounds like I'm doing the 5 right thing and struggling with it anyway. 6 MR. YU: In running the code, if you have extra 7 data where the screen is, you should input that into the 8 code, but if jou don't have that, you probably need to be 9 more conservative. You can run a sensitivity analysis, make 10 sure the screen is catching the plume so that you will get 11 the conservative dose concentration. 12 MR. GNANAPRAGASAM: Also, it depends on the 13 pumping rate. So if you put too small a depth and you put a 14 large pumping rate, it's going to put it on the sites, so 15
) you not only need to look at the depth, but also the pumping i 16 rate.
17 MR. BELLINI: Okay. The other parameters, I have 18 this huge river that the site is on, but it's backed off the 19 river enough so that the drainage area, using the drainage 20 area for this whole big river, which has, I don't know, a 21 1,000 square mile drainage basin up above the plant, versus 22 using a smaller drainage basin in the area of the plant, 23 creates an issue of which should be used. 24 I suppose you could answer that and say, well, use 25 both and try and bracket it again, but I'm not quite sure I ANN RILEY & ASSOCIATES, LTD. Ccurt Reporters 1025 Connecticut Avenue, NW, Suite 1014 Washington, D.C. 20036 (202) 842-0034
92 1 is dispersion, it will be a little less dependent on the J 2 7-s depth. With the on-site code, it's not dispersed in the
/'
3 vertical direction. It is sensitive. 4
. MR. BELLINI: So it sounds like I'm doing the 5
4 . right thing and struggling with it anyway. 6 MR. YU: In running the code, if you have extra 7 data where the screen is, you should input that into the ' 8 code, but if you don't have that, you probably need to be 9 more conservative. You can run a sensitivity analysis, make 10 sure the screen is catching the plume so that you will get 11 ~ the conservative dose concentration. ~ 17 MR. GNANAPRAGASAM: Also, it depends on the 13 pumping rate. So if you put too small a depth and you put a 14 large pumping rate, it's going to put it on the sites, so 15 you not only need to look at the depth, but also the pumping 16 rate. 17 MR. BELLINI: Okay. The other parameters, I have 18 this huge river that the site is on, but it's backed off the 19 river enough so that the drainage area, using the drainage i 20 area for this whole big river, which has, I don't know, a 21 1,000 square mile drainage basin up above the plant, versus 22 using a smaller drainage basin in the area of the plant, 23 creates an issue of which should be used. 24 I suppose you could answer that and say, well, use 25 both and try and bracket it again, but I'm not quite sure ANN RILEY & ASSOCIATES, LTD.
- Court Reporters 1025 Connecticut Avenue, NW, Suite 1014 Washington, D.C. 20036 i (202) 842-0034
( l 93 1 i how your model works in terms of providing dilution either ! . f~g 2 from a very large upstream drainage area versus a small
%- 3- local drainage area.
4 MR. YU: You're talking about the surface water 5 pathway? That drainage area is used to -- it's a dilution 6
' factor in the code for the surface water considerations.
7 MR. BELLINI: Right. Well, I can use a dilution ' 8 factor for a 1,000 square mile drainage area or I can take 9 my local drainage area of five square miles and the result 10 will vary accordingly. 11 Any wisdom on how that might be applied, other 12 than bracketing -- using both, trying both ways and 13 bracketing it? 14 MR. YU: You don't the actual -- ("$g 15 MR. THAGGARD: N,,/ Why don't we move on? That's 16 probably a little bit site-specific. I think it's a generic 17 issue in terms of -- 18 MR. CARTER: I think my comment and question kind 19 of goes to this, really, because I think the commenter from 20 Commonwealth of Pennsylvania and this commenter, too, have 21 some' problems that I think are rooted in the form of the 22 question. 23 It isn't so much what site-specific features rule 24 out use of certain models. I think what really matters is 25
-- and, in fact, I think you can use almost any model on any t
ANN RILEY & ASSOCIATES, LTD.
) Court Reporters k~/ 1025 Connecticut Avenua, NW, Suite 1014 Washington, D.C. 20036 (202) 842-0034 L
1 94 sito if you're willing to be conservative enough and carefu l i 2 enough. ( 3 What really matters is whatever model you settle 4 on, make sure it's representative of the site, and if you're j 5 talking about a 10,000 square mile drainage area and you're 6 i worried about concentration in this little particular area 7 close to your source, 8 that may not be appropriate, it's just not realistic. 9 You have to make the model match the site. That's 10 the key thing. With respect to the Pennsylvania discussion 11 about ruling out certain features, that makes perfect 12 sense if you're going in and proposing to site a facility that is 13 not sited yet. 14 In our applications, we're trying to terminate a
,- 15
( license -- licenses on sites that are already there. I 16 mean, we don't have a option to site the site, it's there. 17 So the question is how do we represent the site for purposes 18 of dose assessment. 19 I think you have to look at the site, look at the 20 model, look at the parameters and make the best fit, i 21 MR. THAGGARD: That was Tom Carter. Yes? 22 MR. BLRKLIN: My name is Rich Burklin. I work for 23 Siemens. Regarding the depth of the well and so on, would a 24 i reasonable approach be to find out what the depth of the ! 25 wells are in the area then you're in and then say -- and
/'N ANN RILEY & ASSOCIATES, LTD. '( Court Reporters 1Q25 Connecticut Avenue, NW, Suite 1014 Washington, D.C. 20036 (202) 842-0034 I
1 95 then do some type of distribution study in order to be able 2
,_ to get, for instance, the mean value of the depth of the -I 3 wells, rather than guessing?
4 We don't want to take the most conservative 5 necessarily. We want it to be realistic. So maybe you can 6 get an idea what realistic is by looking at the wells that 7 already exist in that area and doing something along those 8 lines. So you think the NRC would find something like this 9 acceptable. 10 MR. EID: 11 I would like to add to this that the NRC is contracting INEL to develop a probabilistic code with a 12 different kind of site-specific parameters and the future 13 code hopefully will enable trying to look at the l 4 14 probabilistic analysis using different ranges of parameters. ; 15 l So an example of the sensitive parameter is the 16 depth of the water well and this could be used in order to 17 look at the mean dose distribution, based on RESRAD. 18 MR. MEYER: I have a comment here regarding this 19 subject. I don't know if the NRC has considered this. You 20 guys probably have. But your scenario requires a certain 21 yield. You're assuming that you're going to get so much 22 water out of this well to satisfy the scenario demands. 23 From the perspective of a well developer, a hydrogeologist 24 going out and developing a well for this particular purpose, 25 what is reasonable. What would their requirements be in
/' ANN RILEY & ASSOCIATES, LTD.
kTl s ' Court Reporters
,'25 Connecticut Avenue, NW, Suite 1014 Washington, D.C. 20036 (202) 842-0034
96 1 terms of installing this well? o- 2 They certainly wouldn't put in a screen that's a 3 meter long right at the surface of the aquifer. So if you 4 want to incorporate concerns about reasonableness, that 5 seems like she approach to take. 6 MR. THAGGARD: Dr. Hung, I think you've been 7 holding your hand up for a while. 8 MR. HUNG: I was trying to respond to the question 9 that the gentleman asked on the depths of the well and also 10 the upstream drainage area. 11 The PRESTO model is trying to use the most -- I 12 realize that the depth of the well is very sensitive to the 13 concentration in the well, as I presented earlier. So our 14 PRESTO model is trying to use most of the probable 15 scenarios.
) This is assuming that the well owner brings in 16 the well driller and tries to use that.
17 If you are the owner of the well, how deep do you 18 want to go? So my presentation was saying that most 19 probable scenarios to use that for this. 20 On the second question, on the drainage area, the 21 PRESTO model considered that. So for instance, you have a 22 ten square mile upstream. That means a lot of recharge 23 area. So at the site near the well, you'd have a -- you 24 have to allow for the flow to go through, compared with only 25 one square mile. ANN RILEY & ASSOCIATES, LTD. (/) N- Court Reporters 1025 Connecticut Avenue, NW, Suite 1014 ' Washington, D.C. 20036 (202) 842-0034 _s
. s 97 1
So that's at the -- in the PRESTO model, we allow 2 the user to input this difference. 3 So that will take care of the difference in the drainage area. 4 -MR THAGGARD: Dr. Neuman, 5 DR. NEUMAN: I cannot resist, having at the table 6 the developers of some of these codes, to ask them the 7 following. question, which is kind of rephrasing the question 8 that you have asked. 9 Could you kindly -- I'm a hydrogeologist, I'm not 10 a regulator. Could you explain why these codes or in what 11 way these codes are conservative? 12 MR. THAGGARD: Okay. That's an interesting 13 question. I'll give you a few minutes to think about that 14 while I call on the gentleman. 15 MR. KUHLTHAM: My name is Rick Kuhltham, I'm with 16 Morton Associates. I wanted to speak quickly a little bit 17 about this dilution factor with the well intake. 18 I have always been troubled by the idea that 19 nobody seems to consider that when you pump from a well, 20 that the pumpage from the well in itself distorts the flow 21 field within an aquifer, and that, in fact, those of us who 22 do aquifer testing and whatnot are familiar with concepts of 23 partial penetration and where you see the .affects of partial 24 penetration. 25 When you get a certain distance from the well, you O ANN RILEY & ASSOCIATES, LTD. Court Reporters 1025 Connecticut Avenue, NW, Suite 1014 Washington, D.C. 20036 (202) 842-0034
r 98 1 don't see it anymore. This basically shows us that as you ( 2 move away from the well, that well, when it pumps, will draw 3 water through the whole depth of the aquifer and some of 4 these effects can be influenced by vertical, horizontal and 5 isotropy. But in many cases, as you move away from the well 6 itself, you draw water from the entire aquifer, not just a 7 narrow tone which is contaminated. 8 And I don't really see that these models and these 9 discussions really address that or adequately acknowledge 10 that, and I think that is the reality of it. 11 Of course, it does depend somewhat on the 12 anisotropy that's present. 13 MR. THAGGARD: I think that's kind of the heart of 14 the issue. The assumption you make for the screen length, 15 that affects how much of the aquifer you're assuming you're 16 pulling water. 17 MR. KUHLTHAM: But that's not true, that's my 18 point, My point is that if you have a 100-foot taick 19 aquifer and you have a ten-foot well up here and you pump 20 it, that as you move away from that well, you gradually draw 21 water up. 22 The point is eventually all the water will come up 23 into the screen, it has to, but you bring this water up 24 through and actually you distort the flow pattern of the 25 plume itself. So that the plume, if this happens, will f-(g) ANN RILEY & ASSOCIATES, LTD. Court Reporters 1025 Connecticut Avenue, NW, Suite 1014 Washington, D.C. 20036 (202) 842-0034
99 1 likely narrow and move to the top of the screen over time if 2 this is a continual pumping. 3 So that you would be bringing water from a lot 4 different areas vertical depths and increase the dilution by 5 that mechanism. 6 MR. HUNG: May I answer this question? In the 7' PRESTO model, we realize that -- what you mentioned is 8 significantly true, but when we do the three-dimensional 9 model analysis, we found that -- see, in this particular 10 risk assessment we are talking about here is for one family. 11 MR. KUHLTHAM: Is for what? 12 MR. HUNG: One family. That means that the 13 pumping rate is not that fast. It's very small. So for the 14 risk assessment, when you pump it with a very small amount 15 of water, then the range is really fairly small. So in the 16 l PRESTO model, right now we assume that that would be the 17 i 1 same depth of the screen, but we are in the process of 18 trying to expand that to include the depth that they would 19 draw in. l 20 But our preliminary analysis showed that this
- 21 depth is almost negligible in most of the case. But in the 22 l
future, we would like to include that in the model. 23 MR. KUHLTHAM: Just to address this a little bit i 24 further, a good study, if you want to really evaluate this, 25 it seems, if somebody wants a good paper or something, is to ANN RILEY & ASSOCIATES, LTD. Court Reporters 1025 Connecticut Avenue, NW, Suite 1014 Washington, D.C. 20036 (202) 842-0034
100 1 get a good 3-D or even 2-D model, preferably 3-D, and you 3 2 want "a get your flow lines and develop these patterns, and 3 develop these' patterns even for very small withdrawals over 4 wells and see where your water goes and where your 5 contaminant, where your plume goes and how your plume will 6 move around and, in essence, perhaps even thin as you get 7 towards the well due to other water coming from deeper 8 portions of the aquifer. 9
'If you want to evaluate that effect, you have to 10 fully model it so as to show those effects.
11 MR. HUNG: Yes. We plan to do that. 12 MR. THAGGARD: Either way, you're still left with 13 the question of -- you could do that if you know the well 14 depth you have to work with and you were actually modeling a 15 real well. But I think the problem we're wrestling with 16 here is we're dealing with a hypothetical well, and so you 17 -- 18 .MR. KUHLTHAM: I guess my point is, I really 19 wonder if the well depth is that important, if it's not ! 20 really the thickness of the aquifer that's really the 21~ primary driver here. 22 MR. THAGGARD: Okay. 23 MR. MEYER: Exactly. I'm not positive about this ! 24 point, but I think you're right on that and if you included 25 a more realistic model, that the well depth would not be as
\ ANN RILEY & ASSOCIATES, LTD.
l Court Reporters l i 1025 Connecticut Avenue, NW, Suite 1014 Washington, D.C. 20036 (202) 842-0034
101 1 important. 2 Also, I just wanted to point out that the NRC, 3 this has already been pointed out, but the NRC doesn't 4 require use by licensees of the models that we are 5 discussing here and there is nothing stopping someone from 6 using a more realistic model that incorporates the effects 7 that you're talking about. 8 MR. THAGGARD: Before I call on Henry, I'd like to 9-go around the table and see if we can get any of the 10 gentlemen to respond to Dr. Neuman's rephrase of the 11 question. Walt, you want to talk about D&D? 12 MR. BEYELER: I'd like to be able to just make a 13 l- short attempt at starting to answer that question. 14 I think it's -- we had this discussion a bit at i 15 the program review yesterday and I think it's -- first, I 16 would suggest that the conservatism of the model is very 17 i l difficult to argue for without talking about how the l 18 i parameters of the model were established in general. It's l 19 very difficult to make a sort of structural argument that 20 the model is conservative and that it's also important to 21 bear in mind conservative with respect to a particular 22 performance measure, the dose in this case. 23 Specifically with respect to D&D, I think a case 24 could be made that the treatment of the aquifer is 25 conservative, provided there is a conservative estimate of I 7-s ANN RILEY & ASSOCIATES, LTD.
;'v) Court Reporters 1025 Connecticut Avenue, NW, Suite 1014 Washington, D.C. 20036 (202) 842-0034
102 1 the amount of material that enters the aquifer, and that /"' 2 O) that entire amount of material has been available for 3 ingestion, is brought up from the aquifer. 4 It seems to me that at least that element of the 5 model is conservative, again, provided the estimate of the 6 penetration is conservative. 7 I guess just a second observation regarding 8 site-specific conditions that might bear on the suitability 9 of the model. I think Dr. Meyer discussed the presence of 10 strong heterogeneities being something that is inconsistent 11 with some of the assumptions that are made in the model, and 12 I would, again, suggest that certainly real sites are 13 heterogeneous to some degree or another and that the 14 important consideration is how is the parameter assigned in () V 15 view of that heterogeneity. 16 It seems that it is possible, in principal, in a 17 very heterogeneous setting, to assign a single number for, 18 for instance, infiltration rate, provided that that l 19 infiltration rate characterizes the rare or infrequent ! 20 features that may occur at the site. i 21 So it seems at least possible, in principal, that 22 that sort of a calculation can be done in the presence of 23 strong heterogeneity, but it's important, again, to consider 1 24 the procedure by which the site-specific parameters get 25 established. ANN RILEY & ASSOCIATES, LTD. O's Court Reporters 1025 Connecticut Avenue, NW, Suite 1014 Washington, D.C. 20036 (202) 842-0034
103 1 MR. THAGGARD:
' Let me call on Dr. Whelan, who has 2
been sitting there kind of quiet. How would you respond to 3 Dr. Neuman's rephrase of the question? How do you think the 4 MEPAS code is conservative or do you even consider it to be? 5 I mean, the ground water components in the MEPAS code, do 6 you even consider them to be conservative? 7 MR. WHELAN: The first thing I'd like to say on 8 that issue is that, number one, it depends upon the 9 questions that are being asked and the questions that you're 10 trying to answer. 11 From my experience, and it isn't just with the 12 MEPAS, the codes or this classification of codes, and that 13 is that I do applications, hazardous waste site assessments 14 for a living and I use these codes to help me do these O 15 assessments. V 16 My experience is that the contaminants generally 17 move faster than we think they move and the concentrations 18 generally show up at the wells higher than we anticipate 19 through our assessments that they will be. 20 And even though we have measured soil properties, 21 et cetera, for some reason, mother nature tends to throw a 22 wrench, a monkey wrench in the system and the contaminants, 23 they'll go through preferential flow paths and things of 24 this nature and before you know it, the concentrations are 25 at the well sooner than what you think and higher than you ("
- ANN RILEY & ASSOCIATES, LTD.
Court Reporters 1025 Connecticut Avenue, NW, Suite 1014 Washington, D.C. 20036 (202) 842-0034
104 1 think. 2 I firmly believe that when you go to apply these 3 models, that if you have some site-specific information 4 which you can calibrate these models to, that becomes very, 5 very important in terms of trying to at least capture the 6 essence of the contaminant movement from the source to a-7 potential receptor. 8 With respect to MEPAS, I'm not totally convinced 9 that MEPAS is designed to produce conservative results. As 10 I noted, that if the contaminant moves faster than the 11 parameters that you have defined for that mobility of that 12 contaminant, you're going to find that the concentrations 13 are higher than what MEPAS predicts, and I believe the same 14 thing is true of RESRAD, et cetera, because our conceptual 15 (} site model, how we conceptualize what's happening at the 16 site is actually inappropriate and au such, we're not 17 producing conservative numbers. 18 Now, obviously, you can structure systems such 19 that you do produce conservative numbers. For example, a 20 receptor ingests a source. That is pretty much of a worst 21 case type scenario and things such as dilution, decay, et 22 cetera, generally don't get factored into this. 23 I'm not convinced that these models, when you 24 blindly run them, are conservative and I would be -- I would 25 strongly urge people, if they have monitored data, that they ANN RILEY & ASSOCIATES, LTD. O' Court Reporters 1025 Connecticut Avenue, NW, Suite 1014 Washington, D.C. 20036 (202) 842-0034
r - 105 1 at least try to calibrate their models and this goes to 2 question number three, four and five, and the reason is that 3 that calibration provides you insight into whether your 4 conceptual site model is correct or not or at least if it's 5 close. 6 Secondly, it provides you insight as to whether 7 the values that you're putting in for those input 8 parameters, whether they are within an acceptable range. 9 MR. THAGGARD: Thank you, Gene. Let's get this 10 gentleman, then Jim, and then we're going to have to move 11 on. 12 Why don't you go ahead? 13 MR. NARDI: Joseph Nardi, from Westinghouse. I'd 14 like to hear what people would say -- in our specific 15
) situation, we have a site where we're starting with ground 16 water contamination and I understand some of the models 17 would not handle that. I particularly understand D&D would 18 not be able to handle the situation where you start with 19 ground water contamination.
20 I'd like to hear from the other models how their 21 code would handle that situation. 22 MR. THAGGARD: Charlie, do you want to say 23 something on that, RESRAD? 24 MR. YU: Right. In RESRAD, we allow users to 25 input ground water concentrations into the code, but you O ANN RILEY & ASSOCIATES, LTD. Court Reporters 1025 Connecticut Avenue, NW, Suite 1014 Washington, D.C. 20036 (202) 842-0034
- .06 1
also need to input other parameters, like waste placement l 2
\ time, how -- when the contaminant was placed there, and 3 l RESRAD will use those information to back-calculate I 4
effective KD values and use the effective KD values to 5 further predict future use concentrations in multiple 6 pathways. 7 So that's how RESRAD handles the existing i 8 contamination in the ground water. 9 I also wanted to comment on the conservatism in 10 the RESRAD code. I believe that when we say conservative, 11 we don't want to be too conservative. We are reasonably 12 conservative. We think the assumptions made in the ground 13 water models and other multiple pathways models using 14 RESRAD, I think it's reasonably conservative and we are () 15 16 conservative in the sense that we try to be reasonably conservative on the peak de, , peak concentrations in ground 17 water, for example. 18 If you are conservative in the peak concentration, 19 you may not be conservative at other times, because the 20 other times may not be conservative. The measured 21 concentration could be higher than what we predict at f 22 different times. But we're trying to catch the peak 23 concentrations, trying to make sure that the peak 24 concentration is conservative. 25 And on the other hand, if you are too conservative
' ANN RILEY & ASSOCIATES, LTD.
Court Reporters 1025 Connecticut Avenue, NW, Suite 1014 Washington, D.C. 20036 (202) 842-0034 l t
(. 1 107 1 in one pathway, in the ground water pathway, most of the r' 2 b} 3 contaminant is leached down to the ground water, you may not be conservative in other pathways. For example, waste 4 contaminants atay in the contaminated zone, then that will 5 give you a higher external dose, uptake, other pathways may 6 have a higher dose. 7 l So you need to be reasonably conservative. 8 Otherwise, you'll focus on certain pathway and you're too 9 conservative, you may miss some other pathways and 10 especially important for certain radionuclides. For 11 cobalt-60, for example, it will give you a much higher 12 external dose and if you are too conservative in the i 13 leaching of cobalt-60 to ground water, you may miss the 14 primary pathway, which is the external dose. 15 MR. THAGGARD: (} Before I call on Jim, Gene, do you 16 want to make a comment on how MEPAS handles existing ground 17 water contamination? 18 { MR. WHELAN: The way MEPAS handles existing ground 19 water contamination is you put in a volume, a parallel, the 20 length, width and thickness, and you place that length, 21 width and thickness volume within the saturated zone and it 22 can be at any location within the saturated zone, so it 23 doesn't have to be at the water table surface, and then you ! 24 define an initial concentration associated with that volume. 25 So, in effect, it's uniformly distributed, that 1
)
O ANN RILEY & ASSOCIATES, LTD. Court Reporters 1025 Connecticut Avenue, NW, Suite 1014 Washington, D.C. 20036 (202) 84 2 - C'034
)
108 1 concentration in the volume, and from that, it then -- the 2-code will then calculate emission rates based on the 3 hydrogeologic properties in that saturated zone. 4 MR. THAGGARD: Jim Shepherd. 5 MR. SHEPHERD: Jim Shepherd, NRC. Since you 6 uncovered it, Mark, I'd like to jump to question four, which 7 I think is closely related. 8 In our past practice, where we've required 9 licensees to essentially clean up their sites for 10 unrestricted release, we didn't have a great deal of 11 original contamination and consequently we didn't need a lot 12 of sophistication in our models. 13 I think under the provisions of 20.1403, where 14 we're now allowing considerable contamination to reside 15 on-site -- in particular, contamination in the ground water 16
-- I've had one proposal that I've seen where the preferred 17 remediation and alternative for contaminated ground water is 18 monitored natural attenuation.
19 This comes from the EPA, which is directed 20 primarily at chemical spills, where they can demonstrate 21 some kind of biodegradation. 22 In the case of heavy metals, there is not going to 23 be biodegradation, by and large, and the mass reduction 24 requirement essentially will be met by some kind of 25 migration. So I think we are going to need more ANN RILEY & ASSOCIATES, LTD. Court Reporters 1025 Connecticut Avenue, NW, Suite 1014 washington, D.C. 20036 (202) 842-0034
109 1 sophisticated analyses to identify the actual 2 concentrations, where the plume is today, and, more 3 importantly, where it's going to be in the very long term , 4 and I'm not certain that our existing models can do that. 5 Because we now need -- if we're going to do 6 monitored natural attenuation, how do we know exactly where 7 to model in order to track the plume. 8 MR. THAGGARD: Okay. Well, we're running out of 9 time on this first section here. So~what I'd like to do is 10 go ahead and call on Henry, because he's been up and down, 11 up and down, and then we'll get the other gentleman. 12 We're going to continue these questions tomorrow, 13 as Tom indicated, we've got some time in the program. So 14 whatever we didn't get to today, we're going to go ahead and () 15 16 continue tomorrow. So we're going to come back and have some more discussion tomorrow. 17 Go ahead, Henry. 18 MR. MORTON: I'm Henry Morton. There were two 19 items, I think, that were raised on which I'd like to 20 comment. One relates to, in effect, where within the 21 aquifer you take the water or screen the well. The other 22 relates to the first question under what circumstances can 23 the ground water pathways be eliminated. 24 I think what I'd like to do is speak to these from 25 what -- at least from my experience and relative to trying O ANN RILEY & ASSOCIATES, LTD. Court Reporters 1025 Connecticut Avenue, NW, Suite 1014 Washington, D.C. 20036 (202) 842-0034
.~
7 1 110 s 2 to think of these as residential wells in the more narrowly
\l subsistence or low usage or ordinary usage circumstances.
3 I spent two summers in the 1960s as a well 4 driller's helper and there are some fairly sensible things 5 that I think come out of this experience of drilling . 6 First, with respect to where to screen the well and what 7 aquifer, you basically try to put the well where the person 8 wants it and then you take what you get below, in the ground 9 below that. 10 In the areas that we were working, basically had 11 clay, chert, with gravel beds as the aquifers, were 12 underlain by limestone, bedrock, 13 and then with the fractures or the caving within the limestone, you could get water from 14 within the bedrock. So you had two circumstances of getting 15 water there.
%/
16 The other circumstances were largely sand, 17 sandstone, shale, bedrock. 18 With respect to, first, where to screen the well, 19 in these cases, basically we would try to -- generally try 20 to screen the well over the entire depth of the aquifer . 21' Simply, you want all the water you can get and in many of 22 these cases, the wells don't produce as much as you want. 23 So that was the general rule, screen over the whole depth of 24 the gravel bed or if it's in bedrock, you've set your casing 25 and you're not casing into bedrock, so you get everything
., ANN RILEY & ASSOCIATES, LTD.
Court Reporters 1025 Connecticut Avenue, NW, Suite 1014 Washington, D.C. 20036 (202) 842-0034
c- . , 1 111 within the aquifer at that depth, in the bedrock depth, 2 The other is, with respect to well depth and how 3 to deal with the shale aquifer, I think it's fairly 4 uncommon, quite rare, as far as I'm concerned, that you 5 would get a well generally shallower than maybe 20 or 30 6 feet, all because, first, there is the old issue of the 7 trust on the part of the user that he's not going to get 8 contamination from agricultural waste on the surface. So he 9 doesn't want a shallow well, really shallow, even if he can 10 get it. 11 Secondly, 12 there is the concern for the well going dry good parts of the year. 13 If that well is going to go dry part of the year, it's really not a usable well. So it has 14 to be in the saturated zone, with enough production to get 15 useable quantities of water all year long. Otherwise, it's 16 not a well. 17 Then, finally, 18 another factor is the water quality and the threshold is fairly low. If he's not going to be 19 happy with the water quality because of salinity, because of 20 mineralization, because of turbidity, if you don't have a 21 happy customer, you don't have a well, better get another 22 one. 23 So in many of these regards, there are sensible 24 things that define what's a well, and, of course, the last 25 one is you've got to get the quantity that will serve a O ANN RILEY & ASSOCIATES, LTD. Court Reporters 1025 Connecticut Avenue, NW, Suite 1014 Washington, D.C. 20036 (202) 842-0034
1 112 family and serve them all year long. 2 So there are some fairly sensible things, 3 regardless of what'the regulations might be, with respect to 4 well construction that govern what's a useable well. 5 MR. THAGGARD: Thank you. We'll take this 6 comment. 7 MR. ZHENHUA: My name is Zhenhua, and I'm from 8 Argonne National Lab. I have a few comments, after I heard 9 this morning's presentations. 10 My comments here address Dr. Neuman's questions on 11 conservative estimates. Really, a 1-D model is really not 12 conservative or liberal; rather, it's a simplified model, 13 first, that's my comment. 14 Second is I think one issue -- my background is 15 contaminant transport. So I'm sort of a little bit 16 different from everybody here, mostly concentrated on the 17 screening model here -- is that there are two issues 18 involved here. One is a ground water model. You're 19 modeling ground water flow, and I think that has been 20 concentrated on this morning. 21 Another issue is contaminant transport, which is 22 separate from the ground water flow, for example, and we're 23 talking about KD, and that's one issue here, and that 24 usually is a first order model. l 25 A different model, for example, a diffusion model ANN RILEY & ASSOCIATES, LTD. Court Reporters 1025 Connecticut Avenue, NW, Suite 1014 Washington, D.C. 20036 (202) 842-0034
)
1 113 i depends on the chemical behavior and that has been very L_7 2 explicit in lots of VOC kind of contaminants. One gentleman 3 just mentioned this natural attenuation, that a particular 4 situation now has been heavily emphasized by DOD or EPA. 5 The one particular issue is the modeling as a 3-D behavior ! 6 rather than -- in other words, you have to observe or take 7
-- centinually modeling and monitoring -- modeling and 8
monitoring what's going on in the plume, 9 and I think that is probably beyond the model we're talking about here. Mainly, 10 it's a screening model. 11 That's my comments. 12 MR. THAGGARD: Thank you. 13 With that, I'm going to turn the meeting back over to Tom. As I indicated, we're 14 going to get to the rest of these questions tomorrow. 15 MR. NICHOLSON: Great. 16 , I want to thank all of the speakers this morning and everyone sitting around the table 17 and all the questions from the audience. 18 Just a little bit of logistics. This auditorium 19 does have public access, so the room will not be locked. So 20 it will be open. So, please, take anything with you that 21 you consider valuable. 22 We'll leave now and have lunch. We will start 23 promptly at 1:00, and Walt Beyeler from Sandia will make the 24 presentation. 25 Thank you, i i l ANN RILEY & ASSOCIATES, LTD. l O' Court Reporters 1025 Connecticut Avenue, NW, Suite 1014 Washington, D.C. 20036 (202) 842-0034
6-114 1 [Whereupon, at 12:07 p.m., the workshop was > 2
- 1. recessed, to reconvene at 1:00 p.m., this same day.]
3 4 l 5 6 7 8 9 10 11 12 13 1 14 ) i 16 l l 17 18 , i i 19 20 I 21 l 22 23 I 24 l 25 p-ANN RILEY & ASSOCIATES, LTD. Court Reporters 1025 Connecticut Avenue, NW, Suite 1014 ; Washington, D.C. 20036
~
(202) 842-0034 l
115 1 AFTERNOON SESSION 2 [1:11 p.m.] 3 MR. NICHOLSON: I'm glad everyone is back from 4 lunch. We're going to continue our program now. What we're 5 going to talk about is we're going to talk about publicly 6 available data sources and databases for estimating ground 7 water flow and transport parameters. 8 We have asked two of our contractors who are doing 9
-- Walt Beyeler and Glendon Gee, who~are doing related work, 10 we asked them and they are in the process of finding 11 publicly available databases, we asked them if they could 12 share some of their experience and what they have been 13 learning.
14 So the first speaker will be Walt Beyeler, $4ho is , () 15 16 at Sandia National Laboratory, and Walt is going to talk about the D&D code and the types of information they put 17 into those parameters. 18 Walt? 19 MR. BEYELER: Again, as Tom mentioned, I wanted to 20 review some of the data sources that were used in the 21 parameter analysis for the D&D code, specifically those that 22 are hydrologic parameters. 23 I wanted to also briefly mention some of the other 24 parameters that are used in the ground water pathway; that 25 is, the behavioral parameters that describe the use of O ANN RILEY & ASSOCIATES, LTD. Court Reporters 1025 Connecticut J. venue, NW, Suite 1014 Washington, D.C. 20036 (202) 842-0034 r . . . . .
r; l 1 116 ground water. These aren't strictly hydrological L 2 parameters, but they, in a sense, determine the importance 3 of the hydrological parameters by specifying the amount of ' 4 water that gets used and the different ways that it gets 5 used by the resident. 6 Then I would also like to go through the physical parameters 7 that characterize the ground water system in the Volume 1 8 model, thickness of the unsaturated layer, the porosities of 9 the surface soil lasmr, the unsaturated layer, the 10 saturations of those layers, the infiltration rate from the 11 unsaturated layers into the aquifer, and the partition 12
- coefficients that characterize the unsaturated layer.
13 Just to quickly go over, again, the behavioral 14 parameters. Use of W is drinking water rate for the 15 resident farmer. The data for drinking water came from the 16 US Department of Agriculture's annual nationwide food 17 consumption survey. VDR is the volume of water that is 18 removed for domestic use. Values for that parameter were 19. obtained from the USGS inventory of per capita water use 20 data from a self-supplied water system. So these were 21 annual usage estimates from people who get their primary 22 water from wells, as opposed to general consumptive use. 23 The irrigation rate estimates were obtained from 24 the US Department of Commerce, their farm and ranch 25 irrigation survey. I\ ANN RILEY & ASSOCIATES, LTD.
\'- Court Reporters I 1025 Connecticut Avenue, NW, Suite 1014 Washington, D.C. 20036 (202) 842-0034
r 1 117 I should mention, also, the specific references
~s 2 kl for these parameters are provided in a final slide at the 3
back of the package. 4 The parameters that characterize the physical part of the 5 ground water system include the depth to ground water, the 6 porosities, 7 relative saturations and infiltration rate. The depth to ground water was -- distribution for that was 8 estimated from a compilation of USGS and state observation 9 well reports. 10 The porosity, saturation and infiltration rates 11 were developed from a model that related to soil texture 12 ~ classification to basically impose what seemed to be a 13 reasonable correlation among these parameters, and, 14 intuitively, the porosity, the relative saturation, the 15
-~) infiltration rate would be a function of the type or soil 16 that would occur at the site. So to -- rather than to 17 impose that relationship in some statistical way, we relied 18 on a simple model that connected the estimates of those 19 parameters with the soil classification.
20 I'll discuss that a little bit in a moment. 21 But in overview, Carsel & Parrish and in a report 22 by PNL, which I think Dr. Gee may discuss in a little bit 23 more detail, provide distributions for porosity, for 24 j saturated hydraulic conductivity, as well as for other soil ! 1 25 physics parameters as a function of soil texture ANN RILEY & ASSOCIATES, LTD. i
\- Court Reporters 1025 Connecticut Avenue, NW, Suite 1014 Washington, D.C. 20036 (202) 842-0034 l
l l l j
118 1 classification. 2 The US Bureau of Reclamation provides an estimated 3 percolation fraction; that is, amount of applied ster that 4 ultimately escapes the root zone and becomes deep 5 infiltration as a function of the saturated soil 6 permeability. 7 Those two pieces of information, along with the
'8 application rate at the surface, were used to estimate these 9 parameters.
The application rate at the surface includes 10 both an estimate of precipitation at the site and also an 11 applied irrigation rate consistent with the definition of 12 the scenario. 13 This illustrates the sort of simple model that 14 connects the soil texture classification to the estimates of 15 [ the parameters that are directly used by the model. The 16 soil texture classification is associated with the
'17 distribution for porosity, saturated hydrah>ic conductivity, 18 and, among other soil physics parameters, the exponent B of 19 the Campbell relationship for unsaturated permeability.
20 So the soil texture via the saturated hydraulic 21 conductivity and the relationship that the US Bureau of 22 Reclamation provides gives an estimate of the percolation 23 fraction for a particular soil texture, the percolation 24 fraction and an estimate of precipitation, plus irrigation 25 rate, then determines the amount of infiltration, limited by ANN RILEY & ASSOCIATES, LTD. O Court Reporters 1025 Connecticut Avenue, NW, Suite 1014 Washington, D.C. 20036 (202) 842-0034
119 1 the saturated hydraulic conductivity. '_ 2 That information, along with the unsaturated 3 hydraulic condu::tivity relationship determines the 4 saturation that is consistent with that rate of 5 infiltration. 6 Lastly, the partition coefficient data that were 7 used in the parameter analysis included the compilation of 8 Thibault and others. These were referenced in Volume 1. 9 They provide values of KDs for a number of soils, other 10 estimates of KDs based on concentration ratio measurements. 11 In addition, we looked at a database maintained by 12 the Nuclear Energy Agency, which contained extensive 13 information on soil batch -- on batch tests for KDs, some 14 information on the dynamic column test for KDs. i () 15 16 Again, there are some more specific references for those data sources, which I won't go through, but should be 17 in the handout material that you have and a more complete i 18 list of references can be found also in the Volume 3 of 512, 19 as well as a more extensive discussion of those particular 20 data sources and how they work, how they were used. 21 MR. NICHOLSON: Thank you. Leave that up there, 22 please, while we have a discussion. 23 MR. BEYELER: Sure. I
- 24 MR. NICHOLSON
- 1 l
Are there any questions for Dr. 25 Beyeler? If anybody has a question about the information or ' gg ANN RILEY & ASSOCIATES, LTD. ( ) Court Reporters 1025 Connecticut Avenue, NW, Suite 1014 Washington, D.C. 20036 (202) 842-0034
120 1 how -- I shouldn't say how D&D came up with its information, 2 because it's all values. No questions? Boby? 3 MR. EID: Yes. I have just a comment and a 4 question. The current version of D&D contains default 5 parameters and the default parameters, they respond to all 6 radionuclides, such that you will ensure that the 90th, 7 above all radionuclides, we'have parameters above the 90th 8 percentile. 9. Do you have any plans for modification of those 10 parameters or the approach for selection of the solution in 11 order to account for all radionuclides, or do you have some 12 kind of other approach? 13 MR. BEYELER: I think the approach that would be 14 used in Version 2 of D&D, rather than using default 15 ()) % parameters or all sources, would basically directly look at 16 the distribution of doses for each particular source; so in 17 a sense, bypass the definition of specific default 18 parameters. 19 MR. EID: In other words, that currently you have 20 the dose for certain radionuclides as higher than the peak 21 dose for the individual radionuclide. 22 MR. BEYELER: Well, the dose for each radionuclide 23 is within the distribution of doses for that radionuclide. 24 The dose for each radionuclide may be higher than, for 25 example, the 90th percentile of the dose distribution for y ,) ANN RILEY & ASSOCIATES, LTD. Court Reporters 1025 Connecticut Avenue, NW, Suite 1014 Washington, D.C. 20036 (202) 842-0034
r- . l 121 1 that radionuclide, i f [~')
%-l 2
Does that answer your question? 3 MR. EID: Yes. 4 MR. NICHOLSON: Mr. Meyer, PNNL. 5 MR. MEYER: Thanks, Tom. I wonder if you could } 6 just clarify for us what the distributions that you just 7 presented are intended for. For instance, the precipitation 8 represents nationwide distribution of precipitation, 9 correct? Same thing for irrigation? 10 MR. BEYELER: Yes. 11 MR. MEYER: So in terms of Tul application at a 12 particular site, how would these distributions be used? 13 What were they intended for use as? I understand that they 14 were used in the derivation of the default parameters, but () 15 you haven't gone into any of that. Was there some other 16 use, also, that they're intended for? 17 MR. BEYELER: I think the follow-on use of the -- 18 well, let me start with what the distributions were meant to 19 describe. 20 In a slid. chat Mark presented earlier, the 21 screening calculation is a calculation that is meant to be 22 done with source term information. So the distributions 23 that are used for the parameters are uncertainty 24 distributions for those parameter values, given, in effect, 25 no information about that parameter value. Therefore, the ANN RILEY & ASSOCIATES, LTD. s Court Reporters } 1025 Connecticut Avenue, NW, Suite 1014 Washington, D.C. 20036 (202) 842-0034
1 l 122 1 distributions are meant to characterize possible parameter i l (J 2 values, possible site-specific parameter values might be 3 used there, in a sense, the broadest distribution that -- a 4 distribution describing the uncertain about that parameter i 5 value given no site-specific information about that ! i 6 parameter. 7 So the process of going from that broad 8 distribution to a more site-specific distribution is I think 9 something that is under discussion in the development of l 10 guidance. 11 Clearly, there would be updating of that 12 distribution in consideration of site-specific information. 13 MR. NICHOLSON: A question? 14 MR. POTTER: Yes. With respect to -- O 15 g ,f MR. NICHOLSON: Identify yourself, please. 16 MR. POTTER: I'm sorry. Tom Potter, Radiation 17 Protection Consultant. With respect to KD, which is one of 18 the most important parameters in dose assessment related to 19 the ground water pathway, you mentioned sources of 20 information, the NEA database, and also Shepherd and 21 Thibault. You had mentioned in some of your earlier work 22 that I recall reading that unlike the Shepherd and Thibault 23 work, which was to identify frequency distributions for KD 24 that varied with soil type, when you looked at a broader 25 base of data, you were unable to find such clean separation. [) \_/ ANN RILEY & ASSOCIATES, LTD. Court Reporters 1025 Connecticut Avenue, NW, Suite 1014 Washington, D.C. 20036 (202) 842-0034
123 1 Would you care to elaborate on that a little bit /"' 2 V)- for us and bring us up-to-date with your current work in 3 that regard? 4 MR. BEYELER: There is no current work going on in 5 that regard. Again, we did look ct correlations amongst 6 percent silt, percent sand, percent clay contents, with the 7 KD data, both in the Shepherd and Thibault, and in the NEA 8 database, and did not find a significant correlation. 9 I find that a bit surprising, but that's -- again, i l 10 there are some more details of that provided in' Volume 3, 11 but there has been no additional work on that. 12 I'm not -- I guess as far as a potential source 13 for site-specific data, I'm not sure that that assumption is 14 necessarily relevant in the -- if it can clearly be () 15 16 demonstrated that there is a specific soil classification at a site, then it seems reasonable that clearly KD data that 17 are collected for that soil type would be especially 18 relevant, and whether or not there is a discernable 19 correlation with other types of soil that do not exist at 20 that site seems, in a sense, irrelevant. 1 21 MR. NICHOLSON: Dr. Neuman, University of Arizona. 22 DR. NEUMAN: It is my understanding that a 23 determination on whether or not D&D can be used at a site 24 and accepted is based, in part, on information regarding 25 whether or not there already is contamination in the vadose l Il
~'
- ANN RILEY & ASSOCIATES, LTD.
Court Reporters 1025 Connecticut Avenue, NW, Suite 1014 Washington, D.C. 20036 (202) 842-0034
124 1 zone and the ground water. 2 In order to obtain that information, some 3 site-specific investigation should already have been 4 performed. So I cannot conceive of a situation where no 5 site-specific data are available and yet D&D is accepted for 6 screening purposes. 7 MR. BEYELER: Yes. I guess I'd just remark that 8 the practical condition may never occur. It was, 9 nonetheless, I think, difficult to anticipate the specific 10 amount and type of information that would be required to be i 11 available or expected to be available at sites. So the 12 condition was, in a sense, to structure the analysis under 13 the requirement that there was -- there was no presumption 14 about the amount and type of additional information that 1 () 15 16 would be available. j 1 l I think there is a practical matter and perhaps i 17 folks from licensing would care to comment, that they do, in 18 fact, frequently have a good deal of information about a 19 specific site. 20 MR. EID: Yes. I would like to say that the first 21 block that Mark Thaggard shows in the decision framework 22 starts with the information available on the site. So you i 23 will have some information and typically if you have 24 contamination in the aquifer, you will have information 25 because of monitoring activities. So you will have some O ANN RILEY & ASSOCIATES, LTD. Court Reporters 1025 Connecticut Avenue, NW, Suite 1014 Washington, D.C. 20036 (202) 842-0034
125 1 information to know whether there is contamination and to 2 proceed forward with using D&D. 3 MR. NICHOLSON: Thank you very much, Walt. What I 4 would like to do'now is move on to the next -- we have a 5 logical progression here, and Glendon Gee, Pacific Northwest 6 National Laboratory, will speak about available databases 7 for VZ and ground water transport analysis. i 8 MR. GEE: The disclaimer that I would make on this l 9 i talk'is that even though it says we're going to talk about l 10 all of the databases, Walt has talked about whatever ! 11 geochemistry is going to be discussed and I'll talk about 12 the hydrology. i I 13 I've been accused by my colleagues at PNL, my 14 geochemist friends, I needle them that they never talk about () 15 hydrology and always ignore it when they make their ! 16 proposals, and here I am talking about transport without 17 talking about geochemistry, but Walt did that for me. I 18 So you'll have to at least blar with me on talking 19 about data sets that are available and I'll be talking 20 specifically about things related to measurement of the 21 hydrologic properties related to some of these conceptual 22 models. 23 Here is an example from one of our reports, quite 24 specific, you have a waste site with a cover on it, a soil 25 that's been some kind of designator that someone has ("3 ( ,) ANN RILEY & ASSOCIATES, LTD. Court Reporters 1025 Connecticut Avenue, NW, Suite 1014 Washington, D.C. 20036 (202) 842-0034
e b 126 1 supplied a -- either measured or there's been some kind of 2 Ug generic identification of the soil type, a little bit about I 1 3 the subsurface, sandy loam, saturated zone, and then some j 4 information about the depth of water. That's usually the 5 kind of information that's available to begin with. 6 And the summary slide that I would put up here 7 that you'll probably see later on, since this is courtesy of 8 Phil, you have certain types of information. Some of it is 9 available on the web, some of it not. The generic 10 information come from both regional and national sources . 11 There are hydraulic property information in the 12 \ database from the ARS, the salinity lab. There are other l 13 sources that are identified in our NUREG report, but using 14 this information, one can attain parameter uncertainty 15 distributions, and I'll give you an example of that. 16 What that does with those larger databases is give 17 you some bounding values. You assume that the particular 18 soil or subsoil at the waste site somehow has a 19 representation similar to these large regional or national 20 databases. 21 Stepping down in terms of increasing site-specific 22 information, there are indeed state, local state information 23 that has been now compiled. Those of you who have a soils 24 background know that every state and every county have soil 25 surveys and these have been put into STATSGO and SSURGO. l ANN RILEY & ASSOCIATES, LTD. [~~') \/ Court Reporters 1025 Connecticut Avenue, NW, Suite 1014 Washington, D.C. 20036 (202) 842-0034
127 1 STATSGO is a map-based database that is on the web, that is j-~ 2 more a generic database. SSURGO has more site-specific 3 state information in it. 4 Then, of course, to do the forcing function 5 information, the National Climatic Data Center has the 6 precipitation from literally thousands of sites and tne USGS 7 has the national water information system and a ground water 8 information system. 9 But using this kind of information, one can, from 10 the more local data sets, you can modify the uncertainty 11 distributions and bounding values and improve those 12 estimates, but this is all in the absence of site-specific 13 data. 14 Then, of course, locally, extension service, state 15 agencies, university experts can provide site-specific 16 information. Cities have informacion on percolation tests 17 and other things from building permits and so on. Using 18 that kind of information allows you to modify the 19 uncertainty distributions. 20 Then, of course, there is the site-specific direct 21 measurements. Putting these together, and there are 22 processes that we've identified in some of our reports that 23 allow you to use that information, coupling them with the 1 24 other larger databases to make uncertainty predictions. 25 I'm going to be talking about hydrologic O ANN RILEY & ASSOCIATES, LTD. Court Reporters 1025 Connecticut Avenue, NW, Suite 1014 Washington, D.C. 20036 (202) 842-0034
s l l-I 128
- 1 i
! evaluation and so the input is precipitation. You have ! 2 water storage changes through the vadose zone, through the 3 waste and to the water table, whether you have a cover of 4 some kind and whether it's 1-D or 2-D or 3-D, it all 5 involves these processes that are controlled right near the I 6 surface, and that kind of information is basically available 7 in the data sets that I've talked about. 8 The climatic data is the forcing function. The 9 soils data provides storage and transmission function 10 information, along with the geologic and hydrologic data . 11 The thing that I haven't mentioned much and is not 12 necessarily as readily available is the vegetation, because 13 that is a very dynamic thing, but also very important in i i 14 terms of the hydrologic processes, particularly the 15 evapotranspiration.
)
16 So we want to get information related to 17 precipitation, have that in a data set that's useful. 18 Similarly, as Phil talked about this morning, the net 1 i 19 infiltration, also the redistribution, if that's available, 20 which is governed by the soil properties, and then 21 subsequently, what kind of throughput in terms of drainage 22 into the waste and into the aquifer. 23 The other information that can be obtained from 24 soil conservation, which is also tabular information, is the 25 runoff soil type curves. Then less information certainly a l I i ' g-%g ANN RILEY & ASSOCIATES, LTD. ( ) Court Reporters 1 1025 Connecticut Avenue, NW, Suite 1014 Washington, D.C. 20036 (202) 842-0034
1 129 real, at this point, ,-_ 2 nebulous piece of information that is not necessarily directly obtained is the vegetation, and I 3 will talk about it a little bit. 4 In terms of the climatic data set, of course, the 5 generic information'is obtained from National Weather { 6 1 Service stations and one would typically, in an analysis, 7 8 take the local NWS obtained from the National Climatic ata D Center, simply a look-up, and then the help model, as some 9 of you know, you can simply type in, as a default, the 10 nearest weather station and that provides you all the 11 information, at least in principal, that you need to know 12 about ET, except the plant parameters. 13 So you would want hopefully the longest climatic 14 record, because that gives you more statistics on which to 15 base your estimates and uncertainties. 16 Then if available, and often it's true, you do have some site-specific climate 17 information, but ! 18 typically that information is very sketchy and the climate record is relatively short, generally less 19 than ten to 15 years. 20 So the statistics are not nearly as good. 21 ! The kind of information, 22 as Phil mentioned this morning, includes the water input, the precip, the maximum 23 temperature, solar radiation, wind speed and humidity. 24 These latter four are typically used for potential ET 25 estimates. 9 k/ ANN RILEY & ASSOCIATES, LTD. Court Reporters 1025 Connecticut Avenue, NW, Suite 1014 Washington, D.C. 20036 (202) 842-0034
r l 130 1 In our NUREG report, Phil identified the process 2 in obtaining the information from the National Climatic Data 3 Center. There is a three-step process. 4 Typically, if you don't have any real need for daily information or at least 5 can't justify it, i you obtain these monthly meteorological 6 l parameters, but typically all of those parameters are on a 7 daily basis and in some cases, you can use some kind of 8 l function to estimate hourly information. 9 But you identify the appropriate station, you can ! 10 download the data using an anonymous FTP directly on your 11 computer, and then run a computer code that puts that data 12 into a data file. That's pretty standard things. 13 Again, available in standard water balance 14 calculations, with the HELP code, with I think MEPAS and the 15 other codes that you've heard today, particularly MEPAS i 16 certainly does that. t
'17 Here is an example of the kind of information from 18 one anonymous site.
The kind of information you need, you 19 have precipitation for every month of the year, with 20 standard error, similarly, temperature, wind speed, cloud 1 j i 21 cover, and oftentimes you don't have solar radiation, but 22 you have the cloud cover and you can make some estimates on 23 the solar radiation, humidity and rainy day periods. I 24 With that kind of information, you can generate an i 25 uncertainty in the weather record. So you have -- at this
)
I ANN RILEY & ASSOCIATES, LTD. O Court Reporters 1025 Connecticut Avenue, NW, Suite 1014
)
Washington, D.C. 20036 ) (202) 842-0034 1
I 131 ! 1 particular site, you have variations of over -- coefficients ! (-] 2 V of variation well over 100 percent for almost every month of 3 the year and that kind of uncertainty then should go into 4 your model and leads to uncertainty in the ultimate 5 analysis. 6 Similarly, with runoff, ET and drainage, this kind ! 7 of information is available, and this particular analysis 8 was for Ohio, where.there was a long-term record of 9 drainage. I 10 Now, I talked about the availability. I want to 11 ; talk a little bit about some of the limitations and part of 12 the limitations is in how you use the data. You noticed I 13 talked about monthly records and I'd like to just illustrate i 14 on where this might lead us in terms of conservative 15 estimates or not. UT 16 I don't have a map of the Hanford site to show 17 you, but the USGS performed a regional ground water study, 18 where they took climatic data available from the local 19 weather station, long-term record, so they knew the 20 precipitation variations. 21 They took soil data, where they had to estimate 22 some of that, because there was only one soil map and it was 23 very old. So they had to update it and correlate it with 24 the county maps, because the county data was not available. 25 In addition, they had soil type from aerial surveys. ANN RILEY & ASSOCIATES, LTD. (s) Court Reporters 1025 Connecticut Avenue, NW, Suite 1014 Washington, D.C. 20036 (202) 842-0034
132 1 So they took that information and they basically 2 O 3 tried to estimate the recharge from the plateau where the processing, waste processing occurs, to the river, to the 4 Columbia River. Of course, there is variation in 5 topography, there is variation in soil type and vegetation. 6 The precipitation records were assumed to be, 7 because of the relative uniformity of the area, to be 8 similar. So the precipitation was similar over the 9 landscape, but the soil type varied, as well as the 10 vegetation. That was all put into a model and an analysis 11 done. 12 In that analysis, they tried to estimate recharge. 13
.They did it for the entire region outside of Hanford, as 14 well, and they were calculating these numbers, the amount of 15 entering, the ground water, as recharge.
16 They took a 25-year record and they basically did 17 it several ways. They took the information from -- assuming 18 a long-term average, basically the kind of numbers that I 19 put up on the board before for Hanford, for the months of 20 January through December, as opposed to the actual records, 21 the actual historical record, and I'll just show you the 22 differences in the calculations. ! 23 Basically. in their analysis, from the USGS, they 24 showed that the recharge calculation for a 21-year climatic 25 record had a maximum of 58 millimeters, a minimum of .5, ANN RILEY & ASSOCIATES, LTD. C1 Court Reporters 1025 Connecticut Avenue, NW, Suite 1014 Washington, D.C. 20036 (202) 842-0034
133 1 with an_ average of 12, if you took the actual daily record rs 2 for that 21 years. 3 In the other hand, if you took the long-term 4 t averages, which is the kind of data you get from the NCDC, 5 the kind of information you get on the web, the average 6 climatic record was -- the maximum was 31, the minimum was 7 zero, and the average was two. So a factor of six 8- difference in recharge. 9 The point is that in some cases, this may not make 10 a lot of difference, but at Hanford, where we're trying to 11 calculate transport rates, this 12 millimeters a year 12 represents basically a six-fold increase in transport to the 13 ground water, and in our case, it was important. 14 Another case that was reported by Phil in the 15 earlier NUREG was a case in South Carolina. In this 16 particular case, Phil took a water balance model and he 17 calculated the water balance parameters based on daily 18 average precipitation. These are the sums that were 19' computed for on an average for the year. Infiltration of 20 about 1,121. 1 This is the net water moving through the 21 surface soil, runoff of 84, storage of minus two, ET of 769, i 22 1 with a net infiltration, basically a drainage number of 355. 23 When he then took the hourly data, you can see a i 24 striking difference in infiltration, almost a reduction of 25 50 percent, and the net infiltration dropped from 355 down ANN RILEY & ASSOCIATES, LTD. Court Reporters 1025 Connecticut Avenue, NW, Suite 1014 Washington, D.C. 20036 (202) 842-0034 t
1 134 to 48. One could do this analysis and say, well, it's a
, 2 k
conservative estimate, but we basically were at the same 3 site and the only difference is that we were looking at 4 hourly averages versus daily. 5 The.only point I'm trying to make here is that one 6 has to be cautious when you use this kind of information 7 from data sets and this was from a National Weather station f i 8 data set, right? Correct me, Phil. Okay. 9 But a significant difference in the amount of net 10 infiltration. So if one is going to design a -- if you're 11 going to use this in the codes, and you're using 355 versus 12 48, it could make a difference in the transport analysis , 13 f I just want to say a few things about the database i 14 information that's available. I've already mentioned that 15 the -- that these two databases are on the web, readily 16 available, and from these databases, a lot of information is 17 typically extracted to make estimates of hydrologic 18 properties, porosity and other information. 19 The USDA has been working on a more detailed 20 database, but it's basically under construction. It will 21 not be available for at least another year, but it has what 22 I think is a better estimator of hydraulic properties. From i 23 j these databases, you get soil texture, you get water holding 24 capacity, field capacity, wilting point information, and in 25 some cases, in some cases, you get seasonal water table
/\ ANN RILEY & ASSOCIATES, LTD.
~ Court Reporters 1025 Connecticut Avenue, NW, Suite 1014 Washington, D.C. 20036 (202) 842-0034
135 1 depth. 2 Now, I would caution that all of these databases 3 which are typically used are designed primarily for 4 agronomic purposes. They're basically soil PEDON 5 information that's taken from the top meter and if you're 6 going to predict transport to ground water and ground water 7 is six meters or 20 meters down and you're using this kind 8 of information to estimate the subsoil and you're extending 9 that information beyond what I think is a proper limit. 10 However, for water balance purposes, I think these 11 data sets are good. I would just show you that the 12 Carsel-Parrish information gives you some details where you 13 can get joint probability distributions, so you can get 14 uncertainties in the soil types. In this particular case,
; 15 there are some 1,500 different soil types, ranging from 16 Sands to Clays.
17 But, again, a word of caution, and that is that 18 these 1,500 soil, individual soils were taken from all over 19 the United States. They have different mineralologies. 20 l Even though one may be a sand, it has a different 21 ! mineralogy, and, therefore, it may not necessarily be 22 l i applicable to your site, even though the texture is the same 23 as what you identify on-site. 24 It's, again, from the NUREG, courtesy of Phil's 25 fine work, and, again, from NUREG-6565, these are some of x 1 ANN RILEY & ASSOCIATES, LTD. Court Reporters 1025 Connecticut Avenue, NW, Suite 1014 Washington, D.C. 20036 (202) 842-0034 L
136 1 the data sets. This is the K-SD statistics, but generally 2 distribution functions for residual water, wilting point, 3 field capacity, the B value that Walt mentioned. 4 This kind of information is available from these i 5 large data sets. How applicable it is to an individual 6 site, again, is -- may be questionable. It may be better if 7 you have some detailed information from a county soils map, 8 but at least this information is a place to start.
.9 I would just say in terms of the geologic and 10 hydrologic databases, that indeed, as Walt pointed out, this 11 kind of information as available and contractors and others 12 can get at this.
There is a national water information 13-system, but when I went to retrieve it on the web, I 14 contacted one of my colleagues from Nevada, who is with the 15 USGS, and he had these words to say; the USGS NWIS database 16 includes ground water level information, but my 17 understanding is that on-line availability is primarily 18 handled at the USGS district or project level, and then he
-19 went on to say how to obtain that.
20 And then he referred me to the chief of the office 21 of ground water, who said that's correct, and that l 22 l presently, the availability of ground water data on the web 23 is quite limited and variable among the states; we are 24 presently modifying our national weather information system I l 25 to serve ground water data on the web and plan to have ANN RILEY & ASSOCIATES, LTD.
/
Court Reporters 1025 Connecticut Avenue, NW, Suite 1014 Washington, D.C. 20036 (202) 842-0034
r 137 1 improved national capability to serve ground water data in t 2 the future.
'- 3 But then he goes on to say something that I think 4
is significant and perhaps worthy to note; we have some 5 parameters from aquifer tests stored in NWIS, but many of 6 these data tend to be only in USGS reports. Of course, the 7 quality of hydraulic parameter data is highly variable, 8 dependent on many factors, including the type of test, 9 whether it be multiple or slug tests. One thing I suggest 10 is to encourage contractors to look carefully at such data, 11 whatever their source. 12 That's from what I think is a very reputable 13- sou ca chat indeed we should be cautious about the 14 availability and the type of data that we're obtaining. 15 I'd make one further statement about the i 16 vegetation database. If you are going to use this kind of 17 information, a typical thing, those of you who have run the 18 HELP code or similar codes, MEPAS has similar kind of 19 default parameters, that one can, if you use a vegetation, 20 that you assume that the waste site is gcing to be in grass 21 or some other kind of cover, you simply click a button and 22 that information is available to that extent. 23 I sometimes question whether that's efficient 24 information. There are limited data, regional ecosystem l 25 assessments in some states and some regions. Pacific i ANN RILEY & ASSOCIATES, LTD. ' Court Reporters N 1025 Connecticut Avenue, NW, Suite 1014 , Washington, D.C. 20036 i (202) 842-0034
138 1 Northwest has a regional ecosystem assessment that provides 2 detailed information about vegetation types and persistence . 3 Of course, in the county soil surveys, there are vegetation 4 types. A ceries of associations generally provide the type 5 of cropping, the type of vegetation that is typical, but 6 there is no national database available for this vegetation 7 and you would have to rely primarily on the default values 8 from HELP or MEPAS or these other codes, 9 I'm kind of beating a dead horse here, but I'll 10 just repeat these kind of things. The national weather data i 11 are typically daily values. If you want to get the -- see 12 the dramatic effects of hourly information, then you have to k 13 embed that kind of information into your analysis, because 14 there are -- in some cases, there are dramatic differences 15 in how that precipitation is distributed, whether it runs 16 off or runs into a site. 17 Again, the national soil database is limited 18 typically to the top meter and a half. So if you're going 19 to use subsoil, try to extrapolate that to subsoil, textures 20 and other kinds of information, that's probably extending it 21 beyond its limits. 22 The NWIS is currently under construction and the 23 national plant data is not available. 24 And just finally, we do have large weather and 25 ! soil data sets that are on the web. As far as I know, ! l ( ANN RILEY & ASSOCIATES, LTD. Court Reporters 1025 Connecticut Avenue, NW, Suite 1014 Washington, D.C. 20036 (202) 842-0034
139 1 there's no immediate plans to provide this vegetation
,e'l 2 information in more detail than perhaps is available in HELP 3 or MEPAS.
4 I would just caution that weather and soil data 5 should be used with caution on large data -- and large data 6 sets may be inappropriately applied. 7 The site-specific data are largely absent, but may B be required, in some cases, to verify the use of generic 9 data. 10 I'll entertain questions. 11 MR. NICHOLSON: Are there any questions for Dr. 12 Gee? 13 [No response.] 14 MR. NICHOLSON: Glendon, your comments -- I assume 15 (} that the comments you're referring to came from the USGS' 16 Office of Ground Water. Did he talk much about the ground 17 water atlas that's under development? 18 MR. GEE: I know it's under development. We 19 searched it on the web, but he basically said that their 20 NWIS is still under construction. 21 MR. NICHOLSON: There is quite a bit of 22 information, though, I think, if you go to the district or 23 the sub-district level to survey. Unfortunately, I don't 24 think there is anybody here from the USGS. We invited them 25 here, but unfortunately they didn't attend. But there are a O, ANN RILEY & ASSOCIATES, LTD. Court Reporters 1025 Connecticut Avenue, NW, Suite 1014 i Washington, D.C. 20036 (202) 842-0034
r I ! 140 1 lot of information. 2 There's also a lot of people here from the state. 3 Is there anyone here from a state government that wants to 4 comment about availability of information at the state 5 level? 6 [No response.) 7 MR. NICHOLSON: No one from the state wants to 8 comment. Okay. Are there any questions for Dr. Gee? Yes? 9 FROM THE AUDIENCE: At least in one case, one of 10 the codes has an input for soil erosion rates. Have you 11 seen anything on that? 12 MR. GEE: There are certainly estimates. In most 13 of the soil survey data sets, there are engineering -- I 14 can't remember the exact title, but there are engineering 15 applications and erosion rates are often cited. 16 There are a ratability indexes that can be ' 17 obtained from the county soil surveys. So that is a 18 parameter that can be obtained, at least for some soil 19 types. 20 MR. NICHOLSON: Any other questions or comments? 21 [No response.] 22 ! MR. NICHOLSON: What we'd like to do now is just l 23 l have a very brief group discussion on this topic. There are 24 three questions, if you turn to page, I think it's five in 25 your agenda, you will see these three questions. The people I O ANN RILEY & ASSOCIATES, LTD. Court Reporters 1025 Connecticut Avenue, NW, Suite 1014 Washington, D.C. 20036 (202) 842-0034
~
E 141 1 at the table and people in the audience, I'd like people to ! 2 comment on this. Ralph, could you turn the lights up? 3 Thanks. 4 l The first question is what are the most convenient ! 5 sources of publicly available information on soil and ground 6 water properties for a specified site. I think Glendon went 7 into this to some extent, but he didn't really get into the 8 possibility of finding other information from studies done 9 locally. For instance, going to universities, like land 10 grants universities, or going to the USGS' state agencies. 11 Who would like to venture an answer for this first 12 question? Glendon? What are the most convenient sources of 13 publicly-available for specific -- you answered this to some 14 extent. '( 15 But for instance, at the Hanford site, there's 16 been quite a bit of work done at that site and other places 17 where DOE has large facilities. What can you say about how 18 you'd go about getting that information? 19 MR. GEE: I think certainly for the Hanford site, 20 the information that is available is typically available 21 from research reports. There are data sets on both the 22 geologic property characterization, thousands of wells have 23 been drilled. The granular metric data, lithology and other ' 24 things have been tabulated in what's called a rock-sand 25 database. This kind of information is available. t ANN RILEY & ASSOCIATES, LTD. ! Court Reporters 1025 Connecticut Avenue, NW, Suite 1014 l Washington, D.C. 20036 (202) 842-0034
. e l
1 142 So we would at least for Hanford, we would go to 2 the local source. We're probably somewhat unique in that I 3 sense, 4 that a lot of characterization work has been done on the site, more so than many others. 5 MR. NICHOLSON: The next question basically talks 6 about how to take this national or regional information and 7 apply it to a specific site. Are there any insights on how i 8 l to determine whether it's efficient to model a site-specific i 9 ground water site? l Walt, do you have a comment on this one? 10 MR. BEYELER: Not specifically. 11 I think Glendon i made some good points about there will also be some residual 12 uncertainty about what an estimate of the site-specific 13 parameter is. I think data from, say, the nearest weather 14 station doesn't uniquely determine the appropriate value 15 that modeling would be considering. 16 There are time variations that occur naturally and 17 variations. 18 MR. NICHOLSON: I think one thing that probably is 19 missing from that question is obviously you have to have 20 some conceptualization of your site and the question is part 21 of your conceptualization comes from the national or 22 regional information sources knowledge, not just data, but 23 obviously we talked about this ground water atlas that the 24 Survey has developed surveys, and then you start asking the 25 question on what level of detail do I need to do my b V ANN RILEY & ASSOCIATES, LTD. Court Reporters 1025 Connecticut Avenue, NW, Suite 1014 Washington, D.C. 20036 (202) 842-0034
143 1 modeling, how' appropriate is that information. 2 Would other people like to comment on this 3 question? Dr. Neuman? 4 DR. NEUMAN: My comment is simply that I wi31 be 5 addressing some of these questions in my talk by examples. 6-MR. NICHOLSON: All right. Finally, for which 7 parameters is it appropriate to use generic data and for 8 which ones is it not. Glendon, you alluded to this already. 9 Could you comment further on this question? 10 MR. GEE: Well, I think with caution, you can -- 11 there are now mounting efforts by states and regions to 12 obtain more detailed and up-to-date weather records, in 13 addition to the NCDC. 14 So I think weather information probably is the 15 most -- likely the most appropriate. 16 Secondly, I think if it's a near surface 17 contamination and you have some site-specific information on 18 near surface soils, then I think it's probably a good 19 estimate to use particularly the county soil surveys in the 20 local area where your waste site is located. 21 Beyond that, I think you're making some -- you're 22 bringing a fair amount of uncertainty with you. I can't 23 speak of water table depths and other things, but I think 24 we'll hear a discussion about if you use regional maps on 25 water table depths, you may miss your mark, unless you have ANN RILEY & ASSOCIATES, LTD. C' Court Reporters 1025 Connecticut Avenue, NW, Suite 1014 Washington, D.C. 20036 (202) 842-0034
1 144 1 some on-site data. 2 MR. NICHOLSON: Is there anyone in the audience 3 who wants to comment? Boby? 4 MR. EID: 5 I would like to comment on the use of generic data for a parameter, which is very important, which 6 f is the distribution coefficient or the KD value, 1 7 i It was indicated by one of the speakers that the 8 KD value could be selected based on some generic literature 9 value or all type of soils and there is no correlation for 4 10 that. I believe it may be, in a generic sense, it's okay, 11 but I guess one needs to be careful that the KD is more 12 site-specific and depends on many factors. 13 I do believe there should be some correlation 14 between the soil type and the KD value, experimental work 15 shows-that KD for sandy soil is different than for clay 16 soil. 17 RESRAD lists the tables for different kinds of soils and ranges of KD parameters for different kinds of soils, as 18 an example. 19 So I believe this parameter is very important, 20 it's a sensitive parameter, and I think I'd caution to use 21 this as a generic parameter. 22 MR. NICHOLSON: Dr. Wilrenga, from the University 23 of Arizona. 24 MR. WILRENGA: Tom, in my opinion, one should 25 basically, at most sites, either small or large, take some 1 ANN RILEY & ASSOCIATES, LTD. O Court Reporters 1025 Connecticut Avenue, NW, Suite 1014 Washington, D.C. 20036 (202) 842-0034 i
e s 145 1
-site-specific data and some of the data is collected f-sg 2 relatively simple. All one needs to know, I guess,
(,') 3 certainly one who is interested in the surface soil, one 4 needs to go out with a drill rig, take some samples, have 5 them analyzed for at least particle site distribution So. 6 you know you're dealing with a sandy soil and a clay soil 7 8 and then one can use the national databases and make some inferences about the water holding capacities or the 9 hydraulic properties of those particular soils. 10 11 But one has to know a little bit about it and this at least the minimum that you could do and perhaps also 12 determine, get a feel for the KD value for that particular 13 soil.
'14 So I would think that is the minimum one has to' 15 do, my opinion.
16 MR. NICHOLSON: Thank you. A question? 1 */ MR. BURKLIN: Yes. Rich Burklin, Siemens. I 18 really know very little about this subject, but if you are 19 trying to get a KD value and let's say you have 30 feet to 20 the ground water and maybe have some literature that says, 21 hey, these are the KDs that you might anticipate, do you 22
.only take soil samples at the top or are you supposed to 23 take them 15 below the ground? What do you do?
24 MR. EID: I believe that is a good question. You 25 need to consider variation at the site and you may take ANN RILEY & ASSOCIATES, LTD. (\ Court Reporters 1025 Connecticut Avenue, NW, Suite 1014 Washington, D.C. 20036 (202) 842-0034
146 1 samples, because KD value, depending on the code that you 2 use, you could assume KD value for the soil, KD value for 3 the unsaturated zone, 4 even some retardation also for the saturated zone in certain codes. The D&D code, for example, 5 does not assume that there is retardation in the saturatea 6 zone. So you do not need i to worry about that. 7 So whereas if you have different KD values between 8 the soil and the unsaturated zone, this means you have two 9 values for the KD or retardation. 10 Now, the site variability, of course, the site, if 11 you have a small site, would be not a problem, but if you 12 have a very large site, it would be a problem. That's the 13 purpose of the probabilistic analysis, where you try to 14 analyze the distributions of the KDs and as we've said, if 15 you move to site specific analysis, you need to be somehow 16 conservative, but not excessively conservative. 17 And we do accept the doses based on these 18 distributions, where it is the mean dose rather than the 19 dose at the 90th percentile or the site-specific analysis. 20 MR. NICHOLSON: Bill Dam, from the NRC's Office of 21' Nuclear Materials Safety and Safeguards. He's a geochemist, 22 so set us straight, Bill. 23 MR. DAM: I was just going to make a comment about 24 USGS data, from my previous work at the USGS. They have 25 this NWIS database, which covers water wells and also stream [ ANN RILEY & ASSOCIATES, LTD. Court Reporters 1025 Connecticut Avenue, NW, Suite 1014 Washington, D.C. 20036 (202) 842-0034
147 1 gauged water data, and then GWSI is the ground water site 2 inventory, as Glendon said. But any data that would be in 3 the database would be published, as he mentioned. 4 There's reports that are published annually with 5 all that data in it. But the thing I would go to would be 6 the actual interpretive report, where they've collected that 7 data and interpreted it, the value of the information, 8 rather than trying to get it off the web site. When it's 9 completed, I would go to the actual site. 10 A couple of the programs that they've had, one is 11 the regional aquifer systems analysis, called the RASA 12 program, several ground water basins in the country have 13 done this regional study and then there is the NWQAP program 14 for national water quality assessment program, and that's () 15 16 very useful if you're interested in water quality data. But I would go to the actual report rather than 17 trying to get data off the web site, when it's available. 18 MR. NICHOLSON: Most of these reports that Bill is 19 mentioning are opap file reports. Some of them are water 20 supply papers. There's a variety of them. Obviously, here 21 in the Washington, DC area, you go over to Reston, USGS 22 headquarters, and you can purchase them there, some of them, 23 the circulars are free. 24 As I said earlier, a lot of the Survey has 25 cooperative programs with the states. ! So like in ANN RILEY & ASSOCIATES, LTD . ' Court Reporters 1025 Connecticut Avenue, NW, Suite 1014 Washington, D.C. 20036 (202) 842-0034 i
148 1 Harrisburg, Pennsylvania, the USGS has an office there and 2 they work with the Pennsylvania Geological Survey . 3 So in most areas, that would be probably one of 4 the best places to go first to get your information . 5 Are there any other comments? 6 [No response.] 7 MR. NICHOLSON: 8 What I'd like to do then is move in a different direction now. We'd like to -- now we've 9 talked about the conceptual models. 10 We've talked about the available databases from the national, regional and perhaps 11 related site-specific. . I 12 I Now we'd like to see what it is like to actually 13 go to a site and try to bring some of this information 14 together. A contractor for us at the University of Arizona, i 15 Dr. Shlomo Neuman, is developing, for the Office of O- 16 Research, a methodology. l 17 We are just in the beginnings of developing a I 18 methodology. We're not sure exactly hov far we can go with 19 this, but we think it might have some value to people. It 20 is not, obviously, et this time, fully developed, but 21 perhaps some of the insights that Dr. Neuman has found in 22 trying to apply his methodology to a real site might give 23 some information out. But this is just a research project 24 at the present time. 25 Dr. Neuman? O ANN RILEY & ASSOCIATES, LTD. (V Court Reporters 1025 Connecticut Avenue, NW, Suite 1014 Washington, D.C. 20036 (202) 842-0034
l I 1 .49 DR. NEUMAN: Can you hear me when I speak into 2 this microphone? O) \s- 3 The NRC staff has identified the issue of conceptual modeling and conceptual model development as one 4 that may result in more uncertainty in ground water 5 predictions of flow and, in particular, transport than 6 uncertainty in parameters. 7 So they asked me if it would be possible to B develop a more or less formal methodology or strategy that 9 would help one to identify, based on available data, at the 10 particular site, conceptual models that the site data would 11 support. 12 For those of us who have experience in i 13 hydrogeologic interpretation of data, it is very common to 14 be in a situation where you bring five colleagues to a roomm s 15
\
and show them exactly the same data and each one would 16 interpret them in a slightly or sometimes rather drastically 17 different way. 18 So the question is, can this process of scientific 19 thought and interpretation of data be somehow formalized . I 20 don't think it can be formalized to the extent that one 21 could develop a step by step methodology, saying step number 22 one do this, step number two do that, and so on. But I 23 think that what one can do is try to pull together important 24 steps, put them in a certain order, but then leave enough 25 flexibility for people to understand that these are ANN RILEY & ASSOCIATES, LTD. Os Court Reporters 1025 Connecticut Avenue, NW, Suite 1014 Washington, D.C. 20036 (202) 842-0034
150 1 essentially guidelines and not a step by step procedure. 2 O 3 What I am going to propose to you .s essentially a rudimentary concept. It is at a very elementary stage. I 4 don't know if the NRC will at all buy into it eventually or 5 not. Certainly, at this point, it is nothing but a concept 6 in my own head and I would be very much interested in 7 knowing what you think about some of these steps that I am 8 proposing. 9 And what I will do is paint for you this 10 methodology at a very high level-. I will not go into great 11 detail. To the extent that there will be details, they will 12 be illustrated by a particular example. That example, I 13 think, harks back to the discussion that we just had a few 14 minutes ago, about the publicly available data, what one can 15 and cannot do with this publicly available data, and then 16 what might be achievable with site-specific data. 17 So here is the proposed methodology, in a 18 nutshell. This is kind of the highest level of the 19 hierarchy that we are proposing, and you will immediately 20 notice that I started my discussion by focusing on 21 conceptual models, but what you see here goes way beyond 22 conceptual models and, in far:, proposes a strategy for 23 ground water model development all the way from the 24 beginning, where one defines the problem, and all the way to 25 the end, when one uses the model or the models to come up ANN RILEY & ASSOCIATES, LTD.
% Court Reporters 1025 Connecticut Avenue, NW, Suite 1014 !
Washington, D.C. 20036 ) (202) 842-0034
i 151 1 with performance measures and some information about the 2 uncertainty of those performance measures. 3 4 So I refer to this as a skeletal framework because it is really just a skeleton of ideas at this point, and 5 rather than, as a conceptual model, I will refer to this as 6 the conceptual mathematical model, very much along the lines 7 of what Phil Meyer told you earlier, and that is these two 8 concepts really have to go together. They don't always, but 9 certainly in the ground water arena, they can and I believe 10 they should go into enough about the physics and mathematics 11 of ground water flow that we can cast those in mathematical 12 form. 1 13 l Without that mathematical form, of course, we 14 cannot speak of mathematical models and of quantity 15 i
/~')
kJ predictions.
" 16 So the first step then, and what I will do is, 17 after I show you this kind of general highest level 18 hierarchy, I will go to each one of these boxes and show you 19 some of the sub-steps that I envision beyond, and those will 20 Ima shown by models and ' will not go beyond that into any 21 level of detail.
22 One has to know why one wants to develop a model, 23 what is the context for the analysis, what are the questions 24 one wants to address, what is the problem or the problems 25 that one wants to solve, perhaps something about how one l ANN RILEY & ASSOCIATES, LTD. Court Reporters I 1025 Connecticut Avenue, NW, Suite 1014 1 Washington, D.C. 20036 (202) 842-0034
152 1 wants to solve those problems, how important is it to solve 2 \ the problem accurately, with little uncertainty. 3 That will determine, to some extent, how much 4 time, money and effort one puts into the analysis. 5 The next step -- and for those of you who are 6 familiar with the development of ground water models, I will 7 assume that what I am proposing here is really nothing but a 8 summary of your experience, to some extent at least. 9 The second step is to define what you know that is 10 relevant to the solution of your problem. Now, that 11 involves knowledge about the site, and I will only be 12 speaking about hydrogeology. There are many other aspects, 13 of course, to the problem, but I'm focusing on hydrogeology. 14 So the question is going to be what do we know 15 about relevant hydrogeology on a regional local scale, what 16 do we know about flow and transport processes that may be i 17 relevant to the problem; of course, what do we know about 18 the source term, what do we know about how the source term 19 is mobilized, transported, in the vadose and ground water 20 zones only. 21 And here the availability of data, of course, 22 enters. ! Then once you have collected this information, I 23 refer to it as knowledge-based rather than data-based, 24 because it involves some knowledge of ground water 25 pathology, some knowledge of ground water physics, soil f ANN RILEY & ASSOCIATES, LTD. \' ~s\# Court Reporters 1025 Connecticut Avenue, NW, Suite 1014 Washington, D.C. 20036 (202) 842-0034
1 153 l 1 physics. If you don't have that, you may have a difficulty 2 (O) following some of the steps which I am proposing here. 3 We're talking about site-specific rather than generic type 4 of modeling. 5 The.whole idea here, contrary to picking up a code 6 off the shelf, is that the developer will conceptualize the 7 system rather than accept a pre-concept or pre-developed 8 concept and will select the correct, maybe develop in some 9 cases, correct mathematical model. Hopefully, one would be 10 able to find one off the shelf and apply it in a way which 11 is consistent with his concept and the data. 12 So the next step then is the one that we are, in 13 fact, focusing at the present time more than on some of the 14 other aspects, is the qualitative conceptualization of I~'i 15 hydrogeology and flow and transport within that V 16 hydrogeological setting. And as Phil has indicated to us 17 earlier this morning, and I have just mentioned, this is the 18 stage where hydrogeologists very often start arguing and 19 sometimes coming to blows over just exactly what do the data 1 { 20 imply. l 21 So if there are multiple hypotheses that one can l 22 derive from a given set of data, multiple conceptual 23 frameworks, this would be the place to articulate them and 24 this would be the place to maybe select those among them 25 that make sense, and those that make less sense, in light i ANN RILEY & ASSOCIATES, LTD.
) Court Reporters 1
1025 Connecticut Avenue, NW, Suite 1014 f Hashington, D.C. 20036 L (202) 842-0034 { I l l
154 1 hopefully of available data, and perhaps rank them. 2 Hopefully, if this is accepted, if these ideas are 3 accepted by the NRC, one day we will have a NUREG report 4 which describes this and we have already collected a large 5 number of examples that illustrate various aspects of'this, 6 and we are going to be developing examples, one of which I 7 will very briefly discuss with you. 8 Now, once you have conceptualized things in terms 9 of pictures and descriptions and charts, the next step would 10 be to put things into mathematical language. The first step 11 is not to develop a computer code, but rather to decide what 12 are the physical processes and their mathematical 13 expressions in the form of differential equations, partial 14 differential equations, perhaps representations, we have 15
) seen examples of those today, that one could use to describe 16 these models, these qualitative concepts, to put them into a 17 mathematical quantitative framework.
18 The next step would be to actually develop these, 19 implement these in a computational -- as a computational 20 tool. 21 Now, one of the things that I believe quite 22 strongly is that at least up to about the middle of this 23 box, things can and should be done in three dimensions. The 24 reason for that is that when we are conceptualizing geology, 25 it is virtually impossible to conceptualize geology in one
\ ANN RILEY & ASSOCIATES, LTD.
'('- Court Reporters 1025 Connecticut Avenue, NW, Su te 1014 Washington, D.C. 20036 (202) 842-0034
155 1 dimension. You can conceptualize information from a single 73 2. U bore hole in one dimension, but if I was to drill a well in 3 a horizontal direction, I would see something quite 4 different than what I would see in a well that is vertical, 5 immediately telling me that there is a three-dimensional 6 element to this geology, three dimensional. 7 I think that before we decide that we want to do 8 the calculations eventually in one or perhaps two 9 dimensions, it is good for us to step back and say, now, 10 wait a second, what gives us justification for doing that 11 and what is the proper way of simplifying what really is a 12 complex three-dimensional flow system, what is the best way 13 to simplify it for purposes of embedding it a dose 14 assessment code where computational limitatioas are such 15 that a fully three-dimensional model cannot easily be 16 accommodated. 17 I say easily because I think that many of our 18 notions about what can and cannot be embedded in those 19 assessment codes are somewhat outdated. They are based on 20 computational capabilities that existed ten, 20 years ago 21 perhaps. 22 Computers are developing to such an extent that I 23 can today run easily two-dimensional codes of great 24 complexity on my PC and so can almost _aybody, for $5,000 or 25 less. For $10,000, you can buy yourself a workstation and f ANN RILEY & ASSOCIATES, LTD. k Court Reporters 1025 Connecticut Avenue, NW, Suite 1014 Washington, D.C. 20036 (202) 842-0034
r l 156 1 do calculations in three dimensions. 2 t So I am not as convinced as some are that the 3 limitations to embedding more complex models in dose 4 assessment curves are real. I think that to some extent, 5 they are a relic of the past and perhaps a result of the 6 fact that our existing codes, which were developed years 7 ago, are not geared up for this. 8 Okay. Once a model has been put into a 9 mathematical computational framework and this can now be in 10 one dimension or in two dimensions or maybe three 11 dimensions, as a hydrogeologist, I would very much like it 12 to be compatible with data in two ways. I would like this 13 code to be able to accept real data which are compatible 14 with the complexity or simplicity of the model.
/}
(_/ 15 Most measurements in the field are relatively 16 small scale. They model things on a relatively large scale 17 compared to the scale of a single pumping test, not the 18 speed of the scale of single core samples. So we face this 19 very serious question of how do you take data collected on a 20 small scale and meaningfully build them into a model which 21 is much cruder, has a much lesser resolution. 22 This is a fundamental question that has not been 23 resolved. So as part of our work for the NRC, I have some 24 students doing fundamental work, asking themselves how to 25 resolve this issue. It may very well be that when our l { g ANN RILEY & ASSOCIATES, LTD. (g) Court Reporters ; 1025 Connecticut Avenue, NW, Suite 1014 ) Washington, D.C. 20036 (202) 842-0034
157 1 project is completed in a few years, 2 we will still not have a complete answer to that question, but at least we will 3 have some. 4 Sorry, you cannot hear me too well. Let me know 5 if you cannot hear me. Raise your hand, then I will know 6 you cannot hear me.
?
So one question is how to make sure that the data 8 that you're feeding to the model are meaningful and 9 compatible with the model and then comes the question of how 10 do you know how good your model predictions are. 11 Well, there is no way for us to confirm 12 predictions made over a period of 1,000 years. But it 13 should be possible to confirm some elementary aspects of a 14 model, things that perhaps could be verified either by () 15 16 observation or by deliberate experimentation. I feel very uncomfortable about models which make 17 predictions of things that cannot be observed or measured. 18 If I predict concentrations, I would like to be able to 19 predict them on a scale, spatial scale, and maybe temporal 20 scale, that I can actually compare with real measurements. 21 Water levels, I would like to be able to compare 22 with real measurements. Models that do not do that leave me 23 queasy. I do not feel confident. 24 No matter what we do and how much data we have and 25 how much understanding of a site we have, we will always end (' s ANN RILEY & ASSOCIATES, LTD. Court Reporters 1025 Connecticut Avenue, NW, Suite 1014 Washington, D.C. 20036 (202) 842-0034
158 1 up with residual uncertainty.
~
2 I mentioned in the beginning that there are at 3 least two important aspects of this uncertainty, one which 4 many of us know in many cases how to quantify and there is 5 uncertainty in parameters. If you have a large set of { 6 measured parameter values, you can draw yourself a 7 histogram, fit a DDF to it, and you can say that you know 8 something about the probability distribution or the $ l 9 statistics, and, therefore, the uncertainty associated with 10 that parameter. 11 You can then perhaps propagate it through a model 12 and know something about the uncertainty in the model 13 predictions. 14 How do you do it with conceptual models? When () 15 16 five different hydrogeologists look at the same data and come up with different interpretations, how does one ) 17 quantify it? This is one of the questions that the NRC 18 would like us to answer and we are thinking about it very 19 hard and I even got some very interesting ideas during a 20 project review over the last two days, but, frankly, I don't 21 think that we know very well how to do that. 22 So that's a challenge and hopefully we will, in 23 two years, be a little bit smarter than we are today. 24 And why do we need to quantify the uncertainty? 25 Because the NRC would like to know to what extent can it (~g ANN RILEY & ASSOCIATES, LTD. ( ,) Court Reporters 1025 Connecticut Avenue, NW, Suite 1014 Washington, D.C. 20036 (202) 842-0034 1 i L.
9 159 1 rely on performance measures that the model predicts. It 2 t would be very nice if those performance measures could be 3 given not just as numbers, but as numbers plus an associated 4 margin of error or confidence limit. 5 So that is the ultimate goal. 6 What I will do is very briefly now go through my 7 ovals for each one of these boxes and then, as I said, 8 illustrate some aspects of that by an example. 9 So the first rectangle is the definition of the 10 contextual framework, and I don't think that there is much 11 mystery as to what is meant by that. One has to know the 12 precise reason for the development of the model, what is the 13 nature of the problem, what is the range of potential 14 solutions that one would contemplate, what is the distance
/~N 15 U 16 over which a plume may travel from the source, making some reasonable estimates of site conditions, the timeframe, of 17 course, very often that will be 1,000 years or for 18 decommissioning, at least under the current rule.
19' So there are certain limits on the scope of the 20 problem and on aspects of the problem that one wants to look 21 into and certainly performance measures, the specific 22 performance measures that you want to evaluate would enter 23 right here. l 1 24 ! Once you know that, you have essentially defined 25 the scale of the hydrogeologic system that you want to study ' ANN RILEY & ASSOCIATES, LTD. O.. Court Reporters 1025 Connecticut Avenue, NW, Suite 1014 Washington, D.C. 20036 (202) 842-0034
160 1 and at this point, you may want to decide how deep you go, f- 2
-Do you go to that bedrock and embed it in your model? Do 3
you only look at the uppermost unconfined aquifer? Do you 4 only look at the vadose zone? How far sideways do you go in 5 each direction, just to your boundary, go two miles beyond 6 your boundary? These are the questions that would have to 7 be answered somewhere at this stage of the approach. 8 But, of course, as we have seen earlier, it would 9 depend on what the NRC wants you to do. So perhaps this 10 would be a place to interact with your project managers, as 11 we have heard earlier, and help you define the scope of the 12 problem, 13 The circumstances and the scenarios, are you going 14 to accept an on-site farmer, off-site framer, drinking water 15 scenario, where exactly will you want to calculate your 16 results. 17 Relevant hydrogeologic aspects, what is important; 18 is the nature of the layers in the system important? Do you 19 want to embed faults? Is it a fracture rock? To what 20 extent do you want to map out major fractures? Is that 21 something that's relevant? 22 Those are the kind of questions that would be 23 addressed here. 24 And I think the NRC should really tell you 25 something about the accuracy and the uncertainty that it is ANN RILEY & ASSOCIATES, LTD. / \/) Court Reporters 1025 Connecticut Avenue, NW, Suite 1014 Washington, D.C. 20036 (202) 842-0034 { j i
)
g- . 161 1 willing tc c a in your calculation. So here the onus is
/"N 2 U 3 going to be, ; think, on the agency to tell the modeler just exactly 'aow unportant is this model going to be.
4
' f tt's not so important, there is really no need I
5 to spend a lot of money, time and resources developing it 6 than otherwise there would be. 7 Now, the particular example that I'm going to 8 briefly discuss with you concerns an agricultural center, 9 called the Maricopa Center, about 20 miles, is it, southwest 10 of the City of Phoenix, in Arizona, a very heavily 11 agricultural area, which we think has some analogies to a 12 decommissioning site. 13 There is, of course, no radioactive material 14 on-site, but in terms of the hydrology, there are many "h [b 15 16 similarities to many decommissioning sites, We have some general questions that we would like 17 to address with respect to this site. I'm not going to go 18 through all of them. But one that is relevant to a previous 19 discussion is how well can the site be characterized based 20 entirely on publicly available information. I will show you 21 in a moment what kind of information we've been able co 22 obtai... 23 The next question would be what alternative models I 24 can one validly postulate for flow and transport based on 25 such public information. What ambiguities and uncertainties rmg ANN RILEY & ASSOCIATES, LTD. Q Court Reporters 1025 Connecticut Avenue, NW, Suite 1014 Washington, D.C. 20036 (202) 842-0034
162 1 are associated with these alternatives, and I will point out 2 to one or two that kind of jump at you right at the 3 beginning. They are at the very beginning of this exercise. 4 How can such uncertainties be quantified? I'll 5 indicate one way of doing that. What tools can be employed 6 to do so? What tools can one employ to explore alternative 7 models and their uncertainties? What are the plausible 8 modes and rates of flow and transport in the vadose zone? 9 What happens to a plume when it reaches the water table? 10 How to verify with limited resources, how to reduce 11 uncertainty, and how to simplify the flow and transport 12 representation in realistic and hopefully conservative 13 manner. 14 I'm not going to answer all these questions, but
/ 15 these are the questions that we are hoping to address.
16 Of course, in order to address them, the first 17 thing is to define what is it that we know, what is it that 18 we don't know. So define available knowledge base. The 19 first thing is to assemble data and the first thing we'll be 20 looking at are the publicly available data. 21 Then assemble and acquire the knowledge base, 22 meaning what do you know about what could happen with flow 23 and transport given those data. Then we'll ask ourselves 24 the question could we improve the available knowledge base 25 and its interpretation if we were given site-specific data. f-ANN RILEY & ASSOCIATES, LTD. Court Reporters 1025 Connecticut Avenue, NW, Suite 1014 Washington, D.C. 20036 (202) 842-0034
163 1 Now, site-specific data exists. Pr9fessor Peter q 2 Wilrenga, l Q 3 who is sitting right here, has conducteri very l detailed studies at the site that I'm going to discuss. 4 We're playing a game. We are saying we know 5 nothing about his results. I have a student working on this 6 and he has not read any of the NUREG material, has not heard 7 anything about what Peter has found at the site. 8 In. addition to that, we are planning this NRC l 9 support, some field experiments at the site. So the next 10 step is going to be to bring this site-specific information 11 into the game and see by how much does it change our concept 12 and our model and the level of uncertainty with which we can 13 work. 14 The particular sources of information that we were 15 able to very easily identify, I'll list it for you here, 16 even with the web sites, if somebody wants to repeat the 17 exercise with us. There are various reference maps, some of 18 which I will be showing you, which are easily obtained from 19 the University of Arizona, a land grant university, Tom 20 mentioned land grant universities would be a good source for 21 such information. 22 And some of it is on-line, some of it is not. You 23 need to go to the archives and look for it. 24 Meteorological data, there is a local network 25 called the AZMET that provides a wealth of meteorological p ANN RILEY & ASSOCIATES, LTD. Court Reporters i 1025 Connecticut Avenue, NW, Suite 1014 I Washington, D.C. 20036 i (202) 842-0034 !
)
164 1 information, you've heard quite a bit about that, I will not 2 go into the details. 3 It's a heavily irrigated area. This is an arid 4 region. There is information about irrigation, both from 5 the MAC, Maricopa Agricultural Center Administration, and 6 from the Arizona Department of Water Resources, on a monthly 7 and annual basis, some of that on-line. 8 Irrigated land distribution and soil type 9 distribution, well inventory and ground water resources. 10 Much of the well inventory is on maps and on-line, much of 11 the well information is not. It has to be collected, one ( 12 has to go well log by well log and collect the information. 13 This is one of the laborious things that my student had to 14 do. () 15 16 Sources, Department of Water Resources and the US Geological Survey. And pumping rates from wells, because 17 this is an irrigated area using ground water, so there are 18 active wells in the area, some of that is available on 19 CD-ROM and some of it in terms of records; not published, 20 but archived records that one can get access to. 21 So an example. This little red rectangle here is 22 the area of interest to us and you can see that because of 23 the heavily irrigated site, there is a large number of wells i l I 24 in this area. Now, not everywhere will you find so many p 25 wells, but I believe that almost everywhere you will find 1 ANN RILEY & ASSOCIATES, LTD.
-( Court Reporters 1025 Connecticut Avenue, NW, Suite 1014 Washington, D.C. 20036 (202) 842-0034 b
165 2 some wells. So almost everywhere will you be able to do
, 2 (7 - some and plot some of the types of maps that we are plotting 3 here.
4 The green are irrigated areas, the yellow are 5 unirrigated areas. 6 Soil distribution, we have no site-specific 7 information at this stage about soil properties that we need 8 for flow modeling and transport modeling. So the idea is to 9 take soil textural information off such maps and then 10 supplement it with distributions of parameters associated 11 with these textures out of the NUREG, 12 which I will hopefully have time to show you again, but you already saw some 13 figures out of it that PNNL, Phil Meyer and Glendon Gee here 14 have developed. 15 A Well log information gives us the third dimension . 16 There is, of course, other maps that one could draw. All of 17 this information, by the way, is on a GIS system. We have 18 put it on a GIS system. You don't have to. It's very 19 helpful if you have a GIS system to help you organize your 20 material. 21 Now we go into the third dimension. So here is -- 22 I'm just going to show you one or two of these well logs. 23 It is important to understand, and those of you who have sat 24 on wells or are geologists or are familiar with the work of 25 geologists will immediately recognize that what you see here ANN RILEY & ASSOCIATES, LTD. ( Court Reporters 1025 Connecticut Avenue, NW, Suite 1014 Washington, D.C. 20036 (202) 842-0034
1 166 is already the result of an interpretation by the geologist 2 7- or the well sitter or the well logger. He decided that kl 3 everything here can be lumped into a green area called clay, 4 sand and sandy clay. If you actualij looked at the 5 cuttings, you would find quite a range of soil types, even 6 within each one of these categories here. 7 This uncertainty we have no information about, 8 because we do 1.ot have the original data and probably we 9 wouldn't want to spend the time looking through them and 10 pouring tnrough them in the first place. 11 So we accept'this at face value and when you go 12 just a short distance away, to another well, you'll see 13 this, at a short distance away to yet another well, you see 14 this, and that immediately implies that things are changing
< 15
( in both the vertical and horizontal direction. Therefore, 16 1 it is virtually impossible, based on this information, for ' l 17 us to conceptualize this system, certainly not in one 18 dimension, but not even in two dimensions- properly. It's a 19 three-dimensional system. 20 And all we can do, of course, is subdivide it into 21 laye>;s or try to subdivide it into layers based on this kind 22 of information. 23 ! So then the next step is conceptualization. We i 24 ! have collected -- I will show you part of the information we 25 have collected. ANN RILEY & ASSOCIATES, LTD. ! ()/ s- Court Reporters 1025 Connecticut Avenue, NW, Suite 1014 i Washington, D.C. 20036 i (202) 842-0034
r 167 1 Now, what about conceptualization? And hopefully 2 I have convinced those among you who might not have thought l \' ' 3 that this is necessary, hopefully I've convinced you that 4 this conceptualization of the hydrogeology, the geology and 5 hydro aspect of it should be done and can be done three 6 dimensions, even if there are many fewer wells than what we 7 have here. 8 I have not yet seen a place where there isn't 9 enough information to do some of this in three dimensions, 10 certainly on a regional scale. 11 So you really face the task of conceptualizing 12 this, deciding what is an aquifer, what is an accliclude, 13 how many of those will yc.u have and how are you going to 14 model their boundaries between them, and once you have that, 15 to decide which way are you modeling -- going to model the 16 flow. Is it occurring primarily in the vertical direction, 17 horizontal direction, what? I 18 Well, in this three-dimensional system, it is 19 sometimes very difficult to make that decision. 20 Once you have conceptualized the flow, we'll 21 conceptualize the transport. Then there are many other 22 aspects one can take into account, such as isotopes, 23 geochemistry, if you have temperature data, that can be very 24 helpful. In most cases, that will not be available and 25 people will simply ignore this information, which is fine as
- O ANN RILEY & ASSOCIATES, LTD.
Court Reporters 1025 Connecticut Avenue, NW, Suite 1014 Washington, D.C. 20036 (202) 842-0034
168 1 long as everything else helps you develop the model that you (s \ 2 need. 3 Then comes this question of ambiguities and I will 4 be pointing to one or two of those as we go along with the 5 example. I have, what, 15 minutes, I think that should do 6 it. 7 Identify some uncertainties associated with this 8 and hopefully postulate some alternative concepts and 9 hypotheses and then compare and eliminate -- I will not go 10 through this entire process, but I'll just kind of allude to 11 it. 12 So the first thing you do is to try to develop, i 13 based on this information, first, a two-dimensional picture 14 along various directions. So you draw a cross-section. And 15 here we use a piece of software, commercially-available 16 v software, which costs about $1,000 or so, called Rockware, 17 some of you perhaps may be familiar with. It does the work 18 for you. You don't have to be a geologist. It's better if 3 19 you understand what you're doing, but this program could de i 20 this for you. 21 You give it the well logs and their locations the 22 way you read them off the public record and then you can 23 draw your north-south cross-section. We have drawn these 24 cross-sections through wells which are in the vicinity of 25 the area of interest, for the sake of time, and I will not g 1 ANN RILEY & ASSOCIATES, LTD.
# Court Reporters 1025 Connecticut Avenue, NW, Suite 1014 Washington, D.C. 20036 (202) 842-0034
l 169 1 go back_and show you just exactly where they are. 2 Where the program doesn't know how to connect 3 these, it wil:1 leave it blank. So you will know that there 4 is an open question. 5 And if you look at a west-east cross-section, it 6 will look quite different. The water table is somewhere 7 within this green unit. That's the regional water table. 8 And the program can also draw for you a fence 9 diagram whicn provides -- it's a little difficult to read 10 here, but it provides a three-dimensional picture, and you 11 can rotate it and look at it from various directions, so 12 that you have some idea of what the subsurface looks like. 13 Now, there is very little detail internally within 14 these layers and very often the most important part is in 15 the upper soil layers. Public information of this kind only (} 16 provides you with the soil map, which I haven't used in 17 drawing this. This is the deeper geology. So you still 18 need to supplement this with the soil information. 19 Based on this, you can draw maps, iso-thickness 20 maps for units that seem to be acting as potential aquifers 21 or potential confining units between the aquifers. And it's 22 possible to draw contour maps of the regional water table. 23 Now, the depths of these -- these are actually not 24 depths but actual elevations. This can also draw maps of 25 depths. I didn't bring those maps. This is a 1988 water es ANN RILEY & ASSOCIATES, LTD. Court Reporters 1025 Connecticut Avenue, NW, Suite 1014 Washington, D.C. 20036 (202) 842-0034 [.
170 1 level contour map. I'm sure that all of you know that f"' 2 ground water in an isotropic system flows perpendicular to 3 lines of equal head. So if you draw arrows perpendicular to 4 thene contours, you will see that this is not a simple 5 one-dimensional horizontal flow regime, but it's a rather G complex, in this two-dimensional sense, flow regime, and if 7 I was to draw it at various elevations, and I don't have 8 that publicly available, it would look different in 9 different units. 10 That would mean that I have a three-dimensional 11 system. In some places the water is flowing down, in other 12 places it's flowing up. It's a complex three-dimensional 13 system. 14 So if you want to model it in 2-D, as we are about 15 to do, you will have to decide how you draw your 16 two-dimensional vertical section and the only right way to 17 do that is to draw it perpendicular to this contours. 18 So I will show you in a moment such a section 19 perpendicular to the contours in the vicinity of the 20 triangle, which is the area of interest. 21 The problem, of course, is that this is a 1980 22 map, '88 map, and if you compare it with the 1993 map, 23 things will start looking a little bit different, so you'll 24 have to decide which of these maps. There was quite a t 25 change in water levels in those few years. ANN RILEY & ASSOCIATES, LTD. Court Reporters 1025 Connecticut Avenue, NW, Suite 1014 Washington, D.C. 20036 (202) 842-0034
E 171 1 When are we going to draw this contour? Well, it 2 ( turns out it doesn't matter too much for our site, but if 3 l you wanted something on a slightly bigger scale, it would 4 matter. So transients in this case probably would be of 5 interest. 6 All right. We are ready to start doing something 7 that goes beyond just a concept of the hydrogeology. Of 8 course, one could say much more about it than I did. But 9 the next thing is to try to do something mathematically. 10 The easiest thing to do here is to find a piece of 11 software that's available commercially or publicly to do 12 this and we have used a piece of software called Hydros, 13 that allowa one to study in a vertical two-dimensional 14 section flow in the saturated and unsaturated zone. () 15 We have to decide how we are dividing the system 16 into layers. So we focus, just for the sake of this 17 exercise, on the uppermost two layers in those 18 cross-sections you have seen, in the close vicinity of the 19 site, and we are going to approximate the boundaries between 20 them by horizontal lines, starting to approximate things, of 21 course. 22 In order to input soil parameters, we have used 23 the PNNL NUREG, from which we have drawn PDFs of 24 _ distributions of permeability, porosity, and unsaturated 25 flow properties. There was no information available about O ANN RILEY & ASSOCIATES, LTD. Court Reporters 1025 Connecticut Avenue, NW, Suite 1014 Washington, D.C. 20036 (202) 842-0034
1 i 172 1 how those vary in space, no information about spatial f ( 2 variabilities. So we are going to make some assumptions 3 about that, just for the sake of illustration.
-4 Here is a grid, which I will refer to as a course 5 grid, for the analysis in a vertical two-dimensional 6
section, two layers, water table somewhere here, it's not 7 shown here, and this is a five-by-three meter square grid, 8 meaning these are nodes of a computational grid and each one 9 of these little squares, really a rectangle, is a 10 five-by-three meter one. 11 Then I will show you results with a finer grid, if 12 I have time, and see what happens. 13 So spatial variability. We assumed some 14 distribution -- some correlation, spatial correlation 15 between the data and used a standard geostatistical method, 16 which is actually, in this case, available in the Hydros 17 code. You don't have to go to a geostatistical code to do 18 this. The Hydros code does it for you. 19 You provide it with a PDF, you provide it with a 20 vertical and horizontal correlation scale, and in this case, 21 this code can only do it for one layer. So in order to do 22 this, we had to lump those two layers together. That's a 23 code limitation. The codes can do it layer by layer. 24 So this provides us with a so-called scaling 25 factor for hydraulic conductivity. The actual hydraulic p Q ANN RILEY & ASSOCIATES, LTD. Court Reporters 1025 Connecticut Avenue, NW , Suite 1014 Washington, D.C. 20036 (202) 842-0034 l
f~ . 1 l 173 1 consuctivity, saturated hydraulic conductivity would be this 2 factor, f I don't have a scale here, but it's quite a 3 variability, multiplying the average value. 4 So this is a heterogeneous system, there is a 5 certain pattern of heterogeneity, and we can vary that and 6 see what a difference it would make with respect te our 7 predictions. A very simple game to play. ) 8 And since each one of these patterns is random, we 9 can generate randomly as many as we want. If we generate 10 many, many, that would amount to a Monte C'arlo simulation or 11 many realizations of this. 1 12 So I'll be talking about realization one, 13 realization two. On the coarse grid, it turns out that it 14 does matter whether you use a fine grid or a coarse grid. 15 When you use a two-by-one meter grid, such a this, which is 16 easily done, the patterns are going to be somewhat 17 different. You get a much finer resolution of your 18 heterogeneities. 19 And depending on how you build them into this, 20 your plume may eventually :Ludicate a response or be I 21 sensitive to your choice of grid and year assumptions about 22 heterogeneity. 23 Now, I'm going to show you simulated with this 24 t code, this code does some strange things, such as drawing l i 25 this contours, simulated movement of a plume of a O ANN RILEY & ASSOCIATES, LTD. O Court Reporters 1025 Connecticut Avenue, NW, Suite 1014 Washington, D.C. 20036 (202) 842-0034
174 1 contaminant which is maintained at constant concentration /^T~ 2 right up here. U 3 I will be showing you two stages of contamination 4 or plume evolution, the 200 days and at 7,000 days. Now, 5 let me first mention that when we took the publicly 6 available record of precipitation, irrigation, there is no 7 runoff at the site whatsoever recorded anywhere, and the 8 code has an ability to calculate evapotranspiration for you. 9 It has a plant uptake model built into it. l 10 So we used it and we got no where. There was no 11 plume movement at all. l
.12 In fact, the publicly available data with this 13 feature of the code predicted flow out of the system, no 14 infiltration at all.
/~' 15 V) So we said, well, let's just suppress the i 16 evapotranspiration, and what you see here is the result of 17 suppressing evapotranspiration. So this is the result of 18 irrigation, which is quite a high rate, and precipitation. 19 In the meantime, I talked to Peter, who, of 20 course, knows the site and he is telling me that he knows 21 that, in fact, there is infiltration at the site. He's 22 talking about an irrigation efficiency of between 50 to 80 23 percent, which means that I should perhaps not completely 24 suppress the transpiration, but suppress it down to 50 25 percent or less, but nevertheless, it immediately shows you
*g .' ANN RILEY & ASSOCIATES, LTD.
Court Reporters 1025 Connecticut Avenue, NW, Suite 1014 Washington, D.C. 20036 (202) 842-0034
175 1 that the publicly available information is very misleading, 2 V(N at least if you use it in the way we have, which is not
'3 uncommon.
4 We heard about this water balance approach today 5 quite a bit. That's exactly what we did, with some 6-modification based on the features of the model. 7 So what you will see are some slight nuances as 8 you go from one type of assumption to another. That was 9 realization one, with a coarse grid, realization two, with a 10 coarse grid. 11 Just one realization is a fine grid and the case 12 of homogeneous two layer. The plume is still moving in the 13 vadose zone. Not very important differences so far. 14 Interestingly, this code suggests that the plume () 15 splits when you use a fine grid. That's clearly a numerical 16 artifact of sorts. That's in the cr~e of no heterogeneity 17 in the system. 18 Now, here is what the two-dimensional model 19 suggests happens when the plume reaches the water table, it 20 refracts, because there is a dispersion built into the 21 model. It refracts in a rather smooth way. So, again, you 22 start getting these differences. 23 And here the differences will be a little bit 24 bigger between one realization and the other, not as 25 degreed, has reached below the water table and has moved in /" ' ANN RILEY & ASSOCIATES, LTD. Court Reporters 1025 Connecticut Avenue, NW, Suite 1014 Washington, D.C. 20036 (202) 842-0034
176 1 7,000 days. So the distance -- this, by the way, should be 2 7,000, not 9,000 here. Some differences between two 3 realizations. 4 These could be much bigger if we took the local 5 internal heterogeneity to hr larger correlations, in which 6 case we would literally build lances into the system, 7 something we haven't done yet. 8 Here is the fine grid result. The green is up 9 here. In the coarse grid case, at least in two realization, 10 it flows down here already. So we are starting to see some 11 meaningful differences between these various results, and, 12 of course, the proper thing to do would be to run this many 13 times, all plaucible realizations, all plausible assumptions 14 about the infiltration rates and so on, and then come up () 15 16 with a range, potential values that you could interpret in
-terms of uncertainty.
17 I will not bore you with these pictures and this 18 :is my last overhead. It's going to be a little bit 19 difficult to see, but it kind of summarizes where or how we 20 envision formally now the last part of our methodology, 21 building uncertainty analysis into the modeling process. 22 So the first thing that we would like to account for are 23 scales and the first box here is a question; are the scales 24 of the model -- are the scales of model resolution and the 25 scales of parametricization, how you divide your parameters, r'~g ANN RILEY & ASSOCIATES, LTD. tQ Court Reporters 1025 Connecticut Avenue, NW, Suite 1014 Washington, D.C. 20036 (202) 842-0034
177 1 consistent with the support or measurement scales of the 2 data. 3 If they are, well, then, you can continue and go 4 dire ,y to this box. But if they are not, as is usually 5 the case, you have to grapple with the question of how do j 6 you make them compatible. i 7 I said this question of how to do this is not ' 8 fully resolved in hydrogeology. So you can skirt it or you j 9 can try to do certain things and there is more and more 10 coming out in the literature about how to do it, I think, as 11 I said, we will know more about it in two years. 12 You can either rescale the model to fit your data, 13 you can rescale the data to fit your model. That's the 14 easiest thing to do. There's a little bit more methodology 15 available out there. 16 Then you ask yourself can you treat parameter 17 variability and uncertainty as random fields, and we just 18 had, in the example, where the permeability has varied in a 19 random matter, as a random field within this vertical 20 section. If so, then you can conduct Monte Carlo 21 simulations, essentially the same thing I've shown you. 22 I've shown you two realizations per grid, you can 23 do it 100 times, and then statistics based on that, 24 calculate a range of possible outcomes, and, based on that, 25 a range of possible performance measures and perhaps even p ANN RILEY & ASSOCIATES, LTD. () Court Reporters 1025 Connecticut Avenue, NW, Suite 1014 Washington, D.C. 20036 l (202) 842-0034 { i
i 1 178 try to summarize that .in the form of suitable statistics . l 2 { ?[ . That kind of summarizes it, in a nutshell. l i 3 MR. NICHOLSON: l Are there any questions for Dr. i 1 i 4 Neuman? Boby?
- 5. MR. EID:
Dr. Neuman, I think this is a very good 6 presentation, 7 with lots of illustration to convince about using complex modeling. 8 But there is a question remaining to be answered 9 , about how much characterization that -- characterization \ 10 data needed to support complex modeling and, of course, 11 people may be concerned about the cost associated with 12 collection of such characterization data, because the cost 13 { is not just the models itself, because they are available 14 for free, the cost is the characterization data that is 15 needed. 16 DR. NEUMAN: I'm very happy that you asked this 17 question, because there was another point that I forgot to 18 mention. 19 i All the modeling that you have seen here used 20 publicly available data. There was no site characterization 21 whatsoever. I did mention one piece of information that I 22 received by interviewing a person who knows the site and has 23 characterized it, Professor Wilrenga, and out of that, I 24 learned two things; I mentioned one of them. 25 One of them was that my mass balance was t, ANN RILEY & ASSOCIATES, LTD. Court Reporters 1025 Connecticut Avenue, NW, Suite 1014 Washington, D.C. 20036 ) (202) 842-0034
179 1 completely off the mark, based on the publicly available 2 information data. Clearly, the mass balance approach, the 3 way we have used it anyway, which is not uncommon, simply 4 has not worked. He is telling me that there definitely is 5 deep percolation at the site, at the rate of anywhere from 6 50 to 20 percent -- or 20 to 50 percent, say, of the 7 irrigation rate, which is quite high. 8 By the way, I think it's much higher than 18 9 centimeters at that particular location that's irrigated. 10 The reason I'm mentioning 18 centimeters a year, 11 because that is what the D&D code has generically built into 12 it. 13 The other piece of information that I received 14 from him, which my generic data did not tell me anything 15 about, is that at the site, the water table is not as deep 16 by far as what my regional data told me, but that there is a 17 perched water table only about 20 or 30 feet below the site, 18 and that's where all the water accumulates. 19 Now, we have no idea of what happens to the water 20 below that perched zone, but locally, there is the 21 infiltration rate and the location of the first water table 22 that one encounters is quite different from what the 23 publicly available data told me. 24 My conclusions from that is that publicly 2E available data, even from such a dense network of wealth as ANN RILEY & ASSOCIATES, LTD. 0% Court Reporters 1025 Connecticut Avenue, NW, Suite 1014 Washinc; ton, D.C. 20036 (202) 842-0034
1 180 we have here, is only good for very crude approximation of 2 the site on a large scale and one should be very careful of (A) 3 applying that on a much smaller scale, such as the Maricopa 4 site, which is, say, several hundred meters in terms of 5 kilometer scale. 6 So I personally, from this exercise and based on 7 { my other experience, would definitely support Professor 8 Wilrenga's suggestion that you always collect some 9 site-specific data. I would go a little bit further, I 10 alluded to before, during the discussion, I just don't 11 believe that it's possible to know anything about whether or 12 13 not there is subsurface contamination in the vadose zone and of the ground water, without actually having done some site i 4 14 characterization.
-~ 15 So when you approach the NRC with a statement, 16 yes, we do or no we don't have ground water contamination at 17 l the site, you will already know something about the site .
18 Therefore, in all cases, I'd certainly be interested in 19 1 hearing otherwise, in all cases, you will have site-specific 1 20 data and I think you need to use it. 21 MR. NICHOLSON: Paul Chenoweth, NEI. 22 MR. CHENOWETH: Yes, hi. Actually, the comment or 23 the question comes from your presentation, but probably is 24 directed to the staff. 25 Early on, in showing the publicly available data, i l N ANN RILEY & ASSOCIITES, LTD.
) Court Reportars 1025 Connecticut Avenue, NW, Suite 1014 i Washington, D.C. 20036 (202) 842-0034
7 181 1 l 1 showing where the wells were and the irrigated land, it made
) 2
it very clear that the entire area surrounding your site was j 3 irrigated and there were very few areas around there that 4 weren't.
! 5 Is that sufficient justification to eliminate the 6 irrigation pathways for the subsistence farmer? I mean, I'm 7 trying to get to real practical areas here and I guess I'd 8 like a response. 9 MR. THAGGARD: Yes. Let me make sure I 10 understand. You're saying is that justification to 11 eliminate the irrigation pathway? 12 MR. CHENOWETH: Yes. { l' 13 MR. THAGGARD: I think in this case, it would 14 definitely be justification not to eliminate the pathway. I 15 mean,
} most of the water use in that area is irrigation. So 16 I don't think you will have any justification for 17 eliminating it.
I mean, simply because it's not at your 18 particular site, it's clearly being used in the area there. 19 MR. CHENOWETH: I'm sorry. I lived in Phoenix for 20 a while. The irrigation water came from remote locations. 21 It didn't come from your ground. It was brought in from the 22 Arizona River or the Salt River Project. So the water 23 wouldn't be coming up from your contaminated area. 24 MR. THAGGARD: That's right. But there are 25 pumping wells at the site. f-sg ANN RILEY & ASSOCIATES, LTD. g j Court Reporters 1025 Connecticut Avenue, NW, Suite 1014 Washington, D.C. 20036 (202) 842-0034
l 182 1 MR. EID: Additional comments. I guess for this r^g 2 case, b it is very clear that you could bring water and use it 3 and for irrigation and the land could be used for 4 irrigation. So for that specific case, we have just closed 5 that, adjacent to that not irrigated area, an area which is 6 irrigated and is used for farming activities. 7 So this is support that, yes, there is some use of 8 the farming activities. If that specific area was remote, 9 it is very difficult to acquire water and to develop the 10 soil and make it suitable for agriculture and farming, yes, 11 it could be an argument to eliminate the ground water 12 pathway. 13 MR. THAGGARD: Let me just see if I can answer 14 your question in a generic sense. I think what you're 15
} getting back to is the issue that was raised earlier about 16 the use of surrounding land use information as justification 17 for eliminating a particular pathway, and that certainly is 18 {
something that we're looking at. 19 I don't -- I mean, I can't give you an answer on ' l 20 that today, but that is the kind of thing that we're looking ! 21 and that is the kind of thing that we're going to ultimately 22 put into the standard review plan. 23 I don't know if that answers your question. 24 MR. CHENOWETH: Yes, it does. I guess to be 25 clear, what I was saying is perhaps this wasn't the best ' i ANN RILEY & ASSOCIATES, LTD. (' Court Reporters 1025 Connecticut Avenue, NW, Suite 1014 l Washington, D.C. 20036 ) (202) 842-0034
183 1 example, but for instance, there are areas where all the /*N 2 irrigation water comes from a remote source. So is that 3 sufficient justification or am I following the right 4 approach to follow the logic to say, all right, since the 5 water won't be pulled from the subsurface to irrigate this 6 particular land, that pathway isn't important, other 7 pathways are, I'll calculate those. 8 I'm just trying -- you know, obviously, I'm not 9 looking for a generic answer today, but I'm trying to say 10 these are the thoughts that licensees are going to be going 11 through to try to justify what their specific situation 12 really looks like. 13 MR. EID: I believe that the source in this case 14 is brought from non-contaminated source, so there is no 15 reason to include it. This I am telling you based on a 16 generic example. 17 Now, if the aquifer is good under the source and I 18 the technology allows pumping the water, although you have a 19 deep aquifer, this is something to think about, of course. 20 MR. PARROTT: I'm Jack Parrott, I'm with the NMSS 21 licensing staff. I think one thing we've got to keep in 22 mind is if we're doing these analysis for 1,000 years, you 23 can't look at just what the source of water is today, but 24 you have to ask yourself is the -- is there a source of i 25 potable water at your site and look at that as a potential i O ANN RILEY & ASSOCIATES, LTD. Court Reporters 1025 Connecticut Avenue, NW, Suite 1014 l Washington, D.C. 20036 j
, (202) 842-0034 q l ]
184 1-source; maybe not now, but sometime in the future. 2 MR. NICHOLSON: Shlomo? 3 DR. NEUMAN: In the framework of a 1,000 years, it 4 may be interesting to note that this external source did not 5 exist before the 1920s. In other words, the dams at the t
'6 l site were built in the early 20th Century, toward, say, the 7 first part of the 20th Century.
8 So the picture has changed completely. Imagine 9 somebody went through this discussion even a hundred years 10 ago. 11 MR. NICHOLSON: Paul, did you have a follow-up i 12 question or is it a different question? 13 MR. CHENOWETH: Very brief. l Just the -- I guess 14 the point is are we supposed to anticipate -- I thought we 15 were supposed to use current land use activities.for our 16 planning purposes. So if currently everyone uses irrigation 17 or whatever from a remote source, are we supposed to still 18 look that they might drill down an extra 1,000 feet to find 19 that potable source or do we go with what people are doing 20 today? 21 MR. THAGGARD: I think the answer to that 2
'2 question, Paul, is that there are a lot of issues associated 23 with the use of current land use conditions and not only in 24 terms of the timeframe that you should apply to information, -25 I but as somebody pointed out earlier, the area that you i ANN RILEY & ASSOCIATES, LTD.
O Court Reporters 1025 Connecticut Avenue, NW, Suita 1014 Washington, D.C. 20036 (202) 842-0034
F 185 1 should look at, how far out should you go, so there are a r- 2 ( number of issues like that th<.: need to be resolved and we 3 haven't resolved those yet. 4 But we're certainly thinking about it, we're 5 working on it, but we don't have an answer for you today. 6 MR. NICHOLSON: One more comment. 7 MR. EID: Just to say that there are two extremes. 8 One extreme, we could eliminate completely the ground water 9 pathway and the agricultural pathway completely, and then 10 you say there is no ground water irrigation, there is no 11 agricultural activity. 12 The other way, in between, you could say the ground water 13 pathway -- the ground water is brought from somewhere else, 14 is not contaminated, however, you will have some
/ 15 V} agricultural activity, so it will be just only leaching from 16 the soil to the plant.
So there is the ground water in this 17 case could be eliminated, too. 18 Then the other, of course, possibility, where the 19 aquifer is good, you could use it and then you have -- you 20 pump water from there and then you irrigate, this means you 21 have these two conditions together. 22 So there are different analyses the staff could 23 make to convince the assumption whether the ground water 24 pathway should be included or not. Fortunately, these they 25 do allow, specifically more advanced codes or models, they f~ ANN RILEY & ASSOCIATES, LTD. (v) Court Reporters 1025 Connecticut Avenue, NW, Suite 1014 Washington, D.C. 20036 (202) 842-0034
186 1 do allow elimination of these pathways. (g 2 MR. NICHOLSON: Sam Nalluswami, from NRC's 'd 3 licensing office. Sam? 4 MR. NALLUSWAMI: My name is Sam Nalluswami. I am 5 from the NRC, the decommissioning group. 6 I really appreciate Dr. Neuman's presentation, especially 7 the logs that you presented side by side. That shows how 8 the geology can vary from point to point to illustrate the 9 geological characteristics of the site. 10 In addition to the water resources departments, 11 where you can obtain the well log data, I have used, for 12 example, from highway departments, the state and county 13 highway departments, their logs. They are very useful in 14 getting the geological characteristics of the site. 15 f~T J For example, when the geotechnical investigation 16 is going on, when they do the logs, they always know that 17 when they encounter the water table, they are supposed to 18 note down the water table level. So that is another 19 additional information that would be very useful in getting 20 the water table information. 21 So really that's a very useful tool. But I have a 22 question to Dr. Neuman. From the illustration that we saw, 23 your three-dimensional model of ground water flow model is a l 24 very good way to go in evaluating ground water flow. 25 Also, we have to evaluate the contamination 73 ANN RILEY & ASSOCIATES, LTD. () Court Reporters 1025 Connecticut Avenue, NW, Suite 1014 Washington, D.C. 20036 (202) 842-0034 e
r 187 1 1 transport characteristics and what will be your I ' 2 recommendation with regard to the flow modeling, for 3 example? Would you recommend a deterministic model or a 4 statistical model, both for flow and contaminant transport? 5 DR. NEUMAN: As a hydrogeologist, I have never in my life 6 encountered a situation where the geology is not 7 heterogeneous or virtually every scale of interest. B Wherever you look, whether you look at a scale of 9 millimeters or centimeters or kilometers, you always find 10 variability. 11 On the larger scale, such as the one that I was 4 12 showing in those cross-sections, you can draw boundaries I i 13 between units that appear to be homogeneous and can be 14 called aquifers, but internally they will still be () 15 16 heterogeneous and that is something that I showed could at least to some extent be accounted for through spatial 17 variability models, geostatistical techniques that are well 18 established. 19 And one has to know, of course, some expertise is 20 required to know that, but there are well established 21 techniques. So it can be done. They are not cheap to 22 apply. You need the expertise, you need the time, you need 23 the tools, but it can be done. 24 Because the only information that's available to 25 us about the subsurface is from wells and I appreciate your l O ANN RILEY & ASSOCIATES, LTD. Court Reporters 1025 Connecticut Avenue, NW, Suite 1014 Washington, D.C. 20036 (202) 842-0034
188 1 suggestion about the building boring logs, which I think is
/'T 2 b going to be very useful for shallow information, and from 3
soil data, there is a lot of information missing about this 4 vast volume of three-dimensional material down there. j 5 There is no way that I can see that you can assume 6 deterministically that you know what it is. So, therefore, 7 virtually everything that we do is statistical and 8 stochastic and a simple way to do that is just by 9 repetitions of the kind that I was showing before, the Monte 10 Carlo approach. 11 There are more sophisticated ways to do that, 12 which, in fact, were on my last diagram, but I didn't talk 13 about them, and those are in development. But right now, 14 the Monte Carlo approach. And you don't necessarily need to () 15 16 run thousands of these. Very illuminating to run two or three, just to see the range of possibilities. 17 MR. NALLUSWAMI: Last question. The last question 18 is, for example, the RESRAD model in the RESRAD code, it is 19 very simplified, very simplistic approach. How would you 20 approach combining a three-dimensional flow model and a 21 transport model to get the results for the ground water 22 flow? Combined in the RESRAD code, how do you approach 23 that? 24 DR. NEUMAN: There are two possibilities. Either 25 you have the computational capability to embed it within a gg ANN RILEY & ASSOCIATES, LTD. g) Court Reporters 1025 Connecticut Avenue, NW, Suite 1014 Washington, D.C. 20036 (202) 842-0034 I
189 1 dose assessment code, which cannot be RESRAD, it would have 2 to be a newly developed code, unless RESRAD is modular, in 3 which case it would be -- a modular code should really be 4 your object-oriented code, such as SEDSS, that Dr. Ralph 5 Cady and his contractors are developing at Sandia. That's 6 one way to do it. 7 The other way to do it is rather than using an 8 over-simplified ground water code in which you vary an 9 average permeability a thousand times and give the 10 impression that you have taken uncertainty into account, I 11 would much prefer to run a more complex ground water code, 12 externally, not a thousand times, ten times, and look at 13 that variability, make some simple calculations as to what 14 the variability could be, go beyond that. 15 [v} If you can run it a thousand times, that's good, 16 but I realize that's time-consuming. In other words, I 17 think that a more detailed description of the ground water 18 system, especially for transport, run a few times, so you 19 don't cover all possible ranges, and the results embedded in 20 a dose assessment code, to me, that is much more meaningful, 21 as a hydrogeologist, than an over-simplified code that you 22 run ten thousand times. 23 MR. NALLUSWAMI: I completely agree with you on 24 that. 25 MR. NICHOLSON: Thanks, Sam. Last question before g^)x (_ ANN RILEY & ASSOCIATES, LTD. Court Reporters 1025 Connecticut Avenue, NW, Suite 1014 Washington, D.C. 20036 (202) 842-0034
190 1 the break. Would you take the microphone? 2 b k/ MS. PARR: Elizabeth Parr, Colorado. I think they 3 covered it quite well. I just wanted to make sure those 4 points were made. 5 MR. NICHOLSON: Okay. Thank you. I really 6 appreciate the audience's patience. We've gone a little 7 beyond the break point, but I appreciate the discussion. 8 We're going to take a break now for 15 minutes. 9 So if we could be back in here at about 3:32, we'll start up 10 again. Thank you. 11 (Recess.) 12 MR. NICHOLSON: Before we get started, I want to 13 just make a couple of brief comments. This afternoon, we're 14 going to obviously get into some ore discussion with regard 15 to experiences. There was some discussion before the break 16 about how you go from simple models, like RESRAD, D&D and 17 MEPAS and those, to more complex or how can you use a more 18 sophisticated hydrogeology model in conjunction with what 19 we've called conventional or state-of-the-art dose 20 assessment models, and we'll get into that. 21 But in order to kind of get an understanding of 22 what we're trying to provide to you today, this workshop, t 23 we're hoping, will produce a proceedings and the proceedings 24 hopefully will provide you with not only the information j 25 provided, but references and other material in something ANN RILEY & ASSOCIATES, LTD. O- Court Reporters 1025 Connecticut Avenue, NW, Suite 1014 Washington, D.C. 20036 (202) 842-0034
L I-191 1 that.is a citeable source. 2
) WE also are going to have a proceedings, the 3'
proceedings, as I said, meaning the transcripts, the 4 transcripts should be out in about a week, but there was 5 some confusion. 6 It's going to be at the public document room here 7 in Washington.
~It's not going to be the local ones. It 8
will be just the one here in Washington and it will be there 9 if you want to look at the official transcript of this 1 10 meeting. 11 Now, hopefully, everybody has picked up a copy, if
- 12. you're interested, a copy of the last proceedings we did.
13 The last proceedings was a meeting similar to this workshop, 14 held November the 13th and 14th of 1997, in which the codes, () 15 MEPAS, RESRAD, D&D and PRESTO were presented and there was 16 actually a demonstration, and the proceedings actually have 17 papers in here about those codes. 18 So this workshop kind of builds on that and now 19 we're getting into the ground water aspect, and we hope to 20 produce a proceedings similar to this one. So you might I 21
.want to contact us in about six months.
22-I'd like to now introduce our first speaker. Walt 23 is going to be talking about integrating site-specific 24 ground water modeling into dose assessment. So imagine, if 25 you would, that somebody like Professor Neuman or someone () (m / ANN RILEY & ASSOCIATES, LTD. Court Reporters 1025 Connecticut Avenue, NW, Suite 1014 Washington, D.C. 20036 (202) 842-0034
.]
l l l 1 19J has done a site-specific analysis using some type of a (*g 2 V detailed or more complex model, how can it be integrated 3 into a dose assessment. 4 So Walt will try to go through that for us. 5 MR. 32YELER: As Tom mentioned, I wanted to talk a 6 little bit about integrating site-specific ground water 7 modeling into doue assessmentc. 8 I wasn't going to talk about a particular case 9 study, but rather cover in general some of the issues that 10 would be faced in using a more detailed simulttion of a 11 ground water model in a dose assessment calculation. 12 So I'd first like to give a quick overview of the 13 elements of the default model. 14 I think it was discussed in Professor Meyer's talk. The other models, RESRAD, MEPAS and [J '\ 15 16 so on, have also a fairly simple conceptualization of the ! ground water system. 17 So some of the issues associated with interfacing a detailed model of ground water with a dose 18 assessment calculation would also occur for those 19 simulations, but I was going to discuss this in terms of the 20 D&D model. 21 One of the important issues that comes up is, as l 22 has been discussed, the parameters of the default dose 23 assessment model do not vary in time. Generally, it's a 24 steady-state model and there is also one number that's used 25 to characterize important quantities, such as infiltration
/'
(, ]/ ANN RILEY & ASSOCIATES, LTD. Court Reporters 1025 Connecticut Avenue, NW, Suite 1014 Washington, D.C. 20036 (202) 842-0034
192 1 and so on. 2 So a site-specific model is likely, among other 3 things, to include some additional spatial detail and 4 g perhaps transient behavior. So there is a general question ; 5 6 of how to interface between more and less d? tailed elements of the system. So I'd just like to discuss some possible 7 approaches for dealing with that. 8 As far as the mechanics of doing that interface, I 9 wanted to just briefly mention the SEDSS program that's 10 being managed by Ralph Cady under Research. The goal of 11 SEDSS is to provide for a method of tracking the logic and 12 ! the mechanics of interfacing more complicated or developing l 13 a comprehensive system simulation in a decision support 14 context. () 15 So I won't go into great detail on this. This has 16 been discussed, I think, quite a bit in the earlier 17 presentations. But for the residential scenario, this 18 system conceptualization includes an initially contaminated 19 surface soil layer, an unsaturated zone and an aquifer 20 layer. Water is withdrawn from the aquifer layer, both for 21 irrigation purposes, for drinking directly, and for other 22 consumptive uses on-site, 23 So a central feature of this conception of the 24 ground water system is a recirculation of contamination from 25 the contaminated soil layer through the aquifer and back. 1 ("'T ANN RILEY & ASSOCIATES, LTD. (_) Court Reporters 1025 Connecticut Avenue, NW, Suite 1014 Washington, D.C. 20036 (202) 842-0034 m
p 194 1 We have very simple water balance considerations. The 2
; irrigation rate is specified independently of the 3
infiltration rate and so there is an implied amount of 4 either' net precipitation or evapotranspiration and then, in 5 effect, provides a water balance for that layer. That's not 6 an explicit simulation in the D&D code, but rather an 7 implicit difference between these two rates. 8 Similarly, there is a total amount of recharge to 9 the aquifer, through the contaminated layer, and a total 10 amount of withdrawal that's determined by this independently 11 specified irrigation and total consumption. 12 If there is a discrepancy between those two 13 quantities, any excess withdrawal is assumed to come from an 14 uncontaminated source. Any excess infiltration is assumed () 15 16 to escape the capture zone of the well. But to the extent -- to the total amount of contaminated recharge to the aquifer, 17 those show up in the well, subject to the consumptive
- 18. specifications at the surface.
19 So really any of these elements, there are three 20 sort of large elements of the ground water system it would 21 seem reasonable to consider modeling in more detail, 22 developing a site-specific model for it; that is, the 23 surface soil layer, the unsaturated layer, the aquifer 25 itself. 25 So any one of these or any combination of these
/~ ' ANN RILEY & ASSOCIa*ES, LTD.
Court Reporte s 1025 Connecticut Avenue, NW, Suite 1014 Washington, D.C. 20036 (202) 842-0034
{ . 1 l. 195 1 might consider -- might conceivably be modeled in more 2 detail using the site specific model. 3 The information that is exchanged among these 4 elements, in the case of the surface soil layer, there is a 5 flux of water and radionuclides obviously into the 6 unsaturated layer. Similarly, from unsaturated layer, the 7 aquifer,. flux of water and radionuclides, water and 8 concentration in-ground water are the key quantities that 9 are withdrawn from the aquifer to be used for consumptive 10 use and for irrigation. 11 So the -- again, any of these elements could 12 conceivably be refined, replaced with a site-specific model, 13 The things that might be included in that -- in a 14 site-specific model would include, again, more detail on the 15 (
} spatial variability of the parameters, a consideration of 16 transient effects that might be occurring in those elements; 17 also, additional processes that are not considered in the 18 simple screening models, 19 And some of the additional process, this is by no 20 means an extensive list, that might be included would 21 include consideration of solubility. Currently, there is no 22 dissolution or precipitation modeling default D&D code.
23 The default code includes a linear reversible 24 absorption. There are other absorption models that 25 potentially could be defended. Again, variability or l ((~3
,) ANN RILEY & ASSOCIATES, LTD.
Court Reporters 1025 Connecticut Avenue, NW, Suite 1014 W0shington, D.C. 20036 (202) 842-0034
E . . 1 196 transient behavior in these elements, 2 the default D&D code (~}
\_/ does not include a consideration of the aquifer 3 ;
hydrodynamics; that is, the potential for the hydraulic \ 4 characteristics of the aquifer to limit the ability of the 5 well to produce water. 6 It's assumed that however much water is in the aquifer is capable of withdrawing that amount . 7 There is currently no absorption considered in the 8 aquifer and that's potentially something that might be 9 l supported and included in a model. 10 11 But if one or more of those extensions are made to one or more of those elements,' incorporating the or 12 integrating that sort of a refined calculation in a dose 13 . model involves specifying these summary quantities. In ! 14 other words, if a more sophisticated, (~'\ 15 a more detailed model V of the unsaturated layer would still be required to accept 16 -- if these elements were not replaced -- a single number 17 for infiltration, mass flux in the top, and to provide a 18 summary integrated measure of some kind of flux into the 19 aquifer. 20 ; And so one of the general issues associated with a 21 ; site-specific modeling is, again, accomplishing this change 22 of scales, either going from very crude specification, for 23 example, in the surface soil layer, to a detailed spatially 24 variable transient specification of the unsaturated layer, 25 going back from that more detailed specification to a crude i f - E s ANN RILEY & ASSOCIATES, LTD. Court Reporters 1025 Connecticut Avenue, NW, Suite 1014 Washington, D.C. 20036 (202) 042-0034
E , 1 197 1 specification of the aquifer, for example. 2 So in general, these are some procedures that 3 might be used to do that sort of thing, in going from a ! 4 coarse to a fine scale, for example, from an integrated or a 5 summary measure of infiltration in the unsaturated zone to a 6 detailed model of the aquifer. 7 One possible procedure is to assume that the 8 single infiltration rate is a uniform rate or that more 9 detailed simulation in the aquifer, that's a possible 10 approach, but it seems that there are other assumptions 11 about the spatial variability that are also compatible with j l 12 a single number for infiltration. ! 13 I think it is not automatically the case that a 14 l' single number in a simple element is best represented as a (%) 15 uniform number than a more complicated element, and I think 16 this perhaps goes to the point that Dr. Neuman raised 17 earlier with respect to using a single number to
- 18 characterize a very complex system. It seems that it's 19 possible to do that, but there needs to be very careful 20 consideration of the possible spatial variabilities that are l 21 being assumed, that are being assumed in that process.
22 Going from fine to a coarse resolution, this is 23 maybe less of a conceptual problem. A clear example there 24 would be coming up with a summary concentration from a very 25 detailed model of an aquifer. d ANN RILEY & ASSOCIATES, LTD. Court Reporters 1025 Connecticut Avenue, NW, Suite 1014 84 b3 j
= . 198 1 2 If the number of interest is concentration in the drinking water, then clearly an integrated value or an 3 average value over the scale of the well is an appropriate 4' number to use in the calculation. But those are, I think, 5 just some general observations about the interfaces that 6 need to be considered in incorporating more detailed ground 7 water modeling. 8 Again, we're working on a system that deals with 9 some of the mechanical issues associated with making those ) 10 kind of matches. This is a SEDSS program that Ralph Cady is 11 managing. Just some features of that system are that the 12 models that are used are specified in terms of assumptions 13 that are made about the system rather than selecting from a 14 particular code. The system is designed to allow the user () 15 to describe the assumptions that they believe are 16 appropriate and can support as being appropriate for 17 different parts of the system and the calculation to reflect l i 18 those assumptions. 19 It also includes a data worth calculation to help i 20 the user in assessing the value of including -- of ! 21 collecting additional data, both for supporting parameter l 22 values for an existing model or for supporting a more 23 detailed model. 24 The intention is to allow a wide range of 25 computational tools to be accescible and for this year, i O ANN RILEY & ASSOCIATES, LTD. Court Reporters 1025 Connecticut Avenue, NW, Suite 1014 Washington, D.C. 20036 (202) 842-0034
T 1 199 we're looking at incorporating the D&D default models with 2 the existing 1-D network transport model. l 3 1 That's all I had to say. 4 MR. NICHOLSON: Thank you very much. 5 Are there any questions about integrating site-specific modeling into 6 dose assessments? 7 MR. ROBERTS: Rick Roberts, Rocky Mountain ! 8 Remediation Services. In refining the D&D model, I believe 9 there may be another pathway that would be really good to 10 look at, as well. 11 We talked about earlier where there are sites where the ground water ingestion pathway or irrigation 12 just is not applicable to a site because it doesn't yield 13 enough or it isn't of good quality. 14 And what has to be looked at really is the i (~') 15 V transport of contaminants from the ground water at the site 16 t.o a surface water body, where people uvuld contact the 17 contaminants, and I guess I would ask if adding a surface 18 water component to your ground water pathway has been 19 thought about and if that's something that would be added in 20 the future. 21 MR. BEYELER: Right now, there is an aquatic 22 pathway, as you're probably aware, but the exposure that 23 would occur there is simply through the ingestion of fish. 24 There is, I believe, no calculation of dose due to 25 immersion in contaminate water and I don't know if that's ANN RILEY & ASSOCIATES, LTD.
\- Court Reporters 1025 Connecticut Avenue, NW, Suite 1014 Washiraton, D.C. 20036 (202) 842-0034
T . 1 200 1 something that is under considerntion for inclusion in the (} 2 scenario. I believe that's an element of the definition of 3 the exposure scenario that immersion does are not 4- considered.
- 5. l MR. EID:
t You're right, it is not included in the 6 scenario and it won't be included in the scenario for 7 immersion water, because the model is already conservative 8 and assumes the concentration in the pond is exactly the 9 same concentration in the aquifer, where you're pumping 10 contaminated water, where, in fact, actually you may have 11 this pond is coming from rain water rather than from the 12 contaminated water. That's one of the reasons. 13 So the immersion in surface water is not included 14 as a pathway. r However, ingestion of fish and seafood grown () 15 in the pond is accounted for. 16 MR. ROBERTS: Okay. So you're saying that if you 17 have contaminated ground water in the D&D code, then there 18 is an assumption that an adjacent surface water body has the 19 exact same concentration in that -- and someone ingests the 20 fish from that surface water body. 21 MR. EID: That's the assumption of the model, that 22 you are pumping the water from the aquifer to put in the 23 pond. That's the current assumption in the model. It's 24 different than RESRAD, RESRAD is different. 25 MR. ROBERTS: That may be -- you're looking at O ANN RILEY & ASSOCIATES, LTD. Court Reporters 1025 Connecticut Avenue, NW, Suite 1014 Washington, D.C. 20036 (202) 842-0034
r ._,i 201 i I refinements to the model. Then I guess I'm saying that may () 2 3 be a refinement you would want to put in the future to look l at that transport and using the surface water body for 4 l things other than just fish ingestion. 5 MR. EID: That's a good suggestion. We have 6 l another model, which RESRAD takes into account for this. i 7 Now, developing the D&D model further, I'm not sure if there 8 is a plan right now to make the model more sophisticated,
- 9
'but it could be in using SEDSS, you could link SEDSS dose l 10 impact analysis to another kind of model, which is more 11 appropriate, getting more ground water in a more 12 sophisticated manner and dealing also with surface water.
13 MR. BEYELER: I would just mention it seems there 14 are perhaps two issues. There is incorporating immersion () 15 dose in the D&D model, that is one question, and whether 16 that is considered as part of the residential scenario or 17 not is perhaps a larger question. 18 So in other words, if there is no consideration of that t 19 pathway in the scenario definition itself, then really it 20 would not be included in the model, although it would be an 21 eaay thing to include'if the scenario definition were 22 changed to include that pathway. 23 I don't know if that answers your question. 24 MR. ROBERTS: Thank you. 25 MR. NICHOLSON: Thanks for the recommendation. ). N ANN RILEY & ASSOCIATES, LTD. Court Reporters 1025 Connecticut Avenue, NW, Suite 1014 Waspaton, D.C. 20036
'2) 842-0034
[- 1 202 1 All comments made will be considered. () 2 MR. MORTON: Henry Morton. Historically, we have 3 looked at those pathways and dismissed them as being minor 4 and insignificant. i 5 Basically, what I'm talking about in particular 6 are immersion, swimming, boating, and then the shoreline, 7 shoreline recreation or standing on the shoreline fishing 8 and being exposed to the sediments. 9 The particular history, I think, if you go back 10 through Appendix I, reactor effluent water, and, for 11 instance, look at the output from a code like LADTAP or 12 other comparable codes that we developed and used, they were 13 all pretty consistent, those are minor pathways. 14 Only the fish is really worth considering. () 15 MR. EID: I have a question, if you'll allow me. 16 I believe when we talk about integration of ground water 17 models with dose assessment, it's quite important to talk 18 about location of the receptor, because it will make a 19 difference if you assume the receptor is located directly on 20 the pile of the contaminated zone. I mean, the center of 21 the contaminated zone has some assumptions for these models, 22 or the receptor is located off-site, and this makes a 23 difference. This means you need a more sophisticated 24 transport, specifically if the site area is large. 25 The other question which I just mentioned, the
^\ ,,,) ANN RILEY & ASSOCIATES, LTD.
Court Reporters 1025 Connecticut Avenue, NW, Suite 1014 Washington, D.C. 20036 (202) 842-0034
203 1 site area, it is quite important to take it into
/N, 2 consideration.
G) For small sites, and you assume the location 3 , of the receptor is directly on-site, could be more detailed 4 and sophisticated transport, could be overdone, whereas if 5 , you have a very large site and you assume that the person is ! i 6 located even at the site Psundary, and the site is more 7 complex, you may need to construct some kind of more 8 advanced transport analysis. 9 So I would like to have your comments about these 10 two aspects, the area of the site and the receptor location 11 related to integratine dose assessment with ground water 12 modeling. 13 MR. BEYELER: I guess my observation would be that 14 that question is much more general than the ground water () 15 16 question, in that I -- if I understand, the assumption that the receptor is directly on top of the contaminated area has 17 a strong bearing on the calculation of the dose due to 18 direct exposure and inhalation and many other pathways that 19 are not necessarily just the ground vater pathway. 20 So it seems that consideration of the size of the 21 site is likely to have a much stronger influence on the 22 estimated dose due to those pathways, rather than to ground 23 water pathways specifically. 24 I think that it is -- the calculation of the 25 effect of displacement on the receptor can certainly be f ANN RILEY & ASSOCIATES, LTD. ( Court Reporters 1025 Connecticut Avenue, NW, Suite 1014 Washington, D.C. 20036 (202) 842-0034
i . . 204 1 included in a more elaborate ground water pathway, but it [%)T 2 just seems_ intuitively stronger effect would be through 3 those other pathways. I don't know if there are other 4 thoughts. 5 MR. EID: I don't know -- I perceive that, if the 6 area if the site is very large, you need to have more 1 7 advanced analysis because of the infiltration rates will be 8 higher and the mechanism for leaching could be a little bit 9 different and, also, if the site is more complicated, you 10 have thick layers, for example, you may need to account more 11 for more sophisticated models, specifically when you talk 12 about retardation and dispersion at the same time. 13 So whereas if you have a small site, these issues 14 G could be minimal, rather than important for larger area of Q 15 the site. 16 MR. NICHOLSON: I think we'll end the discussion 17 there and we might follow-up during the group discussion. 18 Gene? Our next speaker is also going to talk 19 about integration site-specific ground water modeling into 20 dose assessment. Gene Whelun vill be presenting a paper. 21 Gene is from Pacific Northwest National Laboratory. The 22 work is supported by DOE and EPA and his co-authors are 23 Gariann Gelston and Karl Castleton. 24 MR. WHELAN: Can everybody hear me? I'm going to 25 sit down here a little bit and kind of work with my little l'\~/) ANN RILEY & ASSOCIATES, LTD. Court Reporters 1025 Connecticut Avenue, NW, Suite 1014 Washington, D.C. 20036 (202) 842-0034 l
205 1 computer. I'd like to note a couple of things and then I (-] 2 have been asked to talk about linking different models 3 together, ground water models, dose assessment models. But 4 in general, this concept of linkage, how do we -- what are 5 the procedures we go through to link two, three, four 6 different models together. 7 I'm going to be talking abvuc that. Walt talked 8 about the types of information that one might transfer from 9 one model to another, the source term to the vadose to the 10 saturated zone, et cetera. So I'm not going to get into 11 that. 12 I'm going to talk a little bit about when the 13 rubber meets the road, actually how do you do this, and 14 we've been involved in developing actually three different 15
} techniques which - and some of them everybody uses. These 16 are real world linkage techniques. It's not vaporware, 17 these techniques are out there in platforms or frameworks 18 that people are actually using and doing assessments with.
19 So this is not a hypothetical concept. 20 Terminology, I'm going to use some very generic terminology, 21 descriptive terminology as opposed to scientific 22 terminology. So if you see things like plug-and-play, it's 23 more descriptive than scientific. 24 All right. The real world is a very complex place 25 and we all know that and any modeling that we do, at least rx ANN RILEY & ASSOCIATES, LTD. (~ ) Court Reporters 1025 Connecticut Avenue, NW, Suite 1014 Washington, D.C. 20036 (202) 842-0034
f . 206 1 in my opinion,
} 2 3
IAndoneofthethingswedo is a gross simplification of that real world. tend to break it up into little components, little boxes, if 4 you will, and then we try to simulate or model or monitor 5 aspecta associated with each of these boxes. i 6 Our world that I have described here, we've broken 7 it up into four areas. One is source, one is transport, one 8 is exposure, one'is impact. 9 Source being the release mechanisms from a source 10 term into the environment; transport being transporting 11 those constituents from a source to a receptor; exposure 12 talks about the ways in which a sensitive receptor can be ! 13 : exposed; and, obviously, the fourth one is what are the 14 impacts, dose, risk, et cetera, to that exposure. () 15 And as we can see, the ground water, which I put 16 in this nice big box, is sitting right here in the middle of 17 everything. So it can interact with the source term, it can I 18 interact with other transport media, it can interact with 19 the food chain and exposure routes. 20 So we can't just look at the ground water modeling 21 by itself, but we have to look at the bigger picture and I 22 understand where it sits within the bigger picture. 23 Many questions have been asked and many questions 24 will be asked over these two days and the answer that you're ( 25 generally going to get is it depends.
~5 k'_s ANN RILEY & ASSOCIATES, LTD.
Court Reporters 1025 Connecticut Avenue, NW, Suite 1014 Washington, D.C. 20036 (202) 842-0034 i l
p . ( l l
- 207 1
It depends on the questions you ask, it depends on (~h 2
- \_) the assumptions you have, it depends on the data, it depends 3
on the level of analysis you want, et cetera. It depends.
- 4
! Can you do this with a more simplified model or a more
- F complex model?
There is no really blanket answer yes or no. 6 It depends. 7 And likewise, when we're dealing with approaches 8 of linking two different models together, it's a function of 9 many aspects. I've listed just a few of them here. The 10 first one is the type of question that you're trying to 11 answer. All right. For example, I want to de a nationwide 12 analysis of 10,000 waste sites and I want to have a system 13 where I link models that address that issue. 14 Well, i that may be a different type of linkage tnan () 15 if I have a single waste site and I'm just simulating it I 16 through a vadose zone, a saturated zone to a well. The 17 second piece o information that might be important to a 18 determine what type of linkage you might want to put these 19 codes together, these models together, is am I importing 20 data or am I exporting data. For example, I run a source ! 21 term model and then I export that information out of a l j 22 system to a very detailed numerical ground water model. I 23 run that detailed numerical ground water model. 24 I import that information in and I link it up to a 25 dose model and have my doses calculated. So it's important
/"'s ANN RILEY & ASSOCIATES, LTD.
(_/ Court Reporters 1025 Connecticut Avenue, NW, Suite 1014 Washington, D.C. 20036 (202) 842-0034
208 1 to know whether the data is being imported and exported or 2 ( v whether it's self-contained. That data is self-contained
.3 within the system itself.
4 The third and fourth items I've listed here, which j 5 tend to crop up as being very important, scale and 6. resolution and time and space and my definition of 7 resolution may be different from others, but it's just a 8 placeholder. 9 Scale, what do I mean by scale? Medium specific, 10 basically a waste site, if you will. Watershed, regional, 11 global. Each one of these scales may require different 12 types of models, may require different. types of boundary
.13 conditions. Resolution, you can have low, medium, high.
14 That's what I use. Somebody else could use something else. 15 High being the three-dimensic.nal finite element numerical 16 model with bells and whistles and spaghetti hanging off. It 17 does everything. l I 18 And then you have medium type models, analytical, 19 semi-analytical, then you have your low resolution models, 20 your hazard ranking system, for example. 1 So each on= of these may require different spatial I 22 and temporal constraints associated with them. For example, l 23 in the medium range, for analytical models, you may be 24 dealing with a plane where the mass fluxes across the plane 25 is uniform, whereas in a numerical model, you may have nodes O L> ANN RILEY & ASSOCIATES, LTD. Court Reporters 1025 Connecticut Avenue, NW, Suite 1014 2 84 b3
209 1 and that mass flux rate associated with each node varies (V 2 spatially and temporally. 3 So these things'are very important when you're 4 i trying to decide how you're going to link these models 1 5 together. 6 There are three approaches that I'm going to 7 outline here. They are approaches that we've dealt with, 8 there may be others. I've placed them in two main 9 categories. 10 The first one I call the traditional linkage 11 approach. That's really tricky. And there we link models a 12 priori by hardwiring all of the connections. We take the 13 models, we glue them together, fuse them together, and 14 they're stuck. What you see is what you get. () 15 16 The second main heading is what I call object-oriented specifications approach. There are two 17 categories under this. There is the master file approach 18 and from the master file approach, this is where we -- where 19 the linkage specifications are developed to meet the needs 20 of the specific models in the system. 21 Then the plug-and-play approach is where the 22 models themselves are adjusted to meet the linkage 23 specifications of the system. So for the master file 24 approach, I've got the models and the system is being 25 manipulated to meet the requirements of the model and then f o (, ANN RILEY & ASSOCIATES, LTD. Court Reporters 1025 Connecticut Avenue, NW, Suite 1014 2 84 I b3 l I' t
1 21t the plug-and-play, I have the system and the models are g 2 being adjusted to meet ( the requirements of the system. 3 We've developed approaches and frameworks for both of these 4 5 and I will talk a little bit about those in a minute. This slide, and I'm sure everybody has seen one 6 version of this or another, this is what I call your 7_ traditional multi-media modeling approach, where your ground 8 water system sits here and you have your dose assessment 9 sitting in these boxes over here, and you've got your source 10 and you've got your transport pathways, exposure routes, 11 outputs, and notice the lines, everything is hardwired 12 i together and like I said, what you see is what you get. 13 More mechanistically, in terms of what does a 14 traditional model linkage mean, I've put this slide 15 together, and think of these boxes as separate or different 16 models. { 17 18 And each one of these models communicate with each other through these little elliptical circles called 19 processors, data processors. You may have one model with 20 sub-routines and these processors could be common blocks, 21 they could be include statements, they could be objects of 22 call statements, they could be literally processors that 23 transfer the information from one model to another model . 24 But as you can see, as I add more and more models I 25 to the system, the linkages become very complex. The reason g ANN RILEY & ASSOCIATES,_LTD. ( Court Reporters 1025 Connecticut Avenue, NW, Suite 1014 Washington, D.C. 20036 (202) 842-0034
~
. 1
)
211 1 is because this system that I have put together is hardwired f) v 2 together. 3 So some of the attributes of this traditional 4 approach are that the models are hardwired together, the 5 data enters the system through the traditional routes, 6 either as a user, I punch the information in, or the 7 information is included in the database which is associated 8 with the system itself. 9 Generally, you have minimal access to outside 10 databases. I just can't say let's go read this database i 11 over here, STORET or something else, and access that 12 i'nformation and put it in my system. You generally also 13 tend to have minimal GIS connectivity. 14 So you don't really -- you can't really () 15 incorporate very well the spatial variability associated 16 with the GIS system, and GIS is going to be very, very 17 important in the future, I believe. 18 In general, what you have is you tend to have a 19 closed form system. I can hit a big go button and this 20 whole system can work because everything is self-contained 21 to a certain degree. I've just heard today that some 22 modifications to SEDSS have occurred, so SEDSS may or may 23 not fit in this category, but here are some typical examples 24 that I feel fit in this category. 25 The second major category in terms of linking
~
(~~')) ANN RILEY & ASSOCIATES, LTD. Court Reporters 1025 Connecticut Avenue, NW, Suite 1014 Washington, D.C. 20036 (202) 842-0034
212 1 models together I call the -- this is your object-oriented
\
1 (~~A 2 specifications approach. What we say is that before we link 3 models together, we actually specify what the linkages, what 4 the formats and what the form of these linkages will look 5 like. And then we adjust the models to make sure that they 6 meet the requirements of these linkages. 7 So I have a model sitting here and this is a l 8 processor which takes the output from this model, reformats 9 it into the format of the system that links these models 10- together, and then stores it such that any other mo. 1 that 11 wants to have access to that information can access it, 12 consume it, reformat it so it fits into that other model. 13 As you can see, with this specifications approach, 14 this spider diagram gets cleaned up significantly, which () 15 means that each one of these boxes can represent objects and I 16 I can take one box out and very easily stick another box in, 17 and the only requirement is that it meets the specification ! 18 of the system for transferring information from one model to 19 another model. 20 Cleans things up. This is very good for QA/QC, 21 for making modifications, for future modifications, say, for 22 example, computers now run so quickly, I want to get rid of 23 an old model and I want to plug in a new model. We're not 24 changing the system. All we're doing is changing the model. 25 So what is the non-traditional master file O ANN RILEY & ASSOCIATES, LTD. (%-) Court Repo'.... s 1025 Connecticut Avenue, M, Suite 1014 Washington, D.C. 20036 (202) 842-0034
213 1 approach? This is under the specifications approach. This
~T is the first category. These linkage specifications are (d 2 3
established to meet the model's needs. So I have specific 4 models, I say.these models have certain needs, and what do I ! 5 do? The system is designed to meet the needs of the model. 6 So how do we do that? We establish a master file I 7 which contains all of the parameters that are required by l 8 all of the models in this multi-media system and of those 9 parameters that are contained in this master file, we group 10 them according to real world objects, like ground water, 11 like surface water, like air, like wetland, et cetera. 12 So we take this master file of parameters and we 13 group them, these parameters according to a particular 14 category that is representative of these objects, those 15 boxes, if you will. 16 Those boxes could represent, as I said, vadose 17 zone models, saturated zone models, source term models, 18 river models, et cetera, dose models. 19 Also important on this is the parameter names in 20 this master file are fixed. So if you want porosity to be 21 N, that parameter name, then it's N and everybody, all the 22 models recognize in this master list that porosity, that is 23 designated by the letter N. 24 And also included in this are the attributes 25 associated with each of the parameters. Is it a constant, 7J01 RILEY & ASSOCIATES, LTD. s Court Reporters 1025 Connecticut Avenue, NW, Suite 1014 Washington, D.C. 20036 (202) 842-0034
V . 214 1 is it an integer, is it a reel, is it stochastic? If it is, 2 what are the characteristics associated with that parameter? 3 1 Finally, the models are fixed into these 4 object-oriented slots, all the ground water models are over 5 in one area, the surface water models are over in another 6 area, et cetera. 7 There are constraints to this master file 8 approach. It would seem nice to be able to have access to 9 all the information by all the models for all the 10 parameters, because now you_ don't have data redundancy. 11 The problem is that the models tend to be 12 hardwired and you have no plug-and-play attributes. I can't 13 pull one model out and put another one in, and the reason is 14 p because that new model may not recognize the parameter names i I 15
%J I have in the system.
16 Second, data and databases tend to be hardwired. 17 Again, the reason is because all the names in the system are 18 fixed. It doesn't allow for different parameter names. We 19 have to fix those. Therefore, it limits the use of legacy 20 codes, which I just covered. 21 It doesn't allow very easily for different model 22 user interfaces. Each one of these models might have its own 23 user interface. Now I need a user interface that is 24 t cystem-wide consistent. And finally, any new applications 25 using different models require retrofitting. (N (s,) ANN RILEY & ASSOCIATES, LTD.
~
Court Reporters 102S '.'onnecticut Avenue, NW, Suite 1014 Washington, D.C. 20036 (202) 842-0034
E . . 215 1 Now I'd like to talk about the third approach, 2 which is the second under the specifications, and that's the
- 3 plug-and-play approach.
' Basically, how this works is you 4 i* end up with a tool bar and there is an icta for each of the 5 main categories associated with the real world. Here I have 6 chemical database, I have source term models, vadose zone 7 models, watershed models, saturated zone models, surface 8 water models, air models, food chain models, dose or intake 9 models, and risk and hazard models. 10 I mean, we can have other icons, but that just 11 gives you an idea. And under each one of these, you could 12 have a suite of models to choose from. Let's look at the 13 source term. Here is an example page, if you will, where I 14 have three source term models and the one that was chosen is l I~ ) 15 the MMSOILS source term model. ! V I 2 l 16 ! People use models not necessarily because they're i 17 the most appropriate, but because they're familiar with 18 them. They like to use what they're familiar with. ? They 19 und6rstand the limitations, the constraints. They can make 20 that model dance and do all the nuances to make sure that } 1 l 21 assessment is done correctly. Therefore, they will want to 22 use their own legacy codes and that's one of the attributes i 23 associated with this plug-and-play system. ! 24 In effect, what you can do is you can pull down ; 25 1 these icons at the top, source term, I know it's difficult ! r (~q-
') ANN RILEY & ASSOCIATES, LTD.
Court Reporters 1025 Connecticut Avenue, NW, Suite 1014 ' Washington, D.C. 20036 (202) 842-0034 l
r< . 216 1 to read, here is vadose zone, here is aquifer, and I can () Gi 2 link them in any way that's applicable to the assessment at 3 my site. 4 Then under each one of these icons, I cen then I 5 pick the model that's most appropriate for the assessment at 6 my site. 7 Eo this is how the plug-and-play system works. 8 These models in particular are adjusted to meet 9 the linkage specifications of the system. The system says 10 if you want to communicate, I'm going to give you the 1 l 11 telephone book. It has all the numbers in it. You're net l 12 allowed to change those numbers. 13 So therefore, model A must produce a certain 14 i amount of information in order to communicate in the system. (m) 15 Model B automatically knows that a minimum set of 16 information exists, so it can go in and consume it if it 17 wants to. 1B You can create additional information if you want, 19 that's not a problem, but there is a minimum set of 20 information that needs to be provided by your model if your 21 model wants to play within this system. 22 This system views all of these modules as common 23 objects. It's all ground water models clumped together as 24 ground water models. And all of these like models, say 25 vadose zone models, have common data specifications.
/~
(T,,) ANN RILEY & ASSOCIATES,. LTD. Court Reporters 1025 Connecticut Avenue, NW, Suite 1014 Washington, D.C. 20036 (202) 842-0034
( 217 1 So they all know what they need to produce and ("'T 2 what they can consume in the system. The reason why we do
\~)
3 this is because we want to minimize the modifications to 4 existing codes. We want to leave the legacy codes alone, so 5 people aren't spending resources trying to change the model 6 so it fits in some system someplace. So you can leave the 7 legacy code alone and use the resources in doing the 8 analysis and the assessment. l 9 The other thing is that with a system like this, 10 and I'll demonstrate it in a second, you can link to other 11 framework environments. In other words, multiple frameworks 12 can now communicate with each other and the reason is 13 because you know what information exists within that 14 framework. () 15 And I demonstrated earlier the easy problem 16 definition protocol. You can visually see your conceptual 17 site model. 18 ! Now, how do we go about linking these, say, two i 19 ! models, three models, or other models together within this 20 system, such that the system understands what your model is 21 and how it can communicate with other models. 22 Well, we require a description file, a DES file, l 23 and it's a very short file. It's not very long at all, 24 which is nice because you don't have to do a lot of work. 25 In fact, we've developed a program in which it can
~
ANN RILEY & ASSOCIATES, LTD. (s_)h Court Reporters 1025 Connecticut Avenue, NW, Suite 1014 Washington, D.C. 20036 (202) 842-0034 L .
218 1 help you fill out and create this description file. In it, 2 I you have to identify the class of model that you have, is it 3 an aquifer, is it a vadose zone model, is it a river model. 4 l You also have to identify the accepcable connections. All 5 those lines that you saw connecting the different boxes, ! 6 i they may not be applicable to your model or the model that 7 you put in here. ! 8 Therefore, you must tell ,he system which 9 connections are applicable. 10 There is also support information, support 11 information being if the model breaks down, who do they l 12 contact? They're not going to contact the system developer, 13 they're going to contact the developer of the model. 14 And then finally, in this description file are all () 15 16 of your input parameters. The names of those parameters and the attributes associated with those parameters, is it a 17 reel, is it an integer, what's the range, is it constant, 18 what are the -- is it stochastic, what are its 19 characteristics. 20 The reason is because you want to be able to do a 21 Monte Carlo assessment with -- have your model included in i 22 Monte Carlo assessment. l 23 If this white outline represents a model that's l 24 l being put into the system, it can be broken down into three 25 basic components; the general information includes your O ANN RILEY & ASSOCIATES, LTD. Court Reporters 1025 Connecticut Avenue, NW, Suite 1014 k Washington, D.C. 20036 (202) 842-0034
219 1 description file, and then all of the inputs that comes into 2 your system, including input through your model-user 3 interfaces or databases that you're calling in, all of that 4 information gets stored in what's known as a global input 5 data file, a GID file, which is outside of your module, if 6 you will. 7 Now, that information is then sent -- that's in 8 the format. That data is in the format of the system. Your 9 preprocessor reads that information and reformats it for 10 your input file, maybe it's a flat ASCII file. That's so 11 you can run your model. You send the output out of your 12 model. Then you have a post-processor that reformats that 13 information and puts it into a format that the system 14 understands.
) 15 And any other 4 01 that wants to communicate will 16 understand what that format is of the system information.
17 So in effect, this is what your model does and there is very 18 little modification that is required by any models that are 19 put into the system. 20 Now, in addition to having models conmunicate, we 21 i are also moving forward to using this type of an approach, 22 the systems specifications approach to linking in databases l 23 i and they can be different types of databases. For example, 24 let's look at scale. I could have a database that provides 25 site-specific information and I go to run my model and my I ANN RILEY & ASSOCIATES, LTD. O* Court Reporters 1025 Connecticut Avenue, NW, Suite 1014 Washington, D.C. 20036 (202) 842-0034 l
220 1 model -- that site-specific information lacks a few of the 2 parameters.
/} I can't populate it.
3 So I can then go into a regional database, pull 4 that information out that's missing, and populate my 5 database for my model and if I need to go to a national 6 level, I can complete filling out the database that's going 7 to run my model. So I can pull in information from multiple 8 types of databases. As long as I meet the data specification 9 needs of the system, all of these models will be able to 10 communicate with any and all of these databases, and this is 11 an area that we're moving into and we've already begun this 12 with EPA Office of Research and Development and the Office 13 of Solid Waste. 14 Where does this all lead us to hopefully some () 15 place in the future, we're looking at setting up a -- not 16 we, but DOE and I know NRC is having discussions with l 17 respect to this, EPA, and DOD in terms of setting up or 18 having a platform set up on a web-based system. And what 19 you can do with a web-based system is you would -- you can 20 have located there the operating software and you can have 21 the databases that are required to operate that software, 22 the software required to consume that data. 23 And for each one of these, you can either have the 24 l software or the data at some central location. So you go up
'25 on the web, you go to Argonne National Labs, they have all i
O ANN RILEY & ASSOCIATES, LTD. Court Reporters 1025 Connecticut Avenue, NW, Suite 1014 Washington, D.C. 20036 (202) 842-0034
~ 1 221 1 the data, all the models, and you run everything there, and Q 2
!a< then you download the results to your system, 3
A second option would be to download the software 4 and/or the data to your computer, which is what I like to 5 do, and then I run it on my own computer and I have the 6 results sitting there. i I can send the information back up 7 to the web if I want. 1 8 And then the third option would be to have the 9 computer models located in multiple locations, at Sandia, at l 10 Argonne, at PNL, at EPA-Athens, and I could have data -- 11 databases located at remote locations, different locations, 12 and I could actually use the web to call up the right ! 13 i computer models and pull in the right databases and operate 14 l (3 this system from your PC, and it -- or from a UNIX system, (_) 15 because it wouldn't make any difference whether you're 16 dealing on the web with a UNIX or with a PC. 17 The primary motivating factor for these systems or 18 these approaches for linking models together is to allow for 19 independent models to be developed, linked and applied l 20 within a single modeling structure. t 21 It also allows for efficient development of future 22 i software, because I can pull out one model and plug in 23 { another. And it also facilitates this multi-tiered 24 assessment approach, one that's -- NRC, that's why we're all 25 here talking about it, and that is that we try to use a ANN RILEY & ASSOCIATES, LTD, Court Reporters 1025 Connecticut Avenue, NW, Suite 1014 Washington, D.C. 20036 (202) 842-0034
F .. 222 1 simple a system as possible that is scientifically () 2 3 defensible and get more complex as the situation requires. Finally, I'd like to note that all the work that's 4 being done is to' support the decision-making process. It's 5 so we can perform scientifically defensible assessments that 6 l are more useful, faster, cheaper, and more understandable to ! I 7 stakeholders and regulators and, more importantly, the 8 scientists and the engineers that seem to think they know 9 everything there is to know about these assessments, because 10 l a lot of times, they do not capture the essence of what the 11 stakeholders are looking for. ' 12 And I want to emphasize this is not academic. 13 1 That's why NRC is here and that's why these workshops are 14 being held. These are not academic exercises. () 15 Thank you. 16 MR. NICHOLSON: 1 Thank you very much, Gene. We 17 have time for just one or two very quick questions. Then 18 most of your questions or comments we will save for the 19 l group discussion at the end of the day. 20
- Are there any quick one or two clarifications for l
! 21 Gene? 22 [No response.] 23 MR. NICHOLSON: If net, we'll move on quickly to 24 the next phase of the program, which is basically we invited 25 industry to make a presentation on their experiences in () ( ANN RILEY & ASSOCIATES, LTD. Court Reporters 1025 Connecticut Avenue, NW, Suite 1014 ) Washington, D.C. 20036 j (202) 842-0034 3 1 1
l 223 1 moving from a site-specific -- to a site-specific dose 2 assessment modeling, Dr. Yim has prepared the presentation 3 for today, but he can't be here, so Carol Hornibrook from 4 the Electric Power Research Institute in Mountain View, 5 California is here to fill in for him. 6 MS. HORNIBROOK: Thanks. I just want to thank Tom 7 for inviting us to do this presentation. I thought I turned 8 the thing on, but maybe not. Do,7s it matter which this was 9 set on? 10 Good. Okay. Obviously, in our presentation, 11 we're not going to address conceptual modeling, because that 12 1 was done earlier, and done quite well, obviously. What 13 we're going to look at is nuclear utility perspectives on 14 site-specific ground water modeling, per se. So what we see ( 15 that applies to us and affects us.
}
16 And the approach that we're taking to this is a 17 couple of steps. One is to determine how important is 18 ground water modeling for a nuclear power plant's 19 decommissioning and their dose modeling, and there are two 20 things we want to look at that. One is the actual multiple 21 nuclide mix that we actually have at the sites. Then, two, 22 we also want to look at individual nuclide cases, because in 23 some instances, people are not finding all of the nuclides. 24 There are just a few that are really standing out. 25 The next thing, once we look at that, then we O ANN RILEY & ASSOCIATES, LTD. Court Reporters 1025 Connecticut Avenue, NW, Suite 1014 Washington, D.C. 20036 (202) 842-0034 I l 1
r,
- 1 1 224 wanted to figure out, okay, what's the impact of site 2
characteristics and here we took three reference sites, and 3 I will explain those iri more detail, but ostensibly what 4 they had to do was with the ability of the nuclides to get 5 to the ground water being high, medium or low. 6 Then once we had these two questions answered from 7 the first two bullets and if the answer was yes , ground 8 water did have an effect, 9 then what we wanted to do was look at what are the major parameters of importance in the ground 10 water modeling, with the use of the D&D and the RESRAD 11 codes. So we're only looking at these two codes in terms of 12 ground water impact. 13 And then do the default input values of these 14 parameters properly represent the industry? 15 Okay. So for the first question, how important is 16 ground water modeling for a nuclear power plant in 17 decommissioning, we did a test case problem and we picked 18 ten radionuclides and we selected these, they actually came 19 from two actual sites, neither site had all of these 20 nuclides in their data, but we thought it is representative 21 of what we find typically at plants. 22 1 But we also added Europium business Europium-152 i 1 23 has been found by Battelle in concrete. So we thought just 24 to be comprehensive and get a good understanding what would 25-1 really be the impacts, we would add that in. ( ANN RILEY & ASSOCIATES, LTD. g N Court Reporters 1025 Connecticut Avenue, NW, Suite 1014 Washington, D.C. 20036 (202) 842-0034 n
225 1 Unfortunately, RESRAD doesn't have Europium-152, 2 O 3 so we used Europium-154, which I think is fine. There is no real problem there. I 1 4 We took cobalt-60 and used it at one picocurie and 5 then we scaled the rest of these nuclides to that to come up 6 with our representative inventory, and we have a good handle l 7 on scaling factors for the industry. So I feel confident 8 that our mix is very representative. 9 The results of this -- I need to find my pen here. 10 l In the ten nuclide mixture, it kinds of shows something that l 11 ! made me feel good. Both in the D&D code and the RESRAD 12 code, the majority of the dose was from external. So that's 13 a good thing. I means that for our mix of radionuclides, we 14 really think on the average, from what we know so far, that 15 it's not going to be that difficult for us to have a good 16 outcome with both of these codes. 17 So that's for the ten nuclide mix. And so what we 18 did was we took this information and we ran a sensitivity 19 analysis on the multiple nuclides and we did it for each 20 code and came up with the following parameters as being the 21 key in this particular analysis. 22 And don't let it fool you, that it looks like 23 there's only four here, we did it this way so it would be l 24 encugh space. It's indoor time, indoor shielding. So 25 there's really six parameters here that are really key in O ANN RILEY & ASSOCIATES, LTD. Court Reporters 1025 Connecticut Avenue, NW, Suite 1014 Washington, D.C. 20036 (202) 842-0034
226 lL the D&D code and four in the RESRAD code. And as you can
] 2 see, none of them really have to do with ground water. So 3
ground water is not dominating our nuclide mix. 4 Now, when we started doing comparisons of the 5 RESRAD and the D&D code, the NRC, the industry, EPRI, we 6 found that there were some nuclides that you did get kind of 7 a real difference in your results. I know NRC is working 8 actively on that right now. The changes to the code are on 9 their way and not too far in the future. 10 But I wanted to show some examples, so you could 11 get an idea of what kind of comparison we did, and how this 12 kind of an analysis shows the impacts. 13 So we looked at strontium-90 and here what we 14 found was that the -- again, you don't see a ground water () 15 component. It's agricultural food, non-aqueous, also in 16 RESRAD, agricultural food, non-aqueous. 17 So the good news is there isn't a ground water 18 component that is showing up in these codes, not with any 19 kind of strong measure. 20 One thing I would point out, though, it doesn't 21 mean that the doses that come out of these two codes are the 22 same at this point. I'm sure when the correction is made, 23 there will be a difference. But right now, it's almost a 24 factor of ten difference. 25 Again, we did a sensitivity analysis to try to see O ANN RILEY & ASSOCIATES, LTD. Court Reporters 1025 Connecticut Avenue, NW, Suite 1014 Washington, D.C. 20036 (202) 842-0034
227 1 which parameters in the code were of importance, just to () i 2 confirm that the ground water was not an issue, and pretty ' 3 much that's what you see .; hen you look here. You don't see 4 ground water, the~ parameters that are important for ground i { 5 water popping out at you. 6 We also looked at cesium-137 and here we do see 7 something is happening. There is a significant difference ( 8 in where the two codes say that there is an impact. i 9 In D&D, it's from aquatic food and irrigated food. 10 In RESRAD, it's from external gamma. And I think most of us 11 would kind of more expect that it would be from external i I 12 gamma. So this does say that there is some difference going 13 on here. 14 And the implications, to us, is that perhaps the () 15 D&D code is, in its simplicity, may be a little too simple 16 and this is an artifact as to why you would see the 17 cesium-137 so high in the aquatic food. 18 We also did this same analysis on plutonium-239 19 and came out with similar results. So it's not just cesium 20 where that happens. 21 Similar results for both codes in terms of the 22 percentages. Again, with the sensitivity analysis, now you 23 start to s<ee ground water showing up, especially in the D&D 1 24 code. XD saturation ratio of the unsaturated zone, i H25 thickness of the unsaturated zone, density of the ! ANN RILEY & ASSOCIATES, LTD. Court Reporters 1 1025 Connecticut Avenue, NW, Suite 1014 Washington, D.C. 20036 (202) 842-0034
f t-l 228 1 unsaturated zone, infiltration rate, all of these are ground 2 water associated parameters. You don't see that really in ! 3 the RESRAD code. You see more of what would give you that 4 external gamma. 5 So to our way of thinking, cesium is something we 6 should be looking at. Here we're being impacted by a ground 7 water, the ground water models within the D&D code. So the 8 next thing, as I pointed out earlier, the second bullet was, 9 okay, how important, if that's the case, how important are 10 the site characteristics when you look at D&D versus RESRAD 11 in your comparison, t 12 i As 1 said earlier, there were three types of 13 reference sites, site A, low potential for ground water 14 contamination; site B, medium potential; and, site C, high 15 potential. This is based on an actual EPA report where they 16 wanted t a look at ground water issues and I really don't 17 want to discuss the numbers up here because it's from the 18 EPA. I didn't select them. We used them because it was 19 actual sites that they had this data on. So we thought it l 20 was kind of a good representation to try to get a handle on i 21 what these differences might mean. i 22 But I included it so you'd know what we did in our 23 analysis. And here, now I realize it's a little -- not 24 complicated -- confusing to look at this, but check it out. 25 If you look at site A, B and C, realizing this is ANN RILEY & ASSOCIATES, LTD.
'O Court Reporters 1025 Connecticut Avenue, NW, Suite 1014 Washington, D.C. 20036 (202) 842-0034
229 1 low, medium and high, if you look at the D&D code and look /N 2 N) at strontium-90, here you get 59 millirem per year, 59, 59. 3 That's pretty much what you would expect, because it's not 4 that water soluble. You don't expect it to be going in the 5 ground water. 6 Same thing for the RESRAD analysis. As I said 7 before, the numbers are quite different, but same thing. 8 You would not expect to see a difference in the ground 9 water. i I 10 i Let's pick two nuclides where you would expect to 11 see a difference. Carbon-14, Iodine-129, again, remember, l 12 l now we've just used one picocurie per gram. We're not using 13 the ratio that I used earlier with the ten nuclides from our i 14 site and trying to put those in some kind of proportion. 15 l [\_)) So this is much higher concentrations than we 1 l 16 would ever find at our site, especially for these two 17 nuclides. And if I could take a step back for a second, you 18 may have noticed that I didn't include Iodine-129 of we 19 didn't include it in our analysis of the ten nuclides and 20 i the reason is that our concentrations are so low that when 21 we've done the analysis, the actual dose resulting from 22 Iodine-129 is in line ten-to-the-minus-three. So we weren't 23 concerned that it was an actual contributor to dose, not of 24 any significance at all. 25 But you will see, with nuclides that are water (~} ANN RILEY & ASSOCIATES, LTD.
'x_/ Court Reporters 1025 Connecticut Avenue, NW, Suite 1014 Washington, D.C. 20036 (202) 842-0034
i 230 soluble, migrate kind of easily through the environment 2 . O (_,) The expected changes in terms of lower numbers 3 increasing numbers, increasing numbers, 4 same thing in the RESRAD. 5 There is a little difference here where it goes up a littl e higher in the median, 6 I'm not sure why that is, but you can still see that there is an increase, 7 because these are more likely to be in ground water. 8 But as I say, realize that these numbers are very high. 9 We'd never get these numbers at our site. It It would be much lower, way lower. 10 ' So what do we see from this? With cesium-137, 11 we 12 see that there is a factor of four difference if you 1ook at . the predicted peak dose and a low potential for ground water 13 contamination, with D&D. But 14 then when you get over here, also with median, 15 then when you get to high, it's a factor of four difference, 16 and that's not what we would really expect. 17 And in the RESRAD code, you don't actually see that and as I said before, 18 it's kind of a similar thing with the 19 plutonium-239 that you see an increase which you really wouldn't expect to see and that RESRAD doesn't show. 20 So now we want to look at the default parameters 21 and these are the same ones that I showed you before,only 22 we've just reordered them and the only difference in the 23 ordering is not any magic. 24 It's just the fact that when it came to these parameters, we wanted to order them in terms 25 of what information we actually had available to us and how ANN RILEY & ASSOCIATES, LTD. ( Court Reporters 1025 Connecticut Avenue, NW, Suite 1014 Washington, D.C. 20036 (202) 842-0034
1 231 much there was. 2 And actually thanks to Sandia and Walt in 3 particular, we were able to get quite a significant amount 4 of data on thickness of the unsaturated zone. There isn't a 5 lot of data for us in these two areas right now and we're 6 actively trying to collect KD information. 7 So my next slide will be on our analysis of the 8 thickness of the saturated zone. In a sense, this is kind 9 of my last slide. Walt was able to provide us with 211 10 independent measurements. Since that time, we've gotten 11 about five or six from the industry and what we did was a 12 13 data fit with a maximum likelihood estimation and this is kind of a typical statistical analysis that's done with 14 hydrogeologic data. What we did was a lognormal () 15 distribution and we came up with a geometric mean of 2.296, 16 and a geometric standard deviation of 1.265, and I realize 17 i that's a pretty significant size standard deviation, but we 18 feel fairly comfortable. 19 We also did a goodness of fit with these numbers 20 and though I'm not a statistician, what I'm told is this is 21 not a bad goodness of fit. The difference between those 22 numbers falls in a pretty good area. 23 Pow, what does this mean? For the 95 percentile, 24 what we would come up with 'ehen we look at the goodness of 25 fit from that data is a 1.24 meter would be the -- excuse me O ANN RILEY & ASSOCIATES, LTD. Court Reporters 1025 Connecticut Avenue, NW, Suite 1014 Washington, D.C. 20036 (202) 842-0034
I . 1 - 232 the parameter that we would get. At the 90 percentile,
'2 we would get almost two meters.
However, in the D&D default 3 right now, it's a 1.22 and I believe, am I right, that this 4 is at the 90th percentile or is at the 95th? 5 MR. BEYELER: That 6 90th percentile. is the default for doses at the 7 MS. HORNIBROOK: At the 90th percentile. So our 8 analysis is not a one-for-one with theirs, it's not too bad, 9 but it's not a one-for-one. But I say it's not too bad in 10 the sense that, remember, when I gave the original slides on 11 this, I also said that I really thought because of the 12 simplicity of the D&D code, that perhaps that was also 13 contributing to make it an artifact to see some of this 14 difference. () 15 So we haven't quite ironed that out yet. In fact, 16 we're still talking to Walt a little bit about it to see 17 what he thinks. 18 But so far, what you can see is for the nuclides 19 ! that we have on-site and the concentrations that we're 20 likely to have, for the most part, ground water is not -- we 21 don't see it as a big issue at this time. 22 When it comes to specific nuclides, I'm assuming l 23 that once some of these corrections get made to the code , 24 that NRC is working on, they'll probably fall a lot more in 25 line with what we're expecting from these kinds of analyses. O ANN RILEY & ASSOCIATES, LTD. Court Reporters ' 1025 Connecticut Avenue, NW, Suite 1014 Washington, D.C. 20036 (202) 842-0034
233 1 Thank you. 2 MR. NICHOLSON: Thank you very much, Carol. We 3 would like to have a few quick quest. ions or comments. 4 MS. HORNIBROOK: Easy questions. 5 MR. NICHOLSON- Easy questions, be nice to her. 6 Boby? 7 MR. EID: Yes, first of all, I would like to 8 commend you on this presentation and the comparison, it is 9 really good work, and I would like just to explain one 10 thing. That the data that you showed, I believe you did use 11 D&D code version 1.0, which is on the web site. 12 MS. HORNIBROOK: Yes, right. 13 MR. EID: Now, we have looked into that and we 14 gave presentations before about the cesium and the strontium () 15 16 and we said that the default parameters in those were used because of the current methodology of selecting the solution 17 vector, because you assume that you have a solution vector 18 for all radionuclides, that they are there. 19 i And I'll guarantee that all the radionuclides they 20 have, the dose distribution above 90th percentile. We have 21 looked at the single radionuclides and to have the i 22 parameters that they are consistent with the single 23 radionuclides and looked at the peak dose, and those numbers 24 are significantly different. 25-So do not be discouraged from these values in the ANN RILEY & ASSOCIATES, LTD.
-O Court Reporters 1025 Connecticut Avenue, NW, Suite 1014 Washington, D.C. 20036 i (202) 842-0034 L___
234 1 D&D code. However, we have numbers using single default 2 values, using default values for single radionuclides, and 3 those they are somehow comparable with RESRAD results. 4 And we will soon, as I said this morning, be 5 publishing in the Federal Register Notice, default table for 6 soil, for alpha, beta and gamma for common radionuclides. 7 Those default tables, many of these radionuclides, 8 cesium and strontium, uranium, plutonium, they changed. So 9 I'd like just to bring your attention, so just look for this 10 comparison and there will be a more valid comparison. 11 MS. HORNIBROOK: Okay. I was trying to give you 12 credit, but I didn't know how much to actually say. Thank 13 you. Are there any other questions? Yes? 14 MR. YU: Charlie Yu. () 15 16 MS. HORNIBROOK: Hi Charlie. MR. YU: Argonne National Laboratory. Carol, I 17 may have missed your explanation on the last slides you 18 showed. 19 MS. HORNIBROOK: You probably didn't. 20 MR. YU: Geometric mean is 2.296 meter. 21 MS. HORNIBROOK: Hold on. Sorry, I don't have it 22 right in front of me. Has anybody got a copy of the 23 presentation? Okay. Yes, 2.296. 24 MR. YU: And your 95 percentile is 1.24 and 90 25 percentile is even higher. Can you explain that a little O ANN RILEY & ASSOCIATES, LTD. Court Reporters 1025 Connecticut Avenue, NW, Suite 1014 Washington, D.C. 20036 (202) 842-0034 ( .
r 235 1 bit? Is that what you get from -- 2 MS. HORNIBROOK: That's a Man-Sung Yim answer. I 3 can't honestly give you that answer. I'm sorry, Charlie. 4 But I can have him call you. He's in Korea right now, but 5 when he gets back, I'll have him give you a buzz. I'm sorry 6 about that. Any other questions? Oh, dear, another one. 7 MR. NICHOLSON: Gene has a comment. 8 MR. WHELAN: This is Gene Whelan, PNNL. I just 9 have a quick question, it's an easy one. In general, we see 10 that the D&D results tend to be, at least from what I saw, 11 tend to be more conservative. 12 Excluding the magnitude of the results and from 13 your experience of doing the comparison between these two 14 models, would you see significant differences in the () 15 16 decisions'as opposed to the numbers? MS. HORNIBROOK: Right now, we look at the numbers 17 because of the regulations. 18 MR. WHELAN: But you make a decision based on 19 those numbers. Would you have different decisions? 20 MS. HORNIBROOK: I would think so, sure. Yes. I 21 mean, at this point, fortunately, one of the decisions could 22 be to use another code and once you've explained why that 23 code is appropriate. But after Boby does the improvement, 24 then I'm anticipating that it will work out. Not work out, 25 I don't know what the right term to use is. It would be .( ANN RILEY & ASSOCIATES, LTD. Court Reporters 1025 Connecticut Avenue, NW, Suite 1014 Washington, D.C. 20036 (202) 842-0034
F . 236 1 good representatio'n of what's there. Reconsidered. () 3 2 MR. ROBERTS: A question. Rick Roberts, Rocky Mountain Remediation Services. 4 On your basic parameter list, you talk about you 5 have a lot of different parameters on saturated zone 6 thickness, KSAT, B parameter, you go through for reach of --
.7 of each of the sites.
8 And it seems you've got your unsaturated zone 9 thickness for site A is 45 meters. 10 MS. HORNIBROOK: This is EPA data. It's straight 11 out of an EPA report. I just picked it because it was based 12 on real sites and I just wanted you to know what I used. I 13 really don't want to try to defend it because it's their 14 stuff. () 15 MR. ROBERTS: Okay. That's fine. What I'm 16 wondering is, on your last slide, you say that the thickness 17 of unsaturated zone should be about 1.2 meters or 1.9 or 18 less than two meters. 19 MS. HORNIBROOK: All I'm saying is that when you ! 20 take the -- two meters, ! 21 MR. EID: i I believe she wanted to say that the { 22 unsaturated zone is thicker than D&D code assumes and the 23 assumptions in D&D code is conservative because the thicker l 24 the unsaturated zone, the more you have retardation and the 25 less that you will have concentration in the aquifer. ! O id s ANN RILEY & ASSOCIATES, LTD. Court Reporters 1025 Connecticut Avenue, NW, Suite 1014 Washington, D.C. 20036 (202) 842-0034
= 237 1 So she is proposing to you thicker unsaturated () 2 3 zone, two meters, based on her analysis, which is the 90th percentile. l 4 l MS. HORNIBROOK: If you look at the cesium dose, { 5 it goes from 28 to 18, when you use these numbers between l 6 1.2 and 1.9, which is two meters. 7 MR. ROBERTS: I guess my question was if you did 8 an analysis, and I'm not sure who did -- I guess EPA did it, 9 but it seems like if you reduce the unsaturated zone to less 10 than two meters, that your dose assessment from your three 11 sites would be much different and your pathway analysis 12 would show much different results. 13 Is that -- you're shaking your head. Have I 14 missed that point? () 15 MR. EID: I agree with you. You are absolutely l 16 right. Depending on the pathways, if you have the direct 17 exposure pathway, it's more significant. This means if you 18 have more leaching of the material, this means you are going 19 to be more conservative if you have thinner unsaturated 20 zone. 21 MR. ROBERTS: Okay. 22 MR. EID: So you are right. Depending on the 23 pathways. Whereas if the ground water pathway is 24 significant, it will be the other way around. 25 MR. ROBERTS: So up front, what pathways are most i 1 O ANN RILEY & ASSOCIATES, LTD. Court Reporters 1025 Connecticut Avenue, NW, Suite 1014 Washington, D.C. 20036 (202) 842-0034
238 1 important, those could really change drastically given a 2 much smaller unsaturated zone. 3 MS. HORNIBROOK: Right, yes. 4 MR. ROBERTS: Thank you. 5 MS. HORNIBROOK: Thank you. 6 MR. NICHOLSON: Thank you very much, Carol. I 7 would like to now introduce Dr. Ralph Cady, who works in the 8 branch. He is both a hydrogeologist and a performance 9 assessment specialist. He will be leading a group 10 discussion on experiences in integrating site-specific 11 ground water modeling into dose assessment and the ground 12 water pathway. 13 MR. CADY: You must have stayed around either to 14 see what I J.cok like or to hear Carol's talk. So I won't 15 presume what the answer is.
)
16 And remember that I have to answer all these 17 questions or I don't go home tomorrow. We can worry about 18 some of these tomorrow, but it would be nice to be able to 19 at least step through a fair number of them. 20 Primarily what we're talking about here is 21 mechanics. It's writing code to essentially link the 22 sophisticated or more sophisticated ground water models into 23 a more traditional dose assessment code. 24 The first question really gets at the mechanics, 25 whether it's a ground water model or the radionuclide t l O ANN RILEY & ASSOCIATES, LTD. Court Reporters 1025 Connecticut Avenue, NW, Suite 1014 Washington, D.C. 20036 ( i (202) 842-0034 I
1 239 concentrations from monitoring. 2 3 We went around the table of the models that are being discussed and there are some that can't handle any of 4 these monitoring data whatsoever, and Gene proposed an 5 engineering approach to incorporating a plume into MEPAS, I 6 would assume within the FRAMES approach, there's a perhaps 7 even more elegant way. 8 MR. WHELAN: There is a model there. 9 MR. CADY: There is a model there, good. All 10 right. In that case, you're talking. What's your answer? 11 MR. WHELAN: We've actually done a lot of linkage 12 of models and I would say that if people want to link a 13 different model in either with D&D or with RESRAD, et 14 cetera, I would say take one of two approaches and, again, rN 15 U i SEDSS may fit into one of these approaches, because the 16 information I heard today is slightly different from what I 17 have been hearing in the past. 18 So pardon me if I got SEDSS wrong. j 19 MR. CADY: I've got a thick skin, don't worry. 20 MR. WHELAN: 21 And that is that I would either do it L the way that we have done it with the plug-and-play
- 22 approach, and that is that we have run more sophisticated 23 models, quote-unquote, outside the system and then actually 24 when you go to run your dose assessment model, you just pull 25 that file in and it's just like that file was created by a t
7- ANN RILEY & ASSOCIATES, LTD. N Court Reporters l. 1025 Connecticut Avenue, NW, Suite 1014 ' Washington, D.C. 20036 (202) 842-0034 I I
r -- . _ , I 240 1 model inside the system. That's one way of doing it.
'2 The second way of doing it is literally taking 3
that model and I'll say hardwiring it into the system. I 4 don't think it would fit very well in the master file 5 approach. So I would say one of those two approaches would 6 probably be the best way to go. 7 And if you're going to change models a lot, maybe 8 the first one, that is this plug-and-play approach. 9 MR. CADY: You've got to take your output from 10 your source term and get that into your transport model or 11 in the case of infiltration, get it into your flow code. So 12 you've got stuff to do at the front end as well as the back i 13 end of this external model if it's outside the framework 1 14 construct. s 15 1 MR. WHELAN: That's correct. And that can be done 16 within the framework construct. Again, like I said, if 17 you're going -- if you're not going to change that model, it l 18 may be just as easy to hardwire that model into your system.
- 19 If you're going to change it, you may want to reconsider 20 doing that, but it's either way would be acceptable, l
, 21 i depending upon, again, down the road, how many changes and { 22 modifications you expect. 23 MR. CADY: How about number two, for the external. I 24 You've got existing data. 25 MR. WHELAN: If you've got existing data, then, i O ANN RILEY & ASSOCIATES, LTD. Court Reporters 1025 Connecticut Avenue, NW, Suite 1014 Washington, D.C. 20036 (202) 842-0034
r . , l 241 1 l again, you can do it one of -- with either way. Right now, 2 with the plug-and-play, for example, you would actually be 3 reading in a file that contains that information. So it's 4 l just like somebody ran a ground water model and produced 5 that file. 6 The flipside to that is you would have to build, 7 for this hardwired system, you would actually have to build 8 a model or a processor that allows you take in this 9 monitoring information and use that directly, skipping, if 10 you will, the transport components. 11 MR. CADY: Another little issue -- you might as 12 well just keep that down. Shlomo acknowledged a sort of 13 nauseous feeling with a lot of these codes because he , 14 couldn't look in and see what the concentration is at a 15 point. 16 I noticed, I believe -- I believe it was in one of 17 the viewgraphs. Well, you also said the same thing. You 18 didn't quite get the nauseous part in, but that you would 19 like to be able to look in and see how your monitored data 20 relate to a simulated data point. 1 21 i MR. WHELAN: I'm not helping you on this one. 22 MR. CADY: Oh, I'll get the question. It may take 23 me a while, but I'll get the question. In what I saw you l 24 present for the FRAMES approach, I didn't see that ability 25 to, okay, tell me what the concentration is at this point. ANN RILEY & ASSOCIATES, LTD. i I Court Reporters 1025 Connecticut Avenue, NW, Suite 1014 Washington, D.C. 20036 (202) 842-0034 I
242 1 I mean, I would assume that that would be something that may
/~N 2
'd. be in there or you certainly would like to have that ability 3 to look at internal variables that typically, in a lot of 4 these codes, you don't get to see. 5 MR. WHELAN: I do have an answer for that. 6 MR. CADY: I know. 7 MR. WHELAN: If we think,back to what I had 8 mentioned, ! I noted that at least in this plug-and-play, I 9 that's the FRAMES concept that Dr. Cady is referring to, in 10 this plug-and-play environment, how that plug-and-play 11 environment is structured is that the models are adjusted to 12 meet the specifications of the system. 13 In other words, this telephone book exists as to 14
'if you run a ground water model, there is a certain minimum
() 15 amount of information you must produce in order to 16 communicate within the system. You can produce more, you 17 can produce intermediate files, you can produce any extra 18 stuff you want. But as a minimum, you must produce -- meet 4 19 the specifications of the system. 20 Therefore, if you want to look at these internal i 21 numbers, if you will, of this model, this plug and play 22 environment allows you to do that. It's just that when yon 23 have other models that come in which were not developed by 24 the same person, they don't know those temporary or 25 intermediate files exist. O, ANN RILEY & ASSOCIATES, LTD. Court Reporters 1025 Connecticut Avenue, NW, Suite 1014 Washington, D.C. 20036 (202) 842-0034
f 1 243 Now, if you wrote both models, you can also read
,s 2 \l those temporary files yourself if you would like. So the 3
system does not preclude you from generating more 4 information for you to inspect and view and visualize . It 5 just tells you.the minimum amount you have to produce . 6 MR. CADY: Okay. Trust me, if I wrote both those 7 models, I'd probably shoot myself. 8 Walt didn't have a great opportunity to go through 9 a lot of the mechanisms, but I know that there are elements 10 of the design in the new version that "n're putting together 11 for SEDSS that address these issues. 12 Is there anything that you can add to the 13 discussion? 14 MR. BEYELER: No. I think it is more of an ( V) 15 object-oriented design and I appreciate that t
- 16. acknowledgement.
I think the old version of the code did, 17 in fact, every combination had to be explicitly constructed. 1B We're endeavoring to get around that. 19 I guess the only point I'd make is that it seems 20 that this is maybe the easier end of the question; that is , 21 if we're interested in going from a complex representation 22 of the ground water system to a dose assessment it's simply 23 a matter of integrating the concentration over the volume, 24 sucking it up in the well and coming up with some average 25 number. ANN RILEY & ASSOCIATES, LTD. Cs
- Court Reporters 1025 Connecticut Avenue, NW, Suite 1014 Washington, D.C. 20036 (202) 842-0034
244 1 I think maybe the more complicated question is th e 2
-')_ other end, going from a simplified representation, how is G
3 that specified in a more complicated ground water system 4 and, more specifically, how do you specify the many 5 alternatives, because you're going from a case where you 6 have one number for infiltration, say, into the aquifer 7 system, what are the possible ways that that might. be 8 distributed spatially given that you're representing a 9 higher degree of spatial in the aquifer. 10 I think that's -- it's a methodological question. 11 MR. CADY: Maybe it's time to move on, because I 12 don't see too many other volunteers here. 13 MR. POTTER: I would -- 14 MR. NICHOLSON: Would you identify yourself? ['5n 25 MR. POTTER: LJ I'm sorry, I keep doing that. Tom 16 Potter. Until we have RAD Windows, which apparently will be 17 available pretty soon, there's -- I have had great success 18 with RESRAD, which, in its simplicity, allows a lot of 19 flexibility. 20 There are two files, the detailed report file and 21 the concentration output file, concentration includes 22 concentration in water as a function of time, and it is 23 quite simple to manipulate data to produce whatever ground 24 water concentrations you want in a way that is conservative; 25 that is to say, you don't deplete your source too quickly, f w) ( j ANN RILEY & ASSOCIdTES, LTD. Court Reporters 1025 Connecticut Avenue, NW, Suite 1014 Washington, D.C. 20036 (202) 842-0034
245 1 things like that. 2 So if you have a sophisticated ground water 3 program that's fun off-line, produce concentrations as a 4 function of time, I can assure you it's a pretty simple 5 matter to make RESRAD produce those concentrations 6 artificially and integrate the results of the complex ground 7 water model. 8 MR. CADY: I think Gene also mentioned that the 9 ability of a knowledgeable user to tweak the system 10 considerably, you've got a vision of what the 11 conceptualization is, but by an appropriate choice of 12 parameters, it's amazing the amount of trouble you can get 13 in or the problems that you can attempt to solve. 14 All right. Let me look at these two, three and ( 15 four. I guess we did try to put them in some sort of order. 16 So let's try to address number two. Gene, how did you go 17 about selecting the particular code that's within the 18 existing MEPAS as well as the current and how about 19 potential futures for FRAMES? 20 MR. WHELAN: I'm going to start backwards, I'm 21 going to start with FRAMES, and that is that the models that 22 were in P"?MES are EPA models, MMSOILS, DOE models, 23 basica 'y '
. PAS and GENII, which is an EPA model now, I 24 guess.
A.1d those really are the first proof of principal 25 models to show that FRAMES is a very viable platform for ANN RILEY & ASSOCIATES, LTD. 0.- Court Reporters 1025 Connecticut Avenue, NW, Suite 1014 Washington, D.C. 20036 (202) 842-0034
246 1 integrating different models. 4 2 And DOD, at WES, has just incorporated in the HELP 3 model'and the RECOVERY model. In the RECOVERY model, they 4 have asked us for no help whatsoever, they just put it in. 5 So it's not a real -- I mean, they have real smart people, 6 but it's not as difficult as it might seem putting a model 7 in to communicate with these other models. 8 In terms of why we chose the models we chose for 9 MEPAS, and RESRAD may be a similar thing, back in '84 and 10 beyond, and even today this happens, if you look at the type 11 of models and assessments and the data that are available at 12 Superfund sites, DOE sites, RCRA sites, et cetera, many 13 times what you end up with is assuming homogeneous iso -- in 14 other words, the data back then were not necessarily there () 15 16 to warrant a more complex model, number one. Number two, , I the computing power was not really there to run on your PC, 17 if you will, running a three-dimensional finite element 18 model, plus the data generally weren't there either, 19 So it specifically met the needs that the 20 regulatory industry has and the regulatory industry is still 21 using that class of model for comparative assessment. So it 22 isn't like we've graduated to just using numerical models , 23 now in all of our assessments. That's not the case. 24 In addition, what's nice about these 25 semi-analytical models is that they actually do fit a nice r~ ( ,g) ANN RILEY & ASSOCIATES, LTD. Court Reporters 1025 Connecticut Avenue, NW, Suite 1014 Washington, D.C. 20036 (202) 842-0034
247 1 niche. Being able to go in and do a very quick preliminary () 2 3 assessment to get a feeling for what type of problems I might encounter, so I can now focus my resources on the more 4 detailed assessments, I know what questions to start asking 5 for those detailed assessments, and, in fact, the 6 preliminary analysis may provide me with the same decision 7 that I would end up with a more detailed analysis. 8 Which isn't to say that the numbers are right or 9 the numbers are better, but the results are such that, say, 10 I get de minimus results and I know that the concentration 11 is going to drop if I run a more detailed code. Therefore, 12 I can show that to regulators and they will say, fine, all 13 you have to do is monitoring or whatever. 14 So that is one or the main reasons why we chose () 15 the models that we did for MEPAS. It's because they were 16 very consistent in the regulatory environment. 17 MR. CADY: I don't know whether Walt, whether you 18 want to address D&D as well as SEDSS, but you have the 19 opportunity. 20 MR. BEYELER: As far as why those particular 21 models were included, I would say to meet customer 22 requirements. 23 MR. CADY: Well, in that case, I'll fill in. In 24 the D&D ground water model, it's really almost a misnomer to 25 call it a ground water model. It's a bucket and you fill ON ANN RILEY & ASSOCIATES, LTD. Court Reporters 1025 Connecticut Avenue, NW, Suite 1014 Washington, D.C. 20036 (202) 842-0034 _ _ _ --- - - - f
f~ 248 1 that bucket with enough water to meet the needs of whatever 2 the scenario is, how many liters per day drinking water, as 3 ( well as the irrigation and livestock requirements, that sort ) 4 of thing, and whatever.comes out of this unsaturated zone l 5 delay function plops into this bucket and you drink it. 6 I So there is not -- Walt alluded to the lack of 7 data in the D&D model as far as the hydraulic conductivities 8 and a lot of these other things that appear in many 9 different models. There is no -- within the aquifer, that 10 sort of thing. 11 So that model was chosen explicitly to be about as 12 conservative as people can think and for use purely as ! 13 screening. 14 In the SEDSS construct, we've chosen NEFTRAN as 15 the first candidate to go into the flow and transport i 16 portion of SEDSS for ground water and NEFTRAN was developed 17 by Sandia, so that helps pinpoint one reason why it was a 18 candidate. 19 The other reason was its ability to handle decay 20 chains, very long decay chains, and NEFTRAN handles them 21 quite well. So that's really my basis or at least my 22 understanding for why it was incorporated. 23 MR. BEYELER: Yes. I think that's true. I'd just 24 add that, again, in the case of SEDSS, the specification of 25 the model for ground water, it's more done in terms of the l ANN RILEY & ASSOCIATES, LTD. fO Court Reporters 1025 Connecticut Avenue, NW, Suite 1014 Washington, D.C. 20036 (202) 842-0034
I 249 l 1 assumptions that are made about the ground water system. I 2 And one of the key. assumptions is clearly the dimensionality 3 of the aquifer. 4
.So from that standpoint the D&D model to NEFTRAN 5
or more sophisticated models are sort of a natural 6 progression to, in effect, a zero dimensional model at the 7 aquifer that D&D represents to the one-dimensional model 8 that's represented in NEFTRAN and later models that are 9 three-dimensional. 10 MR. CADY: Okay. I'd like to take a couple
'11 seconds just to address a balloon that Shlomo lofted before 12 he left, and that was that we currently have all this 13 computational power, why are we stuck with these simplistic 14 codes, like can we do better, shouldn't we do better.
15 How about can we or do we have plans to? I think 16-historically, there has been an interest, at least from the 17 folks at EPA that we deal with on SEDSS, to consider 18 strongly pulling in at least 2-D ground water to SEDSS. 19 MR. BEYELER: Yes. I think -- this is Walt 20 Beyeler from SAndia. I think clearly the capability needs 21 to be there. I think there is a still a question of 22 estimation of parameters for the more complicated model. 23 There are aspects of the uncertainty that you can get around 24 by simulation, for example, but it sort of raises the higher 25 order questions of what is my barrier and how do I deal with O ANN RILEY & ASSOCIATES, LTD. Court Reporters 1025 Connecticut Avenue, NW, Suite 1014 1 Washington, D.C. 20036 (202) 842-0034
1 250 uncertainty barriers. 2 So there are, I think, it's important to have the analytical capability to do those sorts of (~ 3 simulations and it's important to make that capability 4 easily available from a variety of dose assessment 5 standpoints, to solve the mechanics of the integration 6 problem. 7 8 I think it's also very important to think through systematically how each of those complications is best 9 parameterized, where the parameter values are best 10 e 11 established for each of those various degrees of complexity in a way that's consistent with specific assessment and dose 12 concentration. 13 MR. CADY: 14 Has there been any sort of interest at DOE or EPA? 15 MR. WHELAN: The reason why we have FRAMES was 16 just for that fact that we get away from is my model better 17 than your model, is this model better than that model, and 18 19 we let the users themselves make the decision as to what's most appropriate for their analyses. 20 I will say this, that -- and I've developed 21 numerical models, as well as analytical and semi-analytical 22 , and that is people tend to use the models they're most 23 familiar with to start off with. 24 Second, when you do preliminary assessments, you 25 really are not looking at detailed mechanistic aspects ANN RILEY & ASSOCIATES, LTD. Os Court Reporters 1025 Connecticut Avenue, NW, Suite 1014 Washington, D.C. 20036 1 (202) 842-0034
251 1 associated with why contaminants move. All you want to be 2 able to do is to be able to try to capture the essence of a 3 contaminant plume as it passes, for example, a monitoring 4 well, et cetera. That's why I say if you have calibrated ) 5 i information, many times the semi-analytical models, which i 6 { are analytical and numerical solution, they contain both, 7 you can calibrate those models to that monitored information 8 and then to a certain degree, and it's not very far, you can 9 extrapolate to get some idea as to the ramifications 10 associated with the analysis. 11 And what's nice about that is the analytical and 12 semi-analytical models, you can get up and running and 13 calibrate this thing and in a week you have numbers. 14 Numerical models, although with the computing , () 15 power we have, it makes a great access, they are a little 16 more difficult, as we all know, to operate. You've got 17 convergence problems, stability issues you've got to deal 18 with, you've got to put more data in it, et cetera. 19 The more simplified model does not take the place ' 20 of the more complex models. They actually should compliment i 21 the two. And I firmly believe, after doing the site 22 assessments for years and years and years, that a tiered 23 approach is going to be the most cost-effective approach. 24 Now, we have to be very, very careful, though, and 25 that is that if your answer that you're looking for is yes /7 ANN RILEY & ASSOCIATES, LTD. (_s/ Court Reporters 1025 Connecticut Avenue, NW, Suite 1014 Washington, D.C. 20036 (202) 842-0034
..a l
252 1 or no, maybe you don't need the right numbers to come out () 2 order for you to come up with the right decision. in
)
3 But if you are interested in the right numbers, 4 for example, if your value is above this by so many, I pay 5 you a million dollars in restitution, if it's four times 6 higher than that, I pay you twice as much in restitution, 7 then it's very important to make sure you get the right B numbers, if you know what I mean. 9 So sometimes you don't need the right numbers to 10 make the right decision. That's why I asked the question 11 earlier, but sometimes you do. So it's based on the 12 question-that you're asking as to the type of assessment. 13 The other thing is that the simplified models try 14 to capture, in general, the essence of what's going on in () 15 the system. It's an average. Like when you do estuary 16 modeling, you use the average velocity of the tide coming in 17 and the tide going out. You don't necessarily have to model 18 all the tidal effects in order to come up with an idea that I 19 the contaminants are leaving the river and going into the 20 estuary. 21 MR. CADY: We have a question. I 22 MR. LEE: Sam Lee, from USDOE. a Just for the 23 comment. From this afternoon, based on Dr. Neuman's 24 presentation, I totally agree, we have to have a 25 three-dimensional model to try to catch all the complicated L O ANN RILEY & ASSOCIATES, LTD. Court Reporters ! 1025 Connecticut Avenue, NW, Suite 1014 Washington, D.C. 20036 (202) 842-0034
r- 4 253 1 features of underground flow and transport. No question 2 about it. 3 It's a complex situation. It's happening 4 underground. The question is now, refer to your rough 5 equations, how, within this kind of computer power, can we 6 still we have to use this complex model or not. I think so, 7 because the question now is how can we select a complex ! 8 model to be used. Now the question is, it's important, if ! 9 we do not have enough measurement data to select a complex 10 model to be used. 11 So that's why I suggest we have to select at the 12 i site where you can have a very good measurement data, a 13 variable, to be used to be selected where is -- which 14 complex model can be used, then convert or compare, use all () 15 those measurement data are variable, then compare it with 16 the different complex model. Then we can choose wnich 17 complex model to be used. 18 So that's why my suggestion is we have to select a 19 site where it can have a very sufficient data can be used, 20 that then we apply, try to use complex model. Tomorrow I 21 have a presentation to talk about another complex model and 22 I'd like to present it to the audience and see how the 23 comment. 24 MR. CADY: Well, we invite you all back tomorrow, 25 then. p ANN RILEY & ASSOCIATES, LTD. (_j Court Reporters l 1025 Connecticut Avenue, NW, Suite 1014 l Washington, D.C. 20036 (202) 842-0034
254 1 MR. NICHOLSON: Are there any other comments? s 2 '(- ) We'll save a lot of this discussion for tomorrow, also. 3 Mark, did you have a comment? 4 MR. THAGGARD: Yes. I just wanted to make a point 5 of clarification. 6 I think what Dr. Neuman said was not that you had to do a'three-dimensional ground water analysis, but 7 that you need to develop a three-dimensional ground water 8 conceptut4 model and that's not quite the same. 9 So I just want 1 10 to make sure we didn't walk out of here with the wrong information. 11 { MR. NICHOLSON: 12 I think also the issue is that we don't need to do complex models simply for the reason of 13 complexity, that there has to be a very legitimate and 14 relevant need to do it, 15 anc I think we can talk about that tomorrow and that's a good topic. 16 We have two very good presentations tomorrow, real 17 world dose assessments at real sites, 18 and I think that will be a very interesting point to bring up. i 19
-I want to thank all of you for staying here late.
20 I'm sorry we went over past five, but obviously there was an l 21 ! interest. We'll start promptly at 8:30 tomorrow morning. 22 Those of you who want to say something in the afternoon, 23 talk to Paul Genoa and other people, please contact Dr. 24 Ralph Cady, the last group discussion leader, if you want to 25 be'put on the agenda to make a few comments tomorrow i l
' ANN RILEY & ASSOCIATES, LTD.
Court Reporters 1025 Connecticut Avenue, NW, Suite 1014 Washington, D.C. 20036 (202) 842-0034
255 1 afternoon. t { ~2 Thank you. 3 [Whereupon, at 5:25 p.m., the workshop was ! 4 recessed, to reconvene at 8:30 a.m., Thursday, June 24, 5 1999.] 6 7 8
- 9 i
10 1 11 12 i 13 14 l 15 ' 16 17 18 19 20 21 22 23 24 25 O ANN RILEY & ASSOCIATES, LTD. Court Reporters 1025' Connecticut Avenue, NW, Suite 1014 Washington, D.C. 20036 (202) 842-0034
REPORTER'S CERTIFICATE This is to certify that the attached proceedings
/~T ..b before the United States Nuclear Regulatory Commission in the matter of:
NAME OF PROCEEDING: DECOMMISSIONING WORKSHOP CASE NUMBER: PLh'.E OF PROCEEDING: Rockville, MD were held as herein appears, and that this is the original transcript thereof for the file of the United States Nuclear Regulatory Commission taken by me and thereafter reduced to typewriting by me or under the direction of the court reporting company, and that the transcript is a true and accurate record of the foregoing proceedings.
- 8 f\ 0 '
y ' p Jon;Hundley Official Reporter Ann Riley & Associates, Ltd. O no- -
r l AGENDA Q WORKSHOP ON GROUND-WATER MODELING 1 Q RELATED TO DOSE ASSESSMENTS OBJECTIVES: Discuss ground-water modeling used in dose assessments for demonstrating compliance with the radiological cnteria for license termination, and examine technical bases for resolving decommissioning issues. TIME: 8:30 a.m. to 5:00 p.m., Wednesday, June 23,1999 8:30 a.m. to 5:00 p.m., Thursday, June 24,1999 LOCATION: NRC Headquarters Auditorium,11545 Rockville Pike, Rockville, Maryland June 23 8:30 a.m. Welcome and Introductions - NMSS and RES Management 8:35 Meeting Objectives and Review of Agenda - Tom Nicholson, NRC Staff Moderator 8:45 Ground-Water Modeling Issues in Dose Assessment of Decommissioning Sites - Critical Group Receptor, Scenarios & Site-Specific Conditions - g d Boby Abu-Eid, NMSS G 9:15 Decommissioning Decision Framework Discussion on Ground Water - Mark Thaggard, NMSS I 9:40 Overview Discussion of Conceptual Ground-Water Models used in DandD, RESRAD, MEPAS and PRESTO Codes and Information Needs for Each - Philip Meyer, PNNL 10:10 BREAK 10:25 Conceptualization and Calculation of Well-Water Contaminant Concentrations for a Given Plume Cheng Hung, US EPA 10:50 Ground-Water and Radionuclide Transport Model Used in RESRAD-OFFSITE Charley Yu and Emmanuel Gnanapragasam, ANL 11:20 Group Discussion on Ground-Water Conceptual Models & Scenarios - Mark Thaggard, NRC Staff Moderator 12:00 LUNCH l 1
PUBLIC WORKSHOP AGENDA June 23 1:00 p.m. Publicly Available Data Sources and Databases for Estimating Ground-Water Flow and Transport Parameters - Walt Beyeler, SNL & Glendon Gec, PNNL 2:00 Group Discussion on Ground-Water Parameter Estimation from Publicly Available Sources - Tom Nicholson, NRC Staff Moderator 2:20 Methodology for Site-Specific Ground-Water Modeling - Shlomo P. Neuman, U. of Arizona 3:00 BREAK 3:15 Integrating Site Specific Ground-Water Modeling into Dose Assessments - Walt Beyeler, SNL & Gene Whelan, PNNL 4:15 Experiences in Site-Specific Ground-Water Modeling Related to Dose Assessments from the Industry: Dr. Man-Sung Yim, NC State (EPRI) & Other Volunteers from Industry j 4:40 Group Discussion on Experiences in integrating Site Specific Ground-Water l Modeling into Dose Assessments, and Ground-Water Pathway Analyses - Ralph Cady, NRC Staff Moderator I June 24 ' 8:30 a.m. Review Agenda and Announcements - Tom Nicholson, NRC Staff Moderator 8:35 Estimating Ground-Water Parameters for Site-Specific Modeling Considering Uncertainty and Selection of Modelinput Value(s) - Phil Meyer, PNNL 9:30 Group Discussion on Parameter Estimation for Site Specific Modeling - I Mark Thaggard, NRC Staff Moderator 10:10 BREAK 10:30 Experiences in Model and Performance Confirmation and Potential Application to Decommissioning Sites - Ground-Water Modeling Studies of the Characterization of Uranium ( Contamination at Fernald Hsi-Na (Sam) Lee, DOE /EML l 9 l 2 l
PUBLIC WORKSHOP AGENDA June 24
'O 11:15 Ground-Water Modeling in Support of Doso Assessment for the Nevada Test i Site - Underground Test Area Project Ken Rehfeldt and Barbara Deshler, HSI GEOTRANS/ IT Corp.
1 l 12:00 LUNCH 1:00 p.m. Group Discussion on Ground-Water Modeling Related to Dose Assessments - Ralph Cady, NRC Staff Moderator 3:00 BREAK I 3:20 Group Discussion on Remaining Questions Not Previously Discussed Ralph Cady, NRC Staff Moderator 4:00 Closing Remarks - NMSS and RES Management 5:00 ADJOURN l l Agenda Notes:
- 1. Topics and speakers identified are tentative and subject to change.
- 2. The term " site-specific" as used in this agenda means an actuallocation where information and/or data about a ground-water system at that location exists or can be determined using state-of-the-science methods.
- 3. Group Discussions are an integral part of the workshop which will be structured around key technical issues, fol;owing the presentations, and will be open to all attendees.
l l Group Discussion Objective: l l The objectives of the group discussions are to provide the NRC staff with valuable comments i from the principal stakeholders following overview presentations by NRC staff and contractors, and specially-prepared presentations by volunteers from industry, and to determine where technical information is needed to facilitate implementation of NRC staff guidance on site- ; specific modeling. l
Background:
, The guidance being developed by the NRC staff focuses on implementation issues and review procedures for evaluating conceptual models and parameter estimation associated with ground- I water pathway modeling beyond simple screening models. The goal of the guidance is to ensure that NRC licensing decisions involving site-specific modeling are consistent, and to allow for a seamiess transition from screening to site-specific modeling. T 3 t
PUBLIC WORKSHOP AGENDA Organization of Group Discussion: The discussion will focus on the underlying assumptions and justification required to support the conceptual models and parameter selection used in site-specific modeling analysis. The discussions will also focus on "real world" experiences in modeling ground-water pathways and approaches to assess uncertainties and develop confidence in the modeling results. The five discussion periods proposed are: - Group Discussion on Ground Water Conceptual Models and Scenarios - Group Discussion on Ground-Water Parameter Estimation from Publicly Available Sources - Group Discussion on Experiences in Integrating Site Specific Ground-Water Modeling into Dose Assessments, and Ground-Water Pathway Analyses - Group Discussion on Parameter Estimation for Site Specific Modeling - Group Discussion on Ground-Water Modeling Related to Dose Assessments Questions: The following questions were prepared by the NRC staff working with contractors to focus the group discussions. Copies of the questions were sent to the presenters for their consideration. Session 1: Group Discussion on Ground-Water Conceptual Models and Scenarios Session Leader: Mark Thaggard, NMSS
- 1. Under what circumstances can the ground-water pathways be eliminated?
- 2. What site-specific features, processes and events would make the selection of simple dose codes (e.g., DandD, RESRAD and MEPAS) not conservative and/or not appropriate?
- 3. Can one use current ground-water monitoring data to supplant modeling of future l doses? j l
- 4. When is it necessary to go beyond one-dimensional ground-water models (i.e., two and l three-dimensional models)? l 1
- 5. When is it appropriate to model a site as a simple unified layer as opposed to several l distinct soil horizons and/or rock units above the water table?
- 6. How would you model a water-table aquifer that has multiple sources of radionuclides?
- 7. If the ground-water model considers dilution of the radionuclide soi rces, what site-specific and pathway-scenario features (e.g., ground-water wells) snd processes (e.g.,
dispersion) warrant consideration in making the dilutbn calculation 2 Should the ground-water well be treated as a pumping well? O 4
r PUBLIC WORKSHOP AGENDA i Session 2: Group Discussion on Ground-Water Parameter Estimation from Publicly O Available Sources '\d Session Leader: Tom Nicholson, RES
- 1. What are the most convenient sources of publicly available information on soil and ground-water properties for a specified site?
- 2. How does one determine whether the information obtained from national or regional databases and/or sources is sufficient to model the site-specific ground-water pathways?
- 3. For which parameters is it appropriate to use generic data (e.g., national databases and regional data), and for which ones is it not appropriate? !
1 Session 3: Group Discussion on Experiences in integrating Site Specific Ground- ) Water Modeling into Dose Assessments, and Ground-Water Pathway Analyses Session Leader: Ralph Cady, RES
- 1. What are the mechanics for linking ground-water concentrations determined from site- {
specific ground-water models to dose assessment models (e.g., D&D, MEPAS or J RESRAD)? What are the mechanics of using radionuclide concentrations determined from ground-water monitoring and/or sampling in dose assessment models? !
/ \
(h l 2. 3. How was the ground-water code selected? Was it necessary to modify the ground-water model to make it compatible for linkage to the dose assessment module? Please explain.
- 4. How does one test the modifications and linkages to determine whether errors and/or uncertainties are being introduced into the dose calculations?
Session 4: Group Discussion on Parameter Estimation for Site Specific Modeling Session Leader: Mark Thaggard, NMSS
- 1. Given a parameter distribution, what parameter values should be used in the analysis (e.g., mean, maximum or other statistic)? (Are there simple methods for determining parameter distributions for the limited site-specific databases?)
- 2. How can limited site-specific data be combined with national or regional databases to represent the hydrogeologic' system being modeled?
- 3. What kinds of information are available for justifying parameter selection for site-specific analysis? j l
- 4. How does one assure consistency between the conceptual model assumptions and the
["N selected parameter values (e.g., can default values be used for different site I (j representations and pathway scenarios)? 5
l PUBLIC WORKSHOP AGENDA I
- 5. Should parameter uncertainty be factored into the model inputs to determine the need for more detailed site-specific data, and if so, how (e.g., need to examine sensitivity of parameters to dose consequence)?
Session 5: Group Discussion on Ground-Water Modeling Related to Dose Assessments Session Leader: Ralph Cady, RES
- 1. What level of pedigree and/or bench-marking and testing is appropriate for the ground-water codes used in the dose assessment? What would be the criteria?
- 2. How does one determine whether the simplifying assumptions used in the ground-water models are appropriate, and whether they ir troduce unnecessary conservatism, or inappropriate optimism?
- 3. Should independent data be collected to confirm the ground-water modeling? And if so, what data?
(Note: Questions not addressed in the previous sessions, because of time limitations, will be entertained at the end of Session 5 as time permits.) Workshop Products: - Provide input to the Standard Review Plan and development of decommissioning guidance regarding ground-water modeling. - Workshop Summary and Identification of issues for Consideration by the Dose Modeling Group - Workshop Transcript for the Public Document Room - Workshop Proceedings issued as a NUREG/CP of the major contributions - Interaction with the major stakeholders and to provide valuable feedback. INTERNETAccess Information on the workshop, including the updated agenda and proposed topics, can be accessed using the following NRC Technical Conference Forum homepage address: http://techconf.lini. gov /cgi-bin / messages?lic_ term 9' 6
PUBLIC WORKSHOP ON GROUND-WATER MODELING RELATED TO \ DOSE ASSESSMENTS Name Organization E-mail Thcmas J. Nicholson USNRC/RES/DRAA TJN @NRC. GOV e4 curso n as s y 6p ,.,c s ,., p-3 2 -um w.,e _.m._.,
'il.t t Sch wJ t 60 c socie A Pscku4 e en e . c c et hfi cl Nci G hv'tJ'sc G n bc~ t%c% 6iwc <u 'EYau %kv Txchtd micrc4Mel , em hek rh $$ b h. $bLS.au, r.;f la,3,(2nk t r, 1
d';c a - & , w.+ .s ' .a ,:, t. .1. n . u.s , kes a . ,,,,
& /d c 4 at, /dF s: rrac oso @ Sci e6 4, coq 1<'e c ,+ M u o u v O m se. c~c n meanma < ,
, <s f krn m n m 0/ swxr r,,n a wi, ,eup r.-
DeeA6leJ A de/a ee <q G
,le w M lit SFC ' 1r&h Pad 4%;JC ru4 . #en dv d4 r A n Morbi? Arsx, de hdorkee ro/s. ca L IJ A n d b .e,s s m, A Aslec-sa< ,p .cs b L: + < O cs u u #Tc<,,o.-Js2 ucsuu-. t o a &c p,1 %> W A:J4 wec dw.s /wu La ce c.s r a
_A khu=sA &/r E M fo e ri wc' L M . A lb ~yec'q M h%r .M G90 U dea \nt '"% 'h?s c , w % .<., - mwc u c~ ~ w ..,> v .w-Page 1 of 6
PUBLIC WORKSHOP ON GROUND-WATER MODELING RELATED TO DOSE ASSESSMENTS Name Organization E-mail kyrCo 906e.< /bs7td 5'4 b%'tr Senwicer f,ce,,s,,,y yo r t,,et
~g, c 9 u. p s a 5 e au cw c h a rc o..) Cv'sc ~ ''**'~6"Y"'" ^ $tchael A L L Mu maine %kee a;%emm. wn GA~Be 11,r / h ke 4 % ve; DJ'6",$., T1 G K':n tu e c I/Uc n k [ WIl l.C N v cs k n (b a nt !4 ?>9 3 -c 6 to mo >on c_T nk.co.u,J&aby. +alkl,'; *t / ./
t 1 rr/: -'1 MlL~il ,,r'T _ae-!=.e e/ <> wI ! I T fJw A .t N b b W t- 'AbL,cu2GdAM . -
?)G:Ydiv ,54' $ I $ i n O W / / fi ftf 2 ) b&s,h ,m ulUdf Km,
< ?~' Cr/r hHc'N c/ur e- C'r//?"/StC (~'ss/ 4 w
/d>,,o/A e' /1//o.,,' (. . o ,, , . + !C.
, , .<; . 4 . ,_ // . @ <,% .,c ,.,
b k Hn.& 3 f) ae g. ir t C ~\ h >< 6 J-4 G u . ..$,.,,(14 4,p LmEl. hr s 1 i I w C6Cda HMWuMed E t'!U c/wa kp ey n . <v" bnmanue I hnstnsf-n,ac;, Am-nsi Mh n I la/ },L'fuaIa.o L9 h1 %k JJ De uneus.w 1 ha no i s 2o-cs e t JT G4e 1) 2-/Wi GPgj[ 'hh.m 30 erer% 6QU N yct4(Y~L. GawA$6LP M n l' / v? vs&RC r c s .> e n , c . sc2 tlD$ ,.. \* [$( ,
$1 c. MA G I Tt S EW 'S LF b Yry b e.~
4 Met VHic.u O umss uswc w f;!]L tilLCs Yl> k$l lh(Yf^&L'alflll WY? ?- iYc ., N .3 M \\ u. 5 M i ,q _ _ , _. Jou %vn a v/c M JWWWNLb3 eon &El \J MIL L /IV ff S .5 Ye kr emu . Page 2 of 6
l MEETING ATTENDEES 1 Date: June 23 - 24, 1999 l l p. l NAME AFFILIATION PHONE NUMBER 01IG, in x' 9<tigt <.C , o s 96 << 's i << < 9n(~7%% sc<.72'. hNO Olu28 fit f4517) S(IEME% lN-7 % 5Cor D bu ME 801 - 417 - ce s t3, 3 ,// /% r A s > &ch kl ac1 9 n- uw [bn- piado PA DEP- bAl' J t2- 7ts-scnn ftlilit? f. kDSEd7tlAL Ofh)5f*fniS Ind, i 709- SLL b 23 5~ N,d II///s._ / 2)ehs /h r~ -.J, 9 e,', 5k, - c, Lv 9 57W'f kW A/E5 03 773 'W/
- k. OseO Nackr'r toesTim ho ose- 417 arti + c c a..
Ien k'e hU { c/ f //sJ G< ra n s 74) #/S'- J50 3 Sc% tf(CRw k'<d N Rc 9 4 I -4( C- 2_4_f l L ( W ier k CdLC RA 5<3 - A l - S h #1 I Tgit.ip .Msygiu PNN k M - e} M - 7$4 [/.h bc/l $)nMif(* CI)iW9 303- 69:2 - V/29 l
^! Q 3rv1 &db opug lw 80/-WWG%5~3
]A$ $ sl (Ten :6^ T? <- V77
'A&MCaras Der (2cd 7WK Kl
[hd(Io ['ll/ ENC f Alb//X/h 3 0 l- 4/C 7 / 75' 7 h P A h u Cr 16m A hl L- LH J (f?/ (SS6Y Gr'ne /Ohikm 0 h er- psint Ta'i- 3 7 2 -6019
) wd b ee Rsf zyo-4 W_ofg7 1
Page 3 of 6
MEETING ATTENDEES Date: June 23 - 24. 1999 NAME AFFILIATION PHONE NUMBER D.u%sw GA M/4h @w)W- " /* i iE Bava\u '
$dm PJf LA '(901) W Y- Q p W7% tert 6 m cb Mns'WE f.1ah 9ft2-53$
~
$Ew:lukulv 'MK7' ~K d-024nd (LI.C Ar % unk. ArisoM s 2.o 42/-7//9 TIAl .h.k.n. N R C ,!.K) F 9
&I-1,lC-RI6I i e kot.cren Hec /Hs so n - 4ic - 71H_
6esme /%er NPc/223 3 o 1- sir- G .2.t 2. 1, 8, MM u A C-/3 w M SoI-4fr 4(t; . d W Jdh % M2c /Du> M 3 tot-4Ir 7272 3 1lI (i)(+ m NNL / awn 1 ? cit 't (T- Sw c. C 20 " g
)
l
)
l l l l j I l O Page 4 of 6
f' HEETING ATTENDEES Date: June 23 - 24. 1999 k/ NAME AFFILIATION PHONE NUMBER - A+JMA .1&=.kloca WWC_ %( - 6%-W NT$. C[1pm Hwtn t) 5 1~PA zo2 -s'M -920% \ 1:3MtIE Vo&$t.- Lu% w Looso 72 4 722 5924 Ch h y Ya AS/L I M*-2n -((27 N$ bin]U4xs //eut daumw mw db-1729 b>vn AY ./L. KGS fu G iVA (, Ict,lw C-d243-
% d & >s a u :n DOE }o p W6 - O 3 y /
% W1 n or <>12)du- & 7 c<I toy A/k C 301- <) 19- 6 7cw
/$ 4A b,4,' A o/c" ~3 / 4/5 'G2Y9 1 EUEPc0 utc Se\ 96-ce d'
() f h Sc't' bd $ fff( 3c:)-c(/A~ )2 Sb V i [ N O Page 5 of 6
PUBLIC WORKSHOP ON GROUND-WATER MODELING RELATED TO DOSE ASSESSMENTS g T Name Organization E-mail Ws/ N OM NLi)i N.)I? C l rf /4 ( C Q (pt d N[c ,&ct) D I n!n WRC l055 Tvm (6 unc fgy k$l ('WdtI(Rk NbT R16 AA)ST.Od$
% ~ n s. - x.s. s,s.. a. s . s. e nu. c. ,
3k L fDss e0 CAJMW yeso t@,swi , eA u I cme ve & fo hen & SD//c/l2 C 1WS O NKk t/ NW W hk$Nf3 ///f( i swQ ate.Sw j l O i i O Pg. 6 of 6
C e n t a c t G ( e0o 3 B 1b y
)4 E DO R (.r ~
n EU r 1 i ,) ,,%" c5d-g5 Su P CN o8 ub OD NU v1 1 pc bl i SMW un a r W C MA EI T c i l t e e O t f o NS E ad f i c
,E r k AIS R r S e Rs RO M Rt o ah I e ta g e f
ND dop i OIN O us uv i o ,S N D l a goO A GEL l ci t - l ei s n o ao r n i cG N SI yr Mo U J autS DT N I C SE SIG l af N ia eW t E C df r a P f I S o r a T m O i as l t 2 3 inr t ece r e se E-- CU S m S e 2 i aM ,4 i aft n SRE PI S s s i e t na f 1 Foa r t EI T o 9 o i o I n y a ag 9 9 r i m n CCN n d,m e L ip cl FA I I LD Se ee C O at n nm se GS f e en CRE g Ti t OO A u a en rg NU S r d s m1 DPS I i n0 T RE I aC OES t iF NCS oR n SEM 2 PE 0 TN OT
,R O O F
o o o o o o o - C S G E . . . GG AG TG O O E X RE SR HE N N A CCS C M E E M r ioo ON SO EN UE UU RE c mu t L T R P i r NR MN AR U H I C L apc I PD DC I S E l ae DC O t - T I I O U A S gitb WP I W OD N G P O r ie olir AE OA LO P TR NT OS p y m ut S H R F E F SE A T C RO / AR GE _ RfoAb S O I N A O MR PM CM D O C N eSs cit OM P O AO R N H D I et r p eac DA L D LD DE I E M F T t EN CE OL C O O I O oCt r oo i LC AL I O D R N - nn SE BS SN M E S S c I NA I LF EG M L E F Ee xp T S TO I CI E I N L O pt OS YR RS S N G E R ou sa E S I T U P D C ul DS C EE R A O T P OS F I A eM r SM R RS I E T I O T o EE E AP S O C N H Sd E UE E N O O W ce el AN N NR N S M n S T I N DA T T F A S A E (P 3 P I Y a r G EI T L N / RN S )A i I E P S o A I O X U C ( s SI N 1 N N S T E ) MS A 0 G I N
/
P E S L C T O T D A a NU Y F U E A R t h T SE S RO T S T I
- I I L
A O w S PM I a F - AP N O M y E s R G RL O E TE D I I N N 2M F T E ,0 E E R S N I C G I T A R C UA T I A B PI T O T AO N I N RN S G TO EF J U N E 2 ,3 1 9 9 9 l li
e o G e e * * ,* e A P n N G - e O A L A o r T U r A o v E n i c t s u ( psi n e D r R i m e n Daui r a E e D A e- p a d Gt bi t g s o TG r hsa e d t r s HE e f r r a w a wut o i c e EN q a 4ar f e m s t RE u m t e t e 0ys a m 0 e o e e M o AIR i r e e d r ,6 co m d DC _ m 1 r i n p a Na semed b f 2 u s d e I e 0 k t Uy oi l e r 5 e l i OD n 0 i h un c n LO t 0 n w R brg o m o g OS s y g a Eec f r y eu t e n i s GE ( G D a r s e a t w a s Gmss 1o( 5 dDg i n h e c y ml d i t i u e s I CM AO
- e r e 4 r o s LD 4 (
N r
- 9. Gr i
( 0 n - 0 U d i n ) - e t i c s U DE 0 6 R o s c As a 2 5 ( e d n OL I
)
e l u 0i 0 d l e SN E g g G c d ~6 en r m S ,. r EG
- r i
e d e&t l o v n 1 CI S 1 t i f u o 0 R S 4 9 e r ( N mNar i p y/
)
c C I TU 6 i a U nUm aR e s c o n F R EE O3 a n d ( N U R E G t eEr dGs I c e n a r t r o 2 0 RS I AP U E D R - bCe i l s 1 G E 1 4 a sRn o
/ i 4 0
NR T G - a p n 2 DA 4 1
- 9 6 e5r d5i a p : EI 0 4 )
o 1 o t h l a RN I 0 9 n 2f w c 1 N 6 6 o a e 0G
) ,)
s .V r y ) i t e) o 1 s s C F T c i l l R O I o a PM n n AP d d i t RL i o TE n 2M s ,0 E N ST UA BT J PI U N AO E RN 2 3 TO
,1 EF O 9 9
9
I! \l l ll 1; l ll ll ! G e
- e n
( A v e r eer F ss i u r c e r i H E T e
.e Ea T P sreA r a duS l l D G gDthim g ut e ai r o R E do e n s A N i
nda Ewed m l e ne l r e DE soy oif f i e v ed M I OR t ssf e e pc t u o I i t u e mt sd i m b l s ut d e LC O r i t i ofy o ae e e r abo l i GD s sin l n I O n2 a5b e i d o scs o g CS l e f c ir oi n i s A E cm n t h s L or m t e a ere u D M n md t eo a r f c r e t dvi s n e s O O D r l i id S E wou i oyif m r e t i U E L l e c n s a.(0 d r a ina t l d CI N r 2b s l g hmr r ea e R G e5a c r I s e r e nd TI nme i pSd n a o u st t r ao i a 1 0 ES RS r p i l C I a/o l v i o s chc t e at F AU E cy n c dr v i R US
. )e )usi n e t
b a s e n a c ,my t i 2 0 N P DE G d e r o roto n ER e r c o d i o d 1 4 RT r n o s iwc t 0 1 A n ie o 3 ed si t t i si u n or ne m p ( a 0I CI N r t d s nl . io cn e e aoy b F N t s r r r t RG e e r e e w a PT - da n s t s Aah mt i n AO r d ud s i c t r Ldg e RI ere t i c Ae2 ) TM Rb0 : i cs o t e 2 P 0L ot r n d Ae.1 , E ni ,s u c4 dct s a0 u2 SM i i m e s UE t i o a ew BN on ns y c o t o PT s a n hu AA c t d el d RT I u t i TO a o i EN
. y l
l n s ( cO oF n t d' O
)
Il l .
1 ll 1 t l I1lll - G e
- o e
- n P e N sG - - t i hnAv r O A o i i
t r c t s eo T es t e i e s u 5 0 ct r i D m p pn 0 E r u a o e- a ed o1F D i t tg s f r r a cw r 0u w0r m E i coie e anm i a fa r t t e i s M e gl a it l m ce r omhe ur m l e o e o cm d p e r d d err yl e l r hp s ob ya t r me i 1 n ( s u ne l i 0 k sh eyd l 5 t pt r n 0 s h : r a cw u i i 0 n uac m a oo y g aa l y l r t t ei S n l f st I s e cs v s a w nno no 1 ah u i r a o y/ 0 r e e s t e n ar nt i n
)
p 0 e c s ( r de et m nr i U N d i t t er r o ot U on i i pce r e i n o v oac l R s nl c uhs i i m d e E e su bnd d e l nl gg r c cu i d l y G c e i aa d ( er 1 1
- r i
d ol l l t 1 r o 0 t ry r a h iu 4 9 e r b a m np C u ead F 6 i a t nci t
- S v
eu f R a vho fn O5 n d ( N U R a y m iia r ec ovt na v i y/
)
o c t ed
) r e
2 0 1 D , u 4 G E b e mbi t r t n 0
- G e e
n
,e y l
h r e e 3
- n (
4 0 0 1 4 9 l i m awe t l oc T ts E r a 6 6 i n hue s D c i b
) )
a ads l E te a r t e mba d d n d ey r o u d < s b pt r o e s e e
)
a o t : s e h c oco ( c d io t o bm hn n o i tp i ed i t n vl y at i d' e vo en
)
yw l et i as r s p xh ge.(e ~ J U N e n mg ,. E s e 2 3 i v mw h
,1 be O 9 9
9
,e e
r n o f
S c r e e e T iT wnT n yT A i h r ry U A G i e g p i p t i fy s s s e n hlip e s s n g u aca i i c t i r c a u u e n s t el a l na ot a l n m m a r M o a dly y l sdy el o e e l l d G t l , i n o y e u at a gws a s n a s l R r nh s n s u s O a i de s ias t e n i U u m G i t f t t e ( i cier u r e i p e N d .e n f m a m e c f o l n GD z gi l e nt h e s r r e e o ,. t r n t r l i d i t m a r EW n e ca ot o do ea iu e i c s n o i c NA E T s n a u r p g hc A i a no t n e t g l r 0 s RE n a t a ehi n f a i f c 9 s I CR d v r r t r r o n u m o d s hg a i s AM t h e i n a ul a t o i nuf eo e m p u l s e r m p SO a t m i o na r r o v a t SD e e s UE i q n sc e i o d at c c l t i n ML u i e e s v e i n PS i i a s t v n p o e s f r ui t : 6 e ed t r ay a t o u ( TF a r ae
) p e h
t i o r i r r c g e. I OO r e a en d
/
p e- ,. NR e an q zh t a t e SS i nd u 3 / AC n de i f oe t h r b PR t i t n e n c m o i a ht r ea w a ( x PE l l eo t n y e. LE y ot a CI N I in s g m
- n. t ,
o f r e f in i hm t d AN e l t f r i ei n p o e BG a a l I l o at t r q t
) LA f i a u e 5 1 I T N ot c
o no i id f c YA n pn ea r r m L Y t a er r a e ) S i a I m i o d ,e t v S n t a i - a t h i o e J U n a r N e E a 2 ,3 o 1 9 9 9 f O I
li [ uS S e nc c
- e o e dr r c O T
- e r e e e
e o O T oT U n n O d
.e y nl y
n l e u h h e s fh t e h n r e t hi e gn g i n 1 c c n es s f G r o o s i t e st e i t r oa l n M o d d a c i c l oa o u e e t p r s t wy d n d d u r y h eeu o e d is l e d e e a s n r r ni l s w f a f a t i ic e gs a e c n & t e u u d a g -e l e cu r l t l t z l ce t a os D m i n s o c o or s e n dg n i ef o p c e n e n dm c i t a u d e u n d ei o i D l i t a a n i t s n o nn a l t n p r i o c d T g a r i o d n o i sd
- a _
a a t s m t D b n a h p i o e l d m n e a n c s p e d a r a s o _ a t p q e t i d A t e r a u c b l e p h s t i o e a p w a h f e n l ya r m w d i c r e w p i
/ a s a y a t
a t h o b a u s t i r i O7 r s t b t l e e a h h d i t a r e l e e e y d e f a l r s w s e ( u u i o u a t e i t u n l s t yd l e e s h r c t i s d h t h e- a n s o e b c u t l e l u a l s e r s d . p d c e p r m t d r b e a y _ o e e s p - i n v f r n i s i c _ t h e e n u a e d e g m l l y - o T b c p a y f o t i a b t c d o p l e h e o e n p s n a s y l a N t a s n R s d C m i u t J h ) n m U e a t p t N E D a i o o i 2 3 n n n
,1 d s D
O 9 9 9 '
C S o o e m e e e u p r aesT C R c S O nxhh a e dp o e t i h e s e o ou b dc u T-i nsl c i l m i d eo r u e sudo d t y i i c u pn e-r G t r a a t t m E t i e ebe o l ha t e A N am l u tpc c o f f r r E o a ni b ia ot o n S r d n m s R nhmc i t e m i o da c t r I C awpe ( a vt e de o a c P C l aap .e c r n t G E cyt t o t t f i o i u n g i v cra i i g o RR n t sba l c ,. i ae u n OF r ec el t e s y o l
/a r a UO o nwt m l
saio p t l u p r h o t i NR r ih d a u b o r o n DM WA i
.o s ,e l i f z o
i l AN l l i t M y i t e n Te as o ( t a TC h e ps d i s d E E R A i l hu e l s s e mys m u e t r i x t MS oip dct
) b u
e n OS ea on i t t DE l l i o o E S sa L S 8 h n od
,s a
,n h
f c o SM I E O u l dnd en o m n t N T N v a O (T aim o m l soa r g e i D P ont n O )A n a amhe e t i SI E SS cem o i t cn ot a y
, n AU u atc n
l i m SE t ca l u S S E f ol t S F onc p i SO r do MR l stim e i t op i s EI e nu o N NT r st u T. E e a r _ s at i c G r no t e- R idn c t e A J t t s r T U ih N E oe n m s I O s ) N B O 1 9 9 9
C r i t
- i c
* . a O
U l n r G mM M - - - I n e r o uay o i t s u G l d poiiei t E t r nn a r p pu o n ghtxt e i i i f s y c N e l e dceae t e R E p r d e r em a uoslan o cnt it d c e GIR t t t oia ol s e r RC mor ee h w sa gm( ne n e l e p t o OP d y a r i de xm a r UE i ah t s o nixpo s - NR t a wt arepo s d e E DF i on a n dteco ul e c x no n o s r i o p WO on d n soieorsue n n o AR pe s sg d s TM c si ,l l ( iu i t u EA t i or c e o ir ns ( i o r RN nt i n ,l pnedn i g s i n e C i a a aeu og l s S M E s c r c O a i o n tnss t or o e DA l d s ip sn n g a s fi oe/s a ES r o s rr ,n n oi i r L S n dut i o SES p u e hi gsedr t e ( I c paodoe s NS O9 s c e e s s a r t oel sos d dr or i t aisd i
)
/
P a TM OE n r uwcf ci l ) e n t DN a y et ko h r i d her t f w O (T o b a fc do a a S P d s r oiunt r y s EA u e e d o mtaoon dd m e r A )I SS
. t o o cmoo s SS n oi r r n n ,s s c EU m s / e S E u i t a at dn i
n S S l t e a M F t mer idno d i r E N O i i p p o ngko ir T R l e h a r a y s o s i e duoga t ns n d ( cN I u r c c a l sdad u nwn e x oT nE t e- c r wte s p _ d' G o fa a t ,r u o ) R _ t e n ce r s A J r d e ,r p a f u T U N m s i t i wa nae l c r e I O E o ant p N 2 o n t i s s a
,3 r s em o t O 1 9
9 r a gu pl ocr r o h w a 9 n we n )s y s
)
d :
t f C r i t e i c a l mhMPMMpI t n R r G r oea e t ootai t s e o G eb yfddhi a t u E a u one oiiwt f f r p N d emryya e i c R E v ad n d daas ppyd os t e e c GIR t d n at ot cyr peww os hh e r p e RC e e t OP doeayy aa m e l o UE gt ff r n ss o a s r NR d DF r hoa/ e e E o e my ca u s r l sb i l c x WO s es n o p AR niosne dt e f i ad g n o s TM w f- br u d u EA s s i RN a (e isaio t o i r n e i t . eg t n o M E C r ,. e dai ebs g n S d o s c O mo o o est e n e DA of f s n dp s n ES d t ef aoh y i t a L S ehain l e r i S SE e slu s o e i n cs r I g si e ( s NS 1 0 oeetr aic n sso r c el s )
/ TM OE i
t d P e a DN amfc eo ots n me n t h t O (T in n t nr i d i f a w a S P a t u r t t ct i r y EA esloio dis on m e s A )I agos n ns r ( c SS r s o SS ea r a c n EU a ce r n e t S E
)
r d n d' S S im/ t o a r
)
M F Tio hc v r E N O i o i sl a a nsl i a T R sgot n ( I i r f t i t d cN u o r u uet e oT nE a tpsi x o p t t d' G i or r n o R n ceia c l s ) A J t u T U N mpeic o o r e I E yatos n n O 2 3 r r t r N 1 9 9 9 qt ea u l s o O i r e
*
- e O csreA r c mpT ot ah e t
h OPT o e cs e fah e pnom
,s ar da dh sPth ceawo r l E
aisl l f i iw w ea a n ac l A X _ n o ss t dy ser t osy a dmut i o n sl M f ape r o rq do( i r en a P i sypi e du t
- c. L h o ss i l
a orgd i E petbr S r n o ,. r i mi int l i spt e a mhu t t n ky . s hg O e o i t ) aina F hoe a t t egn sn C wdae d Toa afisd d w/ hi l O a o y isi l N e u c s dun t r ,snah a D mt a op h ma ev r o pl t dr h t a I T I y ai t og i a u ehc O br nh h m t aet e N ee sty l e gr ui r S w eih l dap y a of ins cst i F O i met intp o . t h uac l t s R nt hu ag el I r r e u t r r u o P O1 t a ia gq a taft A 1 e ast l T ddted r ie au t i pdeeh a H ue n ncu if oe t z s W nr ho o drear e nb a ad wni a e l A Y r cs nn no ysra o d S t o dn / S m e n het n a r do i r t so C s abe _ sr e a. p ou i is n k u ya at s l t E N
- e. sH ip np b d u A c
i T ei ds g o ee p p R I f i h t wt r mpo o O cufoor ag o r S s ,s r c i t r di st i t umal fa e r o u it a f sg i M O aeamr a p n e uir D t n ad dnc I i ot i nd t hw osu r ul I F np g f r o a'u wa etari u t C ya re t t l A mba i J dcu s e l T U ut r l I cvie iea O N E t i
,s t p mx a p n
a fot c N 2 i y o t er v i S gs ea l
,3 s
O n i 1 dgt r . bu uo es i 9 9 a n ere r 9 iTi n n n hg e dw ehr t o r
- e e * *
- P F E . . .
U ( U e r f o r s t a CRA s e e r es fe U s e G O o s b oev s r t d E N r i t e-l ng a i h e N o m s i diiol t e-t oe f E t e s h u nb a a Nm i t e p r c p s u R e I t h r e t al e e Ue l t C a Ran l a c a t t i i f s sd i h c i E i n A a v c i o t ai et s f i Gva p P e n at c / u P f e a a s o t Cu l d o n b p (e ri p p R w s a l e e. cgc y h Re a O d e y l a 5
- s r A e s r i ,. s a a f l i
c 5o C af m i s n csa i on a 1 f m H u o d s g e m ,i d l 2m t e e E l t e t a s e l e p Ve r S p l ac a ot a s sl iv n F f a i n f o r a l sb f O r a g c f p umi r m 1 o o R a u eao r , l i r _ m r a e 2c s A e n r r m , ten t t e d n e a e a m e r a n a n c r S a m n e s e S sr t s a e t n n b d d e E s p t e sdt a ,3 e b n S i n e a O i 1 s p r f ohl r s n S 2 t h s r o d cy e D ah g M e d a i s udn r o d Gv a E c a c t r s ion o 4o
- i n N h n a gnlot u t a
i b i
- 0r l T r i s u g o 0p y A
r n oir ,6 a s e N na ng t e t ci r i n e o i o aa s2 D t d u n s l nm D ' s s s edd d e S a e a n s tt aa Nt e E n n i )a a Ur L d m d E D e a a t i v Rs C c s e E T o n s p G I o e a - O d s 1 N e f t s r a 5 m h 4 O e u m 9 F ya e ) n I d c t e N c o e r P h s r s a e t a U J n d i T U g i n D N e s t A i C t r e 2 ,3 i s T 1 9 9 b u t i A O 9 o n _ l .\
FM e e
- e e r o or sR C Ce O sd oiM EoA O ae t
f f i v o m n s ut r r t su l txmd ep v o e i p t ci i nl a nw l e e bp s en ao e x
/
o sue t l ix v c ca co f s f f oic eSe id i s n o o teG sm i t t s e e n em r e c ap s .s or bl p s o ne dx l i t eEu S y e t /xn O m c- x as r i c pl ws io oad fmw f M st i t i e t hur s ipt a E ee o x t T bc n e mce ee erl H s n us g s/ ao a g r oD O sn n i l t o of o U u c es i i sdt n pf G i i d e e e co l no H o n m u r i n cr tn dmA wps T S s l t g i t a al s a i p cm i t ee e O n l b ai r xs N d e o c r r l ga n ass nt im M t m n i r dee O O1 e i 3 o s r oo un /c o nt D d i t s p r a E e l e a w s nM L i r r d b ui9 I e i n u et N g r u cn r t f ie G e ral s peh i w a e eni n O i l l s t ot es g r e l d : F r b e i n t sag o w a t M a C O c h ne e y M o e dn r B P n d P o o p u c o e L u A s n N E c n t t e X t e a os i n ur a e S d a m d I P sf i n e T b / fa a d E y s i oc r e t S P s / i o A er s n u J mb es U du N ir E af t a 2
,3 o i
ne c O 1 9 9 9 o f
1 o e e e * - aD . . . - . T sGr aC f avtaTh ss o s r a c eo ci r se sr b ue e P SPSS n nu a t t oi c uf er il r n m o oot ii s r a mys e pm t ut e er ee t p o id sl pc a s i t o i cnis n n etisp r ,o w tg h r t o n ode t t a a ou omdn i nl i at - at e i c a t e t np/o o n si el t n np sat r agn l i r c r ouft d a e e n ds i r umf so r cinp f i i c d o n ere d c fboiNe i Rd d t l i sa eoteah s t e ht Cf un nn y e d op ey o o sop a r l t dr bi st aiecst l oic i m s e at o d cfp a d C ee qa r nie na doat p d ,r r t ehod r O nsr l a m r e N ut fseoc u o c o e spx n vt ee i C ev ne i t cs l n p t d s o isei nt d o i L t y p l u r o ad a e l i e s d u g e sa n U r e n nt i t h n cccl S oo d /o di g ar oorge r I ne l pc a i t w O on dn n a ec eo i t e d or s y m ds nu N i s nr a p c ( ei S ms u se s y r e sp n n po n i; z s s c b
)
tg d A e ni e o a u N t r e ei r ac w ] 4 di u st a t i s i c t i o u l d i n i r o l y ,s l mde ot a er D R q i e n b e t i a na i o d s m E C gnd s t e t n e codo u m ged O ss l d i t en i o d u r e an wudi e i l l M M
- g i
s t t h bg n s f i i n ei v E e dr o t N p h e f l i d g em ci o t wn euo sr D A it r h im t c es ie f e he T cd e l i d t n odc
- n. r I
O dd e ht c e aa t t m i f a ea l eHe pt on N S aa n u g p aa a t e l t r a ot uwn aeg i nn r e uh l v dd ac d b s nw da m ,e r n a o i co a s d e wy dul a u n t e n as esei s y J a e l s d t t e r a r l s r s U r f e N cv o a a s. E oa n r m ms oi nnD e 2 nt i dg de ,3 dv e ef ei n 1 9 9 ie t i o r l i f i da t u Q gn ac 9 ot l n s n t
, t w( O
.O m
J
*l5 e
h ' o D u
\
O o S .$
% m w cc L., eO O O O
- O W . Q c G
(U li a. 1 "O o c cc O u 3 c .9 e ~CU _ . l n (13 O Z l U lll3 - 1 w 00 co Q. u (D G O 3 oW cG X O c t goO D O O Q z
' O g& o_ c C
y .- O (13 .Q F CU
$6 O w C m ..
llll3 e J.;Nk_Q}f_;p,@'-
'. s oey) 6 v.~s?%jy&pd 4
sw . i O4 ' L E$Mhdi ^QWMh@$h == www
. w a z t k M %=Y;ws n su i Eid _
I
B m m a H e t eM * * *
- A W
h l sa f a o RGHGh l t t C i nl e ey ey t g e /w a co r op d i s m ct i o o o a oo h e og y h t a nn t i c t s a mg iy e s c a mp rs o l i i i t s n ns y mab o nt e d e c e d t s p e ae t r w r c t l ine n i b u g p sn i n a l iog g s tr oe t m a o 1 e o p g d r o i c m e c s a l e t r i n ? s u c f a P s t u e c i o r a f i c f e t u N l a r o r t y n e s w h p d e s t t o f u o f N l i : a t i e d o s n a i l L s a b o r a t r o . y
y r f o O o t a r r o e n b b t a t u i o l a p t n d w t u l a t a na o n /o i n N u t u pa t e t s o s e w r p n s e r h t g i g p r o N e gin e c 1 nd r m t i y u e c a r c P f o a n i v vi t
? n -
e s t a l e i o mi re i t d t p i t t n e a O o i r c g n p o u q l m s e i d u p r a t a d c r s u e n o i e i p t v t s s s l ed p i t i ni e n d o e t a u r a r om c at n i a i cd g nd n ml a o u o o a u l au q l ct p c l ob e cyi a a g muiab it i l t a c e mn i r . o e i s t omr ced y t sdi y y r a r eo t it psHSHv a t l c C a y* ae l l h As *
- Mh e O
t W m a t t a B
I u B a C l l e l l Oe t h d F i l t Oe h d F i l
. a z o
n e eOb eOb r I n s r
. R e c e
I n s g f e . i i or f e or n o r v ma a r v ma f o n r a p m& t l i o t i a o t i . t on t a n s on i n s . iA t o a n u c n g l o a l
. l M
i i
) % .
a t
> h e
o n
- C m p
c e p e - o n c a t t t u
- e i u - p a a c
e
- t l l
- u a
a M M - l o - M d - l
- o i d
m Qew l F i e o d O d e i P - a c i f i c D > s e r v a b l r e
- t N -
i o o M n r M M - M s l h w t t h e a t h e a t h a h e a t a t e e - i s m m - m t N a t i at m a t -
- a t o c
n i - i a c c i
- c t a l
a a -
- a s
l l i l - o M M M n o o M - o a o - L l d e e d d e d h _e l l l a 2 l -
- i b
o 1 2 -
- p s
r a t or y -
O D A s R t oi te r S p r E s pe R n a r o r t p D e d v rm . n t e i t c fo a g ei D d u v u n . u dah t o b o d O t n r eh t i t g n w g oO a w u a s n mT S e o r io we r y i i x mE o a ct w af f a m r cRP f r u o n of e . s sf r-o e r i u _ nd ei s q _ on at n b a _ a e s mumde ia et t p , n - e yi nf i mS l p d d- i uA a p l sP i me t e a n mim sE O AM SSOSS
= a u a a
B m a a m a aS l l e F f o sg r eSSnP i sie s t u e-r i i l l a w g f r ie e c /un t u r f ai t f nc i i f i c n ir t f a e r mp pe s e s t ie d p tg a a o n t ic i n n nt i oi o a nt t a f t ni r k t t h hwe r a f s c a )s ote n o c r s
/
t r s i w o t he e c n eg o r f n o d o r e e t f n i t
. - d m dn ie o i t
i o a t miw t o i n i ey n s o e i n ni n p m s t s s e h oisu a c a a n t d t ab s s i n P f u o n g m a a of r t i c c i f wac t r a o e y v N o (f e n n i r t h op c p s l o w e uo r o a l a s t sp r n t N a ee t c e t i o dr e n a t i e d c l 1 a s o b d r y o r a t o e e
y r o 0 b t a r o t y u i i. d n i i o m u t. u t s h e w t e d ht r a e n o r e p c n s) mc i o d t n p a t ni t a i wpo u p t r
)
i T ,g n f nE e n ri i i ut o e + aw t t i c ) r r i g nod yR e b e e pt y l o r ( d u m mi c a o a - l u e c n (t p d r eP y i n e e nc a y y = e t v o d f I b a r r i u ae t h i c : y n c ri i f
- e. et e a o nS p ,
c p g mstr i oC s e r . n v f a sd u n el rai a (S ra o
- t t _
r ot t c u eb s r a i t ef i i nipf f t o pd o s- o ptei g mict n e u vo au r r r a e r C r r a NPREc u aI H a PW*P ne o e a N a a a B
B a a m C a H e*De * * * *
- PL P o l
BATr h a e u n eixWP l sp a r o u e r a a cp r t a m1 is o i t r ot e o k a c k m i b nr s d n e n u e c y i t e e t g e n i t n o n s e : a mp i o ,t n (e s s r s s n t i t t c) be n f f y e o d e e o e ( t c t e c i d f f s v i sy f z i c a e o i e t u p r m co n n t a ro y a n e s t e s o t a h d i v m y hy t a y ) y r i n d d a r r a n t o l u f o i c l u g c x i c P o t a n o i c i f i d u u n c N c n p t h o r t i v i s a u t t w y t e s t
; u r
N s a t a o t o n i i l
- e a
l s d p L a e z b o c o r a f i n r y t o i c e
! l
y r o O , t ra o
- b a
e L n i a n _ o i o o z s i t n a d ; t s t e i t y e w t a r i t v h t r o u t u c u n c p a d i f i c n s n p a i o o c i c t t r i c )y g o ui c t o p s r a a dya p l O o r n a ) yh c d r t s d d i t y t otee s h n r o raf ;n i t a p e e n e at t ub i c n i m s r v i o a e i t ( snff e e z t n y a ct e n n o o c d o rsf o t yff ep n y _ t e c e er sis (e tnx oi s n y oei t v a : et e et c c t u rs i r s eb nd i s rf i b c r pe k u p o mmck otei r i a uh u o aet i t s s a r u r a NTB PWsDD pi i e s PP * *
- l l
e O n * * *
- H U a a a B
B m a a H S a e * * * * *
- P
- P t a u u C sDDbSP l
iy p BT r Dr i s de o u h a ap r p e r t r iais b r uf c o k r id l i c k n mn e d o s a t e s u i cy t e e t De d it v io ic i d( n s e & t i ng s i e s i s r s c z e s c r ac f f y t Ro n o o dh e E t n e er ia c S a f f ng t i v Rm e i c )t e e Ai h i e ( p s Dn a y n t a o r Mn d s t u o a t r r ay s i t s s t r o ' t e) a l o B n g d a s ap i y l n o c d c r i r a e t n mo t u p P a c c i d o w u t i f i c c o e e
- I N n /
o r d n s w t h u o u e c t f r s r t t i v a a N a i n c t i o n y t p s e a l r o w L a b t a t o r a e r t o r . y
y r o O t a r
. o )
b I a. d n i o t a N t s e h w t r o N c mc a P s r t e e O m r a a p i c t e g a r o e l I wg r o e oi p n d wt o ce um f y h h n p r t pt a e eis e l e DDW l h e t O O a a a l l a B 1li
B a a a a l C l e * *
- CsCvF et o mo r n i
o i NDDmid l l e Ua r O sma c Rf Ep ae r s a l u Er t
/ a t s
Ge Orioe h i
/ p Rs o r im e o Co r On db m n Rt-2 s eo nd a
t s 5 D0 t ty c i 6 2 a 3 u i c s a n 3 d 1 d i ai D/E DRe s im m a R ep ni o n E S a d d Rx ur e S As i t c D
/R t p o t n A/ D an o u c c m DME r
a e o e P mp n i c r A et u c a S t ea p e
/ r t P
a y l M s m u i f i c b M o a l c r S d m i N o d O e r t h c I l s o w e o _ S d s d e a e N a t s n , t i d n o n a l e r t o L a q v b o r u i c i a r e y t o r e
~ :
. ~ y _ - ) s t - . i
,- 3 don mt rN i O e s
E
,~
drs t -
*o N e od hf csoUso ere o . =, n cr ab gn n an f
e rell n
" etdriU 4 c HaS
_ O s PaW ACka r C
- s. t n e 4 r
__O "- -{ 1 e e
%CL i
- o g t
gs Wa 4=,ny J *, dy **
~ f.
y
- o,o rg g o 0
%y;.,s=- t Cl owa o X t f ll ua f n ,,dan ' _-"-
o - S-ae FS a,S - c
- f Ah s t
S-
=,,* H g * < ; ,.. Ot na ec
/ . _ _e U eP r A 9
I G , mla uo S oO M J* # Oc4 0t er
- - _m
~
%~ , - o %sw vu S W
' ~ . "Q "
s T4,RN a a i l t a d * - _.v h@ c x .- + ad n d f i r e gr; an s+ i rio S-
- j$: N a t t Ag H, ond Ar a
, :? .
Sp &
- N w co ig g&s.
En 3 e d d n e . o '
.e y h
( i t e iu t tarb na f u o 3" - h s pt r
* ' ' - QT*4 .,
. f a i s o G mim-
. d h ,e T.
- o c,
Ce W k r y R eD un ci h e o h* nP t. o 3 y l c O #mMm n Tia-l e c t
'B
- e *C l
a C aheh r n u - fa e
]# '
k . s?N) 1 i t a r s o e NE Su
, i
- D - n i xg t d
a
- - c s n V h_w@}N%
3 - i k A e t s i oO gu e p 5%.. ee
,l.
%e l a'reB o e 4
ht n o t a du r C i s s n o h g 37WF u O n <. d gt o h-
$w@%
li at l l ei a . C o M .%n.b e c r t a aI r t f, e <- g e
,ys6<
ui p gn. l h ' t g er ?.>- a i . T
-t ai Nce r
g,S p ?. u4
/:
- / /
r Qwdc+-d 4 R o ew e t s y a P y
,,%M e -
N 4.s: E ' t P 4 3 w
,9,:
t t a }_, _ o
- d u
ht P a iM
;..~
i , I' ef $, - p - , q s o l i cf
- c U a ao f n e = -
o l u r e it r r
- f t n kc o s) uu SR
- P _
\ B e n t
awe e s s e t o . P ) 6 e e a t l
> d 0 n vh g % - - '
n9 EreL eP a 1 - dn et , d s . P - l ato i 4 _ V- M he el c a
*w t
ei n e _
, S gai r u t e
- sng
- don rt rgne r t o
e n$oe URW er e ht cr ehley re csdL o ni
}[ o ,,,
0 S st He P e f d 8 s oc o
# er S o tsi c g n
a et$t u ei _, El t aPasCLi l _ . sP a i g W A e t e a v hre t o U e _ hC nd l ePh t o e c l e v L eO t pe r W o ge e - l a C hgstcans yf t
- t s e n G r
7 i H (a dh r oe e i g e p d 8 1 H Sa 1 wg,e p r, l f vd eg oE e
, non are
, o p, gt r u
HGS 3 P p u g, u ni n g p,,..93 g fU i F
,, u.
O C .
.y
1 O O O
880 0 , ,< -
-- ;9;9;9; . . - . . - -
880 10 - l l l
~. '
, ,p , , , _... :l- m _ b
__.m Slightfy Sitty Medum
- Blightly si Aleam.m to Very Rne.___q l
~~~ f to Very Rne Sand Sandtn Dsm to Very Mne Sand l 20 - - 800 l
' d Sligtwy Pebbly CohJ to Medium Sand Pebbly Very Ceees to Cease Sand to Sernfy Medium to Very Fine Pebbie . 580
-Slightfy Pebbly Very -
Coarse Sand,- a # 30 - ,,,,# Sightry Sifty Rne y % - 580 - Sitty Medium to Very Mne Sand 7
~
8 . S!!ghtly Silty Medium
,'.'" to Very Rne Sand
#.,_ ^ _ .r- - 540 p
40 - Slightly Silty Medium to Very Rne Sand l Buty Very Mne to Sandy Salt ]
- 520 g sightty Slity Ceares to Medium Send D 50 - --- s00 E
................................................................T.atg,r,]ay,e,,{,,,,,, -M -
E 60 - sandyorevei
- 460
............ ._7......__ .... . g......y..... - 440 70 - -
Cemented Celeerseus to ~"- m aushey sitty sandy erewel - 420 7........7..... .?... ... oseene w tendium ..-. Send 80 9 . . . _ _ _ _ 9dughly . . _ Commmed _ . . .t. . .e-. . _ _. _.. . ?. . .t . _ - . . . .t. . .
- 400 su --. _ m esseseus 9.. 7......
. . . . . . .?. .7 . sher .. ..anty sendvh- . .. .}. ...f . ... -{ .880 ....
g ..... .
........ ..t 90 Cesrueto.s m.en.
wg.os 360 ses70ee1es.1 Figure 3. Geologic Characterization of SX Farm. The Cross Section Was Taken Through Tanks SX-108 and SX.109 (after Price and Focht 1976). E 3.7 5
% we si m.. Wrt
Figure 8. Simplified Geologic Model Showing the Main Lithologic Units Comprising the Vadose Zone Beneath the 241-AX Tank Farm Nw set o l -30.4 m l o- -ser 16.2 m io - yn Beckml y ,g, (1) 5.2
,_ (
b'. l.. . _ - - i
--':E e - H2 ss+ minor 12.2 m . ..-. ' , ..;;/4. ..' (3) ::;. lgr: .T. . if:P. -iso so _ 4:.* " **j'[ v. ,.j
-l. :.;.!,,.'t...
.. . . . . .. .m-3.0 m 4 ss + 3;,. .; r , .:. .
.. u; gr (4
'ih::r.p ?' ;.n.Y.l.=p=r. :: v.. ". . - ite 40 -
'j 91'i$. s4.
m J9:!. ' .Y
. N.M.
9*. .
.~.-
. MT- i.N,:.
.:p N < ..N..,
51
.b.: . N '.'. o(b'I'.d,.b
..m.- ,e: . .. . .1;,vis .Ti' ..H: .
1 - - - so - 6"1 m $' ,'l'
........... 6.'.d ' . . .. .v ..'T ".#f;+
- d. ..# ,. . ;* '. .. .. . .. . , . . -:. .
*. 4
- 8. j .,.df .g..[.: .~ '. s . ;p,.. * . : .g,s*; a.3",
N[L. .* *,4 ..'t s
- .'r. A ., . . .' ? . p.e.k.. r a. s . ; .M.
. p :. :,;.*:.' % '.' .. . - . .. s' ' .' ':- 1eo
**~
s.*;*r: Wd 20.3r:1.'m:- :*.v&>,.. *h. *
' :. . *:. @****,'0.).:.
- t. : .1.,:n .i .:. .
. y.<. .'e . , .s . : W . ' .:..;'.lY.
. ... z .?:
?:. ;;.ii
~i .*q ::. f
. s::*. \. . . .. ;., ;n. 1.
Q. ..., - ua ro _ **
.'t1'+.
. ,1.
;:.; .i.
.<r:....
m *:. .;::..'.'.y:v
. ; * .;.n:; \..:.,' ,':.::.
..,,.l..,. ,j . , .
, ,i ; W .1; .$ < u v.. .. . . . , -
9.1 m 'i'I' '. i ; .' c' r '
- i ... . . . ;,
.'....c.3.....'..' ....
M 4',C ([$ l'[*. $(('.tfi - t , ; +$ i,' '. t ,' . 9
^
4,'.*
~
3.1 m . A ') ,,,4* M IEJA (to),@ y j ,**t @ '? t993
- 1so to -
Meteet 1. mmel = meters obove moort see level SOURCE: Piepho et al.1996.
-w 24
% 7At-6 E.t,. vtw
Figure 9. Detailed Geologic ModelIllustrating the Distribution of the Main Lithologies Comprising the Vadose Zone Beneath the Center of the 241-AX Tank Farm W E
.' 30.4 m .'
o- - 207 18.2 m to - SP' (1) H1A 3.Sim . 7 m to 30.4 m es (2o) g . ,; _ . ,. --
- 10 0 go -
4.8 ma.5 ....,,,,,,,,,*?S**.'
' *21 ,'.* f.f1A e "
- ' **J .
-- t .(.2e.,a'.;g',. 7,8 m 5m ss+gr(2) ..
,., tortedded as+gr
. i' 8,5 m - 10 0 7 m 4',..*:': s - '
so _ .- si.i
,, ..g: . .' '19 c es (3);. .p......,g..., --
25 33 *
$.5m ,
- 3. m
.!:s:. r .; y y o - " * . . .t. : .
. - 1ro so - i;.V.i.d . ;7.'.i.Y [$U.;.!;;
.:: *: '. ;Y'.Y@.Q.' Q.
" o, .J:).T,l'g.;$ss (5)*.*:[I@,*',j 8:.*J' 11.5 m .
~
~?
At
~
HR 7mi N.solgr( :::f?:.p;;.S.;<:.p.y(.';.n":N:W MM.b'ON,
. _ iso M j.M $9
,w. .'c;;*.5 #' ' [. ? *[kf.%.$;
.u;G: %s,,m .'; c ,'Mt .Y.%., 1 '. '. %,.i.<.,
..~:*....;
,.s.h,i.4... .
n.'. s' .,: , .1 * -*2sz.!*.t- . i .. 1 r. s?. .g- ',
...s
' .# 4** ( }; * ' ; v.;;'..g, g g it-mge so --
nb:. m sc.,>. m . e....;e.q: c..r..:.y n'
*/ #
. .- . b' r:.r..e. ..;. . .%. .n*.;; ,a...
.m :.
. y; .. , .m . :;e. . .
(.. *
)?. .h.'s;.'. +.i,.$d:b.*80. 0: ,. ... ' ' ' ..' -, r.: a6 ~' 140 b ' ' '..., . 6. .
To - . . . s . < : e ; ,W,' .m ; . +
, . . . i
'L, . 1.3 m i,. s . . . , . . , ,
, , .. .; . 7,
, ,., . .. ,, :.,..j. ,.. .,
so -
- r. . . n. ,. . .n.
. - g .. m ., -to m 4.4 m ');~ *5.m.?'.Q l.R.S..'.1 4-x, 1.9 m
-u..~.. .
123.1 mmet (less) n?
+pt..?r7.'*W l . ~,' . _ ' l. &m'; '
W 'm' - me so - Nets mmel - meters eteve m en see i.v.i SOURCE: Piepho et al.1996.
,- - m 25
i ?li' 6 5-e R e s ht C u G c i t r E s e f o R . m o U o go)n d e N nt
> i pap s u
m ui m ( ( l
) e
)
PMr e d f is ue o
< Ao qm m d( e s ' s u
r e
)d e t an g a) l d 3 Q , :h.
r a )e ht a w29 t e 0 8 . p .. - y[i cr eg t 9 e y 1 A m t s 0 y q%F ,c . Rn l ap i n1 e ns rm t m i C e, 2 c o p o l a ' u u .m W M [31 t r )y ap t g m o ne Nn o m p r n 0 0 5 5 1 c 1 m c , r o p 1 2 0 - gn na a pre 2 . ig v n % E i in )2 mS t (o 0 0 1 rn es
)y eim c s e e0 o
c d t
?
di f o RE0 e e e 3;soqp~ nuo r a pp6, e n f ':1 e u6
. I 3
A e r . PA %c r A (s ( 1 e r a 1 0 .ef ( o8~
.s y . c 1 l
h y t ant - t t s fr oa d m ui3t o
=
a S r oe.rv. e e n p
,b 8.m P y a h n t n
.iy r m E^ L: )
h o y I ( , l s a f e s r m W' a. G5 i 4U:.^ u oK me'a c n V ,r d i Ewb s n v j ir s re e nr o fe i sm6[; ig,M[]Mg:Mn D et r y i t
.i 4 Ac t u aW j#g$%[lMqwy der) a (a 8 p i
- d. te s2 t
.n/u 3 If a em n
.ubm dn e e n mea V
1 @MwE{h%)nv. d ijt T. gi t l vi0 ie0 ue5 v c e r Mfg,,a e o si ra tgMy:f}, aZ $ i r C (r (2
< bMu dnseup mea
?
i h Q pn e n ead (L g fNNi3j&&m. N e e rc CV t
% S }s pn. ;
r N s h d l iph(te i6y
?p!p p
5 k l al ui a nr g s Rtn X s n t t ey o P.~t . i a pn lW a ra #$,~;; Sw@' Or f .4yyWQ t s e e mlR8 J1 r
.1 b: - .R x,jpG;?
r p,. nl oa c yd2 u) cd ni os d3 .:j ElM;,
,B i
.m. $l im Cr e l l l
s
;M 1
,s I)WDk y. $ .
iae7 t r i0 l E nla (5 M .$ G Cwfh@Cbm iN@fd<b l 5% j;y. I 2-l 4 2 e r u g i F
{ f ljl ;ll
= s;: ' ;
l- hl
~
O
=
v l
=
n s a i o ht W t a a e P c t i p e f a r i c d e i l S s r c t n P u o ' e n - i o ' g m d e a S R y a I NLI Ah t w a P r I llIIiI 'IIlI I e O t W a e h t f =e3 f o u : )r e n f o d e w i u i t a e fdn " o q t eo l A t. a F 2 a Z t ( ne r r e e ._ in t n x c a m o tu )e " a o tZ sas o 4 w Z n nd d n d c o u e C UaV " o t a r i t a ( y r G u m t a e
" h S c v S r
t e 3 a 2 d E C W _ R U [ MT"d - h " G I F
~
t J c [- - < ' ;; 4 . i i:- . pt : + <
,l
S O FN OO I NT OA I R G TT N AN I L LE E D UC O S C NE MT RN L OM EE y c r A CU C L TM AS ni eA gr Ao o WS D TN P DS
- E g n i t
nd o nI 9 N SA u N AN cd 9 o H e n 9 t U t ma A g 1 N VE t n e OE S y n e P n o 3 O NI s e R G O b Y l t at i a 2 E O AMIG D r g ni P n ed N N O e a U I O T h C mR n 'l J o T TA P O E D i r o v e A NR H S T ET n c i Z OO I I K A L SO U LCF R E OR A R W UE S TT C R PA E N CW NL OL CE O W
~
O S I S S E Y 4 L ~ I C A A N O s A D Rl e F S N P O T U Pd S O Aom T N R G n L E G Nos U OT K I i L a det l S E C N A El u o R O B DcM Ola c N I C l. A Mcd i S S N n m E C L u a C O I T Ao n y N I S E Cb E U R I r
- D R L T ei O Cpm E C E Ape F N F
H R US I O T P D C O - - - - -
i D O A Q }et
-e
. y,,o
,_a, A fn.f;g A
,m,.- c
[:
"$h ,'.
- h.tdh 4
.*a b,e,y ,
1
.- +
., ^
kN N .h' v-
'i[: 3'
- _ ak:.h,
,h, L $9 V lh ' , ji; j,,f x
t g g . y aa
~
p ...
.x . ss.
g: - hhfff !WN 1' g y$$['&ppi y a s ti p
+
w C y s : s"M g
-m[3['rMW[jh.ff
- ? *,
L N y w :q1, ~;
.__ Jw;dl +f .q3 i 3 .
.F x ;
l d- . I
OD 0 N = U C Z c s O R > R k 0 z
+ >
G C t p
- =
Z c s e z K V V c 0 x d C + e = n a A z x
) ,
C d g n Y oB v a n o = L
,l y a y n D A
o : x y . i ( s X C t a
- n l l
l o a a = I u. q V i t r r o x T E - i d n f o f t a E e )t o c 8 R C n / F C a y qk - Ol a B( r a 0
= = n =
K B d e qS i f. I s s R n c r q l I a u u l e o O T M BS o W
O L l l e E D W g O f r e i n p M i u q m u L : S A P _ O A P E h g h g C I T S u o r u o r _ T G ht ht C N t r o t r o A I L p p R E s n s n P D a r a r O T T M - - O
l lfl\! lljil lll l O R C I I F 0
=
U C Q R 0 A k : = I + t p I C e z G V c 0 x U V e = O + z x o R I n
)C d
n a g n o I o V : y
, l T i - s a t n D
a o x y T u q ( i t l l l R i E d a a V n r r O e c - o f o f o P n a )t C . S l 8 y r N B a / C a : e 0 q A s s B( d n c r = = R a R u u o o q q T M B S O l!l
0 L L E , n W i t o c d n G r e e s a N I t a w l l e f i r e w u P l l e hg q wu a _ M ) e o r ht e _ U ht ,
, ht r o n fe D
t P i t r o e
,i u i n
n o p mqa s e H s u D i t t a n l n a n a pi r e G , t r n t r e y r t e e O r e de h t fe n U D , c n i f t i c o i u e p c s O o cu hol l l e q l l a l e F c v , e R e n w t e f o pe
- r w n a o
, t d e e e V
i H r ht t o d d i l a ht ht i
, c a u fr e v wg n p f T. t y n d e f o o c D o o i
t c n i pd ht T n , i d e fe u o mla ht g u y a t a f -
- u t o
p e n R F G ( r r e g p t d l e O f , p , o f CQDVF,DDD p a s P = _ S m r ., N o C A F R O T
~ ~
O T L E W G N I P s M s l e U l e d P d o o M H S M G L i c d U E n m O O D u a R O o n y H M B T r D F e i p T O p m e R S U S O E *
- P P .
S Y N T A R T O - ll
O _ s _ n t n i o e t a L m er t r H e g u s a e e n a D r e n f M o c n o O vh a tg gn S e e r c i t n o a c e M i nle v n i g n v o L c o l ba )s b n o D oe M oi t r mrec pr a N n o ms = n o mn t OU i t a u e mh i t u e mno c O l u i xr a e t
~ l a
u i x c H l c mo v w c a mdn a l o l a R C e s n F c e u ht h o E f o ei t o r e f t b o te r P d t a a r f i u e a e p P o l une c t A q s o u l p ht c u U e l a c c n o pla( r u c M -
- c P -
O) L E D ) O R C V MIF / s OI a TU L E Q S R RA P (
, 9
( H p , x A y SG E L R p LU C
)
V 4 I O OD R / L l
+
OII R l] MT : P g p CT n o ( t P g x E I R i o q MO t u q p AP l o = x E NS S ) Y N A l a c t ( e r D i q e I R t y h M T l a W c n i A OS
~
L O L r
)
l
) c e a n i l m o OW T d l
i r i i n i t t a r SG E l l e m s e n e R IN d wih t p r PP ( e s n de p f i e SM e c l u q a LU l i l e f C I P y w o b l ht a D I d e t n n p OIG n l l i r e m od e g OM U i s D e r c n i t pi e t r n CO e i me I RD M H no n to s o e e r ut o s s i u u s d A T i l t a d a c e An e N T t a ug t s o r i s o yi y t e t x Y R r u e r c t e t i t t n v e ht D O l t e a i l h a a i t g u u g nt t a n r u q q a g n o I P f i i s i C o n r r v e MS n o al a t r c dd i n e t e a t e a rl e n EN C t s pA i g p e ww n s e e r SA n y( i s e D e - - oS c R e CoT I I D C O T W
)
LL O EL DE OW M O N G TI SP EM R ( PUP S I
)
I I KG D oDU n o V OO i t
, n o D MR c n G y i t
a O~ - C H F u ( u V I T f q f E MT o = l
=
AR m , d e C NO r o C o YP F M DSN I M RA E T O S
I Il ) o n i o t l a a r s n t N i o o p
)
O n e c s e u l a s O t a r s e d i s T n o l a v s o p S t n uf o S c e i I e l a C e s d c o R n o v e t o s Rt I h d f o A c e s o c e P ( t e a e ts v o P ht d e ht l l u c i t a e c 6 Ml d e t e a i t v ts a o d e o l a c v r b e l a O o l u v r b o o s n n o o C Ml c a c e s n l i y r Mt c a y s c sa y e L d~ o o a i e l b r n c s n E u t y s l y s e c ml a b y a n o i t D l r a u o a e e v e n n n s a s O B y r o n u y o s e r e r r e s D a s d M e pi v s i e v g y a m i mig e r s i e v l u o c p e
- U*** S O
)
O T O S E R P ( S m T 8 L 1 d _ U n a S ) _ E M B m m R j ==7~ a. U ( u u c c N I 1
/
i t n
/
i t n
$.%:.m a.
= u u S _ 6 O
a , ht 1 5 E _ J: 9;, g n 3 0 C = 0 0
= %
, e ~
3 l N
~
e n E u m u r ( a e e r c l d e o l R E l P f a.{
^
c. s n s m d md o e o n m F e jE:.o0 1 , i 8 u o O F I l p t pt a s b r S T D m ml u e l t l u e E p R a p s w s u P x s e E A R O
s d O l u o c s n e h r c i o e r l y oh i t t a r e x l a s r e v mw ys nl e p e . o s l t e s dl e h c n m od s t e s g o o c o mo vo c i l d e S i c y g s c m l v r ei us a i N mx t el e b ta t a m al e b o yv O I l wer t d n ym p ,p e s si a r mesn l e e x o udp . S - ot s l i c o u e w s ma d c s O U o yc o t ei e s g ys t t a m n s s mlbl L l ut e di n s a e o o c a a s C u n o i l c s a y d a d i sb mo a sei p N c se o m n u v i o l d e dn a n y rd s O d t - u b o a r to o s mltu d-i s e e vb l C e l s p e d d miga u a p e n se e t n o l ps u r si s e m u u o e at u a d o d y b v f e o f r el r t i ob r - md i f n o e a pv e s e x,n i De p i l pm pr u e ul e pl t s 3 m mmo n n s h muse Ap u T o i S c Ao c . c r 1 2 4 O 3
~ / '
O e d i l n ci m udH a ne os T s a yr go iUI ar a t d S po b r a s F a a RleF nL a
&odO nan l
O - Gio eMDA r .a t H N t at r d ne worE d pS s nn a og urA n Y n aR y r o ur C t a r o oT r b L l a a G i t N n o a O N s e n n o g _ r A
/ - N O
s e e r d C o h f o [ s y _D l i
,A R
gl O
~
k f m 1'~ 4 {ER S m1 DS K. a DE AT E - AI NSO R F I RS EC J A RO E SFF 'J L A L RE DL ,
\
A / R dme y S r o E t a r o E b R : L l a a n i o t a N
\ s / o n9 e
n g r A
/ /
O
;s J y
, I ,~
e r K s y
=
ht = (
= v, R,
i n g, t r e t n on po t e s z i o nme e t nde n v t a a hc r r o o O t t a n c m i h r r o e r o w e tui t uts a a ra t t t a wsnt u a i s e ou t t e s ae d n u s mbr i r v eo e et u t t u nr t um y o u r ot e pr t o a r pl i pc a e r o G mo mt mt b L a o s o a w oa l a n Cin Cthfo Cw i N t o a O m m m e n n
\ ' g r
A
n y v, 8-e r s ht = y~ v, i n t r oe pni n s o y nz a t i hc rd e c t l o i h O r t a e e r v w a tu t y t a was c r a t e s ae d D r v n e eo u t ur t um y r o r pfe pr t o a r o b G e mi oq u mt oa L l a a n Ca Cw i N t o a m m n9 e n s ' o g r A
/
O e N n N r X o ' o u d t z e dt f c a n l e c n a l f t u p a r
> o a t u u N i t
b u Pd o l t f a e a s t d - n u M r a l i C u P t u n D e n e i r ht - - - - n 7 O i t r c" 2 z 42 ) o p e 3 O ) A t g R R a s d i D R (
- 4 /
t, t, i n l c = p v D a r u )t + " x e 4 - 4 x t n (z, c" 8 d e = z O z ( e i oS+ ; + d
) 4 d d =, y R t i
l a
+
e, z y r F,+ r c 4 = ( y 4 t, o u '" t c, R D a a t r g x o n n 1 R 7 , 4 b o a e ,
+ R, g
4 L i r " z l a a t n d , c" a e o a P , O 4 = i t a O R m y N e n n
\ ' / o g
r A
- ' x e p 3 P o
2 s 0 G n i. s 0 K K o + + z , c O 0 d 0 0 X e t
=
l u a r x ' d d s f u t u t l f c u a t i, g p s R t n u 2 2 2 u u p ^ V n A V O n 4 e i 2 ht - i, 8 y n C o z
- i t
r e r d d
, =
G o d i D Rs pl r 8 s c , n u i. y z ' a r n o+ c d o V t i o ( z
, S
+
2
?
=
e d a s+ , v R a y a F, d i r
=
C 2 y l c y r, + ' R s t r o u 2 a n 1 r o n e g ,r
+ b a
i o o i. L l a n d r r c d y o i o a P , d 2 i t a R m s N x N l , R i
'gor nG e
n A
/ /
O e n o t r z e o d he t p s e ht t f o n n a r a r ei d u n n t n t a o o d d a s si t s a n n y n a a a u )y ro t t n e emnt nr e a ec n n e e g ht oogf zf s o e ana r r o r n n e po p po O i en haph r r f i s f f i s t t r e t or oyor r o st h e y e yny e p r st e ht it pt i s e t p s s e s f ci c gi n t n o a o od o o ci od a eg r ts o el a epel l l r l t r ( xd ei xt wv n vf ot v a n e u ne e rd r t i t r rd i d i l f udd o oy u pi t o u y l c y s ,nipt s si s c si pt r o u nt a n ng ng t o anal no an a r n en ge e c r o r e r o b o a o or nnt l t vt l L l i d r PpzC ao a - - - a n o a i t a R m a N O N \ dr e n n o A
/ e n n o O n o i s o z i s f r e r e o p d p ) y i s ) t e i s '
- i t
c d A r 2 a r d - e l o l a e u l a - e T, n ' t n x v T 2 i d a s i d g t r ' u ( e n u i o 1 i t A i t A p - _ g u i g1 2
- s 2 n1 2
e n n A o 2 a 2 1 + l 1 ht l o t r t d n d2 n d n T' l i nN a O 2 2 i )i, , a g a a - _ y R s R s t R R n/ r t t n T,' e / o p n/ e t z
/
t, e 1 2 g t t t z s r a v D x r a
~
p o r V m D x n p Q
- 4 p x e p - 4 a
r f o~ _ f o f o p t y y Ti i _ y x e e t i c i t l i t t d t, c c 2 z i l o i, a D o = o a D x y c l e R x l e , l e 4 r o v v t 4 u j F, v a r o n t r 3t t r y t r 1t b a i o o p - o n pe o p
=
L l a
' " n d s F
s g s F" o a n n o n i t a a a a R r r r r N \ ' t t p t n9 e n r A
/
O e N n e , o eht D z ht & d g 2, s e ,n R t et i a r d i nd ai
;!. .j!ll!!j!
u l c tri v _. 2 , t D s D a uod & R s n pb r o & r o & n mu
, 2 g n.
u d R R s R s e nis ahy ;!. .' !ll!!l! ht no t b n ob d D s el e i O i t r r r e e
&n, o p ad R s
p s no di ome i s ;!. .l!' l!!l! n t a e n a l o , t r ad nr b z D e dt a n d i ad c e s ue t R i l c i t r gic i u na o r
;!. .l!' l!!l!
t y r o u nf t t a r o n ociaa D
; b a
o L eus & L l a i d f pi t n ,, n o a I s s u R s i t a O R m N e n n
\ ' / o g r
A
/ N n O e o n , i t a o r _ z t d n - e e t , C a r c z n d O t u & s C a D R s + e c a y ht d B i n D R s X t r o [1 H , c ,c.
+
O p e t & B ) t s A n d U , s - i DI D R ( a r l C n = p x e t u i , P , e n c h d 0 z
=
d i i o v & s + d
)
l d y R " t c a )
+ + 0 z y r
u r t
" =
( y t o a n z c r t ( c" d o o n S+ 1 R s
,o b a
i d e r
=
+ R ,
L l a a n a F, + a c d t 0 o R P i B t a N s v e nO n N ' / o g r A
/
O \ e n o z d t e , j a r I v v u s
, 2 1
D
/
L' t s D a R R , x x s /
/
L x, - x, - t, t, o z , e v m D y C C o ht 4 p X
^
$ )
i n f s f r f r ( _ r R e e O e _ t . r - 0,
- 2 -
o p s L s e R L v v n d i ,
/
t, R s L 2 I
/
a r l c
- c v
, /
t, 2 1 L
, D x L' D
x t u , - D x, + x.
+ ,
z v. e n *c , , L 4 y o o d i i o c
+ 4 C
4 C 4 l d
- x ) M(
c a
) g ' g y r
u r t. A f r e f r f r t o n t (
- e e a r
o i o nx e p e
=
c
=
c
=
c b L a d r = v l a a a y n o R P c i t N a O s e n n
\ ' / o g
r A
) / \
e n o . z l d e s l e -- e r w t a e y e r d e vnehi s en m t u s ng t s i t a s u a r ot e nt n e i s t ep r ua pin i t o - ht nhd ot n mm t a n i s f a oa u i s o ct p n e t gnyo n o r m pnel l o a fc i s xl i r c r e e a pi t c o n e ed o p t of ol yn o s a ai n i m nnt - d e i mt x i o a a r w o s r o mit s r r e me r t r i n d y r e pe e u o v pps pec s nta c t a r o s b n a oi oo a a r nt d t c wto L l a n T A - - - i t o a N m e n
' n \ \ /gor A
l1l '
- / N O
r t e d y a d
)
w z y l , l e ( x w c
, r, i
n r - d .o n = i o ' w t a r O t nn e d V d,, c U n v o c
=
r ,. 2 _ e . d i l c u J y n , r o o t a r i o b d L a a l a R i t n o a N Ox s / o e n n g r A
/ e t N e n r o o e ht e p z s d f ht n t e o n a a ei t r d r u n n d d n t o o n n a a s si t a a )y r t s a a y e mnt o t n e c n n e ht nr oogf e a e r e r g o zf s o e ana po n n r r e poi p htenaph i t r o r e t r e f orse y pt f oyor f yne ty ep i s e p s st r sth ht t e e s i s i c gi s n t n o f ci o lod o o r lod ci a r e g a r s o el a ep l el a t e ( xd xt ei t wv n vf v ot i n d u ne r t i rd t r rd i l udd e o u oy o pt u l f c y s,nipt pi t y r u nt a,o n s si ng s c si ng t o a n o en anal no an b r o ge e c r o r e r o a i L d or nnta l t vt l l a a P r ao pzC - - - n o R i t a N m a \ \ /, g ne e n o r A
/
O
\
?
=
e e y n eht c o ht z g ) t d ,n z et > > > > 6( t e d nd i
,!!llj!l!!I a i ai =
c ri v l t r 3, u t uod F, a n pb s d mu ? - ht e nis y >>>>=>>>> ah ,!!llj, , !l!!I no t b - n F, i i ob d e O t r o s r e r el e ) p ad t p , s s no > > > > S( z n a i di t om ,!!lll!l!!I
=
r l a e e 2, r ad b i t F inr e n n o d i dt a a c z l c t uer nd ?
> > > > = # > > # y r
u i gic i ot e v
,!!llj!l!!1 i
, t o
a n ) , a z,+ r t F, o o nfi ta r b oce ( a i d a L pi t utuS= + a + L l a n o f s s s , i R I , t a F, + N s O N ' v
/ o e
n n g r A
/ ! N e . s nrs O s x oae r , , 5 l ae . x 7 x6 l aay i t s 8 1 nvn5 c u h gr9m1 i s 1 0 f d nio6 2 0 i . o A o uas 0 0 0 u 0 s t i t g r r e 0 0 0 l f oo2 os i nse a 1 t p = = = i n =5 tnle t a c nd o ni s x z d a o oer r f e l ad aaartarT C S f 0 9 E - f o m l e i f o r u{cits 2sE O
, a 2 ,
o p , m o o 0 0 O m o o n . i o t a r t n ) s le i{ e d o c m ) n r e s r o h t o a e y c s a ( l d R e a r e m i T m y a r o ( s o p D A l, o t t a r R s.
, s o m S E e "
r " s e n b e R m" e - L a T
+ - 1- -
l a _ ) ' ' 4jl n
, e ' 2 , o O 0 . i t
o
. =
0
" 0 0
0
,0 0
m o N a go?8! s N \
/o n9 e
n g r A
/ e N O
n n i o i o t s g u r d m n b npe x k i r i r as t 5 4 u s . l ad i x x6 0 d i s ,0 1 ht dt t n00 i nes 8 5 7 1 1 0 t d r 6 2 0 af o nie7 ec 3= ue0 0 0 w w0
%tc t
o ri = 0 0 s r g f ef f uugn s i
= = =
3 2 e n f i edd o na x y Z cKKlor aaa f f i d t of - l i eE - - f o r a gS egjc6 gE O
, 5 4 0 2, 1 o, p , 0 0
0 0 0 4 o*
~n o s #' c ,8 0
i D A m o
, =
t a R o *^ s ,,
=
=
S E ,s "" A s0 g, t r R , "1 0 7n l
=
=; =
=
a _ n 2
=
e *- _ ~ = . "
=
D. ; ) s c 4 r a n o c R - 0e 0 y 0( 4 e l im a r " T y r o ,'. O o
- t p
f /j" a
,L r
> 2 o
m b a e J-L T - - 0 l a n s 2 0 2 s i 0 1 s 0 0 i o _ 0 0 0 = . 0 t a 0 0 o 0 0 0 N ada jb5 oo e O \ s / o n n g r A -
Ground-Water Flow and Transport Parameter q Data Sources used in the DandD Parameter Analysis C/ Public Workshop on Ground-Water Modeling Related to Dose Assessments U.S. NRC, Rockville MD June 23-24,1999 Walt Beyeler Environmental Risk and Decision Analysis Department Sandia National Laboratories Albuquerque, NM Hydrologic Parameters in.DandD
= Behavioral - Controlled by Critical Group Definition
- Uw - Drinking water ingestion rate Vdr - Volume withdrawn for domestic use
)
- IR - Irrigation rate J
= Physical - Controlled by Site Data i
- H2 - Thickness of unsaturated layer
- N1, N2 - Porosities {
- F1, F2 - Saturations {
- I - Infiltration Rate
- Kd - Partition Coefficients I
l Behavioral Parameters I , l m Uw - Data from USDA's Nationwide Food { Consumption Survey for adults aged 20-65 l i a Vdr - USGS per capita water use data from self-supplied water systems t u IR - USDC's Farm and Ranch Irrigation Sun'ey
Physical Parameters O Q = H2 - Compilation of USGS and state observation well reports
= N, F, and I - Correlated via soil texture classification
- Carsel and Parrish, and Meyer et al., define distributions for N, Ksat, and other parameters as a function of soil texture
- USBR tabulates percolation fraction as a function of permeability
- Application rate includes precipitation and irrigation Infiltration Parameters UM
/ d-ifs Y
,O ;2-y Jgg lical e _m a\[ M E +
;E -
Partition Coefficients ; l
= Thibault et al. (1990) c6mpilation i
= Nuclear Energy Agency's sorption database (NEA,1989) !
l n v
l l l s l References
\
e Carnel EF.and R S Pamsh 1988 Developingjoint pmbabihty deswebunons of soil mater retentum characternatus Weier Amcwces Research, Vol 24, ho $.pp 7M,%9 e transport Meyer.forFSDMPD M siess L bdhoMand G W Gee 1997 Uncenamty enalyssa of mfilystaan and subsurface How and NUREGCR4%S e NEA(Nuclear Energy Agency, DECD) 1989 Sorptson Deu Base e Owenby,3 K and D 5 fsell 1992 Climaiography of the Uniwd Sutes No 81 methly station normals of Oceams and Annosphens Admmetrapon Natsonal Climauc Deu Cenuesemperature, preopnaimn, and be Ashevilk N C
- Lh4 TMbauh, D H , M i Sheppard. and P A 5mnh t990. A CmessatCimy=krume=f Atwee ofikfatt Sm/
Canada l'armem Cwpcunts K,JLc Par ate fewrammewaI Assessavnts, AECL 10125 Alomic Energy of ised a UsBR 3993 Dramew Aemenst(revised reprmt)
- USDA 1993 Isralladales /seenikasis es da Saimea resy /M*.aa Asrume=& 1ma/Cmasempram Snrwy
/M*./Ma Report No 871-1, USDA. Consumer Nutnuon Division. Hvansville Maryland e USDC.1994 Commerca.
Hyanaville. Maryland Economics and Sutmacs Admmistauon, Bureau of the Census ensumer huirn
. a
- USGS 69w0s "Esuranud Useof Wsier in the Umsed $utes m 1990,* USG5 Nanonal Cucular lost i
United Swiss."Open Fde Repon 624-H 119 76. U5GS, Denver, ColoradoUSGS l990b
- Annual St Y
l I l 1 l v I 1.
ll, l$ eb
?l[L $
mll$a ~
,$$d,.
l} l rTA
~. a a v
.~
i.0
~
em pb ni sl a ol P G r e a e t c i f id n A nt a D c o aa N n l y b o a r t W s is h wG ses e e s e Uf o O t N & t i l r a t i P 1 i t yZ V o i n p l i a a a n n P _ D
. d d a
c a i f i c b M LiG N o r t h o r a t y e e mr io t u w r e s t o r a t n N y i d o nw a t i o n a l L s a t a e Or b o a t r o r
- y
y r o m E8ay - b t r o a
- a L
l n a s,n dni n n i o ni ost a a st a dni n a i o t ea ast s t a i t u ad t s ea t stea n N a n ad n ad o m r ib t r s s c i mic t i f io i mic t ut f i i t o mic ut if i b t u t s e o i i DE e p b sE ce b se c i r w f i r p iE r p t s h d n t y s s t t s s s t t s s s t t r I n e e i Bi D e e - i - o n f o iBi t e ,s t y ,s t D eBie y , s t t N a n c r sf e nl e ueo i t n sf a u eo e i t n sf a u e o t a r e i f i c s i o Ur Vn ac t r e cVn l a c t r e a c l e c n c a t e gs e cV n e P e c a t einb n U gs inb e n U gs U y c r i p md a u n ae f i yd a n e if d un yd a i nb e f i d o p r h do ot uh u A a ot PB MB MB o ot h M o s > > > ;> t a i i i a i s s d n 5 6 . O' S
, e c
t r e t s f o 5 gGI eSW r u e p n i me s6 o aEx t e o t a m f ocR r r uC/ m s,T G, oeA r CTS St S y l a ,r m e r u s r o n oS oGE f r n u SI,W c et s oic u s a e i t io o eN L vd e a laR r n a S sa f t n mnU meI m s r o mlab C o S /y I otiN r t c oo nDD.C r t a u n f o f I nNA, cl /D a f i i nt a aN r f i n o is ni o s n v e r ir d e s nO Nh t o i d r l a/CO a e n i ic e p e oS niN e g cl o a niG l mtxU r f i c ni t y GeU L o oR f oE
.,e s p
e aa T R g ig S. U el S i n gi
. c s-e sm t
R aS la ene it e N c S o g er L A l e po l e yf _8 m gg) * &k# l l . Tin a B l
e OMH . m VGSC m a a e: l
=
e e o D l e g e o o s' im a i t a t ag t Dt c i a o /ciat D i b n Ha a a Dy d ( s a r Sa t t e a o ro(F N t l ( o ga o e Tg r ie /c r e a c d n DrTn i s s a a g f mt a n F o O s i s u r s( iT r mn ic T r o a s t a n n s io i F s o n n u m n) p s P n s iF a c i f c t s u o r i c i o o in t N n c o r t ) n t i o A h F n t N w s n u )n a l a c y t i o n t i o s i a l L a ) n s t ~yro
y r t o a r o b a L l a n i o t a N t s
) e T h w
dr E- t r o N n o n o c i t o c e r
)no i
t a i f i P c a a i r et a i t p S t at s s n A Ai mre t a r O c o N t s t l a )t v p a 8 e e u i a g w p (E c n ( n e i f o ei r ) e f O/ L t r atet aE)uT s y ma r a t si t cl i pe (n T E b e (W C n mi o( ) T a r a ei t s c t t a e r f iTi o e ad eE N a cta n da pe (y t D i c n e eipMiRSi t c i r pi c/ x r ad n d i i t e - a S e a m a n c teP r MSWH oi u e m e oi l l i l GL S * * * *
- HeG a
C m a s B
i B a e ola e l > t"
* * (
*d PSS m
o S ws i eEST i, l e s l s e a o aDUA t a t i s e l iORT D o r n h tx t e deNG / S G a t a o l l u o a sNO O a wie d r b (m as i t a eg t n p i o o n r a(m r c i t a ep e t ap a d dp aSe i b a t ao n e tc l i f i ag - t i dy oC u p e e r h b n O t s Sa u t i t h t i or a n b im a n i c o a St e t s f o e t u r a d e s r i s f vz eat s o
.e c
y i e r Nno m a a i - r i c n A o r t h w S De fo
! a s e
s S a t
)s U t
n
)
b S a t i a ) A o s n. i t e o gt o r y
y r o t o a r # b a
' L l
a n i o t a N t s e h w t r o N i c f i c
) a P
m e t e e s s s a y a b S b e a ot a s s G t f a a t a I nDb a s e D r r e e a sh yT p t c t t a aDh t a p WWy p r m i g o t e iPg u o r a te u nD ta r n a o e tr e d ot l t y iS t Cb a S a H a / yT /y w / N at gd c ( u C i r o un i el g Si dv a t a eo o o I r e o WdicWGG NIn Lo* l l e *
- e H
a e G m m s B
li oa , a SL i a N m g V e t e a t r n o e e e N i t t a a d a t t i n i o d Do n n Ct a a a o D u - D a t n Ra a y t g e a t b So i b a a o s s in e e Na e Ev A. O g e s c a o i t ay a t i s b o e t e n a c i P Sm u A f i r s c v s o r N y e e t h s s s w e s t N m a t e a n l n t a l L s g a o r 'y
y r t o a r r o t e L b a e l a n m r i t o a s 5 e N e d t e s u ' n e w a po u ht r v d.ba N o y t Se I c oWia i t i f Ni mva i i t c a d P a d e - yit at s mdt o n a i ea n c t d e s mc i i o i p s t y eii g a t a t s L a a ob i m a b a a o t o t ef n i_ ataDgi D D er o ta e r vd a s e s eyD a l i oLH b t t a aS e r - n t e n- a a Watn aiPo D ai toWt c a n aSr tu a n n o iNGsi o o t e i a Sn o a t l i t aNAUcN h l N a a a m a B
a a a a a Oh r S u qeipSacBNHL e l s o 6y a l e u t ep ut t h i d g r m ris ep i i eo dnw e mg mro e m a de w r i c i a a e o e e y nf i nt d a sc dh aio t od e t g l _ ma ar e ie c g r a p r et a n d a c s ed a a n a ds n a d t st e a o s s es s t a s if o n u o t r a s r d i s O oe ed t a v e e r d v a st ag c a e r r ig yf e ms aa t a e o n t a s y o t s e u y i t bu o r u t s sa l n P a eb ed d c t n c i o s i nb e a i o o f i c f e a t n w N o g n p u a r e t p s b f o o w t h n bu r e a r n e o d s s t N a et r cma i p r iw e WW e e t i o ai t t b b n a d y ei l a u L a t ab y s gt e e
1", t O METHODOLOGY OF SITE-SPECIFIC GROUNDWATER MODELING S. P. Neuman O Dept. Hydrology and Water Resources The University of A.rizona, Tucson-O l
RENDERING MODEL COMPATIBLE WITH DATA, PRED4CTION AND ANALYSIS OF UNCERTAINTY l l T ke SCALE (S) OF MODEL Resolution /Parameterizaton CONSISTENT with SUPPORT SCALE (S) of Redevant DATA (Those Denning MMel Struolute, Parameters ForcinD Terms, State Vanab6es) and ther STATISTICS?
, No o
CarVShould Yes Relevant DATA and tenir No f- STATISTICS be RESCALED ]
+ b FR Moder? t Formaty RESCALE Re6evant DATA Formally RESCALE MODEL M FW and ter STATISTK:S b Ft Wdel. Data and theer Stabstes.
L J Possible to Treat No Treat 1 Spatiary Varying Parameters / States as Corre(ated RANDOM > Amtxluity/ Uncertainty FIELDS, and Model as STOCHASTIC 7 Using FUZZY LOGIC. i j 9F k yes Stochaste g
\ ,
Equatorus Usang (Condsbonal) MONTE CARLO SmAation? Ra$uce No e TRADITIONAL Yes r DETERMINISTIC , FORM 7 ir u v Formu6 ate Monte Carlo Numercal Formulate NONLOCAL Interpret Traditional Deterministic 3 APPROXIMATION. Set up MEAN/COVMANCE EQUATIONS Equations as LOCALIZED MEAN Computabonal GRID. POSTULATE and their Closure!Numencal EQUATIONS. Formulate then J Muftwanane PROBABILITY Distribwho ' APPROXJMATIONS. Set up NUMERICAL APPROXIMATIONS l cf Modelinputs. Computabonal GRID. Setup Computabonal GRID. )
+ +
v ESTIMATE Structure, Values and ESTIMATE Structure, Values and l l Generate Mulhp6e RAlOOM Uncertainty of Conditional (Inherengy Uncertainty of Conditonal(Inherently ( REALIZATIONS d WcM trput. Nonunique) Model PARAMETERS. Nonunique) Model PARAMETERS. SENSITTVITY Analysis. SENSITIVITY Analysis. o * + COMPUTE (Cu.JA.,,4 COMPUTE (Condibonal)
. MEAN, PROBABRIrr anser MEAN system Response / Behavior, RES SE/B VIO HIGHER MOMENTS OF System a s mate UNCERMN
, Estimate UNCERTAINTY OF these RESPONSE / BEHAVIOR. th EAN. l MOMENTS.
+
Do NOT
" I CONFUSE Uncertainty of Mean
' Observed /Anbcipa+ed System ;
RW eMdWS@ ! esponse/Behavor.
. . l l SKELETAL FRAMEWORK F@R CONCEPTUAL-MATHEMATICAL MODELING I DEFINE CONTEXTUAL FRAMEWORK l DEFINE AVAILABLE KNOWLEDGE BASE PERFORM QUALITATIVE l CONCEPTUALIZATION AND SCREENING OF HYPOTHESES 1 ir I PERFORM MATHEMATICAL CONCEPTUAllZATION AND i QUANTITATIVE EXPLORATION OF O HYPOTHESES 4 RENDER MODELS COMPATIBLE WITH AVAILABLE DATABASE PREDICT SYSTEM BEHAVIOR UNDER UNCERTAINTY PREDICT HYdROGEOLOGIC PERFORMANCE MEASURES'UNDER UNCERTAINTY l O
% 4 DEFINE GONTEXTUAL FRAMEWORK O 4
( PURPOSE OF MODEL/ ANALYSIS 3 I 1r c HYDROGEOLOGIC SYSTEM m it. (CIRCUMSTANCES AND SCENARIOS ] NT HYDROGEOLOGIC ASPEC if DESIRED O RELIABILITY (ACCURACY / CERTAINTY) OF ANALYSIS l O
\
VAlkA { 4 .
. ?)
O . C ASSEMBLE Sitt M@ie$$AL B4fA] . . . . - .. > A%SEMN[lldN@UIRI KNOWLRB$i BAtt . IMPROVE KNOWLEDGE BABE AND ITS INTERPRETATiDN O O
4 4 O ISSUES OF CONCEPTUALIZATION AT MARICOPA AGRICULTURAL CENTER (Analogue of Decommissioning Site)
= How well can site be characterized based entirely on publicly available information? )
e What alternative models can one validly postulate basec on such public information? !
- What ambiguities and uncertainties are associated with these alternatives?
e How can such uncertainties be quantified? e What tools can one employ to do so? l e What tools can one employ to explore alternative models and their uncertainties? e What are plausible modes and rates of flow / transport in the vadose zone? e What nappens to plume shape / concentration when it reaches the water table? !
- How to verify with limited resources? !
e How to reduce uncertainty? ' e How to simplify flow / transport representation O in reaiistic/ conservative manner?
O' i n SO~URCES OF INFORMATION Reference Maps
/ University of Arizona Library l / www.srnr. arizona.edu/nbs/ gap /nbiidata.html Meteorological Data l l
/ Arizona Meteorogical Network (AZMET) l / ag. arizona.edu/azmet Irrigation Data ,
/ Monthly: .MAC Administration ;
/ Annual: Arizona Dept. of Water Resour.
/ Both online O irrigated Land and Soii ryge oata
/ Az. Land . Resource Info. System (ALRIS)
/ www. land. state.az.us/alris/htmls/ data 2.html
! Well Inventory and Groundwater Records
/ Ground Water Site Inventory Database (GWSI), Az. Dept. of Water . Resources
- Well Logs and Construction Data
/ Az. Dept. of Water Resources (ADWR)
, / US Geological Survey (USGS) Pumping Rates
/ CD from ADWR; MAC O
1 i
PERFORM QUALITATIVE CONCEPTUALIEATION O ^"a ac==="'aa or aveora==== (ON REGIONAL / SITE / SUS $1TS SCALES IN 3 N CONCEPTUALIZE l HYDROGEOLOGIC UNITS / FEATURES c CONCEPTUAllZE FLUID FLOW m u NCEPTUALIZE CONTAMINANT TRANSPO u PETUALIZE HYDROGEOCHEMISTRY ISOTOPE HYDROLOGY l v CONCEPTUAllZE TEMPERATURE AND HEAT FLOW ^) IDENTIFY CONCEPTUAL AMBIGUITIESIINCONSISTENCIES NTIFY DATA AMBIGUITIES /UNCERTAINTD POSTULATE ALTERNATIVE CONCEPTS / HYPOTHESES ( COMPARE / RANK ELIMINATE HYPOTHESES]
PGAF0ftM M&fMEMATISAL i 00NORPTUAbWATION AND O a"^""" EXPLORAfteN OF HYPOTNESEs k' ARTICULATR GONUPTUAL MOD ALTERNATIV84 MATHEMATICALLY k_ PARAMETE P t NCERTAINTIES 1 IDENTIFY CONCEPTUAL-MATHEMATICAL AMBl0UlTIE4/ UNCERTAINTIES l 0 - POSTULATE ALTERNATIVE CONCEPTUAL MATHEMATICAL MODELS EXPLORE 8ELECTkD ALTERNATIVE 8 ANALYTICALLY / NUMERICALLY 1r POSTULATE / EXPLORE ADMISSIBLE SIMPLIFICATIONS / ABSTRACTIONS OMPARE/ RANK /ELI INATE ALTERNATIVE
, = - ? m ; '
e-
; } !)l
! ; i,I 4 1; m
E e nl i l l n 2 a t a e o S . e r SMR i wig t a 5
]t -
- w g e-
. A i s .i r
r M-L~ W
- a p
s I _ o i c nm pa et.. r :nr S
=,R:nwy$wMMq .N nmx:mdg j'
- s m
J . r 1 " a .e M%.tw:x%e7_4g 9 k Mm b.NeMmwn@bwgO a wm yw. "i. s W i q i... q e w ee
,p M ~ ,%g E 6 m c fmp=d^y l i
M gvg ,h F u f ; i u e W:fx .n.k n r%Q.h= g4 3 , . uB :f+ $pi 0 ey;,?-7gwgsw.,4.';.,ha 3,: gum,,ec. ,h F .keI JMg
; 1_
i .
=p4 -
n p$$Wn% Q . hb
- w. wwn- _e,gm Mwn,d o $w. p=l'
; ,. ; r '
dyJ' -
,A .e
$s i
t % Ra nnp,y.',h a-h Gww. n u . ; m
]v s*,eg$n.M".p
,v h
f b i 9n 3 ynsq a.2w9 t o N. - r .- t 1g3vfM4 s !g m.; I9 n
.yMms~-
i d &',y:p;;.g te' 7
~ cx 5 _
d - NgiW. n i-1 p a l. l km%pN.}g ^sc44.y u. _4
,~ .
n 4t~c.
~ _
e.- o
,s:wj.
,t _v _
i . t a g -m N.
. sa y%.t mMgr...g s -
5gy.&- eqqs i r r 4%m a l vp g ag jq% y - - i
%w
- g ,% 0 d *
- .mj&;y94~+{x gy _
n n. w ,g w'?.
- f1<&
a rrm r 3 7B 2
'h
. p .
s e g,ng Lg4 ,ggG. ,
.m;yp.,~.$n. _ -
f ~ . l l os S@- . x =% s . e gge yg ' n~2g6y% w y, :y~3 gm. , , . - $k . w OU w e gw% Beg pa.l mgk@s e
- vw pM , . j:. ,q..
o m:p-u. Ss.gX. n h . - t -[3- '.wgi ' 2 T
. ?
- a. n4~
Lwp 6
*e 4
^
<OT .
,s
{. 5
1\
;!l !
li li
, i 1; .!- i'
,.I
,i.i:
E nU_ i s et rl e5r t zat e SARSw i t e N S
/
A s . W
- s. 4 O s:-' s
' , ~
= .
.J . .
f c:
^? >.' - .
f' -.
- a. ,.3 . , ~-
l f "~ :.. s m n
,_'u.
,d'.y%:. m:s.y a m m:a-r* .f _'+. N" R:j. e .
w n r
, ns e 5 ? 9 ..;-
.- Q e.'
lQ:2;+ s i sf
~
s eihdv l e a O. i gf 'fnyL;; ggp ~% i M B s Lf s,~Nhg'Rgrt 3a#g. .
~
8 h~.y .agsy a p mT J wt!rmQ,s%;p:@aag b 3
"- Qe iu o Mh hU O .-
_~ fg
~
c .aM fk zf ;+y b nn
.; ~
i . r . a,MYtJ $ i
,m k kn%
+y * -
nv a mE .Lt -
.Qy:t u U 5)@yu
?
M 35 b i - s b. f 7 p 4 o 2-
,E &M' ,,:e
+V ,
n .E k F,d. b.
. r ~ .
i o .r
.5 ir}
dgb . t ~ ? _% u 7 - y@':1 $n,Q 3 3; . b i eih4 wW M,;Q. #. i g t = r t R t k ~. - i D s affRm7-
~$ E m
t yk g 0__ l i mhhnw.s%ky mh d$ r r
^f.k .
e o nU i d-i l S Wy' s 9Mhliw6 m.$ 8 I
- 2 S
f
;* ,v f
s :W .
.- M .
' , 4 -
i - ' .
- i .
c . . wMG M
$ a.d pr $
, w W jA v
k 1,?
O The Well Log of D-04-0418 CDD I 1,180 _$dbani(10f 4 Water Level: 1049.00 (1988), clay, sandy clay (45) 1098.00 (1993)
+ Unit: Feet 1,130 sand (20) clay, sandy clay (50) 1,0B0 =
sand, sandy clay (75) 1,030-geo- _ _ - - _ Clay (60) 9%
O The Well Log of D-04-04 20 CDD
+ Waw uvel: 1033m (1988),
N Sandy Loam (10) 1069.00 (1993) 1.180 sand, gravel, clay (70) . Unit: Feet clay, sandy clay (118) 1,080 O sand, sandy clay (30) 980 clay (117) 880 O
l O The Well Log of D-04-04 20 BCD. 1,1eo Sandy Loarn (S) + Water Level: 1084,00 (1988), 1113.00 (1993) sand, gravel, eley (26) + Unit: Feet clay, sandy eley (SS) clay, sandy clay (80) l n.a.-i l bh
+u i I sand, sandy clay (40) 980 clay (160) l 660 l
l l
4 0
..4 s
k kl 'l 6 - I g li mwn n o
~ .. /S :
e g
. m . c.;
/ \ :i f(w:; ;s'N O 'in t i.?.
xw > . .. 4 O ' Zi
- g 5 ,s ,_ _ . Si.
( gs $ . ;..' - o c-Z yl_ f Q :y.2
; gg 3p'n iI 3 4 :.
; pg o .: i .
4 y,r.,...,.,. . 13 ; . k g ma n ;- ry. . k . g , 51%. R )% ,, U . *? '* t'.y *: '. s : N k ,.~ ' 3 t. Y ~ h. M m ; :. y w .Ik I'. a :- va gl. Ih325ei AM%,1$ .
' ~
5 h l I t i 1 l i n 1 il I t t I l t i i I 8 a R W f*
. a.
P
l I ii o y a: a s g .. k 1 O > = - -
N x'N'N i O , j
; s ~. / f , / /
o '
/
/
/
e , e O
h c sa (. - 3 i e i= i l g ==ahakas.sna 5% .... .3 i b $ hhkkkk$E 8 -
<l3REE t. 3 ;
_. ..;__,%,~~......... , .gc_. s ~ 1, ,,- -- , l {-. 6 l' e i a
\
o r.k6m ,Wui h\ ' e i e i
.e j
-l '
- @f W
) Ii E'/'g(@hfy,kh(?
d %Ql ' E w
-g. . l- .
c i a_ 3e~s,, C y-
.3 s e-4g %
i~ --QW t ;, wy xn u,[of
} ::3 g ; e ,.
3 g 3 g gs' ,
.s a
C
+ .
i@ l.,1 le - O i1 ,
-f-'-~-~g[.gt4 l *}
T 1* m yx g i m .[ 3h6, N ' 3' ej
. h
( o U$ i ~!' -6 i c ;&cthe ', 'x . a. k,!
~-
' t.
iy hh Ik hMM'- n w w;Js -[e %?* i; .g;sa 41, _2
% NSI i'
em pg ,ngy r - g4 v bjk ,
?
w I
@ hs p%@g$dM as n -s
;Rst:%hfd
. ', . ., e..
4 J - Eb-r_
- - ~ __. -. - -.
-- -,.-- - - -- .. ~ w- ---s------
.-a
< > u.)
I < y
; as p e z m ! 3 o 1 ; ,. ni gjf,Insas-g
- i 1111disillh
?
essee DEI 5 BIER l
'I
' ',, .1
[* ., Y!s
.m
;JmmS$f-3 ",? W , - .
t
'7 l8.f g is 5 [
e Y, hINw g(In U .,, - khh
.;. - t Ehh egy re,h.kN5 h
,3 ; pr~ -wi w)
Mg Qgfjybup 4ft:i }, m w "..qy
.E e Q$;9A a W g;>,,m;;
p.p npir . kf .'.
-- l- '- ~ l shy e new b' ?hp b !hn h
?' ara m .g,, .w a '
b JU$ht~7d .kr 5 e. m ne} 6fd '/ upp W Lglj'%f! e; Lle y'TU :w v y o ji ' ii;f pj:l& fa O Q w w 9%w.r& g 3,#
. s
'\ ,t e
,yy
- sp ,
!I camWE #
C g Mk f
- e. R*M, n%,. ..
's nQ <gD'N us:),
1 ($:/ yQ'T:wh C K#1q . I ' eA4X.9 ';sh ' F W i .k *k
',,Y,Y,' % .
-h ,
i p;?k}\W ,- *^^' . l} I N( *1i l t -.
, 1.; .;
ag - G w , ..
~
t g I Q.
- 1 3
O w C O O
lI ,15! i i! i!i; lm!
. ; .I! (2 :; .; ili e _
s r t e e
- a Or .
w e _ 3 t _ a 9 _ f ole w0000 8G234 _ sE r r ue f bm 32358M0G000 9 10000 1 1 1 91t1 1 1 1 11 a 1 t a tota s d. o 0 0 0 0 nwt r ef r1 235 Dl 000D 0023 o o3zie0000MM0711 .
&S t _
C9ASc1 a 1 1 1 1 11 1N N f ~ Sg r y
. a u ' gg;g~glM I m
'1: pI 3 .
$' :,ii!I f.
9
.m. .
9 . 1 .s ? v y3,
/ -
o e' ' . x0
.,- t-N v
N af' i n _ $' Q@ a p e j
, , g"
)
_: f tM s e r n7 ._ ' l a i M
$s
. 4
. z C .
b 1 m A -
< ~
e.
~ %h
.N M
<k' m
_ ..u rs " A-n i e. i A m j ., 5 ,",- t i 2 1 n 4 , t ;. ct . b- ! u Pp1 ,IMw.'g;.'[jr. ;h' d
- f ; .
k l[ r ,p U ; g g 0 A y
.[ 3c ,s<Ja:l}' m_.#
s' aJt y?; g U * - I y A g n#_ -' f o c 4M g o oe i t s l e
$ c,y v
e 1 L r t e a W 1
\
Hli _ O i
'f j . -
d i
~
r . . s e G ( ys ar Da e ,- 4 1 8 0 Co ,w 2 0 . 1 9 sT r o. t e n t ayo ai * - nLt 8 Oisi t o aul a r o e a z , t g o s o t t e s s 4 nnR o e e - c g = 2 no or g o Cet o. o e B 6 2 . M O
o t
' %,m M
sd yi 2 1 I r aG D k m 0 1 I 0 e . n 0i 8 0F 7 , 4 0I t s 6 are
- 0I onyiLt oa e
4 01 a s r m 2 . 0i t Y-n r ef o y 0 ci 0I _ nn . oU L C 2 0 fl\fft
O . 4
~, , -e .
1 A 1 g bQ - O .
*i'l b
I
~
l 2 3: i g g_ N. __ l g g& . C-t e i O
- i jil o
d i r sG y a e 6* Di n 0 F,o 0 s w
- 0 r 7 eT y -
t aa no
+
o i t nLi os ael t a ni z . i ? i i a r e a
! '?
,'{-$ .
I. ]1, $b . . t neR ne -
?
eg 's
=
c o nr i
~ ,
oe Cte p H
~
=
2 p o ll
O W V 4 04
- 4 =-
e
** G .:: -
$ Q O
= ._
s .
. o-
,p N ..
. .; z. .
(..._., ' g p's . o= g O. 'w%. .j..f7 71. g" - .. Qf :....,. ..
. . , c u .
e= 1 O
o e d i r , sG y . a e m DaF 0 ,o 0 s w 0 r 2 eT y m aa n t O o _ oos nLi a t
~
u7 i e. . t aoi l r e a t neR ne - eg , c nr o , oet Ce H . 2 , x o
e
- f e f i
s Im i = 1 . , [ g 3 y a . 8 .
- g. u. :
uan , O
g 9 O - o de s ! Ik y' . O ~~ 3 .: -
,g e.
ea ' gGs
' a yN .
g 4 O P
4 = 0
. 1
.f l
l
. : :s ,- l n - .
. .8
(
. a- 1 1
U=
# , l Ei H .t
,\'
'h Q h -
a u . 8 . a a=
$s!; ,
E N m) O
6 t.
- O
.i I .
b=
, s O
i . w 1e $...s ~~' c 3 u . N 1 O
O e
- . i. . S L, (iaa ) 4 y j a
y pyg -
; .. s.
<> y xp ,.
yt q \ '- 4 llc . ,r . 3..; w \ g 3,l .
. 8
., J s j.
,f-
* ; 4 4 w ,
fy- . Y wh c
,- g v 1 .
e .= . s ,
,e .)
h . E ' 5 lyk ' h
" 4 ?: fA s' ;q.
>g .-
^
y4 y en 4'
- a. .s ..
k . , b
.y ..
.Q."
4
- f i e 1
,,g
/. p
$ / \
,.:4 y .,
l '[: .
,( '
l l i l
e )
- O %
l l 0 - 1 ' e - g . A
' \
1 r a O :
- 1 o
y ~ t i i v - t - c . ud di r nG o e 7, Ce s c ra , . i o , W l uC a r ,
/
~s-h dy o o HT w, '
p ( f f t oon rt o a i z
, , M ,l n 4c ci l aa - e F e gR p ..
4:s lP n i X. i . 4 l a _ i.- c , s S . O yy s
+.[y
- O
J_ 1 . , s _
.t y
t i i w t - c diud - r .. nG o Ce s .
- ~
j i e r a L , r g de o - aC %:~ . J A r n:h y
~'
dyn e - s n , H On f g , oo rt i a , j_- 5 i
]gs . lJ
%h o z t
cl aae F R i e w' 7l . g f . n D- -. i
- l : .G' .
a S c / -
?
}
r -
% )' ,, .
p o, _
O :: ::: :: : :~ .:: :
///////f7/ / / /////////!
//////////V //////////!
V / //////// / / /////// //!
/////////////////////!
/////////////////////!
)
r
/////////////////////!
e /////////////////////! t /////////////////////! s a /////////////////////! n /////////////////////!
/////////////////////!
3 /////////////////////! x /////////////////////! 5 ( /////////////////////! s /////////////////////! r /////////////////////! e /////////////////////! y /////////////////////! a /////////////////////! L /////////////////////! o /////////////////////! O T w
/////////////////////!
/////////////////////!
/////////////////////!
f
/////////////////////!
o /////////////////////!
/////////////////////! _
d /////////////////////! i r /////////////////////!
/////////////////////!
G /////////////////////! _ e /////////////////////! s r /////////////////////! a /////////////////////! o ///V/////////////////!
/////////////////////!
C ////V////////////////!
///VVV///////////////! -
///VV////////////////!
/////////////////////!
O
NUREG/CR-6565 PNNL-11705 Uncertainty Analyses of Infiltration and Subsurface Flow and Transport for SDMP Sites Manuscript Completed: September 1997 Date Published: September 1997 Pr-red by I ' feyer, M. L Rockhold, G.W. Gee
~
Pacific Northwest National Laboratory Richland, WA 99352 T.J. Nicholson, NRC Project Manager Prep red for Divisien of Regulatory Applications OITice cf Nuclear Regulatory Research U.S. Nuclear Regulatory Commission Washington, DC 20555-0001 NRC Job Code W6503 nau
'o...*
3 i 3 I i 0 9 4 I.
. 0
- 0 I .
_ i 5
. 8 i 0. ] l 2
1 _
. 0 'n -
> 0 ']
a m c > c
. / > /
. 7
, 0 c
'm [ 0 'm 1
- . 0 [c
. 0 [dn y
a a l S I 8 C 0 r
. 6 r l o
- i
- 0. o f i 0f
. 0 t t n
n e -
- . e t t
n n o i 6 o
. 5
, 0. Cr g
- 0. C
. 0 er
. 0 e t
a - ta
. W i 4 W
. 1 l a
g
- 0. a l
, 0.
. 0 du i 0 du i
l i s , s a . e e m . R , 2 R r , 0 o . 3 n g , 0 a t , 0
. 0 e l
e -
- l l
2- -
- _i. _i -
9 0 0 o 0 4 7 0 8 6 4 2 0 g 4 3 2 1 1
%R yE yW gk 53 i
. 0 .
- i6 ,
. 0 ,6
. , 0
]
,5 *
, 5. 'l
- m
. 0 m c
, c
/
- 5. '/cm
. 0 'm c
,0
, [
i5 [ y
. 0d - ,
a
. n , l C _
a , S r _
. 5 r
, 4 f o _
i 4. o f , 0 n t
. 0t e n ,
, t e , n t
n , o C
. 0 o ,
-i. 4.
0 r C , r te
- e ,3 a
. t , 0 a -
, W W - ,
d _
- _ . 5 ,
e _ .3 d e , t
. 0 a t - ,
a r
- _ . r , u
- u - .
t a
,0 2. S
. t
- a l
. 0S a ,
l a -
- ,3 n ,
m _ - . 0 n , o r o - N , N .
. 5 - _
, 1
- - - ,2 : - _ : ,0 0 -
- - - _ - - - - 4 3 2 i 0 7 6 5 4 3 2 1 o g }fm
%A yE ym
et 4:' g et
'o~
i i -o i 1
~
H
- i _
i !, T ia - ie , 0 i
-3)2
! g -
- e 1 ,
- - c m
01 - I
, h
! l -
i
! o$ ! I
. 5 l EM -
i
; i
' -01 6
~
3 ! , u ; i
- y
!r
\ ,
i
~ *kx
~
~
~
- r. ! ! -
g*l !
.! i 7
d w w i, i. I
~
mE o: { .5 . l; l
- a i m -
MMMd m n - o I- o
, e r y l e
8 8 % 8 j jpd trop!dw3 Jpd t uop!d w3
". c.
o
~_ l l
- l
- Ie- i
== n _. " 6 "E l i
~ o "E -
"% l
- 2 .
E 1
,z i
's i , _
d^ l t 3 , l
' I j i b
- ,z -
}
~6E 'm2
_U b;;: -8 \
> ; c
- 3 l
. - m l
~
- 5. - Y l
- e i
-91
, 'SS
; i O
3 - . i' s i . d E e -
', Z j -
n . ii3 6 giiiig . il4 eiiliiei(6 l'l l e n m N - C' g, ,,i : iiii>>'i e jpd teop!dwg Jpd (sop!d wg I
b r s ( Integrating Site-Specific Gound-Water Modeling into Dose Assessments -- Topies Public Workshop on Ground-Water Modeling = Technical Considerations for Model Updating Related to Dose Assessments {
- Elements of the default model U.S. NRC, Rockville MD , Reccnciling spatial / time scales June 23-24,1999 = Logistics Walt Beyeler
- Assumption-based model specification in SEDSS Environmental Risk and Decision
- When are site-specific models useful?
Analysis Department Sandia National Laboratories Albuquerque, NM l
/~
)
1
)
Elements of the Default Model Interfaces Between Elements i
. 1 i
I a) I i j ,r l I,*T,N[ ~
~
, y e
l l Flus Rases of Water aufRatoonucluks thmaanned gm ,,,,
' W ' ",[
i Rationucluks
' t' \ Flux Rses oJ Worer m Well daA*) r . 0.N.$" D ' _,.. gh,,A,.) !l ""# #
. 4 4 *i '
- p. -
- l l,I V,
j
O V Possible Model Refinements Changing Model Scales Beginning with N'UREG-5512 Volume i Models
= Solubility Limits
= Non-linear or irreversible sorbtion = Coarse tofine resolution: assume constant or unif rm behavior at the finer scale?
= Spatial Variability
= Fine to coarse resolution: average the fine-scale
= Transient Flow result to calculate an effective value?
= Aquifer hydrodynamics, chemistry
=...
1
)
Support for Site-Specific Modeling Sandia Environmental Decision Support System (SEDSS) j i oModels are defined via assumptions about the i
. system i
! oHelps compare the value ofmore sophisticated
! i modeling to other options, and guide aquisition of ;
j supporting site-specific data l j cDesign will allow a wide variety ofmodels to be ; j accessed i cDandD models and I-D network transport model ! will be available in forthcomming version I l I ! /^ }\ I
O - t o y r o c n n t i fi a i o r o cgs t s b e pl n n t l e n L a i e G o Se nt l a e d m nl e n t oss i at s r a i t o O SiMe r s G aC N a ge l
,r s n a t
s ia nt A l aK e w t e ed aws r gd o h Wa n h t r o e nD e N t u n c e I no r
~
G i f i c a G . P O ~ ~
g i n l g p n g i k - i nma i nes k n s l r o id? l od f nl s f h oae
? st sg d c n o s ci n w nim i i n r o als?
h els h t~t ao n . e cne i i t cd e e s od om e u mmo er m mms er s e e Q h t t et a nt e e h t a s s a e w-r w ad mr s ad e s t n s t n o h aue osh au d o r s r o W g a W g t. e e . L
O n l d t n g s o as i l l a e o t ut c c cgo n e-i c t adda u i i a p l u nivp n o min ec r m p o aim d i d hr a c I P I nI a CaC R o N e r u e r n n t u i o o l a c c s t a i t ne rs a u o p l s e eo t n r a g tD x h n n x DC t E I E S I O i a d t r e o er /c ea i msnap i A r ce at fr a uW dr ne ut oa f cfl a rsS uo ri l t T r S GrW ub Su u S M o n e i t o d c a i s y n r u i z n aw ro e l l o i t p elF a , v S l o s u g O O m m u V S g ~ m u _m ~
e h f c o a e o r p p y t p A e h t t a e g s a a t e D k a. g n i c i n O i i t L d r n r nh c pi too f o i o a x u t E e s o l s ep&o r s c up gR e a t n p e qa S m e f oet i g r e n&& o e r e a pl i u pk yn a m q Tl i I mci S T e e eee R . u 9
O n i o L A B O t L l u y G o s n c L A N O e e I G E R d R - h n E D se ep L A C E H S R j
]f S E jf MeD T A
W g h jh ng C h M I F in t U I D I C i E E h l al a M P S $ ec r (( (f rS w M U H G oh I D I w E H M Ct xw i N _ i O r I T t . U L O a O S E . M R
s l s sl s $ l e : ss n e e d h nd oeio od d o s ciet o mno at oc an am c M ki r i f l efd i i c n g i nt c picd p eo ea n L eA p m p s, i n s s n k : n s c e n h o n ei gf gi o i L cc a o ac i a t a k ek i l t np c r a a r ol ni sl i n o p gf i cl i t - u f h Ai p n i c
- e eh m O r eh pcmc m c ei a o a o a gwS o o c o rt rf kadr r d pd po p p s p nat epeAe Ah i
L h n s yi n A l yi elei ib l l al s ab r F n Orte s aPh s m oi rt - e&l i b t s e i t ioc t s gau t s i d r j eae map r l ut j sd ea y apb r S tao . . O e e .
l a O h t s u p
&ch ni t a gl a ml o eap t
c c t u u o HE chm i a O or y p p s A t e n n g u o o l a e l aca t o i t a i t ms n R l a s e eo mt r n i e h g t xD eo DC l e r u s I n I n E d o ? > o p * > M E x y e d e' e e e O i d a s y br n r o l i 0a Sud c se o n-do aZ t t t a r n uo Z V a e a w
- i A S S m h +- m i t -
t a c r
'e.
l P i . r ey e y t aya u t r hpwa d y na a c era at w w dw r w M o p sh ot a l eh va t fr ah SuWt a nh ua t l s mP OP P oP r a n t A G a n T r -
- i4 ;
i o d t i en s s Npio c ,c d s pa t "t a t i nt s ai r ns t i a r . t u p um Sr "R n ie mt c oi ie gt ec r l it ao s r aa Ra t np O T I n ro ef sn
" e l a
e t r na oh
/ r ea ih t
te ce oe U I
~
R CC-SC-PR ~ ~
. ,!\\ l,!\ ::
X CD
!\\ l!\ ll \
C \\.\.:ll.\,..:ll l : J :! \ l I\ \ll '.),' l ll
\
G) h u \!! l'\\ 't ll\,l V
"O !\
O l ?.. Q, l\ ,ji \ l \ K Aif$,V:' o i:k W l: ,\ 8 (U l'
. ey C n " =i BD o '
8i
.M li ii O wax ns s (U
L \ b \ O e
O s. , l SD e AA d AR o l a RS d s E M e n e DR i r i o s E , g wid t a MM n dr a b a 1 C 1 L M i k r t n at a U S i L h h D M X , e g r u e. y S O t o a o d t i LT L r s v I O i h t s h t OA c nt uit S C a e m O c mM , o n es o e nt eM M I r p o p tsl et n s , R i s o y1T p yf m s a sC S3 A o e SS l c st r a cI cS mAM r PI a r AG oEM n ot ed F M o s es n: l l a a d -
, O e at e mmeSI SS i l i
t d d t i n i n s a o a u ii ODE ER O T r MDoMMCSP r l ~ ee eeeo ~
g
;9y
;R c,:i _
s . ns ~ g. ol ) i e tda o c iM f i cg n o e pi n i t a c S kn i f i c e p di L S r e o s nt o t s e c ihe o r p s e r c ) a l u Oa t d o gy)= a D M t o r c ep ,Ej j
=
j p -
' /
bA C O
~ .
O h c s
~s e
m d l r a d a o e n w o r e n n h g l
'a l
a r e p 's s i u t n) r p l e r o e c. g d e h ct n A ot e m m l t a , n s,dr ei l e t e a r s e
, i mr,t o c o
c t s i e a m xp a o F mp a ai r , l s r l ol a n r mc s s p d e t e d t g n t e e gd e u o r t n e c e s i e i G )r O h n mls (t i r a i l b s i t a a e d at a a rd s i o-M a a r . a n o eh , i c t s o c pmfncDe e i o oibvj e l e n et s t r a sl e mid , n , o n ni F o r ah sit syi eit d fet r o i o i t e a t teh tt mewt uat l n aqd ei ui n t c s s ni b rd r o a e i l i f i ay ee t l x d i CMS e sr e t a c u f a c sg
,i f
a r p aa si f s h d e d te y,t i l r r e ( a e e r T e sed mob e e c e s n s abi t my me i gl
- mtc e l n k al a a ad r n aed rj o
O o i L nt s ul EMAmPt l e ab P oM ~ N ee o e e e
s t s s e O n t e r o s a c i u s ef r a r b i r ed o e t t t mc a t nl s a nyI e n y a r a c r d e o o l P t eg eleU e s m g C & e d mf l t n
- gi r
a o d s eit i t u wra e ol a i h p s M u f o c l P d u t t r t a r t o a n ts n p e r O o r Nh e ei r mr e e c n e r p r fl ei f e o i d e af ,f c u p r sd i ed i e q e weaf r i A - s of r o r ef ol i z a r s dr n e a ta b we r wu o oit s l i i h oh F al t l l l v a ed r a;a oi c r ad t s t t l p o te ot t e l s n a e a sb n u nl u pa
- s. s c d t ei tr ef i w
a oa ot of e O i MDDaDDN M eee eee . [l)
4 4 O
. c U
(U O u d Q.
~
<C O >
(U
~
od cn J l i i O
y.
'A .
I, ATT
-l t i .
& 6 y B5 . . . ;
m e us s 4 'o Ml ; F- Tifl lE 1 R1 I p V
! 7M i TE o i T i_. -
iI E o i ecn f f edr vaa ,Sfdl r gul sSu u e c ecle o au er op it f m o r s e E i nsl,eLsMlu , s a cemn t l al swIOre l l e r uf nSt i r e e i mdoMwot nit a ui di r r t a sacMdS o :s f l o:jen uS s e n nl e o s f sa y d w e )s hoLreo i ww n t t u bU dnn,gO pd rI l l l t oh o a p M u o o S e e ). aap r g oit M bbe e n 0 0 0 a ss t c (W r igd seMyoc rds em,p r asa l ug di n r o' e - i o i t l dt en u i u de a a n uI e o n sst enop n
.r l eh Tt s nt o
ie
.aluo ns r
Mo ow mfo r l l n o z es ee t e M e ek c ri mi n e e amuli sd o - c l u l r se Th t n da d e e p aih d pt l e o - o cge t citaiv - u at ee ooer snofo
.e l
M e c Sm r n sgt o a a alamcu qp a wr o uta e a u vae r osdh St ieat r go n v e S r u o N O I T S e s, tsiet Lt snd I Oaen nt t m h shti a ,i n adth e.. or Ss Lddg I Ot a yh e e/ n t i oS S P Swmu cn i , Sia c L I Mfo dro s( s a ese Mee dc a I R n gl e O C Ms ur ud r,e mel c MmDLe l e S M M S E D ee hp Tt ymiuMwapp o pd e n oa t s u at sha si r hn Tu e e- r u s d is n o o ) Ar io sn s mis Ei s M lm l P( a u t in nE a nt m 7 man 2- t ai 4 nm oa e 1 t Cn cr U u M do o eC S W n id S et e 2 f T c S r
- Du-r em o
~ UC AA PP EE so SS' MM ~
\
{ %c+z.tr,.n. awc g
.D
,m cw.
i< y%a.,, 9,ug hqq.gu-. m=w# w. nww me.%
,gg.:j r % .vu w-jW'f. f 7
- +& ;,.r f T'% q M
~ffQg ymc.wcgnymwgn f} *Q*f p+f j s kG j n ~e w.
- xVn WpA , M.yw33-9 c
,,fy .
.w,% ""' , mM Bio ~bi-g mg.fpty ei >p &c.,y;aniC g y,WJ,,;cg .,.%;gm: 4,7.s Mmn..q.q,eay 3; .
,4 [s.A @ .p.;.4,.9:p.co
.a.y.
k.a . m. m5
,Q Whq%$w<3
%e;Q .~ MN, bQWWwM,,n.p:g%% :;@% y,.u f M 'U c. q,.bb. .,p
%,Q . VWh ~ t %%.W., " b MW~ w- @f ##F,. % mrf *%: : 25 lr,.?r;4 %.1;",,7*p%. WWW, fM--u.C y@.. .r'. ,1:,,y ps* Whf,W.w 0
&O.a%...i.
f:y w gu&.%nW.a WW J.. %n.mSpp M.W:W
- n O R.
mY$%m;r.W;&W +~v %i%.%.,g.n. +j.,Qt a...._Y ' ' ,M ,rn . Q N.b 4 w- . w Q;.um:g . %dsk W 4 +:m.d+ g-w;u w.??N g., W e?+t %: a . x :.w Y s w m e% ^ M. ~ .v.e% - . -*.1&wnm%.u.a;.p,+#ghety%.mp%n%wp.s.
. .w e.
+ &e w v.m:n.wm- na n:.9h.c bm m ~M
.pm.
-m.~
. a, w y... e ,. m,a s y;sn g m .
w
- v. ..
w.s, . : y y 4'
- Q,e'
,,QR Qt.. e~ w%.f <t. M.v:w. k.wnM g l Q ;*f
~ f.q@ga.p.,.f.'g:.W.a 5,y pp ,Q Q. g&p y ,
OYbIkhbh_ }$s i $m.kf,k. f,;,ff hh;.C m.1, n hh
.g&k hh_ , b% hh?h. - h?. N
-m. ,M. w h y m.g s wy.n :4.n.w. a . . o , , . a
-r :
,n.mw ?u.%e%g.p. W up~m :;. m h ,we.e .v p $ g g ,9 . w My f"k@.,p'$;n. -I M M: w nM N Qyy.p;;$;.4 w,m;
~
ns ,
+3 4 m ag s
M2ddk. R m .M,y@a@w .c%w%u.W ., . E.m www,;u} w
- n. . p 4.hg~.%3 .ww . % e io..dj.u, .m w4mwrr o.w% g,a,g.,s.y n-
%.pm, .,,
% n,u n ,a ig u Mmm , 6 .w 5v.wa.;y .u m .a,as a m.:
y t g w,.g,,n.4;,g m m . w s.3 ., y.p , ., ' yv.: W. " ' j. gu#g" fq
^
g , . . . .w v?J.u @, j h.
., C k&Q e & )Maab nf.fM d?.}E 4 LQ k m. W.X g4@g"WP,($h:.a%,d&"
4.dif8ti@ a M;b WL%u, g,nw m.:m..n**c x,: Thp.W,* e.e w Ah. Qm>: k;m.m n-p.F+u 4 W m & n, w-p a r. pay.v n.-Q
- s. m: p'y'4*C 44
@k f.f'hl wmm ?...$4
- m<;n e f.o a m., < .
;w .%,g s
X x~ w1 T ..
- e. .,v,w EJ&mi o
4.wo.'p;q;F.c mg g.m,, 4.m.,,;e,
%m
,s:
n
. . . G $%,.k%'&eQ 4.& ,d> s.
4 .s..
.- g5&
b &s%
- gyx WQ ' HlNi5
- M*Qr . , :f s.s&*s y{&Q. n.-s.p!.%..s%, . M,Mh(g,w%k&.'m
. mWW;irsq.am %.&#:..W h k ,
- y. ^
v%- u ww r
- r. n
.. m.,
MMP EW g$yfd!?
$ d % g ,% W M l $a,e$MN A- MW% h knQ%u,m. dlqMIi$$fld@m.v.:~.w.ow & .w. .
Mv. w%nw. %g 4yae.w$M.~x E a . g wr;; ~ +p
~. $N -
L
, dc.4rst W%w. . - - t~uyv.rm .: v$ s 7~ we.e e# :; ewe. .
.mm 9 n u.g e m . yw.aw ,m mp...
y N. l
,~,v p L, m
m,: e& . v.a .. A . w A nr$n :..w.,a. _x w . m.hk , Mhk-s Np$hh.+Nk$h A hdh$h.+~kIk1).k. ay- w e Q a e +% ;$g r"' A* **T' v &, y g&&a , +; W ,y,g QMS W ?!W p6~%1.u e." M:2g m&s.n.4.,t: m w' ' e s@ s v' u 4a<..-4 pg w@
.i &w&,V%MM m? <*wgfMntka m.
y.W: +, 4.; t;s+ n 1yg.g . x hen 42 7 paw; y p r:.M,,j.ge w w m.mQM mx w,, . h.4#n' '
- w. O gy ar. s, m
w*m .q<w,e.m.p 3.u.g as. x.: . m .a. c\ 8 %e s,. w w. we
- m. e m n .: p sc w m .:.u.,q
_ n %gm,J.d y
.u;,
- w. p.
w .nre;.s.xe._,;.: o .
- z.n --
1 p.,%ea ., .
, g ,A,-
,. e
*,g % - ty, >
(, '
- 4. f pr 3.N .y-. "y( , s o. . . ,3,
(., , .=y,Q. w,m.u,,p ,,,, ,g - 3. ,y ; dM .' %*...u.. 7N g y %g, g .p;. >z
%. v.e ,
( 81 '**' o p. , w.g ,0 4.m.v g gy;7.;n. ., . .f. ,
.p! % g;, g.. u' sd 4w '.m Q.?' w:- ~ - b.. r
; f. .-
s.M ^- v Q':) .. e Q. 4 w ;; .&;q. 6 Y.p W .9 @?<,w.. .s.[.f d @ryro.M.w@$< M g.:
. @eU,
.i . o.
, m.wpg/.">-w,W,,s. :m 3.- e
,4Ai c.4. . . ,p@. $,.+W .a.n,, . : w Wh,oy.. n 6. . : .~ '
M,j%jo
.x
-3 ..
s O*+e'i+$.- c,#. q.3,l sgr, g' ?q.s.. M:.'aW
,4 , .c .
fa . w ,. e f*y, *,,,r-tie.gM J $fY4 )- (3 x 4 .
.4.+.
ut b.y e.9a,'>%' 7v. asp me.ia. . 4 ,c4 q e,u #
.uw"y 5;.v ,,,
e t .
- p. .'. e .,,r.
w ,, .ee- .w % whwW
% %@.g? D?.ZO.); '.1M 41,.QJ46 w,me.3 .W.G'g4 C. W ,, e . A y.y klT. aw N;n.
s yp"f[y Nm Q.j's.Myp m%e., w a . ww%w.m. Q,, s
- ,# -Vrf.d. r 4w :: w..
w w y w8e pM.tnM a b;,y @m %n@.pp%ghm. .n r G M p M www M M w$SQ:v.w b, 7.1]po%:php. p p ".1 w
- m. _
w .w w.m.yy\pam,.n.y w w..u e,7A w m m.y n.y . .m.n n y e. s
.v.. c O.w.
. .~n .g~w.m.L..w
>~4X?%mB :myy1 u ce.
Ae.~ , n,.,,, w$ . . ,,
.hp.u.e, .
a..
% .- A..
,?
3: g s';i . 's
^^
v 7, ; 6 s Q;%'% t+,;Q.wn y'.e;~.ff. G%%.::, an- %<:i,; y .*N : y 3,t .y'-r.s -)& zn..p';yCrnp{w:if.Q s pp >.3.v % e.,k: ?y; Ay
~. w p.
g A;.p%p s.
.;..s;?.g;m(
9;g;yA* M. , .A. dw%.;a2p :.wwjau.3 :Wa au,st
, .- wc.n m.:ee .
y ' g)..m.m;yg uf q;sM :y . ~; t. .. g.rs.~y u M. .& a Q+-p q:wk w QvvW. (Wu ;-Q.%?orqw n y,r. p;f,,, g .>.q.; s.se.w &.faXi,g w np.y;g ne. q~qqu a ~.. 5:;y 4"p. ..y,m,m. n&s i 'm : pv y
. . @. ckn% n.
- dy N * $rd E QEM,m, 4'")
-g M,w**Y**i'(51 N?l kh.I NCfMMM
[;y.,:.
,9
. M.pw,.aA., { 7 4,.'4*f/:fMk 8g4' ,_ %.y y g .
e0 ,* fg%@w . .a '6'
'W 7%.e9.m.[~h:m%'h* e,y **Th .ffM* # 2
%.62h ,g,
- ~,,", ?.:t&g&e. hm :3sp'rr&g..<W>
M,.ww'A ~ 4.%,.9
- 4. W z ,$ y. m e.et w w. % p. % c .4isb w.=4.+g&.2.,$ q m
ee m.re%
,n
' ~
v%f:m w M: msa
~:-
%,o.v. +:p ;.. <.
w
~= j; ,r. g m.~ gte '
yny u +m.yh[y.(*
.s n @m
, .-h.sb ..c;S%.Leup w.
nm. n;; c. m,: . cx.q., g; g4 .O w .-n p S. m W--" e . unma N. u w s w ,:.:p: @g
,.v. .
cg.H. W. y:.'*-tp.wr
%.u,v i
W,-u J7 .w e.: ap%LmM.j:+e
>g3. Q ow g.n.s. w g+ .c p w: .O-.n .x -}
- y%n qs,t:wm '? . ~m9.c;; gas.d:,:g,;.y, v,w %
- c. .,;M
- I r.
c-
.4
,,m,y ] p.wi ,I
~
. gerN;;i em:q:3,m . .w w
{ n.t.::w,,,
- p w e, e.. 7. hg . e.e a ,q., s...m.; p..g a. = .
w 0 : - h:m,d, w 43 .m.. v U9, => .d e.
- w. .*g/ *c< r. -
. J'rE
~3 6 , vpc m.n<- 9%.<q,:w;W.. % ..%. W .+. M r 4 e
Hri E A yn ; O: I
- g.;7 Q(p .[;rp';
Q ;r. .. g;,. ami'.!W bv
. ~
, , . p.r
, ., ' 3.;,9p; g,g,;m).p;Qu:' .. j .d Q ,,y) ,3:5 c n.
# %'.' - A;Tgg,. 4,4. ;>%p Qh g.@i.m i;
/^#dAr*.
- m4;6 w.yh.;,;,nh ::.S4;M, 7 *. rib. g r f.h
,t %. , K9 N>,M.m.%...,,,m%q; W45&,QtWy; W ~. A tJC Q t4; 9 M M yL *-
m w:2
.-NM M %.?... % m.. M:-
W
. m tWyi, 1m-?
I%: 1 C. . a; ,,wks.: a.e.,A.,o, cwee u. . W d.*t.S. n n~ % .'f.~. M.L . r. a %... wu. <, *:Q: W ,L,We.m..we>.
. n. 4^
4 4
O l s e s s do e t e s n l u m u i o d o n g b i t a mits r i c " i t s t f i t s e e x n t c e A e pj"el ck i
.t o m _
h S b r s nl o c e o fo n o r c g n o a a o nt s i oivtn o o k mo a e r O r p i l n mic t i f k r p n p t o ai o o e ci fcd owi A e sic t mas eplamemf n i i y o e s a e _ l a t t l u tai n r mf r d - P sd u ai e m j o d mh e -
& d a m n s o o t l
b g l l sla mit o e r o r p l u d e wmqku y s P oe o e n ~a i MVCRL i - E O eeeeee ~
l i e F i c t n s a io l s h c t e t o d p o s n c m r t n a S t
)
r e s n E i at
,o h
o c D r et h
)
l e e g f b a wa i i n u a q s tu r h a n c r g ,o e n e g) u gt c x
.i eio e c .
o r ( e e n g
,t
. a s
r / t i nt e e , e n l h l o . mt r e as ec t d c (e fo m (t ri o e s n a s e g mlb eI r sr ee t t i a mt as uc a f ot p a o r P eba _ k s e n p tuathmir t r a i n . s c e p pn a cl u n . e ac L l CAF i S I eeeee '_ l
f o t l e u n d O s e I l a r I p n r _ M o l e _ e _ n u n u _ _ ~ s _ e _ U R d G_ o
,e" M
l l i F a _ _ h c n t a i o _
,B '
t _ _ a _3 _ p m r l e _ r s o H> o s s d e o Qs O f e _ _ e c n _ _ c C o 'ro I Y / l i r e p d e d S
]UIJ ___r e o t n E vM P M 6q a D l _ _ ,
a _ _r n
<__{% \
i o _ t ,, a /W ~ t r u t e 9 u p s U
}
q S E t p us e MOl m. o A/i Rup F F t O C I n
mO s )?,}$ _ s l e ~-
~
~-
d _ n M o '
~-
~-
1 __ _ o
'( [_ ' ~
i - ~ t ,
~
~
~
a , ~ i f i c 3 C O* c e p i n o t S aa t c af i l Dic e S e p O d o M - r o s e s s e c o s r a e s fj p OW f a a b b a t a a t t a D ' D
=
a c l a
, J D ei c f
i n o C t ie Sp S i R g e - 9
.~ .
O r u al o y e t a t at r o o D n t m e d r e gc n a e r n n aol otu ao i s i np i s mta wm mta sm e e eocr cflod oo cfl oc r o i t t c . e is vy A . S e - O i. i - cd Ae s e r r ea a u rB w o y e t u- b f t l ot r a o t o - u e S n t m FW g e n iaol c n a e e d ot r u r ao n t i np i a r mta wm mta er ocod oocfl oc pfl r o O O . . . ' o ~ .
g e u r et n l h t u r a e P c d r a t n ut wm c nar t f ss f oor dd e ,S S o e p neg s s r p ek n r Ae s pnl i u d t oA eietL cs d ,d u e r an I n d oF e F o r emf i p oT gt i o ol e - nl e f t n a seg n l c e p i t i f wvi o D eS ml i t ai c p u v l l a oM i t e Ae n ef l op tbi a ot h v o M y S h t e n wD i t o ya rl m r l "s d t n i t a aP f o dl ee i i et i cl m& i t opi c i l a p ff a r P "M AE F P e oe '
~
l l [
O s s l b e e c a o r d n P sa t t g ns n e re md i k a sn M - sU e n sd s n Y i oAa s e R i cl eb r e' A Dip s O M h e na ee - fh M t t D eC - U r o plye r S l p as i u i ca S f E i t nr
,C I
aiteeM _ l o csE GSaDF y m,C A r rl a mf ofuA r et O i P r es PUN o ~
~
se o - ~ -
~
t O Nuclear Utilities Perspectives on Site-Specific Groundwater Modeling Carol Hornibrook, EPRI Man-Sung Yim, Contractor to EPRI Approach How important is GW modeling for NPP decommissioning dose modeling?
- Multiple nuclide mixture case
- Individual nuclide cases How important are the site characteristics?
- Three reference site conditions What are the major parameters of importance in GW modeling with the use of DandD and RESRAD?
Do the default input values of these parameters properly represent the industry? {} -
O' How important is GW modeling for NPP decommissioning dose modeling? A test case problem: 10 Nuclide mixture selected (Co-60, Ni-59, Sr-90, Cs-137, C-14, Eu-154, Pu-239, Pu-241, Am-241, Cm-243) Scaling factors used to mimic the actual inventory of radionuclides. O Results -Important pathways 10 nuclide mixture case DandD .
- External 84 % - Agricultural (non-aqueous) 13.7 % - Aquatic food 1.6% - Irrigated food 0.48 % - Drinking water 0.28 %
RESRAD
- External 96.4 % - Agricultural (non-aqueous) 3.6%
1 2
O Key parameters ofimportance Multiple nuclide mixture case
. DandD
- Indoor time and shielding
- Outdoor time and shielding
- Fraction of human diet from the garden
- Human consumption of root vegetables RESRAD l
- Indoor time
- extemal gamma shielding j
- Outdoor time
- Area of contaminated land i I
O Individual nuclide cases Sr-90:Importantpathways
- DandD
- Agricultural food (non-aqueous) 99.8 %
- Irrigated food 0.15 %
- Drinking water 0.04 %
- Aquatic food 0.04 %
3
- External gamma 0.02 %
RESRAD '
- Agricultural food (non-aqueous) 99.6 %
- External gainma 0.29 %
- Soil ingestion 0.08 %
U, 3
1 i l l 1 0' Key parameters ofimportance Sr-90 DandD
- Fraction of human diet from the garden
- Soil to plant transfer factor (leafy vegetables)
- Human consumption of milk
- Plant to milk transfer factor
- Human consumption ofleafy vegetables RESRAD
- Soil to plant transfer factor
- Human consumption of fruits, vegetables, grain O
Individual nuclide cases Cs-137: Important pathways DandD (
- Aquatic food 66.4 %
-Irrigated food 13.6 %
- Drinking water 1.99 %
- External gamma 0.49 %
- Agricultural food (non-aqueous) 0.19 %
- RESRAD
- External gamma 83.5 %
- Agricultural food {
16.4 %
- Soil ingestion (non-aqueous) 0.06 %
4
i
- 1 l
O Key parameters ofimportance Cs-137
- D et>d D '
- Kd I
- Saturation ratio of the unsaturated zone I
- Thickness of unsaturated zone
- Density of unsaturated zone
- Infiltration rate
- Human consumption of fish
- Fish bicaccumulation
- RESRAD
- Indoor time
- External gamma shielding factor
- Outdoor time O
How important are the site characteristics? DandD vs. RESRAD Three types ofreference sites
- Site A: low potential for groundwater contamination Site B: median potential for groundwater contamination Site C: high potential for groundwater contamination s
O 5 b
O Ikusic Pararnter Values forSks A, B, and C 1.Rrcmtniraalarrhmeraataxe g(g g mud arr Ass mDr) bpruitst had pvuury dhne stilsamrm in.htm m pmerv frv4) siteA sihycby 45.7 32.6 104 0492 Q2 7.6?o2 ste B kmfudy 15.2 1090 4.9 Q435 Q2 1.7&>l km/shinkirg usegsedcbv sitec adgawi 3Q5 4WD 4.05 Q395 Q2 4.0fol
- 2. Rrs:nraalare as wr x uin,> aug w aam.
nmuv siteA sh&mtantgtkintus 7.43e3 Q395 Q2 siten sarhuutmsae ISbt Q395 Q2 sieC sadgawhdtiamlimsue 22bt QJ)5 Q2 O\ DandD vs. RESRAD (1 PCi/g) Comparison of the Predicted PeakDose site A site B site C DandD RESRAD DandD RESRAD DandD RESRAD - Sr-90 59.3
)
0.731 59.3 0.731 59.3 0.731 Cs-137 2.03 1.76 2.04 1.76 8.03 1.76 Pu-239 13.5 0.174 13.6 0.1 74 86.3 0.174 Co-60 6.87 7.69 6.87 7.69 6.87 7.69 C-14 0.119 0.254 0.17 3.09 2.1 2.62 1-129 3.88 0.127 29 0.127 272 1.64 6
i
~
( Do the default input values of these parameters properly represent the industry? Key groundwater model parameters (with DandD)
- Thickness of unsaturated zone
- Saturation ratio of unsaturated zone
- Density of unsaturated zone
-Infiltration rate
- Kd Test analysis.
Thickness of unsaturated zone Raw data of 211 independent measurements available. {
- Data fit with maximum likelihood estimation.
Results: Lognormal distribution geometric mean = 2.296 geometric standard deviation = 1.265 (Goodness-of-fit, Kolmogorov-Smirnov test: Dn=0.0549 < 0.0936 (l.36/sqrt(N))
- 95% = 1.240 m
- 90% = 1.963 m
. DandD default input = 1.22 m 7
O Distribution of unsaturated zone thickness data
'f
(! II Il lll ....::&
/
J . . - . - .-- 3-O 8 .}}