United States Nuclear Regulatory Commission - Protecting People and the Environment
Home > NRC Library > Document Collections > ACNW > Meeting Transcripts > 2000 > 120th Meeting - July 27, 2000

120th Advisory Committee on Nuclear Waste (ACNW) Meeting, July 27, 2000


                       UNITED STATES OF AMERICA
                     NUCLEAR REGULATORY COMMISSION
                                  ***
                  ADVISORY COMMITTEE ON NUCLEAR WASTE
                          120TH ACNW MEETING
                                  ***

                                   Nuclear Regulatory Commission
                                   Room T2B3
                                   Two White Flint North
                                   11545 Rockville Pike
                                   Rockville, Maryland

                                   Thursday, July 27, 2000

               The Commission met in open session, pursuant to
     notice, at 8:31 a.m., THE HONORABLE DR. B. JOHN GARRICK,
     Chairman of the Committee, presiding.
     MEMBERS PRESENT:
               DR. JOHN B. GARRICK, Chairman
               DR. GEORGE W. HORNBERGER, Vice Chairman
               DR. RAYMOND G. WYMER
               MR. MILTON N. LEVENSON.     STAFF AND PARTICIPANTS:
     DR. JOHN T. LARKINS, Executive Director, ACRS/ACNW
     MR. HOWARD LARSON, Acting Associated Director, ACRS/ACNW
     MR. RICHARD K. MAJOR,, ACNW Staff
     MS. LYNN DEERING, ACNW Staff
     MR. AMARJIT SINGH, ACNW Staff
     DR. ANDREW C. CAMPBELL, ACNW Staff
     MR. MEDHAT EL-ZEFTAWY, ACRS Staff
     ROBERT A. NELSON, NMSS, Division of Waste Management, NRC
     JOHN T. BUCKLEY, NMSS
     STUART A. RICHARDS, Office of Nuclear Reactor Regulation,
     Division of Licensing and Project Management, NRC
     LARRY W. CAMPER, NMSS, Division of Waste Management
     CHERYL TROTTIER, Office of Nuclear Regulatory Research,
     Division of Risk Analysis and Applications
     TOM L. FREDRICHS, Office of Nuclear Regulatory Research,
     Division of Licensing and Project Management, NRC
     WILLIAM R. OTT, Office of Nuclear Regulatory Research,
     Division of Risk Analysis and Applications, NRC
     PHILIP D. MEYER, U.S. Department of Energy, Pacific
     Northwest National Laboratory
     GLENDON W. GEE, U.S. Department of Energy, Northwest Pacific
     National Laboratory
     SLOMO P. NEUMAN, University of Arizona, Tucson
     PETER J. WIERENGA, University of Arizona, Tucson.                            C O N T E N T S
     ITEM                                                    PAGE
     OPENING STATEMENT                                       216
     STATUS OF THE NRC'S DECOMMISSIONING PROGRAM             217
     HYDROLOGY RESEARCH                                      281 .                          P R O C E E I N G S
                                                      [8:31 a.m.]
               CHAIRMAN GARRICK:  Good morning.  The meeting will
     now come to order.
               This is the third day of the 120th meeting of the
     Advisory Committee on Nuclear Waste.  This meeting will be
     open to the public, and today the Committee will hear an
     overview from the NRC Staff on the Decommissioning Program.
               We will review a project by NRC's Office of
     Nuclear Regulatory Research on Hydrogeological Model
     Development and Parameter Uncertainty; continue our
     preparation of ACNW reports; and finally discuss matters
     related to the conduct of Committee activities and matters
     and specific issues that were not completed earlier.
               Howard Larson is the Designated Federal Official
     for the initial portion of today's meeting.  This meeting is
     being conducted in accordance with the provisions of the
     Federal Advisory Committee Act.  We have received no written
     statements or requests to make oral statements from members
     of the public regarding today's session.  Should anyone wish
     to, please notify a member of the staff.
               It is requested that each speaker use a
     microphone, identify themselves, and speak clearly and
     loudly.
               We're going to hear about the Decommissioning
     Program, as I said, but what's of great interest to the
     Committee today is that we're going to get an overview of
     the entire program.  This might be the first time we've done
     that.
               And Ray Wymer, the Committee Member that's our
     expert in this area will take the lead in the questioning
     and discussions.
               DR. WYMER:  Thanks, John.  He has already put me
     on thin ice now by calling me an expert.
               [Laughter.]
               DR. WYMER:  This is part of the meeting I have
     been looking forward to, actually.  We're going to have a
     discussion, I think, of the integration of the three parts
     of the D&D program.
               We've heard about the site decommissioning
     management plan a number of times, and we've heard about the
     license termination plan, and we've heard about the modeling
     studies, all separately.
               Now it's my understanding that we're going to hear
     it in an integrated fashion, and I'm interested to see if
     the whole is greater than the sum of the parts, or whether
     the integration is really just the three parts again.
               Bob Nelson, I think, is going to lead off here,
     and I trust he will introduce the succeeding speakers, so
     let's commence.
               MR. NELSON:  Thank you.  Good morning.  My name is
     Bob Nelson.  I'm Chief of the Facilities Decommissioning
     Section in the Decommissioning Branch of the Division of
     Waste Management.
               Joining me today in this presentation are Stu
     Richards, Project Director for Decommissioning, Office of
     Nuclear Reactor Regulation; and John Buckley of my section.
               We welcome the opportunity to discuss the
     Decommissioning Program with the Committee.  This briefing
     is intended to provide the Committee with the status of the
     Decommissioning Program.
               A similar status report was provided to the
     Commission in SECY 00-094, dated April 25th of this year.
     The presentation based largely on that paper, with some
     added information.
               Copies of the paper and this presentation are in
     the rear of the room for the members of the public.
               The Staff will provide a briefing to the
     Commission on the Program on October 23rd.
               The function of the Decommissioning Program is to
     regulate the decontamination and decommissioning of material
     and fuel cycle licensees, power reactors and non-power
     reactors, resulting in the ultimate goal of license
     termination.
               A broad spectrum of activities is associated with
     this program, as described in Attachment 1 to the Commission
     paper.  In meeting the briefing objectives stated here, we
     will discuss each of these principal activities.
               This slide just provides a brief summary of the
     Decommissioning Program that I will describe in more detail
     in subsequent portions of this presentation.
               On July 21st, 1997, the NRC published its final
     rule on radiological criteria for license termination,
     commonly called the License Termination Rule as Subpart E to
     10 CRF Part 20.
               NRC regulations require that materials licensees
     submit decommissioning plans to support decommissioning of
     their facility, if such is required by a license condition,
     or the procedures and activities necessary to carry out the
     decommissioning have not been approved by NRC, and these
     procedures could increase the potential public health and
     safety impacts on workers or the public.
               NRC regulations also require that reactor
     licensees submit post-shutdown decommissioning activity
     reports, commonly referred to as PSDARs, and license
     termination plans for LTPs to support the decommissioning of
     nuclear power facilities.
               The Staff is currently developing guidance for the
     Staff and for licensees to use in reviewing and developing
     these plans and other information submitted by licensees to
     support decommissioning of their facilities.
               Some of the more importance guidance documents are
     listed on this slide.  A more complete listing is supplied
     in Attachment 15 to the SECY.
               Since there are some acronyms used here on the
     slide, I'll point out that MARSSIM, as you are probably
     aware, is the Multi Agency Radiological Site Survey and
     Investigation Manual, and Reg Guide 1.179 is the Standard
     Format and Content Guide for license termination plans.
               Material and fuel cycle decommissioning activities
     include regulatory oversight of the site decommissioning
     management plan sites and other complex decommissioning
     sites; implementing the Commission's direction under
     Directive-Setting Initiative 9, by initiating a pilot study
     for performing decommissioning without submittal of a
     decommissioning plan;
               Undertaking license termination file reviews;
     performing financial assurance reviews; providing West
     Valley oversight; interacting with EPA and ISCORS,
     inspecting SDMP and other complex decommissioning sites;
               Maintaining the computerized risk assessment and
     data analysis lab or CRADL; and Office of Nuclear Regulatory
     Research providing data and models to support performance
     assessments.
               You can see that we have a rather complex program
     encompassing those elements.  We also have extensive
     interactions with other agencies, and some of those are
     listed here on the slide.
               I'll discuss West Valley a little later in the
     presentation, so I won't dwell on it here.  ISCORS are the
     Interagency Steering Committee on Radiation Standards, and
     we work extensively with EPA through ISCORS, principally, to
     resolve issues related to regulation of radionuclides.
               This interaction is necessary to avoid unnecessary
     duplication of regulatory requirements.  Principal ISCORS
     activities are carried out through its subcommittees which
     report back to the full ISCORS Committee at its quarterly
     meetings.
               The current Subcommittees include Risk
     Harmonization, Mixed Waste, Recycle or Clearance,
     Decommissioning/Cleanup, NORM, and Sewer Reconcentration.
               ISCORS produces an annual report, NUREG 1707.  The
     1990 report is currently in publication and should be
     available soon.
               ISCORS also maintains a website at ISCORS.org for
     members of the public to have access.
               The principal activity currently going on now with
     the EPA is the negotiation of a Memorandum of Understanding
     concerning the two agencies' interaction regarding cleanup
     of sites.  That activity is currently ongoing, and the MOU
     has not been finalized at this point.
               In addition, the Staff has been interfacing
     extensively with EPA Region I regarding the review of
     license termination plans submitted by Maine Yankee.  We
     anticipate similar interactions with our review of the
     Connecticut Yankee Haddam Neck LTP which we have just
     received.
               Similar to our support of the development of the
     MARSSIM or Multi Agency Radiological Survey and Site
     Investigation Manual, we are participating in an interagency
     working group that is developing MARLAP, or the Multi Agency
     Laboratory Accreditation Program.
               Finally, we routinely work with various state
     public health and safety organizations, other federal
     agencies, and tribal organizations in conjunction with our
     safety and environmental reviews of decommissioning plans,
     license termination plans, and other license submittals.
               I'd like to introduce John Buckley, who will
     discuss our oversight of SDMP and complex decommissioning
     cases and our terminated license review.  John?
               MR. BUCKLEY:  Thanks, Bob.  Good morning.  My name
     is John Buckley.
               As Bob mentioned earlier, one of the main
     activities in materials decommissioning is oversight of the
     SDMP and otherwise complex decommissioning sites programs.
               As indicated on Slide 7, there are currently 29
     SDMP sites or complex sites; 26 of those 29 are actually
     SDMP sites; three are complex decommissioning sites.
               Twenty-three of the 29 sites have already
     submitted decommissioning plans.  The Staff has reviewed and
     approved 13 of the 23 decommissioning plans which we have
     received.  Of the 29 sites, the Staff expects that by the
     time we get all DPs in hand, 11 of the sites will request
     restricted release.
               DR. WYMER:  These are all reactors?
               MR. BUCKLEY:  No, these are material and fuel
     cycle facilities.
               And 11 of the 29 current sites may be transferred
     to Agreement States by the Year 2002.  One site will go to
     Minnesota, possibly, and ten to Pennsylvania.  To date, the
     Staff has removed 31 sites from the SDMP.
               Detailed information on the current 29 sites in
     presented in Attachments 3, 4, 5, and 7 of the Commission
     paper, so if you need additional information, that's the
     place to look.
               In 1990 the staff began a review of all previously
     terminated license files to assure that the licenses were
     properly terminated and that the sites posed no threat to
     public health and safety or the environment.
               NRC contracted with Oak Ridge National Laboratory
     to conduct this review.  Oak Ridge reviewed approximately
     37,000 terminated license files.
               Of the 37,000 files reviewed, regional inspectors
     have verified that 38 sites are contaminated with limits
     above the unrestricted release limits.  A listing of the
     contaminated sites and their decommissioning status is
     presented in Attachment 6 of SECY-94.
               The regions have almost completed their review of
     all the terminated license sites.  Approximately 120 sites
     are left for review by the regions.  Of those, 29 are loose
     material sites and approximately 92 are sealed source sites.
               I will turn this presentation back over to Bob.
               MR. NELSON:  Do you have any questions on those
     two slides before we move on?
               DR. WYMER:  How many reactors are there, just the
     two that are --
               MR. NELSON:  We will get into those.
               MR. BUCKLEY:  The reactors come later.  This is
     just the materials.
               DR. WYMER:  Okay.
               DR. HORNBERGER:  What is a loose material site?
               MR. BUCKLEY:  I'm sorry?
               DR. HORNBERGER:  Loose material.
               MR. BUCKLEY:  It is a nonsealed source.  Anything
     -- everything not a sealed source is a loose material.
               DR. CAMPBELL:  I have a question.  What was the
     release limit you used, is it the new LTR 25 millirem limit?
     In other words, did they do an analysis of these 38 sites to
     determine that they exceed 25 millirem?  This says above
     unrestricted release limit.  The unrestricted release limit
     is 25 millirem.
               MR. BUCKLEY:  But it has changed.  So some of
     these sites were released prior to --
               DR. CAMPBELL:  So this is the old standard.
               MR. BUCKLEY:  Some are under the old standard,
     some are under the new standard.  Attachment, if you look at
     Attachment 3, it will provide you an indication of which
     criteria is being used.
               MR. NELSON:  To the SECY, Attachment 3.
               MR. BUCKLEY:  Correct.
               MR. NELSON:  To the SECY.
               DR. CAMPBELL:  So some of them, they actually did
     assessments and determined that the amount of material on
     the site would exceeded 25 millirem?
               MR. BUCKLEY:  Correct.
               DR. CAMPBELL:  And then some of them you used so
     many picocuries per gram contamination levels that were the
     old standard?
               MR. BUCKLEY:  Correct.
               DR. HORNBERGER:  And the sites with 25 millirem,
     these were basically field surveys that demonstrated that
     they exceed, or was it a computation?
               MR. BUCKLEY:  I think what the regions did, and I
     don't want to speak for them, but they went out and did
     field surveys.  If they found something that looked like it
     was beyond twice, three times background, then they flagged
     it.
               MR. NELSON:  Any other questions before we move
     on?
               [No response.]
               MR. NELSON:  The staff continues to implement the
     Commission's direction under Direction Setting Issue for
     DSI-9.  There are three facilities currently in the pilot
     program, as identified on this slide.  The purpose of the
     pilot program was to look at, exam the feasibility of
     licensees performing decommissioning without a prior
     approved decommissioning plan.
               The pilot study was approved by the Commission in
     a staff requirements memorandum dated June 30th, 1998.  We
     last reported to the Commission on this topic in
     SECY-99-160, dated June 22nd, 1999.
               We plan to meet with each of these licensees this
     fall to determine how well they have done, where they are in
     their cleanup and what lessons we could learn from this
     activity, and provided a status report to the Commission in
     January.
               For example, the first site, the Westinghouse pump
     repair facility in Cheswick, is nearing completion of its
     activities, and they plan to submit a final survey report
     and request for release of portions of their site very soon.
     The other sites, I don't have a current status on, but we
     will be getting that as we move towards our fall meeting.
               Another aspect of our decommissioning program is
     review of financial assurance.  The staff routinely reviews
     financial assurance submittals for materials and fuel cycle
     facilities and maintains a financial assurance instrument
     security program.  This entails review of decommissioning
     cost estimates, as well as the financial assurance
     instrument, whether it be a letter of credit, a standby
     trust agreement, et cetera, to verify that the instrument
     meets our requirements, could be executed if necessary, and
     that the cost estimate is reasonable for the activities
     anticipated.  Routinely, between 40 and 60 financial
     assurance submittals are reviewed each year.
               Also, under this area we are consolidating our
     financial assurance guidance which has consisted of a
     standard format and content guide, Reg. Guide 3.66, and a
     Standard Review Plan for decommissioning financial assurance
     submittals, NUREG-1337.  The consolidation of these
     documents will be into our new decommissioning Standard
     Review Plan so that we have fewer guidance, individual
     guidance documents on the street, and most of the guidance
     will be contained within the Standard Review Plan, a module
     and an appendix to that plan, that we briefed you on, I
     believe earlier in the year.
               NRC's decommissioning responsibilities for the
     West Valley Demonstration Project and the West Valley site
     are specified under the West Valley Demonstration Project
     Act.  Presently this activity includes prescribing
     decontamination and decommissioning criteria, reviewing
     draft portions of the EIS for decommissioning, and closure
     of the site, reviewing safety analysis reports prepared by
     DOE, and performing periodic onsite monitoring of project
     activities and records to assure radiological health and
     safety.
               The Commission's draft policy statement regarding
     decommissioning criteria for the West Valley Demonstration
     Project and the West Valley site was issued in December 1999
     for public comment.  The draft policy statement specified
     NRC's license termination rule as decommissioning criteria.
     We are currently reviewing comments received both from a
     public meeting, as well as written comments, and are
     preparing a follow-on submittal to the Commission.
               It is currently the staff's plan to brief you in
     more depth on this in your October meeting, at which time we
     will have prepared a submittal to the Commission.  We will
     discuss that with you at that time.
               In addition, we have developed a Commission paper
     on stakeholder involvement, because of the many stakeholders
     and complexities of that activity, and that is currently in
     its final stages of management review.  It should be going
     to the Commission shortly as an information paper.
               The Office of Nuclear Regulatory Research
     continues to provide data and models to NMSS to support
     assessments of public exposure to environmental releases of
     radioactive material from site decommissioning.
               Research has provided NMSS with data on
     radionuclide solubilities that will be used to assess
     releases from ore-processing slag; data on degradation of
     archeological slags that will be used to assess the
     long-term performance of slags as a source of radioactive
     contamination.
               As you may be aware, many of our SDMP sites
     processed ore containing radioactive material, and as a
     result have large amounts of contaminated slag onsite.
               We will also produce guidance on characterization
     of decommissioning sites containing these slags, and provide
     documentation on unsaturated zone monitoring strategies for
     use in review of monitoring proposals for licensing actions
     concerning decommissioning and waste disposal facilities in
     unsaturated media.
               We assisted in the development of technical bases
     to support selection of site-specific parameter values for
     estimating flux and transport in dose assessment codes.
               We are developing a probabilistic version of the
     computer code, RESRAD, and modification of the Sandia
     decision support system to allow multidimensional
     groundwater pathway analyses.
               Now, I'm going to move into the reactor
     decommissioning area, and I'll provide an overview of the
     NMSS involvement, and then I will turn this presentation
     over to Stu Richards, who will discus NRR activities.
               From the NMSS side, reactor decommissioning
     includes our project management of technical review
     responsibility for decommissioning two power reactors.  NRR
     has project management and licensing oversight for 17
     decommissioning reactor facilities.
               The program also includes implementation of the
     plan developed in response to Commission Direction-Setting
     Initiative Number 24; development of standardized technical
     specifications for decommissioning; conduct of core
     inspections and project management for all licensed,
     non-power reactors.
               NMSS has project management and technical review
     responsibility for the Firm I and Peach Bottom Unit I power
     reactors; NRR has project managing and licensing oversight
     for 17 power plants that have either submitted DPs or their
     equivalent, or PSDARS.
               We have Attachment 11 and 12 to the SECY paper
     which provide status information on each of these plants.
               The NMSS principal activities in support of
     reactor decommissioning are summarized on this slide.
               [Pause.]
               Technical reviews of license termination plans are
     currently underway for Maine Yankee, Trojan, and Saxton.  An
     acceptance review is in progress for the Connecticut Yankee
     license termination plan.
               NMSS has supported public meetings for the first
     three, and will provide similar support for Connecticut
     Yankee.
               In addition, we will be conducting confirmatory
     radiological surveys at these plants in support of the
     license termination process.  Such surveys are currently
     scheduled in August for the Maine Yankee and Trojan plants.
               Project management responsibilities for the Peach
     Bottom and Fermi plants for NMSS have been limited to
     processing minor amendment requests.  It currently does not
     represent a major workload for us.
               In the guidance development area, NMSS has
     recently published -- finalized and published NUREG 1700,
     the Standard Review Plan for License Termination Plans, and
     we are supporting NRR in development of a variety of
     guidance documents related to decommissioning.
               Are there any questions on that before I turn the
     presentation over to Stu Richards?
               [No response.]
               MR. RICHARDS:  Good morning.  Can you hear me all
     right?  I'm Stu Richards.  I'm the NRR Branch Chief
     responsible for the decommissioning power reactors.  I'm
     also responsible for the operating reactors in Region IV.
               I'd like to talk a little bit about what NRR does,
     and hopefully clarify this integration function here.  Just
     a little bit of background:
               When a power reactor shuts down -- well, let me
     back up.  The rules are written for reactors to shut down
     with some forewarning, so the rules are written such that a
     plant enters the normal end of life but recognize they're
     going to go through decommissioning, and they have time to
     plan for it, and go through a normal process.
               Unfortunately, virtually all the reactors that are
     presently in decommissioning, didn't get there that way.
     They found themselves in some kind of a problem and ended up
     going into decommissioning on fairly short notice.  So
     there's a little bit of a disconnect between the experience
     we've had to date and the way it's supposed to really
     happen.
               Hopefully that will change in the future, but when
     a plant does enter the decommissioning phase of operation,
     the first thing they want to do is, they want to make the
     transition from an operating facility to a decommissioning
     facility.
               And that is primarily NRR's role in the
     decommissioning activities.  We have the front end, we take
     the plant from an operating unit into decommissioning, and
     really NMSS has got the back end when it comes time to
     terminate the license and determine what to do with the rad
     waste.  That's their end of the business, and there's a lot
     of work in between where I think we both do a lot of work
     together.  So I spend a lot of time talking to Larry Camper
     and Bob Nelson, and we actually are, I think, pretty well
     integrated.
               There is a lot of transition right now in the
     decommissioning world.  And I'll try and give you a little
     flavor of what's going on, and then hopefully answer your
     questions.
               Again, the primary interest of plants that enter
     decommissioning is to get rid of all those operating reactor
     rules that apply, because, you know, when they shut down,
     they still have a full Part 50 license that applies to that
     facility.
               The way it has worked in the past is that
     licensees have had to come in to NRR and request amendments
     to their tech specs or to get exemptions from the Part 50
     requirements that they don't think should apply to a
     decommissioning plant.  That entails a lot of paperwork and
     analysis on their part, and also takes up a lot of the
     Staff's time to review all that work and to issue approvals.
               We've been doing that on a site-by-site,
     case-by-case basis, so it's quite work-intensive.
               About a year ago or year and a half ago, we
     thought, hey, it's time to stop trying to do that.  Let's
     look at putting together what we call the integrated
     rulemaking and try and put together a process where plants
     can go through that phase and get relief from operating
     reactor requirements without having to provide us with
     paperwork and analysis and without us having to review it;
     just put it in the regulations.
               In order to take that step, we felt we needed a
     single, solid technical basis for backing up these
     regulatory changes.  So we put together a technical working
     group to take a look specifically at the risks of storing
     spent fuel onsite in the spent fuel pool.
               And that study has been going on now for a little
     over a year.  That group came out with a draft report early
     this year.  We got comments back from the ACRS, from a
     number of public stakeholders, and from the industry, and
     those comments are trying to be resolved with a date of
     August the 31st to come out with the final risk report.
               But the idea here is that the primary event of
     consideration is the zirconium fire in the spent fuel pool.
     If somehow you drain all the water out and you get the fuel
     to start to burn, you have a very significant event, because
     you don't have a containment building; you can have multiple
     cores; you have a real problem.
               On the other hand, the probability of that event
     occurring is very, very, very small.  The draft report said
     it was less than three times ten to the minus six; at this
     point we can't say that's incredible and take it off the
     table, but it's pretty close to that.
               So, at any rate, that's a little bit of background
     on what we're doing.  The integrated rulemaking, the slide
     lists the five areas we're looking at.
               What we did is, we picked out three or four of
     these areas as areas where the industry felt they could get
     immediate financial relief, if we could do something to make
     that transition easier for them.
               Emergency preparedness, safeguards, and insurance,
     cost the industry a lot of money, and they feel that when
     they shut the plant down, they should no longer have to meet
     the operating reactor requirements that incur those
     expenses.
               So we're taking a hard look at those three.  The
     backfit, if you go to the backfit rule for decommissioning
     plants -- or, actually, the backfit rule, period -- it's
     really written for operating reactors, and if you put on a
     legal hat, you could say this rule doesn't apply to
     decommissioning plants; it's not written that way.
               Decommissioning plants felt like they should have
     the same protections as an operating reactor on the backfit
     rule, and the Commission told us to treat them the same, and
     then when we get around to it, to put into the rule, so
     that's why that's there.
               And then operator training and staffing, again,
     when a plant shuts down, they still have all the training
     programs and staffing requirements of a Part 50 license.
     You know, they want to get away from that, they want to
     transition to the minimum staffing they need in order to
     basically maintain the spent fuel pool.
               So, we have put together a rulemaking plan.  We
     submitted that to the Commission at the end of June.  That
     is SECY 001-45, I believe.  That paper is now public,
     although the Commission has not voted on it.
               We asked specifically that we be allowed to
     release it to the public in order to get prompt feedback.
     So we are waiting for direction from the Commission on how
     to proceed on that.
               The second major bullet on the slide talks about
     regulatory improvements, and what we're talking about here
     is, beyond the first five issues that we were looking at, we
     wanted to -- or we proposed to the Commission to take a look
     at all regulations in Part 50 that apply to decommissioning
     reactors, and our proposal is to modify all those
     regulations in some form so that we have one section of the
     regulations that applies to decommissioning plants.
               As it presently stands, you know, you're searching
     through Part 50, trying to figure out, hey, what in here
     applies to me now that I'm shut down?  So we thought that it
     would be proper to try and bring those altogether into one
     place.
               We owe the Commission more information on that by
     September 15th of this year, and there's a resource issue
     there of whether that's the right approach to take.
               The industry has a different view on these two
     items.  They have recently come in with a letter to us that
     said we think you ought to do this in one shot, just put it
     into one rulemaking, and they felt that we could do the
     whole thing in 24 months.
               Personally, I disagree with that.  I think that's
     unrealistic to think that we can overhaul all the
     decommissioning regulations in 24 months.
               And I feel that, you know, if you take that route,
     you end up with everything being held hostage to whatever
     the one hardest issue is.
               So we had a workshop earlier this week with the
     industry out on the West Coast and challenged them to make
     sure they are asking for what they really want because if
     they want it as one package it is likely to drag on longer
     than 24 months.
               On the other hand, there are no reactors right now
     that appear to be entering the decommissioning phase, so you
     can argue that what is the rush?
               The next bullet talks about the generic
     Environmental Impact Statement for decommissioning.  That
     was last done in 1988 and we are doing an update on that.
     We are working closely with the Environmental Protection
     Agency and the industry.  I might note that specifically
     this update is going to consider entombment, rubbelization,
     and partial site release, which is selling off pieces of
     land before the plant actually enters the decommissioning
     phase, or the final license termination phase.
               We are also responsible for guidance documents.
     Bob talked about some of the ones that NMSS is doing.  We
     provide guidance documents for the post-shutdown
     decommissioning activities report.
               When a plant enters the decommissioning phase one
     of the first things they are required to do is to put
     together this overview document that describes what they
     intend to do in the future as far as decommissioning the
     facility.  We also provide guidance to the inspectors.  It
     talks about a handbook for the inspectors that we are
     putting together and various NUREGs.
               One I would like to mention is that we do put
     together a Frequently Asked Questions NUREG that we find to
     be very helpful.  We have a lot of public meetings and it is
     nice to be able to hand out a book that says, hey, here's a
     lot of the questions you might ask and has the answers in
     it.
               We are also responsible for the Decommissioning
     Inspection Program while the facility is under the
     responsibility of NRR and work with the regions to define
     that program to make sure it is implemented properly.
               Bob mentioned it before, but just to make sure it
     is clear, we have a memorandum of understanding between NRR
     and NMSS and what it says is that NRR is the Project Manager
     for the facility until all the fuel is out of the spent fuel
     pool.  When it goes to either offsite, which right now it
     isn't doing, or it goes to dry storage, at that point NRR
     transfers responsibility for project management to NMSS, so
     again on the big picture we are primarily concerned with the
     facility entering decommissioning and making that transition
     to truly a decommissioning facility, getting them through
     the changes in the tech specs and the regulations.
               NMSS of course is focused on what to do with the
     waste, final surveys and license termination.  There is a
     lot of overlap there and we do spend a lot of time talking
     together about it.
                    Flipping on to the next slide, the next slide
     is really just a status of facilities, the power reactor
     facilities that have entered decommissioning, twenty-one
     reactors between 1963 and 1998.  Two have completed decon
     and dismantlement.  That's Fort St. Vrain and Shoreham.  We
     have got six that are undergoing active decon and
     dismantlement, nine that have chosen the safe store route
     where they are going to let the facility sit for awhile
     before they decide what to do with it, and we have four
     facilities that have a combination of storage and then
     decontamination and dismantlement.
               That completes the two slides I have.
               DR. WYMER:  What actually constitutes going into
     decommissioning?  If two have completed decon and
     dismantlement and if Maine Yankee is now coming in to talk
     to you about the two, are they not entering decommissioning?
     What does that mean?
               MR. RICHARDS:  Well, the two are the two that have
     had their licenses terminated.
               DR. WYMER:  That's already done.  Okay.
               MR. RICHARDS:  The six -- let me back up to the
     options.  You can -- when you shut the facility down you can
     enter SAFESTOR, basically bottle the facility up and let it
     sit, you have to complete the process within 60 years, or
     you can enter direct and active decon and dismantlement of
     the facility, such as Maine Yankee is doing.
               DR. WYMER:  That isn't entering decommissioning?
               MR. RICHARDS:  Yes, that is.  All of these are
     entering decommissioning.
               DR. WYMER:  Oh, I thought you said earlier none
     were entering decommissioning.
               MR. RICHARDS:  No, no, no, no.  I'm sorry.
               DR. WYMER:  Oh -- no new ones.  Okay.
               MR. RICHARDS:  This is a summary of these 21
     reactors.  The idea here is we have got some that shut down
     and enter decommissioning, active decommissioning, taking
     the facility apart right away.
               We have got a number of facilities that shut down
     and elected to put the plant into a SAFESTOR condition.
               DR. WYMER:  Okay.
               MR. RICHARDS:  And we have got some that have a
     combination of both.
               DR. WYMER:  I thought I heard you say no reactors
     are entering the decommissioning phase.
               MR. RICHARDS:  No, I'm sorry, I meant no new
     reactors are on the horizon.  We don't have any reactors
     right now that are operating that we know are scheduled to
     enter decommissioning.
               DR. WYMER:  Okay.
               MR. RICHARDS:  That is because of the deregulation
     of the industry and I think the license renewal and the
     purchase of facilities.
               DR. WYMER:  Okay, my apologies.
               CHAIRMAN GARRICK:  Isn't the whole issue of the
     schedule for a repository and the success of dry cask
     storage, for example, going to have a major impact on what
     constitutes decommissioning and the transition, say, between
     NRR and NMSS?
               Well, supposing Yucca Mountain doesn't come about
     and simultaneously that dry cask storage works out to the
     satisfaction of everybody and the public picks up on this as
     the solution, at least for the time being, and the time
     being could be 100 years or so -- that would change things a
     whole lot, would it not?
               MR. RICHARDS:  Well, actually, right now I think
     both the industry and the NRC are carrying out their
     business based on the assumption that facilities are going
     to use the dry cask storage option simply because even I
     guess on the present schedule Yucca Mountain is pretty far
     out in the future.
               CHAIRMAN GARRICK:  Yes.
               MR. RICHARDS:  There are a large number of
     operating reactors that already have dry storage.  There are
     a number of decommissioning plants that either have it or
     are moving in that direction.  There seems to be a
     recognition that if you want to get on with decommissioning,
     the thing to do is to license a dry storage facility under
     Part 72 and then just decommission the rest of your facility
     and get rid of the Part 50 license so all you have left is a
     dry storage ISFSI facility with maybe a very small staff of
     people care of it, and, you know, I am just speculating that
     it appears --
               CHAIRMAN GARRICK:  What if somebody comes along
     and says, look, since we are in a dry cask storage mode and
     we have fuel onsite, we would like to take advantage of the
     fact that this is a high level waste storage site and store
     other kinds of waste onsite that would be beyond the 25 MR.
     Are these kinds of proposals feasible?
               MR. RICHARDS:  If there is somebody here from the
     Spent Fuel Project Office, it might be best for them to
     answer, but I know there's facilities that are planning on
     storing the greater than Class C waste --
               CHAIRMAN GARRICK:  Right.
               MR. RICHARDS:  -- in dry storage casks with the
     spent fuel actually.
               CHAIRMAN GARRICK:  Right.
               MR. RICHARDS:  As far as the low level waste, in
     order to terminate the license, they have to get that
     material offsite and meet the 25 millirem ALARA criteria, so
     again in order to terminate your Part 50 license you have to
     decontaminate, remediate the site to that license
     termination rule criteria, but you can rid of the Part 50
     license and still have your dry storage over here separate.
               CHAIRMAN GARRICK:  I guess what I am asking is a
     realistic appraisal of the situation.  Does it suggest that
     quite possibly a lot of these will not get out from under
     Part 50?
               MR. RICHARDS:  Well, they have that option.  They
     can still --
               CHAIRMAN GARRICK:  But you don't have a sense of
     how this might play out as far as --
               MR. RICHARDS:  Right now I think most facilities
     plan to go dry storage with a Part 72 license that are in
     active decommissioning, trying to terminate their license.
               They are going to decontaminate the site to the
     license termination rule criteria, terminate the Part 50
     license, and you will be left with dry storage under Part
     72.  I think that is what they are doing.
               Larry, am I wrong on that?
               MR. CAMPER:  No, I think you are right.  There are
     also economic incentives to do that too, because of the
     different categories of licensing.
               I mean one of the things that prompts a movement
     from an operating facility to a facility in decommissioning
     is a categorical change in licensing fee, so that is clearly
     a motivator.
               As Stu has said, the trend is a movement toward
     isolated storage onsite because of the high level waste
     repository problem.  There will be probably a storage of
     greater than Class C but beyond that, no.
               I think frankly any movement or any move by the
     industry to store waste other than that or to collect waste
     from other sites would probably not -- certainly would not
     be met with political receptivity and would pose a number of
     challenges for us as well.
               I think what you are seeing now is what you are
     going to see for the foreseeable future.
               The other thing that is interesting too, and as
     Stu mentioned, we just came back from a conference at NEI
     out on the West Coast, and there was a time when we were
     anticipating more reactors moving into decommissioning than
     we are.  We can see, of the ones listed up here, we can see
     four more out there coming -- Humboldt Bay, SONGS I, Yankee
     Rowe coming back, and Big Rock Point.  Those are the ones we
     look out and project that might be coming along in terms of
     staff work for decommissioning per se.
               CHAIRMAN GARRICK:  That is the big thing that's
     changed over the last couple of years --
               MR. CAMPER:  Exactly.
               CHAIRMAN GARRICK:  -- is that license renewal
     expectation has gone from 4 or 5 percent up to maybe 80
     percent or 85 percent.
               MR. CAMPER:  Absolutely and of course the industry
     is very excited about that, as you might expect.
               CHAIRMAN GARRICK:  Right.
               MR. CAMPER:  But I think what you are seeing now
     is what you are going to see certainly for the foreseeable
     future in terms of how isolated storage is being handled.
               The Yucca Mountain Repository question, you know,
     who knows?
               CHAIRMAN GARRICK:  Okay.  Thank you.
               MR. LEVENSON:  Can I ask a somewhat general
     question, since we are on the idea that you have a crystal
     ball and can look ahead?
               Do you have any perception -- the U.S. is spending
     a fair amount of money to subsidize silo storage of spent
     fuel in the former Soviet Union countries as something that
     is significantly cheaper than dry cask storage.
               Do you have any perception that might be coming
     into the U.S. picture?
               It initiated here.  It's what has been done at the
     EBR-2 Reactor for 30 years.
               MR. CAMPER:  I can't comment on that.  I don't
     know enough about that trend to comment whether that will
     ever materialize.
               MR. LEVENSON:  You have heard no discussions?
               MR. CAMPER:  No, I am not aware of any.
               MR. NELSON:  I apologize for not introducing my
     Branch Chief, Larry Camper, Chief of the Decommissioning
     Branch, Division of Waste Management.
               MR. CAMPER:  We know each other.
               [Laughter.]
               MR. NELSON:  Moving on, as a result of our recent
     organization, NMSS Environmental Review responsibilities
     fall under the Environmental and Performance Assessment
     Branch of the Division of Waste Management.  However, those
     activities are budgeted under the decommissioning program,
     so I am including them here to provide a complete
     description of our budgeted program.
               The activities in the environmental review area
     include preparation and review of Environmental Impact
     Statements, or EISs, review of Environmental Assessments
     prepared by the Staff.
               Presently it is estimated that EISs will be
     prepared for the following SDMP and complex decommissioning
     sites -- the U.S. Army Jefferson Proving Ground; Dow
     Chemical Company; SCA Services; Michigan Department of
     Natural Resources; Mallinckrodt Chemical; Shield Alloy
     Metallurgical Corporation; Fan Steel; Kaiser Aluminum;
     Sequoyah Fuels Corporation; the Babcock & Wilcox Shallow
     Land Disposal Area; The Moly Corp., Incorporated Washington,
     Pennsylvania facility; and Whitaker Corporation.
               Three of these have already submitted
     decommissioning plans for restricted release.  They are Fan
     Steel, Sequoyah Fuels, and Moly Corp Washington.
               It is our practice to develop an EIS for all
     restricted release submittals.
               The others that I mentioned we anticipate either
     are or may submit a decommissioning plan calling for
     restricted release.
               The Branch will also prepare an EIS -- I should
     say that of the three -- Fan Steel, Sequoyah Fuels, and Moly
     Corp. Washington -- the Sequoyah Fuels' EIS is under
     development, the draft EIS.
               The Branch will also prepare an Environmental
     Impact Statement for the West Valley site.
               Environmental assessments must be prepared for
     most other licensing actions including approval of DPs
     involving unrestricted release for SDMP and complex
     decommissioning sites.
               The Environmental and Performance Assessment
     Branch reviews all the EAs that we develop.
               That's all I plan on saying about that activity.
     Are there any questions in that area?
               DR. WYMER:  Just a comment.  There is going to be
     an awful lot of restricted release sites around the country,
     and mostly it is the horizon that you can see to is 100 or
     200 years, something like that.  And the question comes up
     of, how about after that, you know?  What sort of plans or
     safeguards or what is in place to make sure -- I know that
     you have financial assurance considerations and you have
     various governmental involvements that have to be in place,
     if that is appropriate?  But, still, the horizon seems close
     compared to the duration of the risk.  Can you say anything
     about how comfortable you feel about that or what -- how it
     is handled?
               MR. NELSON:  Well, we are just -- restricted
     release is certainly a new approach to decommissioning.  We
     haven't done a restricted release, completed a restricted
     release approval under the License Termination Rule.  So, in
     some respects, we are learning as we go.
               I need to point, though, restricted release is
     that.  Our regulatory oversight would cease at the time we
     terminated the license.  So there would not be, for example,
     inspections.  There would not be reports from the licensee
     to us for review.  We would have to be satisfied that the
     licensee's plan for restricted release met all of our
     requirements and that we could -- and met the License
     Termination Rule such that we could cease regulatory
     oversight of the site.
               DR. WYMER:  That is the crux of the problem right
     there.  How do you make yourself comfortable with the fact
     that everything has got to be okay after you turn it loose?
               MR. NELSON:  Well, to some extent the same way we
     -- to some extent, exactly the same way we do for
     unrestricted release.  The dose limit is different but we
     still have to do a very similar assessment for restricted
     release.  Of course, there are differences, but the
     similarities are that we have to do -- we have to assess the
     dose assessment performed by the licensee and conclude that
     the resulting dose is a reasonable estimate of the dose.
     Whether that is above 25 or below 25, the analysis approach
     is largely the same.
               The real difference that we need to look at are
     the institutional controls and financial assurance that are
     put in place to keep the dose under 25.  That is the big
     difference.  We haven't reviewed one of those before.  And
     so we have developed some guidance in the Standard Review
     Plan, but that is going to be a significant area of our
     review.  In fact, we believe it is so significant that we
     will focus our review for those cases on financial assurance
     and institutional controls before we begin the technical
     review.
               We plan to do a phased review for restricted
     releases so that we can satisfy ourselves that those
     requirements, that we have some confidence that those
     requirements were met before we go into a significant
     expenditure for technical review.
               I don't know if I can get more specific on what we
     were looking for in a -- I think that the real, I don't
     think is as much the cost estimate.  I think we can
     reasonably estimate cost.  The question is, what mechanisms
     would we authorize other than, say, a transfer to DOE for
     long-term oversight?  What other mechanisms might we
     consider for institutional control?
               MR. LARSON:  That is on our agenda for October.
               MR. NELSON:  Yes.
               MR. LARSON:  And you can get a sense of Ray's
     interest in this topic.
               MR. NELSON:  Larry.
               MR. CAMPER:  Let me just add one thing to that.  I
     was going to comment on this later.  Getting back to your
     question about the integration of all these activities.  We
     have the restricted release scenario in the regulations, and
     it requires certain criteria, long-term durable controls,
     responsible third parties and things of this nature, as well
     as design considerations that go out beyond the timeframe
     that you were mentioning.  But in all of this, what is
     emerging for us, we think is a big problem is finding, is
     for these licensees to identify a responsible third party
     that will be in this for the long haul.
               As a result of that, the staff is working on two
     things.  One is we have had previously, and are currently
     finalizing working arrangements with the Department of
     Energy where DOE would be a cooperating agency on the
     Environmental Impact Statements for these restricted release
     scenarios.  That is something that there were some exchanges
     that went on between the two agencies back two or three
     years ago.  Now, some of these sites, Sequoyah Fuels, for
     example, is at a point where it is time for the DOE to
     emerge in that role.  We are working that issue.
               But, secondly, and more importantly, and I think
     this kind of gets back to your integration comment really,
     one of the largest challenges we face in decommissioning is
     to ensure that we find viable third parties that will
     oversee these sites for the long haul.  We are finding that
     while the regulations currently have what we think is a very
     good mechanism in terms of institutional controls defined
     within them, for licensees to find a responsible third party
     to step up, whether it be a state government, a local
     government or even a private entity is problematic.  And,
     therefore, we have grave concerns, as I think I pick up from
     your question as to, what can do this in terms of long-term
     stewardship?
               There is a mechanism under 151(b) of the '82 Act,
     the Nuclear Waste Policy Act, that allows sites to be
     transferred to DOE.  We will be talking more about that in
     your October meeting.  But we are starting to --
               DR. WYMER:  Transferring to DOE doesn't --
               MR. CAMPER:  I'm sorry?
               DR. WYMER:  Transferring to DOE doesn't change the
     nature of the problem at all, it just gets it out of your
     backyard.
               MR. CAMPER:  Well, what it does do, though, is it
     provides a mechanism for long-term stewardship.  I mean I
     think our regulations, in terms of the dose criteria, the
     requirements that have to be in place for a restricted
     release scenario to occur, are sound.  The issue I think,
     though, is who will be that third party for the long haul?
               DR. WYMER:  Yes.
               MR. CAMPER:  DOE, of course, is structured to do
     that.  They have the infrastructure, and they have a
     stewardship program that would seem to be ideal.
               DR. WYMER:  It is embryonic at the moment, but
     they do have one.
               MR. CAMPER:  Right.  So, we think that that is a
     pathway that needs to be explored more aggressive with DOE.
     And the staff has had a number of management level
     discussions with DOE managers about that.  There will be
     communications that will take place between the two
     organizations in the near future.  So that is an area where
     we think, in terms of the big picture, and the overall
     integration of activities, that is a key part of that
     puzzle.
               DR. WYMER:  That helps.  Thanks.
               MR. LEVENSON:  Let me just ask a nit question
     about that.  Is it to DOE or is it to the legal entity of
     the federal government, since there's motions afoot to
     dismantle DOE?
               MR. CAMPER:  Well, I wouldn't begin to comment on
     what is going to happen DOE as far as anything of
     dismantlement.  I can only tell you that right now, we are
     working to solve and make sure that the institutional
     controls, you know, methodology currently set forth in our
     regulation, the LTR, that DOE seems at this point in time,
     under its stewardship program, to be a viable pathway.
               Now, there are conditions for those transfers.
     And I dare say that DOE will be concerned that all those
     conditions are met, no cost to the government, for example,
     on those conditions.  But as far as the long-term prognosis
     for DOE, I couldn't --
               MR. LEVENSON:  No, I wasn't asking that.  I was
     just asking whether the transfers is legally to the federal
     government, or specifically to DOE?
               MR. CAMPER:  It is to DOE under 151(b) of the '82
     Act.
               DR. WYMER:  Thanks, Larry.  That helps.
               Sorry.  Go ahead.
               MR. NELSON:  That's all right.
               Any other questions on that topic?
               [No response.]
               MR. NELSON:  I would like to move on then to some
     of the efforts that we have taken and plan to undertake to
     enhance the effectiveness and efficiency of the program.  I
     will provide a more detailed description of the streamlining
     and rebaselining initiatives in a few minutes.  But over the
     past few years, we have also placed a significant emphasis
     on guidance development to assist licensees in complying
     with the License Termination Rule and to aid the staff in
     providing consistent and efficiency review of licensee
     submittals.  Examples of this activity include the
     Decommissioning Standard Review Plan and the License
     Termination Plan Standard Review Plan.
               In the next fiscal year we plan to conduct a
     complete review of all decommissioning guidance and
     consolidate that guidance as required.
               To support these efforts, we have conducted a
     variety of public workshops over the past year-and-a-half as
     input for our guidance development effort and we plan to
     continue to do so.  An additional workshop, for example, is
     planned in November.
               In support of our efficiency and effectiveness
     initiative, we have implemented the streamlining objectives
     summarized on this slide.  Further, the staff is
     incorporating strategies to achieve performance goals that
     the agency has set in its strategic planning process.
     Examples include focusing on resolving key issues such as
     institutional controls for restricted release, which we just
     discussed in some depth.  Other focus areas include partial
     site release and rubblization, for example.
               We conduct stakeholder workshops to seek licensee
     and industry and public input, and we have ongoing efforts
     to enhance our Standard Review Plans to make ourselves more
     efficient.
               In September of '99, the division began to
     rebaseline the materials decommissioning program to
     determine the current status of each SDMP and complex
     decommissioning site, and to develop a comprehensive,
     integrated plan for successfully bringing these sites to
     closure.
               To facilitate planning, site status summaries were
     prepared as of the end of -- or as of December 31st, '99,
     and these were developed for each SDMP and complex site.
     These summaries for each site are included as an attachment
     to the Commission paper.
               These summaries indicate the status of each site
     and identify the technical and regulatory issues that could
     impact the removal of the site from the SDMP or the
     completion of decommissioning.
               For those licensees that have submitted a DP, we
     have developed a schedule based on that submittal.  For
     those licensees that have not submitted a DP, we have also
     developed schedules, but those schedules are based on
     information that is currently available to the staff and the
     decommissioning approach we anticipate the licensee to take.
               The comprehensive plan includes identification of
     all the major milestones associated with management of the
     site.  We have done that using project management software
     and have produced an integrated Gantt chart for each site.
     An example of one such chart is included in the SECY, but
     the charts exist for all sites, and we are managing to those
     milestone charts.
               In addition, for the License Termination Plan
     reviews that we are receiving, we are doing similar --
     developing similar schedules and managing to those
     schedules.
               The program is not without its challenges.  We
     have talked about some of them today.
               The License Termination Rule is a dose-based rule,
     and one of the challenges is to assess the dose from a
     released site.  To assist in that effort, we've developed a
     technical basis document for dose modeling that is included
     in our Standard Review Plan for Decommissioning.
               And the intent of that is, again, to provide
     guidance to both the licensees and the staff on an
     acceptable approach to do dose assessments.  That should go
     a long way to assist us in that effort.
               But even with that guidance, it's new to a lot of
     our staff.  We have for many years, released sites under the
     SDMP Action Plan, which didn't really require dose
     assessment, so getting our staff up to speed, trained in
     doing this, is going to be a challenge for us.
               The release of solid materials is equally a
     challenge.  We don't have a consistent regulatory basis for
     releasing material.
               We use guidance that we have developed, and this
     issue is clearly before the Commission, and until we have a
     rule of some type, we will still have to use what guidance
     we have, and often do a case-by-case review of licensee
     requests to release solid material.
               We've talked already about restricted release
     cases and I'm not going to dwell any more on that unless you
     have additional questions.
               Under the topic of innovative performance-oriented
     approaches, it's important to note that the rule -- that our
     License Termination Rule is a performance-based rule, and
     that licensees are going to find and propose innovative ways
     to meet that rule.
               Things such as rubblization is an example.  We can
     anticipate similar approaches or other approaches in the
     future that we're going to have to address.  These are not
     cookbook solutions, and will require some effort by the
     staff to review.
               Partial site releases:  Again, under the SDMP
     action plan criteria where you have set numbers you have to
     meet, it was an easy determination that you had cleaned up a
     site and therefore could release it.
               We now have a dose-based rule, and you have to
     look at a partial site release, and one of the questions
     that arises is, to what extent does the site that you
     propose to release contribute to the dose on the site that
     still is under license?
               So that --
               DR. WYMER:  Or vice versa?
               MR. NELSON:  Or vice versa, exactly.  And, of
     course, finality, with the ongoing disagreement between the
     Agency and the EPA over acceptable decommissioning criteria,
     finality remains a question that the licensees have to
     consider in their decommissioning approach.
               Just one final word before I conclude the formal
     presentation:  You asked about integration of the program.
     You've heard all the pieces described here.
               To assist in that integration, we have formed the
     Decommissioning Management Board which meets every two
     weeks.  It consists of appointed members from the Offices,
     NMSS, NRR, Research.  OCG attends, and these meetings have
     set agendas, issues that are brought up before the Board
     receive a coordinated review.
               An example is Standard Review Plan for
     Decommissioning.  Each one of the modules of the Standard
     Review Plan was reviewed by the Board before it was
     published.
               This Board, as I mentioned, I think, very
     effective in providing the integration needed and the
     guidance needed to the staff on addressing these issues.
     It's an effective forum.  Their meetings are well planned,
     the agenda set, action items assigned, and followed up on at
     subsequent meetings.
               This has gone a long way to bring the integration
     necessary to bring all these complex pieces into play.
               That concludes our presentation.  We welcome your
     questions or comments.
               DR. WYMER:  Thank you very much.  I like your list
     of challenges.  Those are about the ones I guess we would
     have come up with here since we've had discussions on this
     in some degree of detail or other.
               Let me ask you about a couple of specific things:
     One is the status of rubblization.  I know that Maine Yankee
     has come in, and I know that they have sort of backed off a
     little bit from it now because of the State of Maine's
     requirements that are more stringent than the NRC's, I
     think.
               And where does that stand now?  Do you have a
     sense?
               MR. NELSON:  I can tell you where the Maine Yankee
     review stands.  Maine Yankee has submitted and LTP and they
     have asked us to review the LTP that has been submitted to
     us, which includes rubblization.
               They have told us that they plan to submit a
     revision to their License Termination Plan.  We don't know
     the extent of that revision; we don't know what they're
     going to change.  And until -- whether or not they remove
     the rubblization concept is a question I can't answer.
               That plan is tentatively scheduled -- the revised
     plan, they have committed to try to submit a revised plan by
     the 31st of October, and until we get the revision, we won't
     know what's in it.
               But in the meantime, we are proceeding with the
     review of the plan they have submitted, and that review is
     ongoing, and at such time that we get a revision, we'll
     review the revision and make adjustments as necessary.
               So rubblization is still in the -- still part of
     that concept and we are reviewing it.
               DR. WYMER:  The second question has to do with
     clearance.  We know that Secretary Richardson has put a hold
     on everything with respect to releasing materials,
     especially like the nickel from the K-25 plant, which causes
     a lot of problems to everybody.
               But that then -- the DOE then, as I understand it,
     turned it over to the NRC and said, okay, give us a ruling.
     And NRC turned to the National Academy of Sciences and said,
     write us a report.
               What's the status of the report?
               MR. NELSON:  I'm not personally involved with the
     clearance rule or the release of solid material rulemaking
     efforts.  I'm looking around to see if there is anyone from
     the staff who might be able to address that.
               MS. TROTTIER:  Where's my microphone?  I'm
     Michelle Trottier, Research.  We are expecting a proposal
     from the Academy by week's end.  I think that's tomorrow.
               Their Governing Board has met.  They have reviewed
     what their staff put together based on what we submitted to
     them in our Statement of Work, and so we expect to be
     beginning that shortly.  But it has not commenced yet.
               DR. WYMER:  It will take a couple of years before
     you get the report?
               MS. TROTTIER:  It probably will.  I do think that
     what the Governing Board was concerned with was the
     timeframe.  The Commission wanted a very short turnaround on
     this, and I suspect that what we will see from the Academy
     will ask for a little more time.
               But nonetheless, it is not a simple issue, as you
     might guess.
               DR. WYMER:  Okay, thanks.  Let me ask, John, have
     you got any specific questions?
               CHAIRMAN GARRICK:  Well, I am trying to figure out
     what the real technical issues are here, so that we can be
     helpful.  Larry had mentioned that one of the primary
     problems was the identification of viable third parties to
     oversee sights, et cetera, et cetera.
               I agree with Ray that this last viewgraph is
     pretty much on target with what we have identified as some
     of the issues.
               But when you talk to a lot of the facilities
     people and the reactor people, in particular, and ask them
     what are the primary problems associated with
     decommissioning, more often than not what comes out of that
     is an identification of the handling of low-level waste
     materials, solid materials, as being the one that has the
     greatest amount of uncertainty surrounding it, or at least
     if you follow the regulations as they now are, the options
     seem to be quite limited.
               So, the release of solid materials or the handling
     of low-level waste materials product from the
     decommissioning of facilities apparently continues to be a
     major factor, particularly if you think about economics.
               The safety issues do not seem to be particularly
     significant.  So, I would ask you the same question, and
     that is, what do you really see as the principal problems
     with our primary interest being on the technical side,
     associated with an effective regulatory program on
     decommissioning?
               Are they pretty much what you've touched on here?
               MR. NELSON:  Well, I think they are.  The
     challenges slide, I think, are the principal challenges that
     face us.
               In some respects, we're on a learning curve.  We
     have, in addition to the dose-based rule, we have MARSSIM,
     which is a new approach to site surveys, so we are in some
     cases learning as we go there.
               MARSSIM -- we are -- for example, the staff is
     used to in a decommissioning plan, getting a very detailed
     final status survey plan as part of the decommissioning
     plan.  Under our old guidance, NUREG CR 5849, that was
     possible.
               Under MARSSIM, however, you need to know -- you
     need to have some information about the post-remediation
     condition of your site to determine the number of samples
     you need to take.
               So in some respects, you can't submit the same
     detail, final status survey plan to the staff up front
     before you decommission.
               So, that's something we're trying to gain some
     comfort with, and therefore, what do we expect in the form
     of a final status survey plan under MARSSIM?
               That's an area where, for example, we're learning.
     I wouldn't call it a significant challenge, but it is a
     learning curve.
               And as the industry is learning MARSSIM as they
     implement it, I think that implementing MARSSIM is a new
     approach, and, therefore, I think, a minor challenge.
               But it's an implementation challenge, and I
     wouldn't call it a real technical challenge.
               Outside of those that I had identified, I think
     that those are the real challenges that face us.
               The dose-modeling, that would encompass things
     like rubblization, because, you know, you have a different
     source term with rubblization.  And so you need to look at
     --develop a site-specific dose model of leaving that
     material behind.
               MR. CAMPER:  I would add to that by answering that
     there are four things that I see on the horizon that the
     Committee could be of assistance to the Staff on in the near
     term:
               The first is, you will be provided in the October
     timeframe, the Staff's proposed Decommissioning Criteria for
     West Valley.  That will come to you as the Staff is
     providing the proposed final decommissioning criteria to the
     Commission about the November timeframe.
               We've briefed the Commission's staff recently on
     West Valley.  We owe them another status report in
     September, and one of the things that they specifically made
     it clear to the staff on, was that they were strongly
     interested in seeing ACNW input on the decommissioning
     criteria.
               Now, we have a very tight schedule, and we're
     going to react as best we can, given that schedule, between
     October and November when the Commission expects to see the
     paper.
               And I know that that timing is not necessarily
     consistent with your normal process, but we're going to be
     having some interactions with you on that.  So that's a big
     one on the horizon, as I see it, near-term.
               Secondly, as Bob pointed out, we have put a lot of
     work into going back and looking at RESRAD and D&D.  We work
     closely with Research in doing that.
               We try to make those modeling approaches more
     probabilistic in nature, to remove some of the conservatism
     of the default parameters in those.  I think that at some
     point, having the Committee take a look at those modeling
     codes in current terms, post those adjustments, and seeing
     if you think that we have done all that we can do to make
     them as probabilistic as possible, and to make sure that we
     have the appropriate level of conservatism in the default
     values, I think that's something that would be of value to
     us down the pike.
               Thirdly, on this issue of clearance, Stu and I
     just came back from this meeting with NEI, and I will tell
     you that this clearance issue is very, very big on the minds
     of the reactor community that is undergoing decommissioning
     or wanting to clear materials from their sites, whether they
     are operational or in decommissioning.
               They see it as a very large question mark.  They
     think it has the potential to impact the costs that they
     project for decommissioning.
               A number of them are coming into us with
     innovative approaches.  For example, Big Rock Point met with
     us recently and discussed a pending licensing action in
     which they want to be able to clear some materials that
     would end up in landfills that would be a disposal.  They
     want to do that not under a 20.202, because they have
     concerns about the material that would ultimately go to
     landfills still being characterized as being radioactive in
     nature, so they're pursuing a licensing action, and that's
     sort of an interesting twist.
               I think that at some point, as we look at some of
     those types of licensing actions in the months to come, I
     think coming to the Committee and sharing with you the
     technical basis for which we are making those decisions, and
     getting your input on it, and opportunity to put your
     footprint on whether or not the technical logic seems to
     make sense or not, would be of value to the Staff.
               The finally, of course, there is the one I
     mentioned earlier, that being institution controls.  We'll
     talk with you about that in October.  I think it will be
     interesting for you to hear a bit more about the role, the
     possible role of DOE and what we might do further in terms
     of other pathways or approaches for dealing with the
     institutional controls issue.
               So those are the four that I see in the near-term
     where I think specific feedback from you would be of value.
               CHAIRMAN GARRICK:  Thank you, that's good.
               DR. WYMER:  George?
               DR. HORNBERGER:  I was particularly interested in
     your reviews of the License Termination files, these 37,000
     files.  And so if I do the simple math here, it looks as if
     you have got about 1 in 1,000 where you released a site that
     maybe you shouldn't have.  So, you know, a Type 1 error rate
     of about 10 to the minus 3, and NRC probably wants to do
     better than that.
               And I was just curious, have you gone back, or do
     you plan to look at those 38 to see what lessons are to be
     learned in how you do your job?  In other words, what went
     wrong?
               MR. BUCKLEY:  The 38 cases that were identified
     that were found contaminated, I think what happened was that
     there was a lack of documentation in the file.  In many
     cases in those, there was no final survey that was done.
     Those tended to be the very old licenses, the ones that were
     terminated early on in the late '50s and '60s.
               Oak Ridge found that the licenses that were
     terminated from '85 to present turned out to be very good.
     There were very few mistakes made.  I believe that number is
     one or two, as opposed to the remaining licenses that were
     -- the sites that were contaminated.  So, most of the
     mistakes that were made were made early on, and from, I
     would say roughly '80 forward, the results were quite good.
               DR. HORNBERGER:  The other thing that occurred to
     me is that it might be interesting to apply D and D to at
     least the 38, just to see whether or not DandD would have
     kept them in.  I am just -- it is just an interesting
     perhaps exercise to apply DandD to a test case, or a series
     of test cases where you have some interesting historical
     information.  And they are probably not terribly complicated
     sites, although I don't know too much about these SDMP
     sites.  Just a thought.
               DR. WYMER:  Milt?
               MR. LEVENSON:  I have got a couple of questions.
     In the area of both rubblization and clearance or free
     release of material, do you perceive the technical part,
     that is, monitoring what is there as being a significant
     part of the problem you have?
               MR. NELSON:  No, I don't see that as a problem.  I
     think that the -- we haven't seen a problem to date in
     monitoring, and I don't know why it would arise in the
     future.
               MR. LEVENSON:  The basis of my question is that
     the committee visited some decommissioning sites in Europe
     in May and they did a couple of interesting things.  They
     don't call it rubblization, but they have solved the
     hypothetical question about how do we know there isn't some
     activity inside that concrete that might come out, by just
     running it all through regular aggregate crushers to a
     relative small size and run it under counting equipment on
     conveyers and make real sure there is nothing inside before
     they dispose of it.
               And they have a similar sort of philosophy with
     metal things.  It has to have only external surfaces that
     are monitorable so that things like structural steel if
     there is riveted sections, they cut that out.  Say maybe we
     don't know what is inside between the two plates.  So they
     feel much more comfortable and have more public confidence
     on releasing things when they can say every bit of surface
     is external and has been surveyed.
               I wondered if, when you get into the areas of
     release, whether anybody had thought ahead about how do you
     assure yourself that the monitoring really does tell you it
     is okay?
               MR. NELSON:  Well, I will answer that in two
     parts.  Right now our release criteria are surface
     contamination oriented.  We don't have volumetric release
     criteria.  That volumetric release criteria question is the
     crux of the release of solid material technical issue.  What
     are the criteria?  And then what -- how do you measure that?
     And that is clearly a technical issue that hasn't been
     answered.
               To date, if a licensee wants to release
     volumetrically contaminated material, we would have to
     analyze that on a case-by-case basis and do some type of
     dose assessment on it.  But I don't know that we have an
     issue at this point with measuring it.  It is, once you know
     what it is, what criteria to apply, and how do you make that
     assessment?
               DR. WYMER:  It is my understanding that the
     Division of Research has issued a contract or subcontract to
     study the problem of measuring volumetric contamination.  I
     wonder what the status of that study is.  Is there anybody
     in the audience who can speak to that?  Here we go.
               MS. TROTTIER:  Actually, we have two contracts,
     and these are follow-ons to work that we did for the License
     Termination Rule.  One is with EML, the Environmental
     Measurements Laboratory, and the other one is with ORISE.
     And I will make an offer to you, because Commissioner Diaz
     has expressed a lot of concern about this issue,
     particularly the concept of detectability, and that, you
     know, he makes a true statement -- with enough effort, you
     can detect anything.
               So what we are doing next month is briefing him on
     these contracts.  And it might be worthwhile for the
     committee to hear the work that we are doing, maybe at an
     upcoming meeting, so that, you know, we can give you the
     status.  I mean these are about a year into their work and,
     in fact, they are going to present their current status at
     Research's Water Reactor Safety Meeting in October.
               So, any time that there is time on your calendar,
     and if it is of interest, we would be glad to bring our
     contractors in and just give you a brief synopsis of the
     progress that they are making to date on this issue.  It is
     basically volumetric measurements.  You know, what is
     capable, what is realistic?  Costs, all the factors
     associated with doing this kind of measurement.
               DR. WYMER:  Well, this is important enough, there
     is enough money at stake that it probably would be worth our
     while, if we can fit into our schedule, to hear something
     about it.
               MS. TROTTIER:  Right.  I understand that.
               DR. WYMER:  So we will talk about that.
               MS. TROTTIER:  Okay.
               DR. WYMER:  Thanks.
               MS. TROTTIER:  Oh, Cheryl Trottier from Research.
               DR. WYMER:  Let me ask around the table if the
     staff has any questions?  I guess we have time to invite
     audience --
               CHAIRMAN GARRICK:  Well, I want to comment on one
     thing or ask a question, because there has always got to be
     one off-the-wall question, and I will provide that.  I am
     kind of curious as to whether or not this whole business of
     contaminated sites is being looked at by some systems -- or
     from a systems oriented perspective.  And I am not thinking
     just of radioactive contamination.  I am much more worried
     about the arsenic contamination in Silicon Valley, for
     example, and what goes in those kind of industries than I
     am, at least from a public health and safety standpoint,
     than I am from these, a large number of these sites.
               It just seems to me that the opportunity here is
     fantastic for some creative systems engineering of looking
     at these sites in the context of what makes sense to make
     use of so-called restricted sites.  And I guess my question
     is, and I know that the NRC, I have been told many times the
     NRC regulates, it doesn't solve problems, so it is probably
     outside your scope, and that is why I put it in the category
     of off the wall, but I wonder if there is an attempt on some
     sort of an interagency basis or interagency/industry basis
     to take a look at the kind of activities that are necessary
     to have as a fundamental part of a society's infrastructure?
     Many of them involve hazardous materials and what-have-you.
               I can't believe that there wouldn't be a way to
     optimize this process in such a way that every acre of these
     sites could be put to very good use even if they are left in
     a restricted state, if we had at our disposal the ability to
     manage all activities and all hazardous operations and
     manufacturing facilities, and process plants and
     what-have-you.
               Is there anything like this being done at a high
     level that gives people assurance that these are not lands
     that are lost to mankind forever?  Because the truth is
     there are many operations that are far more hazardous than
     what we are talking about at most of these sites.  And it
     just seems that some creative geographer or somebody could
     find very extremely effective ways to have these sites serve
     mankind and take and in the end save the need for siting
     some of these facilities in what are currently pristine
     environments.  Are you participating in anything like this?
               MR. NELSON:  Yes.
               CHAIRMAN GARRICK:  Are you aware of any such
     investigations or studies?
               I just think we are very uncreative as a nation
     when it comes to this kind of an issue, and that radiophobia
     has created such a barrier to the use of locations that have
     had radiation related activities that we have become stupid
     about effective utilization of our resources.  And I am
     curious if the NRC has ever contacted in any -- to be
     engaged in any of these kinds of studies.
               MR. NELSON:  I am not aware of any interagency
     studies that may be going on.  My suspicion is if anyone
     were doing it, it would be EPA.  So I am not aware of any.
               DR. WYMER:  Let me ask if there is -- oh, a staff
     member.
               MR. LEVENSON:  Use the microphone and identify
     yourself, please.
               MR. FREDRICHS:  I am Tom Fredrichs, I am in NMSS
     in the Division of Waste Management.  And as far as reusing
     some of these sites, I don't know if the NRC is involved in
     any interagency sort of things, but we are cooperating with
     some of the licensees to reuse these sites, some of the
     reactor sites where they are going to repower with a fossil
     fuel facility.  And that is part of the partial site release
     challenge that we have, to be able to get these sites put
     back into productive use more quickly than if we would have
     to wait for the entire site to be -- the license termination
     process to go through.
               So, I mean it is not a nationwide effort, but we
     are at least making some small steps in that direction.
               CHAIRMAN GARRICK:  Thank you.
               MR. LEVENSON:  And sort of in that connection, the
     question I was going to ask, and it is still relevant, is,
     what are the restrictions on restricted release?  What does
     that really mean?  What options are there for use of the
     land?
               MR. NELSON:  I think that to a large degree
     depends on a couple of things.  One, the nature, amount and
     physical arrangement of the material that may be left
     behind, and the licensee's own intent for that land.  If
     there is a large amount of material buried onsite, but the
     site boundary includes some buffer zone, I would imagine
     that part of that site could be put to some other productive
     use, maybe a park or whatever.
               So, I don't think there are any restrictions on
     restricted use.  Restricted use doesn't mean build a three
     foot concrete wall around the facility and have an armed
     guard standing by.  It means that you have placed -- that
     there are restrictions on how you can use the site and those
     restrictions would have to be based on some of the factors I
     mentioned.
               But, ultimately, it comes down to what the
     licensee wants to use that property for.  Normally, it would
     remain -- the title would remain with the licensee, or the
     ex-licensee once the license is terminated, and so they
     would have to make the proposal on how that land would be
     used or not used after the license was terminated, and the
     legal mechanisms they would put in place to ensure that that
     was the only uses that the land would be put to use for.
               MR. LEVENSON:  Yes, I think that the sites
     probably, John, generally fall into two categories, those
     that were power plants.  That means they have access to
     cooling water and media that makes them good for power
     plants, and they are probably people who will seriously
     consider reusing the site.  I think the more difficult one
     are the nonpower plant sites.
               CHAIRMAN GARRICK:  Yes.
               MR. LEVENSON:  Which might not have motivation the
     same way as for power plants.  Power plant sites are at such
     a premium that somebody is going to repower them.
               MR. NELSON:  Well, let me say we know of no power
     plant that is planning on a restricted release at this
     point.  The only restricted release proposals that have been
     submitted or discussed with us have been from materials
     licensees.  So, I don't think -- at this point the reuse of
     power plants is not an issue.
               MR. CAMPER:  Yes.  I would only add to that, I was
     going to comment sooner or later, if you look at the power
     plant sites, I mean they really are pristine in the final
     analysis.  I mean typically there is not a groundwater
     problem.  And even though License Termination Plans have
     been submitted designed to meet our standard of 25 millirem
     and ALARA unrestricted, when it is all said and done, as a
     result of the scabbling and decontamination takes place, if
     you look at your final site survey results, you are probably
     going to find those sites in the order of a few millirem.
     So, they really are quite clean.
               It is the more complicated process sites or
     material sites that pose problems, not the least of which,
     of course, is groundwater in some cases.
               DR. WYMER:  Yes.  DOE has a whole bunch of
     materials handling sites, as you know, and they also face
     the problem of restricted release and greenfield release,
     and some of the solutions, just for Milt's -- answering
     Milt's question, some of the things that they are
     recommending range from wildlife management parks to parks
     for children to manufacturing sites.  There is almost an
     infinite spectrum of things, pretty much as you have
     indicated.  It depends on the imagination and the wishes of
     the people that are on site.
               MR. NELSON:  Right.  And we have seen some variety
     in those material sites.  For example, Jefferson Proving
     Ground site in Madison, Indiana, they fully plan to use that
     as a game reserve, and to have a portion of the site
     available under the Fish and Wildlife Service, another
     portion to the Air National Guard, but the DU contaminated
     portion would be restricted access and would be retained as
     a game reserve under the Fish and Wildlife Service, that is
     their proposal.
               In a number of other cases, probably the greatest
     majority of those 12 sites are looking at some from of
     onsite disposal of large volumes of waste and some type of
     an impoundment cell.
               So, there is a variety of approaches and, again,
     the nature of the restrictions will have to depend on those,
     you know, the three or four criteria identified earlier.
               DR. WYMER:  Okay.  I would like to give the
     members of the audience a chance to participate, and I think
     maybe in order to start, I would like to see if there is
     anybody from the Nuclear Energy Institute who would like to
     comment on this, who is in the audience, since I am sure
     they have been involved in all of these areas one way or
     another.
               I guess not.  Is there anybody else in the
     audience that wants to make any comments?  It is a passive
     group.
               [Laughter.]
               MR. NELSON:  We just gave an excellent
     presentation and answered all their questions.
               DR. WYMER:  We appreciate it.  Thank you very
     much.  I think we picked up 20 minutes to work on letter
     writing.
               MR. NELSON:  Thank you for your time.
               CHAIRMAN GARRICK:  Thank you.
               [Pause.]
               CHAIRMAN GARRICK:  I think what we would like to
     do, and we don't need to be on the record for this, is
     review the one remaining letter that we have.
               [Recess.]
               CHAIRMAN GARRICK:  We'll come back to order.
               The next item on our agenda is hydrology research.
     The committee member that is going to lead this discussion
     will be George Hornberger.
               DR. HORNBERGER:  Great.  I am not prepared to
     lead.
               [Laughter.]
               DR. HORNBERGER:  Okay, so actually you will recall
     that we have had a number of discussions related to research
     being conducted by the Office of Research and this is a
     continuation on it.  In one sense we are lucky in that we
     overlapped with a workshop that Tom Nicholson put together,
     but it was held simultaneously.  We are unlucky that we
     couldn't attend it, but we are lucky that we have some of
     the people who are in town because of the workshop, not
     because of the ACNW meeting, willing to stay over and come
     to the ACNW meeting, although Lynn might say that they came
     to the ACNW meeting and Tom was lucky that they were here
     for his workshop.
               [Laughter.]
               CHAIRMAN GARRICK:  Are you suggesting they
     wouldn't come to the ACNW meeting?
               DR. HORNBERGER:  We have presentations from
     several people, and Phil, you are going to go first?
               MR. OTT:  First, with a little introduction.
               DR. HORNBERGER:  Oh, okay.  Bill is going to do
     the introduction.  Sorry, go ahead.
               MR. OTT:  Bill Ott from the Office of Research.
               My first remark is there is no luck involved at
     all.  When Tom was planning the symposium I essentially told
     him let's try and do it in conjunction with an ACNW meeting
     so that we can have some overlap and meet with you.
               CHAIRMAN GARRICK:  You have made us feel better
     already.
               [Laughter.]
               MR. OTT:  We are actually attempting to come to
     you more often and help you at the end of the year be
     prepared to write something on the Research program.  A
     couple of months ago we had Linda come in.  We've got Shlomo
     and Glendon and Phil and Peter here today.  We hope to do
     this at least once more, maybe twice more this year and
     bring in some other contractors to meet with you.
               Today we have got the PIs that are essentially
     running the hydrogeology research in the program.  We have
     projects with both Pacific Northwest National Laboratory and
     the University of Arizona.  We are basically looking at
     uncertainties related to modeling of hydrogeologic systems.
               Specifically, we are looking at parameter
     uncertainty, primarily at PNNL, and in conceptual model
     uncertainty.  They represent two aspects of the program.
               The work at PNNL is directly responsive to a user
     need.  It is the work that was requested of us by NMSS.  It
     is currently being used by NMSS and is being used by the
     Office of Research in terms of helping to improve models
     like RESRAD and DandD.
               The work at the University of Arizona is not user
     need originated.  It is essentially what we call
     anticipatory research and was initiated because we conceived
     of a potential problem in the future dealing with
     assertions, allegations, whatever by opponents of licensing
     actions, that we just have the wrong conceptual model, how
     do we deal with the uncertainties associated with such
     contentions.
               That's really about all I wanted to say.  I wanted
     to point out the difference and indicate that we are trying
     to do this to help you guys with your Research report.
               Oh -- one other thing I wanted to mention.  There
     are other activities going on in the conceptual model area.
     The National Academy should come out with a report this fall
     on the symposium that was held about a year ago out in
     California.  This fall at the AGU meeting in San Francisco
     there is going to be a special session, which I believe was
     organized by Shlomo on uncertainties in modeling and he may
     mention more about that during his talk.
               Without further ado, I will introduce Phil Meyer,
     who is one of the PIs on the PNNL project.  Phil?
               MR. MEYER:  Thank you, Bill.
               My name is Philip Meyer.  I would like to just
     make sure that everybody realizes that my colleague, Glendon
     Gee, is here as well and will be available for question
     answering.  Because the research symposium was held, we also
     have a lot of additional material.  Should that come up in
     questions we'll be pleased to provide that.
               I am going to be talking in pretty broad, overview
     terms here in a short presentation and we will have more
     later, if needed.
               Like Bill said, the motivation for this research
     is that it was undertaken in response to a user need from
     NMSS and in the broadest of terms our work is intended to
     support the development of guidance for the termination,
     license termination process.
               The general background for the problem is that the
     dose assessments that are used in the determination of the
     safety of the site rely on simplified models in a lot of
     cases.  The models are formulated in addition with fairly
     limited site-specific data and these are characteristics of
     the analyses that the Staff with NMSS are very concerned
     about and that were uppermost in our minds in addressing the
     issues.
               Given that these two conditions hold, the result
     is that when you make predictions of dose for comparison
     with the regulatory standard that those predictions are
     uncertain, and our research is intended to provide the NRC
     Staff with improved tools to address that uncertainty and to
     quantify it.
               I'd just point out that this consideration of
     uncertainty is consistent with the risk-informed approach
     adopted by the NRC.
               The research objectives of our project were to
     document a method for assessing uncertainty and in doing
     this we have extended previous methods that we have
     developed for the NRC in previous project work on the Low
     Level Waste Program and also for SDMP sites.
               The emphasis here is to try to provide practical
     tools, so will work fairly closely with NMSS Staff to
     provide them with tools that they can use and tools that
     they need as they proceed through their guidance
     development.
               In addition, they specifically asked us, the NRC
     specifically asked us to look at three codes -- DandD,
     RESRAD, and MEPAS.  These are three dose assessment codes
     that you may or may not be familiar with that are used by
     their licensees and by the NRC Staff.
               That is not to say that any of the methods that we
     have worked with are not applicable to other codes but we
     are specifically directed to look at these codes.
               The relationship of this project to other
     projects -- there's work that was mentioned this morning
     regarding modifications and improvements to the DandD code
     and the RESRAD code.  Those projects are going on at Sandia
     and Argonne National Laboratories and information that we
     developed on generic probability and distributions for soil
     hydraulic parameters has been incorporated in those two
     codes recently in terms of the default parameter
     distributions that have been implemented in the latest
     versions of those two codes.
               In addition, we are collaborating with Drs. Neuman
     and Wierenga on some of the work that they are currently
     doing that they done in the past.  The monitoring project at
     the Maricopa site in Arizona gathered a fair amount of data
     which we are trying to use in our test case applications
     that we're currently working on and the current project on
     conceptual model uncertainty has a lot in common with our
     work, and we have been trying to interact with Dr. Neuman on
     that.
               Just some brief background so that I can be as
     complete as possible.  You might ask yourself why hydrologic
     uncertainty is even important, and the reasons are fairly
     straightforward.  One is that when you calculate dose you
     find that the dose is very dependent upon the hydrology, so
     the calculation of dose is sensitive to your conceptual and
     parameterization of the hydrologic components of the
     problem.
               In particular, with the relatively simplified
     models that we are looking at, there's two basic terms, the
     source term -- which is a function of the amount of water
     that enters the waste -- we refer to that as net
     infiltration, being the water that passes below the zone of
     the roots, where it would be possibly taken back up through
     evapotranspiration.
               The other thing is that the travel times with
     these models, and not only these models but in reality, the
     travel times of the contaminants are dependent upon the
     net -- the net infiltration as well.
               So the dose is very sensitive to the hydrologic
     terms and in addition the hydrologic terms are difficult to
     estimate accurately, particularly given limited data at many
     of these sites, and in addition to limited data the
     properties, hydrologic properties, often vary spatially and
     temporally, so it makes them -- with limited data you have
     possibly greater uncertainty about what the parameter values
     should be.
               I have already mentioned that the codes are using
     fairly simplified models for relatively complex processes.
               Just a couple examples of the kind of variability
     that we're talking about.  This is some data from the Las
     Cruces trench site on saturated hydraulic conductivity.
               It is well-known that saturated hydraulic
     conductivity varies spatially significantly and you can see
     here that the variability on this plot, which was a number
     of meters in dimension, is several orders of magnitude.
               On the right-hand side here is the data from the
     USDA facility in Coshocton, Ohio that shows the drainage
     from four lysimeters located at the same location
     essentially and you can see that there's a lot of
     variability.  This is annual drainage from lysimeters.  You
     can see there's a lot of variability between lysimeters and
     also a lot of variability over time.
               In terms of having a short record, if you have a
     record of data at a site it is going to tend to be short.
     You have to try to represent the long-term average process
     at a site.  Using a short record you are liable to be
     inaccurate in that estimate.
               Just an additional example to illustrate spacial
     variability -- this is a site on the Hanford site where
     Glendon is currently conducting some experiments, and there
     has been some experimental work in the past.  This is just
     an interpolation from bore holes, bore hole geophysical logs
     of the bulk density at this site, and you can see a lot of
     variability and you can also see that although at this site
     you might interpret the bulk density as resulting in a
     layered process, in fact it is not perfectly layered and the
     experiments conducted at the site have illustrated that a
     lot of the water -- waterflow and transport -- at the site
     from a point source injection does not travel downward but
     in fact moves laterally.
               And that's important because a site like this, if
     you wanted to model it with a code like RESRAD, MEPAS, or
     D&D.  You would have to assume that the flow and transport
     was one-dimensional.
               In fact, here is a schematic from the RESRAD
     documentation, illustrating some of the approximations used
     in that code.  And I list some of the important
     approximations over here.
               As I said, you have to assume that all flow is
     one-dimensional, and, in fact, that it's steady state, and
     also occurs under a unit gradient.
               In addition, the code only allows simple layering,
     and within each layer, the properties have to be homogeneous
     and isotropic.
               And in addition, the transport curves, absorption
     process curves is modeled linearly and in an equilibrium
     state.
               And in addition to those, with any code, you have
     numerical approximations to the equations that you're
     solving, and in this code they may be, because of the
     simplifications, may be more severe than others.
               DR. HORNBERGER:  What does isoptropy mean in one
     dimension?
               MR. MEYER:  Well, of course, there is no isoptropy
     in one dimension.  I'm just pointing out that you don't have
     an option.  If you have a site where you have clearly
     anisotropic properties, you can't represent that.
               So, the basic procedure for assessing uncertainty
     is to first off, assess the code's conceptual model and
     identify the hydrologic parameters, which is not too hard to
     do with the simplified codes, since the parameter set is
     fairly limited.
               One of the issues is, due to the simplification,
     the codes use lumped parameters, as I have already
     discussed, and that has implications for an uncertainty
     assessment because in that case, you're looking for, in
     terms of the parameter uncertainty, you're looking for the
     uncertainty in that lumped parameter.  You're looking for
     the uncertainty in a parameter that has to represent a
     fairly large area of -- a fairly large domain.
               And that both helps you and it hurts you.  I helps
     you in the fact that you can use -- that there are sources
     of information for the parameter probability distributions
     that are probably more valuable because you're looking at a
     lumped parameter, but at the same time, that parameter
     really represents and effective parameter at the site, and
     so you have difficulty in interpreting measurements that
     might be made at the site in terms of what the proper
     effective parameter for the scale of the model should be.
               We have worked to identify sources of information
     for parameter probability distributions, and I mentioned
     those previously, that some of that information has been
     incorporated into RESRAD and D&D, and we have also looked at
     ways to combine generic information that might represent
     value of a parameter obtained from data gathered across the
     country and applied to a particular site.
               We call that generic information, in that it's not
     site-specific; it represents an average value for, say, a
     particular soil type for measurements made across the
     country or within a larger region, and contrasts that to
     site-specific data which is gathered actually on the site.
               In terms of an uncertainty assessment, you need
     some way to combine those two types of information, and we
     have provided a plausible method to do that.
               In terms of the actual analysis itself, Monte
     Carlo simulation is used.  The codes, RESRAD, and MEPAS both
     have built in Monte Carlo shells that are available to use
     with those codes to get probabilistic results, and D&D,
     they're currently working on that option.
               One other thing:  We look at sensitivity analysis
     and statistical correlation measures to try evaluate
     parameters that are critical to the analysis.  Those
     parameters would be useful to probably gather more
     information on them, that sort of issue.
               So, as far as the status of the project, one of
     the tools that we have provided to the NRC staff is a
     simplified water budget model that gives you a transient
     estimate of the components of a water balance in the near
     surface.
               The code uses or tries to use available climatic
     data and a very simple flow representation in order to be
     applicable at the majority of sites with the limited data
     that we anticipate to be available.
               It has been delivered to the NRC staff, and we've
     got a water resources research paper that is currently in
     press.
               This NUREG 6656 provides a variety of information
     that goes into more detail.  I believe the Committee all got
     a copy of that.
               It goes into more detail on what I've talked about
     and the methods that we're currently using.
               I was out here in February of this year, and
     provided an all-day training session to NRC staff, and we
     are currently applying the methods to hypothetical test
     cases.  These are intended to demonstrate the application of
     a lot of things.  They are a lot easier to understand, once
     you see an actual demonstration.
               And this test case, we actually got a lot of the
     information from the NRC staff in terms of the source, the
     characteristics of the source, and the scenarios that we're
     looking at, trying to make it as applicable to their license
     review process as possible, their termination review
     process.
               And there is a variety of analyses that we're
     looking at.  That report on the test case application should
     be out by September.
               Some of the remaining work that we've identified
     are problems that have come up in terms of -- as we go along
     with the analysis -- are proper definition of effective
     parameter values.  I've actually had in my presentations, a
     number of licensees come up and talk to me about their
     particular sites, and this is always one of their issues, is
     that they've got a scattering of data and they're trying to
     apply one of these models.
               And they don't know how to reconcile, and what the
     best way to reconcile the data is, with the respective model
     that they're trying to use.
               There are a couple of issues involved there, only
     one of which is effective parameter values.  The other one
     is something that Slomo is going to talk about in a minute,
     and that's whether or not they should be or how to decide
     whether or not their site fits within the framework of the
     codes that I have been talking about.
               Another issue is the general applicability of
     Bayesian Methods which are used to combine the site-specific
     information with more generic information on soil property
     distributions.
               CHAIRMAN GARRICK:  Are they already in your
     methods, the Bayesian Methods?
               MR. MEYER:  Yes.  Well, there is one particular
     method that is documented in one of the NUREGS that we put
     out that we have been working with.
               CHAIRMAN GARRICK:  But that bullet doesn't mean
     that you're questioning their applicability?  No?
               MR. MEYER:  I'm not questioning the applicability.
               CHAIRMAN GARRICK:  Okay.
               MR. MEYER:  Establishing the general
     applicability, under what conditions should they be used and
     when should you --
               CHAIRMAN GARRICK:  The answer to that is obvious;
     they are generally applicable.
               [Laughter.]
               MR. NELSON:  The question is how to apply them.
               CHAIRMAN GARRICK:  Yes, that's right.
               MR. MEYER:  We had a little bit of discussion
     about this yesterday, the pitfalls or the issues to look out
     for.  One of the other issues that has come up is the
     interpretation of uncertainty assessment results when you're
     using simplified and/or conservative codes.
               That's an issue that I have just started to talk
     about or just started to think about.  I haven't talked
     about it too much yet.  But it seems to me that there's a
     little bit more to the issue of when you're using a
     conservative code when you apply an uncertainty assessment,
     what exactly the uncertainty assessment results mean.
               It's clear to me what they mean when you're trying
     to represent a site as realistically as possible, but if
     you're working on a conservative analysis, it's not always
     clear cut, and I think it deserves a little bit more
     investigation.
               Another issue is the relationship between
     conceptual model uncertainty and parameter uncertainty for
     these simplified codes, and hopefully that's something that
     Slomo will be able to help me out on.
               There is a variety of statistical correlation
     measures that have been proposed for use, and I haven't
     really seen any good guidance on how to use those,
     particularly with respect to this particular problem.
               In any event, this project is scheduled to be
     completed next year in October, and so we will have whatever
     work done at that point.  That's all I've prepared.
               DR. HORNBERGER:  Thanks, Bill.  Questions?  Milt?
               MR. LEVENSON:  Yes, have one.  In assessing the
     uncertainty for simplified models or conservative models, et
     cetera, do you find the uncertainty distribution as
     significantly different?
               That is, I would guess that in a simplified model,
     the uncertainty might be symmetrical, whereas in a
     conservative model, it's highly unlikely that the
     uncertainty is symmetrical, and, therefore, you're getting
     into evaluating uncertainty and you need to connect that
     somewhat with the basic way you got the number in the first
     place; is that right?
               MR. MEYER:  I'm not exactly clear on what you're
     asking, but in general, the uncertainty tends to be -- the
     uncertainty in dose tends to be highly skewed, in my
     observations, so that you -- the distribution of dose has --
     tends to have a few values that are significantly larger and
     you've got a few cases in terms of your Monte Carlo results
     that result in very high doses, significantly higher,
     perhaps, then the median dose.
               MR. LEVENSON:  Yes, that comes out of the Monte
     Carlo, but the question is, if you've used kind of a
     simplified model, do you see a different skewing than if
     you're basically starting with something that's a very
     conservative model.
               In a very conservative model, wouldn't you find
     fewer high doses in the uncertainty range?
               We tend to use uncertainty as though it was
     symmetrical and independent of the number to which you apply
     it, and I don't think that's really correct.
               DR. HORNBERGER:  Perhaps I'll try to help clarify
     that.
               MR. MEYER:  Okay.
               DR. HORNBERGER:  The idea might be, if you have
     your log K distribution and you use the realistic log K
     distribution, even in a simplified model, then more or less
     you're taking into account that you could be wrong on both
     sides.
               If, on the other hand, you took a, quote/unquote,
     conservative case and said, well, I'm going to not use that
     hydraulic conductivity distribution; I'm only going to use
     the very high end, then you would be injecting some
     asymmetry into the uncertainty involved because it would be
     much more unlikely that you would have higher values than
     you chose.
               MR. MEYER:  Yes, and that's my point out worrying
     about the interpretation of the results of an uncertainty
     analysis when you're using conservative models.
               That's my concern, is that how do you -- it's not
     clear to me how you actually interpret the result, if you
     have a conservative model?  The parameters really mean
     something different than instead of really meaning the way
     that I tend to interpret a parameter uncertainty.  It's that
     it represents a degree of belief in the actual value of that
     parameter.
               So if you're trying to represent some extreme
     values or more unlikely values, at the same time, trying to
     interpret the results as this is my degree of belief of the
     way the dose is going to be, those two seem to be
     irreconcilable to me.
               MR. LEVENSON:  I think I would agree, and that's
     why I raised the question.  There is a tendency to make
     certain assumptions that uncertainty really defines that
     things could be worse.
               And if you're doing conservative calculations, it
     may be that all the of uncertainty is on the other side.
               MR. MEYER:  Right.
               MR. LEVENSON:  I'd encourage you to pursue that
     effort.
               DR. HORNBERGER:  Ray, do you have anything?
               DR. WYMER:  No.
               DR. HORNBERGER:  John?
               CHAIRMAN GARRICK:  In your applications, have you
     begun to develop any feel -- or in your trial applications
     -- for how the contributors to uncertainty are distributed?
               For example, how much does the retardation
     phenomenon contribute to the uncertainty?  That's an
     interesting question, and, of course, it's very
     isotope-dependent.
               But that's why I qualify it by saying in your
     applications, what have you learned about the principal
     sources of uncertainty?
               MR. MEYER:  In terms of these simplified codes,
     because they are simplified, the processes, the choice of
     processes as to what's going to actually have the most
     influence on dose, and, therefore, probably contribute to
     most uncertainties, is fairly limited.
               And I mentioned the net infiltration rate having a
     strong influence on the dose.  The distribution coefficient
     also is extremely important for exactly the same reasons.
     One is that it largely controls the release in these models.
               The controls are released from the source, so it
     determines how much stays up in the source, in which case it
     could be good or bad, depending upon whether you have a
     pathway available to the surface that's independent of the
     release from the source in water.
               And the other thing that the distribution
     coefficient controls is the rate of transport via water.
               CHAIRMAN GARRICK:  Right.
               MR. MEYER:  So those are really the two dominating
     factors in terms of the parameters that I have looked at.
               There are also some parameters that are particular
     to some of these models.  In RESRAD you have to decide how
     you're going to -- or what the parameters are that govern
     the well location and the pumping rate from the well, and
     those are sort of related to the exposure.  It's something
     that's difficult to predict how someone in the future, how
     the well -- some hypothetical well in the future is going to
     be configured.
               And those parameters are pretty important also,
     that's my experience.  Some of the other parameters that you
     might expect would be pretty important like the saturated
     hydraulic conductivity, because of the simplifications in
     the models themselves, tends to be less important, but maybe
     not realistic.
               DR. HORNBERGER:  Phil, as you know, if I look at
     D&D, in particular, that code, as you know, it was really
     developed as a screening tool.  Now, whether you believe
     that's good, bad, or indifferent, that really was the intent
     when it was developed at Sandia.
               And I'm just curious.  In your work, obviously now
     your work and others' work on D&D is moving toward using
     this screening code in a more realistic context.
               Do you see any problems in doing that?
               MR. MEYER:  I definitely do.  And part of it's the
     issue of applying a screening code in not so much in --
     well, I mean, there is an issue of applying it in a
     site-specific case, but I think that that's more of a
     standard hydrogeological problem, deciding whether any code
     is applicable at a particular site.
               There is nothing unique about that.  I think the
     D&D code is fairly limited.  It was intended to be limited,
     have limited application that should be conservative.
               So, that's separate from the issue of trying to
     apply D&D using a probabilistic analysis, and that's the
     issue that we've been talking about, and I have some strong
     reservations at this point about that.  That's sort of my
     intuitive feeling, but I haven't done any technical analysis
     to try to clarify that or make it clear to other people.
               DR. HORNBERGER:  Okay.  The other question that I
     have is, clearly, as you said in not just D&D but RESRAD and
     MEPAS as well, you have to wind up, if you're going to use
     these in a site-specific case, defining effective parameter
     value, because they are effective parameters that are lumped
     models throughout.
               So, you've done some work on how you might define
     these effective parameters.  What I'm curious about is that
     in some of these sites, the evaluation is to be over fairly
     protracted periods, long time periods.
               And my guess is that some of these effective
     parameters might depend upon the climate forcing that you
     have.  And if your time periods are long enough and your
     climate changes, have you investigated how effective
     parameters might not be constant parameters?
               MR. MEYER:  Not really.  I mean, one of the issues
     -- well, no, I'd say I haven't really looked into that.
               I think that, just thinking about it, other than
     the forcing term from the top, the amount of water that
     you're applying to the system, I can't think of anything
     that would change.  I mean, the soil properties are going to
     evolve over time, over a thousand years, and I don't know
     how much they would change, but probably relatively little.
               So I haven't really looked into whether or not the
     effective value of the properties of the soil or a KD term
     would depend upon the amount of water you're applying to the
     system.  I don't know if --
               DR. HORNBERGER:  What about rooting depth?
               MR. MEYER:  Rooting depth only comes in -- let's
     see, in RESRAD, rooting depth only comes in in terms of the
     external pathway -- not the external pathway, but the plant
     uptake.
               So, you know, those are issues that are -- it's
     really hard to deal with those.
               DR. HORNBERGER:  Those are unknowns.
               MR. MEYER:  Because you're speculating about
     future conditions.
               MR. OTT:  Thanks, Bill.  We have asked the
     investigators to be reasonably brief so that they will be
     available afterwards for a group discussion so you can quiz
     them again.
               So, both Phil and Glendon will be here after Slomo
     and Pete get done.  But with that, I'd like to introduce Dr.
     Slomo Neuman from the University of Arizona, and Dr. Pete
     Wierenga.
               Slomo will give the general overview for the
     project.  The project also involves some test cases and
     using some datasets, one of which is actively being
     developed at the Apache Leap tuft site, and Dr. Wierenga
     will speak to that during the presentation.
               MR. NEUMAN:  Good morning.  I do appreciate the
     opportunity to talk about our research with the ACNW with
     respect to the last two days' workshop.  I can tell you that
     I have gotten away from it with one full page of handwritten
     suggestions, ideas and comments that I find very useful for
     my future work.
               One very important element of the work we are
     doing is peer review, so any review and any comments that we
     can get from people with various backgrounds as this group
     would be, we would welcome.
               The project that we are working on is formally
     titled Testing of Groundwater Flow and Transport Models.  I
     am collaborating with our outgoing Department Head of Soil,
     Water and Environmental Science, Professor Peter Wierenga.
     The work has both theoretical, computational, and laboratory
     as well as primarily field components, so what we will do is
     I will present a very short overview, hopefully short
     overview, of what we are doing and then ask Peter to tell
     you a little bit about one of the major field studies that
     we are presently conducting in support of this work.
               The object of our study as defined by the NRC are
     models of groundwater flow and transport as they impact
     performance assessment not only -- I understand that today
     you are focusing on decommissioning -- decommissioning
     reviews would be one aspect where these kinds of models
     would have some importance, but we are asked to actually
     consider the broader context of decommissioning, uranium
     recovery, low level and high level waste, so really it is
     the gamut of groundwater flow and transport modeling or
     modeling aspects as they impact virtually anything that the
     NRC is doing where groundwater plays a role.
               The motivation for this study was that a number of
     years ago NRC Staff has identified conceptual models of site
     hydrogeology as a major source of uncertainty in performance
     assessment in any of these contexts.
               I will not have time to define conceptual models,
     but let me just say that I understand behind the term
     "conceptual model" any hydrogeologic interpretation of site
     data.  You go to a site.  You collect whatever geological,
     hydrologic and other information you can, and you face a
     problem in that the subsurface is very complex.  You have
     very limited information even when the database is large
     about that subsurface.
               The system is an open environmental system and it
     is in the nature of open environmental systems of this kind
     that they are given to multiple interpretations.  In the
     language of modeling they are given to the promulgation of
     multiple conceptual models, so the objective as defined for
     us was to recognize this and to develop as well as test to
     the extent possible experimentally -- this is why we are
     running an experimental component -- a method or a strategy
     for, first of all, selecting the correct model or range of
     models, interpretations for the available data, evaluating
     them, and hopefully coming up with some idea of how to
     assess the uncertainty associated with multiple
     interpretations.
               Now I consider the NRC to have been prescient in
     its recognition of both of these problems, conceptualization
     and associated uncertainty, because very shortly after their
     request for proposal has come out, others have formally
     started recognizing that conceptual modeling is indeed of
     major importance for environmental issues
               In particular, I thought that I would mention to a
     recent National Academy book called, "Research Needs in
     Subsurface Science," which was focused on the EM Science
     Program within DOE, and it lists four research emphases.
     The second in the list is conceptual modeling.
               They propose to in fact do basic research in this
     area.  They think that such basic research may eventually
     bring about the development of a tool box or methodologies
     and the NRC in fact is hoping that we will come up with a
     toolbox which focuses on new ideas, very strongly on
     groundwater heterogeneity because the subsurface is so
     complex and variable, scale dependence, and various aspects
     of uncertainty, so we are conducting our field work as well
     as much of our theoretical primarily in areas that encompass
     this range of issues.
               One of the very first things that we were formally
     asked to do as part of this project is to review the
     literature, and that included a good number of
     representative NRC environmental impact and other types of
     reports, but more so the general hydrogeologic literature
     and this is how I would today, from today's perspective,
     characterize some of the main conclusions that I have drawn
     from this literature survey.
               I would start by saying that virtually all, with
     very few exceptions, attempts at groundwater modeling, not
     only by regulatory agencies or the DOE but virtually by
     every practitioner today, as well as very often by academics
     is to adopt a single conceptual model and once you decide
     what interpretation of the existing data you are willing to
     accept, you then run with it.
               When we say that we are dealing with conceptual
     models, I think it would be fair to say that we are dealing
     with the most important and fundamental aspect of modeling,
     the very first step, where you decide what is it that you
     are going to deal with and how are you going to deal with
     it, the conceptual framework?
               I will refer to it later in the conceptual
     mathematical context as the structural framework, the nature
     of your equations, not so much the parameters that enter
     into them, but just how do you write your equations, what
     processes do you include, what area are you going to study,
     what inputs are you going to adopt as being significant to
     your problem.
               What perhaps surprised me to some extent, and
     perhaps less so, was that I found myself focusing at the
     beginning primarily on what a statistician might call Type
     II modeling errors, and that is the adoption of an invalid
     hypothesis, in the language of the statistician -- for us it
     would be a model by not rejecting a model that perhaps is
     not entirely supported by the available data.
               You may have a draft methodology report in your
     hand which contains not all but some of the many examples
     that I have collected, and many of them in the regulatory
     context -- not all of them in the regulatory context --
     where this arises.
               And so I have literally focused on that very, very
     heavily until now.  What is wrong with some of the models
     that people have been using?  Why is it wrong, and what can
     be done?  How can this be remedied?
               I really think that everyone, including the NRC,
     needs help in just recognizing this problem.  So I would say
     that Type II models are the most pernicious type, and for a
     scientist, they are always the worst, because no scientist
     wants to be caught with a criticism saying your theory is
     wrong.
               Less so are Type I modeling errors, which also are
     a result of selecting a single conceptual model.  Those
     types of errors arise due to the fact that even though the
     model you have selected may be justified by the data that
     you have used to come up with it, there may be, and very
     often there are, alternative models which are equally
     likely, which you have not considered.
               So you have, by rejection, essentially -- by
     omission, rejected possible valued alternatives which may or
     may not be important to the final performance assessment
     questions that you're asking, but that has to be
     established.
               Type I errors are important to or can be analyzed
     using existing statistical methods, much more easily than
     Type II methods.  So, our approach at the present is to look
     primarily in a subjective manner at this type of error and
     try to introduce quantitatively uncertainty due to the other
     type of error.
               It becomes very apparent, primarily because of
     Type I errors, that by working with a single hydrogeologic
     concept, a priori, and never considering anything else, one
     would, a priori, underestimate uncertainty, because of
     under-sampling of the model space, but even of more concern,
     at least of concern to me, is the possible statistical bias,
     if you wish, introduced by Type II errors when one chooses
     an invalid model.
               So, I was very concerned with that for quite some
     time and still am.
               CHAIRMAN GARRICK:  Do you distinguish between the
     issue of invalid model and the misuse of a valid model?
     Under sampling could be a misuse of a valid model.
               MR. NEUMAN:  Not necessarily a misuse, but by
     selecting a model which is appropriate for the task, given
     the data that is available, all models that we use are
     conditioned on the information available to us.  The
     information is never complete, so everything is conditioned
     on what you have.
               By not considering other possibilities, you may be
     excluding outcomes of a performance assessment that you
     should not be excluding, so the outcomes that you do
     consider may be valid outcomes, given the information that
     you have.
               But there may be other valid outcomes which simply
     never enter into the picture.
               CHAIRMAN GARRICK:  That's what you mean by the
     alternative?
               MR. NEUMAN:  Then we mean exactly the same.  And
     in the report, there are quite a number of examples, one of
     them, for example, being the issue of what causes the high
     hydraulic gradient on the Northwest side of Yucca Mountain.
               And I present in the report in your hands, a
     step-by-step analysis that I have done of -- I don't
     remember how many -- seven, eight, nine, quite different
     conceptual models that have been proposed by various people
     for this.  And I rank them according to my own
     understanding.
               And it is this kind of ranking that I hope we will
     be able to incorporate into eventually a guidance document,
     which is what the NRC hopes.  I don't know how successful
     such a guidance document can be.  It can never encompass all
     possibilities, but that should be an example of what we hope
     to do in terms of Type II errors, in particular, in other
     words, eliminate all models that are not fully supported by
     all the available data, and rank the remaining ones in terms
     of this the most likely or most plausible in light of the
     data and so on and so on.
               So, this is systematic conceptualization, as well
     as an attempt to introduce uncertainty into the analysis,
     due to less than ideal models.  A question which is of great
     importance to those who use simplified models -- of course,
     the NRC, as well as other regulatory agencies, as well as
     the DOE, do rely on multimedia models in which the
     hydrogeologic component -- and I cannot speak to other
     components -- is greatly oversimplified.
               So the question then arises, and, in fact, has
     been raised earlier, how does that simplification impact
     both the choice of the model and the uncertainty associated
     with it, the Type I and Type II errors?
               One of the things we will try to do is to come up
     with a formalism by which a complex hydrogeologic
     environment, as it is implied by the available data, could
     be simplified by appropriate averaging techniques, whether
     it is averaging in space, or averaging in time.  Our
     preference is averaging in probability space, because it
     maintains some of the information about spacial variability
     that is so important to geologists and hydrogeologists.
               But it certainly is not the only way in which this
     can be done.  But one of the conclusions from my literature
     survey was -- and I was surprised to hear this gentleman ask
     earlier about conservative complex models versus less
     conservative simple models, because what I keep hearing from
     the NRC licensing staff, and what I keep reading in the NRC
     documents, and not only NRC, is the notion that simple
     models are, a priori, conservative, simple and perhaps
     generic.
               And I find that not to be the case in many cases
     -- in most cases I have examined.
               So it's kind of a little bit going the other way.
     Here is what we are then attempting to do:
               We are attempting to develop a methodology which
     addresses the issues I have just listed.  It definitely does
     focus on the subsurface hydrogeologic aspect of things, so
     it doesn't cover everything.
               The two elements are conceptualization and
     uncertainty.  Our work is generic in that it should apply to
     the entire gamut of groundwater-related modeling activities
     that the NRC is concerned with, but we hope that it will be
     of practical use.
               I hear today that the NMSS is not finding the
     material we have so far released as being of direct
     practical use to them.  I am surprised to hear that in light
     of my review, because I think that some of the examples I
     have included should already have been of use as examples of
     how one should attempt to eliminate Type II errors, which,
     in fact, can be found in some documents.
               But we will try, and this will require more
     communication between us and the NRC over the next year or
     two.  We will try to make it sound even more practical by
     developing step-by-step guidelines to what one should be
     doing.
               We are very much interested, and I think it's very
     important that our research be supported by real site data.
     And this is one reason why we are applying, testing some
     aspects of the ideas and the methodologies that we are
     developing, on real sites.
               And the ultimate product should be a systematic
     framework for identifying and quantifying uncertainty, once
     the systematic logic for conceptualization has been
     completed.
               As Phil Meyer has pointed out, we benefit from
     collaboration with PNNL, and here are some examples of
     actual collaboration:  Exchange of databases, learning
     methodologies, especially the Bayesian Methodology, which
     one of my students has used, the one that Phil Meyer has
     developed with Glendon and his colleagues.
               It was already mentioned that Mary Hill of the
     USGS and I are organizing a special session for the Fall
     meeting on predictive uncertainty of groundwater models.  I
     was hoping originally that it would be a special session,
     not only on predictive uncertainty but also on conceptual
     models.  That was vetoed by the AGU committee that looked at
     it.
               They consider conceptual modeling to be a nebulous
     concept, not clear just exactly what it means, and very
     little work has been done.  I forgot to mention, but that
     was one of my findings from the literature survey.
               There is virtually nothing in the hydrologic
     literature.  I shouldn't say nothing, but virtually nothing
     about conceptual modeling.
               It's starting to come out.  This book that was
     mentioned earlier, which is supposed to come out will soon
     be perhaps one of the very first things about it.
               And we hope to eventually transfer this
     information through some kind of a workshop to the NRC
     licensing staff.  We envision, at the current budget rate,
     to complete this in about two years.  If the budget changes,
     so will this change.
               Now, I will not go at this point into the details
     of this overhead, because I think I'm taking more time than
     I should, but I will mention that there is an established
     methodology in groundwater modeling -- I will refer to it as
     the traditional approach.  This is the most recent and
     sophisticated representation of that traditional approach.
               There is an approach that one can use with
     existing tools and concepts and very often does use in
     groundwater hydrology where one postulates a deterministic
     model, then postulates an uncertainty model for the
     parameters.  Sometimes, if there are monitoring data,
     optimizes these parameters, these prior parameters to come
     up with posterior parameters and associated uncertainty, and
     then propagates the uncertainties through the model.  There
     are various techniques to do that.
               We hope to transcend this by incorporating in this
     methodology, ideas relating to structural model uncertainty,
     the conceptual mathematical aspects, and ideas which are so
     important in a geology where things vary on a multiplicity
     of scales beneath your feet, no matter where you go.
               There are issues of resolution of information lost
     by averaging.  There are serious issues of scale, and
     uncertainty or randomness due to an inability to measure
     everything beneath our feet.
               These issues to which I refer to as stochastic
     elements, because they are typically treated in a stochastic
     manner, these two aspects, we are working on, incorporating,
     piecemeal, because there is no way we can cover the field,
     into our theory -- or into our methodology, I should say.
               So that would be the first bullet on this next
     overhead, integration of structural and stochastic aspects
     into the traditional uncertainty analysis that is presently
     available as a tool, perhaps has to be put into a single
     document, but we are finding that that doesn't always work.
               One of the requirements at this stage is to come
     up with effective parameters.  I'm so happy that that
     concept has arisen earlier.
               These effective parameters encounter a severe
     problem of scale.  An effective parameter for this volume is
     not generally going to be the same as an effective parameter
     for this volume, and an effective parameter for one type of
     hydrogeologic variable is going to be different than for
     another.
               So, we find that the traditional approach may
     sometimes not be the best, and so we are also developing and
     looking at potential non-traditional ways of incorporating
     those two aspects into the methodology.
               I will finish my part of the discussion by just
     alerting you to this.  There is a series of two overheads.
     I will not go through this unless you ask me to go
     specifically through some aspects.
               But it lists for you, work to date, part of which
     has been completed, and part of which is ongoing.  And under
     each element here, work element, you will see whether it is
     ongoing or completed.
               I did mention the literature review which resulted
     in the letter report of March, 1998.  A draft methodology
     letter report, which we hope will evolve into a NUREG,
     hopefully rather soon, that was published last November.
               Our field laboratory -- we have more than one, but
     the active field laboratory now is the Apache Leap research
     site, and I remember talking to you about that last time we
     met.  Probably the members of the Committee are not entirely
     the same, but nevertheless, I'll be happy to go back to that
     discussion.
               I will only mention, and then stop, that the past
     work that was done at the Apache Leap research site, which
     is an unsaturated fractured rock site, was done with
     high-level nuclear waste as the focus, though the work is
     entirely generic, and most of the work that my group has
     been doing there -- of course there were previous groups
     also under NRC support that have been doing work there --
     most of our work was associated with air testing and
     characterization of the rock, as well as study of scaling
     properties and so on, using air.
               All of that has now been completed.  It has led to
     a certain conceptual framework that we feel comfortable
     with, and that was an evolution, and I will be happy to
     discuss that, if that is of interest to you.
               The question we are asking now is how relevant is
     all this air testing and the results obtained from that, to
     water flow?  And that is of concern, not only at Yucca
     Mountain.  We considered this site to be a sort of analog of
     Yucca Mountain, but it would be of interest to any site
     close to the surface in which a fractured porous rock -- it
     doesn't have to be tuft; it could be sandstone; it could be
     chalk -- is encountered.
               For that purpose, Professor Peter Wierenga is
     running at the site, water and tracer experiments, and I
     would now call on Peter to tell you a little bit about those
     experiments.
               DR. CAMPBELL:  Before Pete gets started, I want to
     apologize that we don't have his viewgraphs.  We will get
     copies of his viewgraphs after we're done today.  And if you
     accidentally in the audience got a set of viewgraphs that
     are missing -- these are two-sided viewgraphs, and the ones
     that were handed out, if you got a set that's only
     one-sided, throw it away, and there are some two-sided
     viewgraphs in the back.
               MR. WIERENGA:  Thank you very much.  It is a
     pleasure for me to present my material here at this
     distinguished meeting.
               The cooperators after the program, I am listing on
     this overhead.  This is in addition to myself and Dr.
     Neuman, we have a post-Doc and a technician and a retired
     scientist from the USDA, Mr. Rice, who is on an hourly basis
     also helping out on this project based on his extensive,
     very extensive field experience in this area.
               The sites are located in southern Arizona, or
     central Arizona, more or less.  And we have talked today
     about two sites.  The site of my talk today is the Apache
     Leap site.  That is about the same elevation as Phoenix.
     Here is Tucson and here is the border with Mexico, of
     course, and there is the Maricopa site that was -- we did
     earlier studies, and hopefully will continue additional
     studies.
               This is a regular alluvial site, and this is a
     fractured rock site.
               The fractured rock site was used for earlier
     studies by Dr. Neuman and his students, and they did the air
     permeability studies on this site, and they installed these
     bore holes, vertical and standard bore holes.  There is
     another one somewhere here.
               And they did air permeability studies between
     those bore holes.  They cleaned up part of the site, and
     found a large fracture here, and then built our infiltration
     studies over this site, so this is Plot Number 1, 2, 3, 4,
     5, 6, 7, 8, 9, so we established nine plots, each three by
     three meter for a total plot size of nine by nine.
               And then we installed various observation points,
     measuring devices, drilled holes and installed measuring
     devices below the surface of these plots and then flooded
     the plots with water.
               We kept a constant head of water on the plot.  I
     show you a few slides, additional slides of that setup.
               The devices that we used to measure the water
     below the surface of the plots were basically three devices:
     One is the neutron probe, which gives us a relative count
     rate which is related to the water content or the degree of
     saturation.
               We used a tensiometer which measure the matric
     potential in the rock material.  The matric potential is
     related to the energy with which the water is held in a rock
     material, and then we used a solution sampler that is used
     to extract the pour water from the rock material which we
     then take to the laboratory to analyze.
               So these three devices were installed on each plot
     at different depths.  This is a viewgraph of a tensiometer
     which is a porous body that is filled with water and then we
     have -- basically we have here a pressure transducer and the
     pressure transducer has it's leads to a data logger at the
     top, and the negative pressure that we measure then with the
     pressure transducer after is in equilibrium and the
     surrounding material is recorded, and that gives us an idea
     of the degree of saturation or the matric potential of the
     rock material.
               This is one phase of the installation of the
     material.  This is the solution sampler.  There are two
     leads to the surface, and we apply vacuum on those, and draw
     the samples through this porous body into the stainless
     steel and then apply pressure and force it to the surface,
     so we have those at five depths below the surface.
               Here is a further version of the project.  Here we
     have completed the -- we have a concrete barrier around each
     plot and these are the various devices in each plot, and
     these are the supply tank, and each tank is connected to one
     plot and by measuring the level of water, the rate of
     waterfall in this tank we know how much water has
     infiltrated to the surface of this plot.  It's a fairly
     simple setup.
               DR. HORNBERGER:  Did you tag the water at all,
     Pete?
               MR. WIERENGA:  Now we have lately tagged it with a
     bromide tracer.  We would like to do additional work on it
     but I will show you some results and show you what comes out
     of it -- but we could tag the water in each individual plot.
               DR. HORNBERGER:  Right.
               MR. WIERENGA:  This is a later completed version.
     We also built a structure over this -- this is our
     instrument trailer -- so it prevents rain.  Of course, the
     surface of each plot, there's a floating cover on each plot
     so we have no evaporation losses and whatever goes into the
     plot we know pretty sure that all of that is infiltrating in
     the plot.
               DR. HORNBERGER:  It looks like Biosphere 3.
               [Laughter.]
               MR. WIERENGA:  Thank you.  It's hot too.  It is
     warm and in the winter that is nice to work on it, and in
     the summer it gets awfully warm, a little too hot.
               There is a great deal of variation, variability --
     not variability but spatial variability in the infiltration
     rate.  This is the cumulative amount of water that was added
     to each of those plots of the first 200 days.  We are now at
     about Day 250 so we started this experiment just before
     Christmas, started flooding it, and so you see in Plot 3 and
     9 we have fairly high rates but at Plot 5, for example, we
     have only a rate .036 and these are in between.
               The rates are fairly steady but here they drop off
     a little bit and these are also steady but then they
     accelerate.  Why that is?  We don't really know that.
               In a mineral soil you would see, of course,
     initially you would have a very high rate going into a dry
     soil, but as time goes on you would see a slowing down of
     the rate and sometimes a great deal of slowing down of the
     rate but in this fractured material this is not happening.
               The data that we have from this is this is an
     example of the relative count rate measured with a neutron
     probe.  The relative count rate, as I said, is related to
     moisture content.  As you see, as time goes on at 30
     centimeters the water content is increasing and so is it at
     55 but to a lesser degree.  The water content was much
     slower at 30 centimeters initially than at this depth,
     therefore it has a lower count rate because of the drying
     out of the profile before we started the experiment, but
     then gradually wetted up.
               This is the deeper one at 450 centimeters was
     initially already quite moist, and there you see that it is
     30 and 150 centimeters.  It kind of reached the same
     relative moisture content, so this is kind of the behavior
     that we observed with the neutron probe.
               It is a very slow process.  I thought the
     experiment would have already been over, but this is a
     long-term experiment because the permeability of the
     material is not fairly high.
               Now in a vertical direction, you see a great deal
     of variability.  You barely see -- you see the water
     infiltrating.  You could see some, in this particular plot,
     some response maybe down to these depths.  These variations
     are mostly due to uneven absorption of water from the rock
     material itself by the dry bentonite that we packed around
     the neutron probe.  We have to pack everything with
     bentonite to prevent preferential flow along the observation
     devices down to the subsurface.
               Another set of readings is from the tensiometers.
     This is in Plot Number 9, an example of the tensiometer data
     that we get with tensiometers.  As you can see, there's a
     couple interesting things.  Initially at the surface it was
     dry so you get a very negative matric potential, but as the
     water front arrives it gets fairly wet, so it wets up and it
     becomes closer to saturation.
               The saturated soil of course, the matric potential
     is zero, and that 1 meter is following and then the 2 meter
     is following.  The 3 meter is fairly steady.  Initially it
     was wetter because that became -- we packed the tensiometer
     cups in a wet silica flour, and that finally diffused into
     the surrounding rock but here it stabilized and then the
     water front is arriving and it gets closer to saturation.
               The 5 meter depths again was a little water
     because of the construction of the tensiometers'
     installation, then it's fairly flat and nothing -- the
     waterfront clearly has not arrived there.
               Now then, is this a good example or is this a bad
     example?  In the next slide I have plotted all of them, all
     of the data from all of the tensiometers, and you see a
     large variation in their behavior.  While this one is
     arriving fairly fast, Plot 8 is responding fairly fast, this
     is at 50 centimeters, Plot 7 is really a laggard and it took
     almost -- what is this? -- 100 days between the arrival
     between one plot versus another, so it is not a fairly even
     process and that is what you quite often see in the actual
     world in the hydrogeological setting.
               Now this is at one meter.  Again this is arriving
     later, but the behavior is also slower.  It is slowly
     wetting up, so it looks almost like it's matric flow and not
     fractured flow in this sense.  Otherwise in fractured flow
     one would see a rapid increase at some depths, at some
     tensiometer, but here we don't see this, but again quite a
     difference between the arrival of the wetting front at that
     depth.
               This is the three meter one and again we see here
     slowly wetting up here, here it was already.  The moisture
     content was relatively high already at the 2 meter depth,
     but here we see for example the 6 meter one is increasing,
     the 2 meter is fairly steady but then starts increasing, so
     I am feeling that the moisture front has kind of arrived
     after about 100 days at this particular depth, by and large.
               These are the tensiometer data.  Now of course
     after about 200 days, we added a bromide tracer to all the
     plots, to the water in all the plots, and so far we have
     seen only appearance of the bromide in the subsoil of the
     Plot Number 8 at 1 meter, and the amount of bromide was 30
     parts per million, I believe, and so we see an increase in
     the bromide concentration at this depth at, after day ten,
     but we did not see any bromide at 50 centimeters, although
     we have a device there, and we did not see bromide at any
     other points except at 5 meters it is appearing now -- I
     mean at half a meter in Plot 5 and last weekend, this past
     weekend, my people told me it was up to 10 parts per
     million, so this data shows fairly irregular behavior of the
     water flow, but much more so of the tracers, and I think
     this tracer is, in my opinion, clearly some indication of
     fractured flow in this plot and perhaps this one also.
               Maybe we will see more of it as the waterfront and
     the tracer front is moving deeper in the profile, but so far
     we have not seen any response to the tracer.  What we'd like
     to do also to learn a little bit more, have a different
     tracer on each plot.  The problem with that is the cost of
     analysis goes up a little bit, and I don't know whether we
     will be able to afford it, but that would give us even more
     insight in the behavior of the tracer through the subsoil.
               So if I can -- by the way, the bromide behavior is
     also clear from the washing out of the salt.  The salinity
     of the subsoil is relatively high, so when the salinity
     decreases, the bromide increases so we have two independent
     measures that it is not an analysis problem.  It is really
     actual data.
               So far we have regular variability in infiltration
     rate.  After flooding 200 days more or less you reach the 3
     meter depth and we have not yet reached the 5 meter depth
     except in one plot.  It seems to break through.
     Breakthrough of bromide has occurred at Plot 5 at half a
     meter and at Plot 8 at one meter, but no bromide observed at
     other points and the early bromide breakthrough in Plot 8
     indicates fractured flow.
               That is what I wanted to tell you about this
     experiment.  Hopefully, if we continue this, at least maybe
     another half a year or a year, constant flooding, then we
     will learn a little bit more about the importance of
     fractured flow at this point.  So far, I feel I wasn't
     totally convinced that we would see fractured flow, but my
     conceptualization of that plot was incorrect and I must now
     see that we do have fractured flow at this site and that has
     to be accounted for in the models even though maybe it is
     happening only at one point, but one point over 9 x 9 is 81
     square meters.  If you multiply that over a larger area,
     then it becomes quite often -- quite clear that this kind of
     behavior is very important for environmental impacts.
               Thank you for your interest.
               DR. HORNBERGER:  Thank you, Pete.  I am sure we
     have some questions for Slomo and Pete.  I just would -- for
     those on the committee who perhaps don't recognize it, I
     should point out that we have a phenomenal concentration of
     expertise on the vadose zone processes in semi-arid and arid
     conditions in this room.  Woe be the world if we had a
     disaster in this room today.
               [Laughter.]
               DR. HORNBERGER:  That is not their only expertise
     but I know the committee has some interests in vadose zone
     processes in arid regions, and I just want to remind you
     that you can take advantage of the expertise in the room.
               Questions for Pete or Shlomo?  Ray, do you want to
     go first or do you have to bug out?
               DR. WYMER:  I have to bug out.  I know very little
     about the field but I do know Shlomo and I am impressed with
     the clarity and lucidity of it and the very high conceptual
     level that you have tackled her and the information you have
     presented.  That is about the extent of what I can talk
     knowledgeably about.
               MR. LEVENSON:  I just have one comment in response
     to Shlomo's question.  My question was really not as to
     whether simple or complex models were more conservative.  My
     question was the distribution of uncertainty should not be
     the same for conservative and nonconservative models.
               MR. NEUMAN:  If I may just make an observation in
     that respect, one of the attributes of relatively simple
     models with fewer compartments, larger compartments that is
     often cited is the fact that the variance of the parameters
     reduces as the size of the compartment or the averaging, the
     level of averaging increases, and that is definitely true,
     so you could say uncertainty goes down, but on an overhead
     which I think you may have but I have not shown I make the
     observation that at the same time the mean of the
     distribution moves as you average, and so by averaging you
     are moving, shifting the entire distribution -- yes,
     indeed -- you are reducing the spread, but you are removing
     the mean, and unless you are aware of it and factor it in,
     you end up with a bias, and I am more concerned with that
     bias -- in other words, the reduction of the uncertainty.
               MR. LEVENSON:  That is really the issue that I was
     trying to address.  Very good.  Thank you.
               DR. HORNBERGER:  John?
               CHAIRMAN GARRICK:  Well, I am not going to talk
     about the earth science component of this but one of the
     things that this committee has spent a lot of time advising,
     if you wish, the NRC on, is that analyses ought to be done
     on the basis of getting a handle on results that are as
     realistic as possible.
               This debate of uncertainty analysis and
     conservative uncertainty analysis makes no sense to me.  If
     you are talking about an uncertainty analysis the only time
     it makes any sense to me is when it is uncertainty about
     what you consider to be a reasonable model and a reasonable
     investigation.
               A set of distribution functions that are called
     conservative to me don't have much meaning.  A set of
     distribution functions that are a direct result of your best
     attempt at a realistic appraisal of a parameter or an
     aggregation of parameters does make sense, and then the
     concept of conservatism is something you apply in the face
     of that information, but I was just curious about some of
     the strategies that you are adopting here as far as these
     dose calculations are concerned in terms of the uncertainty
     analysis and I think it is essentially an oxymoron to talk
     about uncertainty analysis and conservatism.
               To be sure, the distributions ought to be your
     best shot at what you believe your state of knowledge is
     about a parameter --
               MR. NEUMAN:  I am extremely glad to hear you say
     that because that is exactly my philosophy.
               CHAIRMAN GARRICK:  Yes.
               MR. NEUMAN:  I am willing to accept from
     regulatory agencies sometimes the need in the face of lack
     of information and lack of ability to develop a fully
     realistic model to err on the side of conservatism, but I
     can think of many examples where the word "conservatism" as
     you suggested has no meaning and my favorite example, which
     is a very simple one, when you are propagating a solute --
     it could be a contaminant or otherwise -- through a column
     of porous medium it will depend, the transport will
     depend -- for inert tracer will depend on the velocity and
     on the dispersivity of the medium.
               If you change the dispersivity, you make it small,
     you get a sharp front and it will take longer for the
     contaminant to reach the edge.  One could say that by
     therefore increasing the dispersivity, one is being
     conservative because one will get a sooner and earlier
     breakthrough.
               Well, but at the same time, you will get a lower
     peak because you have increased your dispersivity and in
     that sense you are not conservative.
               So I couldn't agree with you more that the notion
     of conservatism often doesn't make sense, and I would
     perhaps add to this that I think it has been grossly misused
     and is being grossly misused.
               CHAIRMAN GARRICK:  Well, the point is that there
     is a point beyond which you obscure the truth.  You obscure
     the validity of the analysis and what the evidence for that
     analysis can support.
               MR. MEYER:  I just want to make a clarifying
     comment, since you seem to be addressing some of the issues
     that I raised.
               The issue is not so much the use of conservative
     distributions for parameter values.  The parameter values,
     the distributions of those that we have used in all of our
     analyses, are not conservative parameters.
               They represent the best estimate of the knowledge
     about that parameter at a particular site, and in the face
     of limited data.
               However, the models themselves, the implementation
     of the processes, which in the codes are intended to be
     conservative in some way, and so you're taking those --
               CHAIRMAN GARRICK:  Well, this is --
               MR. MEYER:  You're taking the parameters that may
     not -- that aren't conservative, but you're applying them in
     models that may be intending to model the whole situation in
     a conservative way.
               CHAIRMAN GARRICK:  It was designed that way
     because it's a screening tool, and so I understand that,
     yes.
               But what we're trying to preach here is, let's
     understand, on the basis of the evidence, what can be said
     about a parameter, and then if we want to apply conservative
     values or criteria to that, we can do that.  But let's start
     with knowing what our best shot is and what the real answer
     is.
               MR. MEYER:  Right.
               DR. HORNBERGER:  Slomo and Pete, you've introduced
     a very vexed question, this whole issue of structural
     uncertainty, if you will.  And now that you've changed -- or
     Pete's changed his mind about his conceptualization, okay,
     at the Apache Leap site.  He just told us that.
               But I'm curious now, given that, does that just
     mean that you change your favorite conceptual model, or does
     it mean that you actually have several different conceptual
     models that you would apply now?
               MR. NEUMAN:  Do you want to start, Peter?
               [Laughter.]
               MR. WIERENGA:  Well, of course, one should always
     be open to different conceptual models.  And focusing on one
     conceptual model that things will happen, like you have in
     mind beforehand, that is not the right way -- not the right
     way to conduct science.
               So, I think that, while I discussed it with Slomo
     this morning, he thinks that a continued model with the
     possibility of having fractured flow in the -- with a spot
     of the -- model, is probably the way to go.
               And I would agree with that.  It looks to me that,
     you know -- and I'm not an expert in fractured rock, but it
     doesn't really matter so much, where that fracture is, but
     what does matter is that the water of the tracer gets there.
               And so maybe it's more important, how many of
     these fracture one has per surface area that contribute to
     quick transport of contaminants from the surface to the
     subsoil or to the groundwater, but not necessarily, you know
     -- we really don't have to know precisely, the description
     of the pathway; what we need to know is what is the chance,
     what is the -- how many of these pathways are there per
     surface area, to do the modeling?
               And maybe Slomo has a better answer to it, but,
     yes, I did change my position.
               MR. NEUMAN:  Not better, but I just want to add a
     little bit.  And that is that we have been working at this
     site for at least a decade now.  And our concepts have been
     gradually shifting.
               My own work until now, as I said, was associated
     with air flow.  Well, we are now pretty much convinced that
     in order to properly characterize air flow under the
     conditions of our testing in the past, one could justify
     taking this fractured medium and conceptualizing it for
     modeling purposes as a heterogeneous continuum in which
     permeability varies, air permeability varies form point to
     point, according to a random field model.
               And we have come up with a particular fractile
     representation of that, but that's kind of a secondary
     issue.  So a continuum stochastic representation, random
     field representation, seems to work for as far as describing
     the heterogeneity that we could observe, based on our
     experiments.
               Now comes Peter's experiment, and low and behold,
     I do not think that the particular air permeability model
     that we have developed is fully going to explain the air
     flow -- the water flow.
               We don't know, though, because our air tests were
     down below three-meter depths, and his water has not reached
     below three meters yet.  So it's extremely important that we
     continue this test for at least half a year to a year to see
     just how deep this will get, and hopefully so it gets into
     the domain where we do have air permeability data and see
     what happens.
               But even more important than that is what Peter
     showed you about the difference in behavior between water
     and tracer.  It is obvious, I think, based on his
     preliminary results, and likewise obvious if you compare the
     air and the water results so far, that a conceptual model
     and associated mathematical model that may work for one
     phenomenon, may not work for another phenomenon.  And that
     is the really interesting thing that is happening here.
               This is why I think this idea is so interesting.
     It also raises some extremely interesting scaling issues.
               And so as far as I'm concerned, if I may enter a
     plea here, in the past we have been working for the NRC at
     other sites.  One of the earlier sites was in saturated
     granite at a site called Oracle.
               We were able to extract a lot of very useful
     information from that.  One of the conceptual changes that
     that site has instituted, in my own mind, is that I started
     that work believing that one should be able to carefully map
     in three dimensions, the distribution of fractures in
     granite and then use that geometric information, plus
     geologic understanding of the fractures, to make predictions
     about flow.
               And the conceptual shift that has occurred in my
     mind was that that is not a viable way to proceed.  And I
     now have collected a huge set of data from other sites which
     supports that.
               But the plea:  What has happened is that as we
     have completed our -- the easy part, the hydraulic testing
     part, the project's original plan called for additional
     tracer testing.
               But we were stopped in the midst, and we were
     never able to verify that what we have learned about the
     site regarding hydraulics, would also apply to contaminant
     or tracer transport, because the budget was cut right there
     because Congress shifted its interest onto Yucca Mountain,
     and so everything had to now go into the vadose zone, and we
     were dealing with the saturated zone below the water table.
               It took the Department of Energy and others years
     to discover what we have said at the beginning, that Yucca
     Mountain does have a saturated zone, and one should not
     focus only at the shallow area, but the original concept was
     that contaminants will never reach that deep.
               So, these concepts are changing and evolving with
     time.  And my plea is the following:
               The same thing happened to us with air
     permeability testing at the Apache Leap.  The original
     program was to continue with gaseous tracer tests.  And
     Walter Illman, I don't know if he's in the room or not, who
     conducted the pneumatic tests for us, was all set up with
     his technician to go ahead.
               But then the budget was cut for that, and so we
     have never had a chance to do tracer tests.  And just think,
     if Peter did not have a chance to run tracer tests in
     conjunction with his water tests, how less rich would we be
     about our understanding?
               How much more limited would we be in our ability
     to see what is happening at the site?  I really think that
     it is absolutely essential that there be more continuity in
     the funding of projects which the NRC considers, a priori,
     to be meaningful; otherwise, let's not even start funding
     them.
               But if you fund them, fund them continuously so
     that a good amount of information of this kind can be
     gained.  So in this particular case, I would just simply
     suggest, let's make sure that Peter has the budget to run
     his experiment for at least another half a year to a year.
               DR. HORNBERGER:  To the extent that the ACNW
     controls the Office of Research Budget, we will dot hat.
               [Laughter.]
               MR. NEUMAN:  I know you don't.
               MR. WIERENGA:  I would like to add, though, that I
     feel that, in general, the NRC research staff has been very
     supportive and very generous in the support of my research
     and Dr. Neuman's research.
               But there are things outside their control, also,
     like it is out of your control, it is also out of their
     control.
               And I'm really grateful for all of the support
     that I have had from the Nuclear Regulatory Commission to do
     this work.  I would be happy for the support.
               MR. NEUMAN:  All I can say to this is amen, and,
     in particular, Tom Nicholson really deserves, I think,
     accolades for the way he has been supporting our research
     all these years.
               DR. HORNBERGER:  Before I leave and let you off
     the hook on my question, I'll let you guys off the hook, but
     I wanted to see if I could put Glendon on the same hook,
     because I know that he has a lot of insights developed from
     lots of work he's done at Hanford.
               Do you have any insights on this whole issue of
     structural model uncertainty and how we should handle it for
     vadose zones?
               MR. GEE:  Well, I am very interested in, and
     endorse Slomo's approach.  I think it's not been done
     before.
               Certainly Hanford is just beginning to come to
     grips with these kinds of problems.  Charlie Kincaid, who
     basically is doing a lot of the modeling for the Hanford
     site was here yesterday, but unfortunately he's not here
     today to perhaps provide some insights.
               But basically I think he actually came for the
     two-day presentations so he could listen and learn and
     hopefully integrate some of the things that are being
     planned, and benefit from the NRC research in this area.
               I just think it's an extremely important aspect of
     the modeling that has not been looked at before, and I'm
     grateful that NRC is interested in it.
               DR. HORNBERGER:  I just have one other general
     question:  I don't know if any of you -- well, I'll throw it
     out and see if any of you want to answer it.
               One of the things that we get asked, that is, the
     ACNW, in evaluating, if you will, the NRC's research program
     is to what extent the NRC should be involved in doing
     research, because, after all, one argument could be made
     that it should be up to the applicants for licenses to do
     the research.
               The other argument is that NRC has to have
     expertise, capabilities to do reviews.  Do you have any
     insights for us on how you would weigh in on such an
     argument?
               MR. NEUMAN:  Can I try?
               DR. HORNBERGER:  Slomo?
               MR. NEUMAN:  On this question, of course, I've
     been working off and on for the NRC for quite a number of
     years, and this is a question that arisen almost on every
     occasion where the research program was reviewed one way or
     another.
               I cannot speak for the need for research by the
     NRC and its contractors in areas other than the areas I am
     familiar with.  But it seems to me that the issues that the
     NRC is facing are of tremendous complexity, and, in
     particular, the issues that I'm familiar with, which pertain
     to hydrogeology, groundwater being a major transporting
     element in the environment of radionuclides and other
     contaminants, actively or potentially.
               The earth is such a complex system, and we know so
     little about it that I just do not believe that it is
     possible for either a regulatory agency such as the NRC or,
     say, an agency such as the Department of Energy that might
     apply for a license for high-level waste storage or other
     entities that would apply for license that entails analysis
     or requires analysis of subsurface processes.
               I just don't see how the NRC could possibly do
     this work in good faith and with competence, without having
     a good understanding on its staff, an up-to-date,
     state-of-the-art understanding of as much as possible
     relating to these processes, as they pertain to the task at
     hand.
               It so happens that other agencies, and certainly
     private groups, are not going to do the kind of research
     that the NRC is currently supporting.  It's unique.
               Even the experiment that Peter is running now is a
     unique experiment.  You think about it as a simple
     experiment, but it is far from simple, but it's absolutely
     unique.
               The Department of Energy has botched similar
     experiments at the Idaho basalt site, and I can assure you
     that when we concluded our work at Oracle in saturated
     granite, I paraphrased to the NRC, Churchill by saying never
     have so few done so much in so little time for so little
     money.
               [Laughter.]
               MR. NEUMAN:  And I can say that again.
               [Laughter.]
               MR. NEUMAN:  And so I really think that despite
     the fact that perhaps the NMSS does not always see where all
     of this is leading in terms of their practical needs, I
     really think that it is of relevance, and I certainly would
     think that it is a very good thing for the NRC to do some
     research of this kind, and perhaps more than it's doing.
               MR. WIERENGA:  Also there is really not much
     possibilities in this country to get long-term funding for
     field research, and unfortunately, we don't have right now,
     the structure in the government so that we could easily do
     more integrated work.
               I see that we also need to do long-term work, but
     also more integrated work.  For example, scientists from the
     National Labs could participate in an experiment that we do,
     and take other measures of, let's say, how water and how
     this process behaves in the subsoil.  I don't see that much
     happening in this country, and we don't seem to have the
     mechanism for it.
               But certainly we need to do some longer-term work,
     and the NRC is in a position to have more continuity in
     their program as opposed to an NSF grant that takes three
     years, and that's almost impossible to do the work for the
     type of budgets they have.
               You cannot do field work for that, it's
     impossible.
               DR. HORNBERGER:  Phil?
               MR. MEYER:  If I could just second what Slomo and
     Peter said, but I also want to offer just a little bit
     different perspective.
               I have attended a number of public workshops in
     which there are a lot of licensees in attendance, and both
     from the questions and comments that they make in the
     meeting, and also from my personal interaction when they
     have come up to me and told me about their sites and what
     the issues are, that they are, number one, strapped for
     expertise in addressing the hydrogeologic issues at their
     site.
               And, number two, that probably it's related; that
     they are desperate for guidance, and they look to the NRC to
     give them guidance on how to analyze their sites.  I don't
     think that it's just that they want a -- at least that's not
     my impression, that they just want a cookbook so that they
     can quickly get on with, you know, the processes, but that
     they seem to be genuinely concerned about doing a good job.
               And in my experience, the work that the Research
     Office has supported has offered a lot to the licensing
     staffing the time that I have been working with the NRC.
               DR. HORNBERGER:  Thanks.  Well, I have a host of
     detailed questions, but I think they will have to wait till
     Slomo's AGU session.
               Thanks very much.  Thank you, Mr. Chairman.
               CHAIRMAN GARRICK:  All right, I think that what
     we'll do is recess now for lunch, and then come back and
     continue our ACNW report-writing session.  I should announce
     that we are targeting to adjourn at 3:00, and I think this
     will terminate our need for keeping a record.
               [Whereupon, at 12:35 p.m., the recorded portion of
     the meeting was adjourned.]
Page Last Reviewed/Updated Monday, January 27, 2014