Opening Remarks
B. John Garrick, Chairman
Chairman Garrick reviewed the board's report to Congress.  The new U.S. Department of Energy (DOE) transportation, aging and disposal (TAD) canister system design needs further study, but is a promising system to avoid excessive fuel handling. DOE needs to determine the compatibility of possible TAD canister designs with the capabilities for storage, handling and transportation options at each reactor site.  Integration of the TAD approach into the waste management system. DOE may benefit from the review by outside experts on this.

The Total System Model is a good evolution of DOE's thinking of the total system for waste management.  Studies of the natural system continue to help explain the characteristics of radionuclide transportation in the natural system.  DOE needs to develop a better understanding of repository performance.  (See entire NWTRB report at   http://www.nwtrb.gov/reports/05letter.pdf.)

Fundamental understanding is essential.  The board will evaluate DOE's understanding of radiation doses based on waste forms and the natural system.  Movement of various radionuclides through the saturated and unsaturated zones is of great interest to the board.  Scale is an issue in modeling--small amounts of radionuclides versus the size of the natural system--grams versus mega tons.  Understanding this leads to greater confidence in the overall waste management system.

Presentations
J.R. (Russ) Dyer, Chief Scientist, U.S. DOE, Office of Civilian Waste Management (OCRWM)--"Program Overview "
The Atomic Testing Museum, where this meeting is being held, is a new facility that captures the history of the program that had significant impact on science and technology in the U.S.

Several organizational changes have occurred since last November at OCRWM.  The office of the director is still vacant pending Senate confirmation.  Paul Golan is acting director.  A new organization chart will go into effect in the next few months.  It takes away the east/west distinction of the previous approach.  Golan has 13 offices reporting to him.  Various offices are in charge of study, design, license, build, operate, prepare, transport and dispose (future).

Sandia National Laboratory is the lead laboratory to support OCRWM, under the purview of the chief scientist. 

The budget approved for Fiscal Year 2006 was $500 million with $50 million reserved for reprocessing.  OCRWM is studying which agency has authority to spend this $50 million.   This is well below funding requests.  There has also been a 1 percent rescission across the board. 

Plans are underway to simplify surface based repository designs based on the transportation, aging, and disposal (TAD) canister design, which entails no more bare fuel handling.  Final approval of TAD within DOE is pending.  Integration of this approach with overall system as asked for by the board is DOE's goal.  Actions are underway to prevent problems with research processes through better validation, oversight and internal documentation.  DOE believes the canisterized approach is vital and is making all efforts to evaluate it.

Q--Will a simple canister design be created from scratch or will it build on existing designs? 
A--No approach has been settled on.

Q--What is your timeline? 
A--It is still under development.

Q--Why haven't you decided on scientific priorities yet? 
A--80 percent of the program is well understood but the approach is not yet finalized.

Q--What about transfer of risk to utility sites based on the TAD approach? 
A--Discussions with utilities are underway and will tick up over time.

Q--Recommends a senior person be hired with materials expertise in terms of the waste package. 
A--Appreciates the suggestion.

Q--Will the design of the vault be changed based on the TAD approach? 
A--Short-term, no; more study is needed for the long-term.

Q--How will you absorb the budget cut and what time lags are anticipated? 
A--Identify the critical pieces of work and work on those.  Conceptual design of TAD will drive work when complete.  Trying to keep important issues from being pushed off the table.

Q--A concept similar to TAD was rejected in the past.  It would be worthwhile to explain why the previous decision was wrong.
A--Agrees, but can't address it now.  There was a report on the multi-purpose canister that might be revealing.

Q--What is the status of the total system performance assessment (TSPA) approach?  What is the fit with the global effort to solve the world's waste problem? 
A--Second item first--reprocessing may fit into this, but can't speculate.  On TSPA, an evaluation is ongoing since a new TSPA will be required based on the new EPA standard.

Q--Will there be a chief scientist at the lead lab, such as that at WIPP? 
A--It's an evolution.  Administrative processes are being set up. 

Q--How do other initiatives like reprocessing impact the OCRWM program?  Will OCRWM suffer?   
A--Similar issues were raised 5-8 years ago relative to transmutation.  Geologic repository is needed under any approach.  There may be an urgency issue though, with a possible impact of short term priorities.

Michael Ryan, Adjunct Professor, Medical University of South Carolina  "Conservatism, Non-conservatism, and Uncertainty in Dose Calculations--Risk Informed Dose Calculations." (PowerPoint available)
His views presented today are his own.  NRC 1999 report on risk informed performance is a benchmark report.

The goal is to highlight the key risk contributors.  A range of tools produce a range of results--extreme bounding analysis, bounding analysis, sensitivity studies, on-off calculations and calculations and probabilistic risk analysis.  Risk informed approaches can be used to determine releases and interactions.  Exposures to humans can occur by inhalation, ingestion and skin absorption.  He discussed various exposure scenarios for each of these.  For ingestion, dietary intake is a key factor.

He concluded by suggesting that a broad spectrum of approaches be used to make risk-informed decision.  Know the foundations and limits of these approaches.  Challenge reference factors.  All have strengths and weaknesses.

Q--Only a few radioneuclides are of concern at Yucca Mountain--Iodine 129, carbon 14, etc.  Should the focus be put on these? 
A--This is a worthy approach.

Q--Have you looked at the assumptions within TSPA?  Would a scorecard of analysis be useful? 
A--Not familiar with TSPA, but risk informing approaches are essential.

Q--Which approaches are better than others? 
A--Look at the individual case and see what is most appropriate.

Tim McCartin, U.S. Nuclear Regulatory Commission, "Implementation of a Dose Standard after 10,00 years" (PowerPoint available)
Mr. McCartin gave background on proposed Part 63.  NRC is adopting the Environmental Protection Agency's (EPA) new standards for doses that could occur after 10,000 years.  NRC will estimate doses for nuclear workers, as EPA has done for the general public.  NRC will also come up will a value to represent climate change after 10,000 years.

Important radionuclides over the years of repository use include americium 241, plutonium 240 and 239, Neptunian 237 and Technician 99, iodine 129, uranium 234, with various accumulations and/or decay over the timeframe.  He showed an illustrative dose estimate over one million years.  The waste packages are assumed to fail after 60,000 years due to corrosion based on their assumptions.  In the long-term, Neptunian 237 is the longest lasting radionuclide.  Doses will peak around 150,000 years at 8 mrem and drops at 200,000 years to 4 mrem.

Mr. McCartin also discussed climate change.  EPA said that a constant climate after 10,000 years could be assumed.  Key values include temperature, rainfall and deep percolation of water flowing into the repository horizon.  NRC studies expect wetter climate over the time period.  Historical vegetation patterns are used in conjunction with current conditions to estimate the fraction of precipitation that deeply percolates.  Ranges for future deep percolation go from 13 mm/year to 64 mm/year.  The mean value is 32mm/year, six times greater than current rate--more infiltration due to greater wetness and coolness in the future.

Comment periods ended last year.  NRC will finalize its standard after EPA finalizes its standard.  He responded to several questions related to how he calculated the doses.

Jens Birkholzer, Lawrence Berkeley National Laboratory
Ernest L. Hardin, Bechtel SAIC Corp. (BSC)
"Mass of Water Seeping Into and Out of Drifts Over Time" (PowerPoints available)

Jens Birkholzer
Studied water seepage in the geological "drifts" that hold canisters of nuclear waste - similar to what can be expected in a Yucca Mountain repository.

He defined key aspects of seepage testing:

Seepage - dripping of liquid water
Seepage rate - mass per time
Seepage % - ratio of seepage rate divided by percolation flux
Seepage fraction - # of packages affected

Various processes affect seepage, including canister shape, permeability, etc.; drift temperature, excavation, and degradation.  He has found that there is a tremendous impact from the heat of the nuclear waste package on seepage - if the environment's temperature is above the boiling point (100C), it vaporizes the water and therefore eliminates most real seepage problems.  His study predicts seepage during thermal periods for different locations in the repository.

Mr. Birkholzer explained his study methodology and calculations.  Conducted ambient seepage testing and calibration with liquid release tests.  Found that a "barrier behavior" affect results in seepage levels being less than the amount of the liquid injected.  From this data he created "look-up" tables showing different results for different categorizations (based on permeability, effective fracture capillary strength, percolation flux, and flow focusing).

He also studied affects of drift seepage depending on whether the drift was still perfectly intact or whether it had collapsed do to seismic events.  Collapsed drifts showed higher seepage.

Overall seepage percentage estimate over time - 2 percent seepage up to 600 years, then 5 percent.  98 percent of all water would divert overtime.  (Mean seepage rate approx. one drip per 50 seconds.)

Conclusion: Seepage is a function of location and time (heat level).  Seepage predictions in TSPA are based on science and are realistic, conservative.

Q - Do you plan to revisit this process, but looking at a longer time-span, in the future?
A - Period after 2,000 years - assume stays stable, so won't have to change.

Q - TSPA suggests 98 percent of seepage diverted?
A - 98 percent of seepage does not pick up radionuclides (only 2 percent does).

Q - Will you predict seepage data for all test sites - qualitative agreement?  Consider looking at other analog data?
A - Qualitative analysis is still ongoing since it is very hard to come up with quantitative numbers (only certain of locations and wet spots - 10 percent could show wetness).

- They are looking at analog data (e.g. from caves in Spain with similar geology to Yucca), and will continue to do so.

Q - Why is your data based on an abrupt climate change?
A - It does consider a drastic change, then a slowing with decreasing temperatures and less thermal impact. 

Q - Whose work justifies the official decision to calculate the spacing of waste packages at 81 meters?
A - Don't know, but his work supports this decision.

Q - How does this differ from NRC calculations?
A - Tim McCartin: NRC used much more simplified thermal calculations.

Q - Shows how rapidly an unsaturated zone responds to climate changes.  The look-up tables seem to show that higher permeability and capillary strength result in less seepage.  Has DOE found any rocks that fit these characteristics?
A - Most rocks do and tend not to have seepage.  They show a range of possibilities because of future climate change uncertainties.  Want to be conservative.  Don't expect instantaneous/dramatic change in future climate.

Ernest L. Hardin, BSC
Explained how the mass flux of water seepage into and out of drifts overtime is calculated.  Assumptions include that seepage impinges directly on top of the drip shields, and if those shields have breaches, that the seepage then drips through to the waste package (WP), where it snakes around the package and gathers with other seepage drips at the bottom of the drift.  Found seepage in approx. 24 percent of WP locations.

Effect of evaporation or condensation are limited to the availability of moisture - and the drifts are expected to be thermally impacted for at least 10,000 years. 

His study looked at a drift condensation model in three stages: first - at temperature above evaporation level, then - at half evaporation, half condensation levels, and finally - at full condensation.  Calculations assumed a steady state at 1,000, 3,000, and 10,000 years.  Uncertainties included the distribution of the liquid saturation and the ventilation of air volume under the drip shield.  The study found that condensation is required for transfer of radionuclides - he argues for circulation in drifts.

Q - Could condensation occur directly on the drip shield or WP?
A - Yes, under the drip shield or WP.  Drift must have high saturation and be non-ventilated for this to happen.

Q - At post-closure, isn't Yucca proposed to be non-ventilated?  Why use ventilated data?
A - Barometric effect, water will take away nuclides and gas can get in and out.

Q - What is the time required for packages located at the end of a drift to cool?
A - WP near the outside of a drift will cool within 200 years - inside, approx. 1,000 years.

Q - One presentation shows heat reductions bringing vapor into a drift, one shows it flowing out?
A - Different assumptions of boundary conditions account for differences in affects on vapor.

David C. Sassani, Management and Technical Support (MTS)
Robert L. Howard, BSC
"Mass and Activity of Key Radionuclides Potentially Released from Waste Forms, Waste Packages, and Drifts over Time." (PowerPoints available)
David C. Sassani
Source-Term Model Concepts: Within waste packages there is a basket of materials including metals, alloys, and waste that can be degraded and released.

There are two types of waste package (WP) - commercial spent fuel packages, and dual commercial and HLW packages.  His presentation focuses on commercial packages and testing on uranium dioxide fuel pellets which show uranyl alterations through phases of drip tests.  He looked at waste degradation models, dissolved concentrations, and then radionuclide transport.

His spent fuel degradation models envisioned several scenarios:

• Instantaneous release from gap fractions
• Chemical and temperature changes affecting dissolution rates
• Cladding failure - after waste package breached, clad split grows as more externals flow in.  (When outer shell breached they assumed the inner barrier would instantly breach in same area.)
• Radionuclide release due to corrosion - based on temperature, degradation lifetime.  Saw longer lifetimes at lower temperatures.

Drip tests at Argonne National Laboratory on spent fuel pellets provided a good indication of what to expect of degradation in a repository.  It showed a long expanse of time for neptunium dioxide (NpO2) to precipitate.

Additional studies were conducted looking at x-ray absorption, neptunium/uranium ratios, neptunium/plutonium peakage, and neptunium behavior and control factors.

The Science and Technology Source-Term Targeted Thrust Program with the DOE is focusing on:

• The changing conditions within drifts over time
• Identifying critical processes within each time interval
• Targeting radionuclides that are the major contributors to dose

Their projects include study of:

• Spent nuclear fuel dissolution mechanisms and rates
• Formation and properties of U6+ secondary phases
• Waste form and waste package interactions
• Integration of in-package chemical and physical processes

Q - Your model shows substantial scatter of dissolution rates.  Is this broad range a consequence of uncertainties in numbers?
A - Temperature and the oxidation state of carbonate play a large role.  The surface area they found falls within regulation requirements, but they are far from equilibrium conditions.

Q - Are there plans to move these results forward toward a new TSPA?
A - Future of the Science and Technology program is uncertain.  Up to this point, intent is to investigate and find scientific answers (rather than regulatory ones).  This data may never get rolled directly into regulatory models, because those are reasonably conservative.

Robert L. Howard, BSC
Looked into the Yucca Mountain radionuclide inventory and found eight, tracking their activity over time.  Found that Strontium 90, etc., had high activity, but degraded quickly.  Whereas some radionuclides with a high inventory (e.g. U238) had low activity.

Expressed concern about thorium.

Tests saw no waste package, drip shield failures.  Damage to waste package (WP) - tight cracks with no effective flow through cracks, diffusive only.  Releases from drip shield also diffusive only.  The cladding assumed no barrier function.

WPs located on geologic faults produced uncertain results.

Looked into the transport process of radionuclides in terms of spatial domains (waste form, WP, invert, EBS-unsaturated zone).

Found that when WP breached (axial splitting), spent nuclear fuel (SNF) degrades into a porous rind.  Rind saturates quickly and completely below 100C.  Saturation in SNF corrosion products is from absorbed water and is a function of relative humidity.  The diffusive area is a product of the fuel rod length, split opening, and the number of failed rods in the waste package.  HLW glass diffusive area equals the sum of the initial surface area of the five HLW glass logs placed into the WP.

Corrosion transport due to internal components degrading after a WP is breached is a function of the mass of stainless steel and carbon steel available and their respective degradation rates.  Their study did not account for the consumption of water by chemical reactions.

Ground motion modeling cases were done for three potential hazardous occurrences (nominal early failure of WP, igneous intrusion in drift, seismic ground motion) to determine the extent of damage and transport. 

Q - Did your study show that there needs to be water inside the cladding for a swell to result and a release to occur?  Does this mean that there will be no splitting if water does not get inside the cladding.
A - Yes.  It is usually mechanical damage that affects the cladding.

Q - Does your model take into account cladding already damaged when it arrives at Yucca?
A - Yes.

Q - Your study shows that even if it does fail, it would need to be at a time when the drift is below 100C for degradation to occur?
A - No, degradation occurs, but no transport does.

Q - How many of the radionuclides will reach the biosphere?
A - Scenario dependent.  This is a TSPA question, which he doesn't have results for.

Q - Stresses on packaging are occurring, but not penetrating, in first 10,000 years?
A - Yes.  A lot of uncertainty with seismic event though.  Drip shields would make localized corrosion on WP unlikely.

Q - Residual stresses possible?
A - Yes, WPs can bang up against one another and leave residual stresses.

Q - Does the WP experience rate equal about 1% failure?
A - Yes.  This is not a fixed parameter, however.  It could be less than 1 or as high as 1.2 percent.

Q - NRC gave dose profile for long-term performance, including peak dose.  TSPA results were very different.  Why?  NRC model was illustrated and well-caveated, but still, difference explained to be based on release rate.  When are you going to have a source-term that you have confidence in?  Sensitivity of long-term peak dose?
A - Have a source-term now that we have confidence in.  By their pure nature though, TSPAs will change over time.

Q - Project timeline?
A - Science and Technology program looking at secondary phases and package types.  3-4 year program.  Could produce future TSPA.

Q - Relative to failed fuel assembly WP - would hope DOE/industry would put SNF/HLW in zirconium devises to reduce that 1% failure number.
A - No requirements for utilities to do this and don't anticipate they will.  If releases are dominated by seismic or igneous events, should focus on those.

Bruce Robinson, Los Alamos National Laboratory
Bill W. Arnold, Sandia National Laboratory
"Mass and Activity of Key Radionuclides Potentially Released from the Unsaturated and Saturated Zones over Time" (PowerPoints available)
Bruce Robinson
Flow of radionuclides in the unsaturated zone is through fractures in the rock and through matrix flow through more solid rock.  The scientific basis for this is based on several observations at Yucca Mountain and elsewhere. Process flow and transport models are informed through data sets by direct calibration or by consistency checks.

They use a dual permeability particle tracking model with probabilistic travel time delays to account for sorption and diffusion.  Numerous radionuclides were considered.  Advective transport was studied with three scenarios--present, glacial transition and monsoon.  The water table rises in the future due to a wetter climate.  Parameters are probabilistically defined to propagate uncertainty through the unsaturated zone model.

The radionuclides take on various forms while moving through the unsaturated zone.  The model predicts the distribution of arrival times of radionuclides at the water table over the entire repository footprint.  "Normalized breakthrough curves" are used to display the data.  Colloidal species travel most rapidly as they go through the fractures.  Other radionuclides tend to absorb more into the rock and move more slowly to the water table.  The infiltration scenario has a dramatic impact on results.   Sensitivity analysis indicates a low to moderate amount of impact of the breakthrough curve due to uncertainty. 

Varying the diffusion coefficient has a large impact on the curves.  This is captured in DOE's TSPA model.  For the discrete fracture model, a large impact is also seen--dramatic delays in transport.  Travel time is much reduced in fractures versus the matrix, as might be predicted.  The travel time increased along with a greater spread in distribution of arrival times if most waste packages fail as opposed to only a few failing.  But these will vary by radionuclide.

A variety of clarifying questions were asked.  He talked about the inherent nature of these experiments as being on a small scale, as compared to Yucca Mountain, and limited to an observation period of usually five years, the tenure of a graduate student.  Experiments are designed to address key questions.  The process models give confidence in assumptions used in the TSPA abstraction model.

Bill Arnold, Sandia
The model for saturated zone flow has not changed much.  The model is a 3D site scale flow and transport model to simulate radionuclide transport in the saturated (water filled) zone. To date, 200 simulations have been done for radionuclide transport in the saturated zone. 

Advective flow of groundwater is through relatively limited volume of fracture networks.  The groundwater flow is more uniformly distributed in the porous medium of the alluvium.  The model runs show that uncertainly in advection encompasses a significant uncertainty overall relative to the flow of water.   Radioactive decay is significant in that there will be no transport if it has disappeared.

Relative to matrix diffusion, there are uncertainties in travel time due to the types of rocks and fracturing and chemistry issues.  Various sensitivity studies showed varying levels of matrix diffusion depending on the radionuclide. 

Another test involves sorption--how a radionuclide absorbs into the surrounding environment whether it be alluvium or the rock matrix.  In a run with neptunium (Np), the uncertainty in Np sorption encompasses a moderate portion of the overall uncertainty.

There are many parameter uncertainties related to groundwater flow and geology, and radionuclide transport. 

Q--Will some radionuclides be released to the environment within 20 years? 
A--Need to be careful in interpretation.  Only 4 out of 200 runs show release to the environment within 10 years.

Q--Where should you focus your work to eliminate uncertainty? 
A--It's somewhat subjective. 

Q--Is most of the flow in the fracture channels? 
 A--Yes, but the data is not entirely definitive.

Public Comments:
1) Sally Devlin - Compliments every report, improved science over the past five years.  TSPA - licensing off for another two years (she created and sent report on it).  Resents modeling - need reality.  Have nothing for canister (need better alloy than 22), no design for mine.  Start testing on canister in mine immediately.

2) Grant Hadlow, Chemical Engineer - Valuable resource in John Garrick.  Bad news - 20 years of junk science.  Murder of Paul Brown - gangsters got a large portion of money for Yucca.  Can't be bribed.  No material can stand decomposition for 100 years.  Treat Yucca as a depository.

3) Additional slides from Ernie Hardin's presentation requested.

Mark Nutt, MTS
"Management and Technical Support Peak Dose Sensitivity Analysis" (PowerPoint available)
Looked into a simplified model to evaluate Features, Events, and Processes (FEPs) that influence peak annual dose of radionuclides.  This is NOT a compliance model.  He acknowledged that the board has expressed dissatisfaction with the lack of result curves, but these are provided in the report.

Mr. Nutt used documentation current as of early 2005 and historical Yucca Mt. info regarding post-closure repository performance to consider FEPs over a 10,000 year post-closure period.  The simplified model includes representative FEPs that could potentially affect the peak dose (FEPs that have either a minor or no effect on the peak dose were generally not included).

Study assumed that slow and/or infrequent degradation could come from engineered products (drip shield, waste package, and pallet), but that these are of secondary importance to primary degradation modes:

• Slow stress-induced degradation in emplacement drifts
• Infrequent stress-induced degradation incurred with seismic or volcanic events

The evaluation found that the vast majority of screening justifications applicable to a 10,000 year period are also appropriate over a time period that covers the peak dose.

Key assumptions in his study included: a constant climate state, an assumed seismic event causing drift collapse over the period of peak dose, and that general corrosion is the only corrosion-related process.

Study's key findings:

• The peak annual dose depends on a nominal degradation process.
• Seismic events will occur, but are not expected to have a significant effect on peak annual dose.
• Igneous intrusions may influence dose prior to the onset of significant WP failure due to general corrosion, but is not expected to have significant effect on peak annual dose

Sensitivity analyses - Factors that affect peak annual dose:

• Infiltration rates and percolation rates through the unsaturated zone have a minor effect on the magnitude of peak dose.
• Drift seepage has a significant effect on the magnitude of peak dose.
• WP performance (general corrosion rates) have a significant effect on peak dose estimate - including magnitude and timing.
• Drip shield performance has a minor effect on peak dose.
• The natural barrier functions as an effective barrier for several key radionuclides over the period of peak dose.

Official previous NWTRB meeting transcripts and agendas available at: http://www.nwtrb.gov/meetings/meetings.html.